COVID-19 Questions and Answers (updated daily, Part 1)

I am a trained public health professional. Colleagues and friends frequently ask me questions about COVID-19, so I started this daily Q&A to respond. It helps keep me informed and sane. Hopefully these updates are helpful to you too! [These answers are my own and do not represent any organization I’m affiliated with.]

NOTE: On 8/27/2020, I ran out of space here, so I started a PART 2, which is online here.

Q&A for 8/26:

#Partisan Divide

Question: I’ve said how much I like reading these. Here’s my contribution on the first paragraph of yesterday’s post. Mask wearing is definitely politicized (what hasn’t been?). However, I’d remind anyone who would listen that pulling correlations/trends from grouped data is tricky business (e.g. Simpson’s Paradox). Related to your referenced Gallup survey for example, since so many more of the people who live in cities are democrats (order of 2 to 1), one could expect Democrats to wear masks significantly more than Republicans for no other reason — much harder to social distance, more contact with strangers, etc. The same explanation — demographics in cities — can be made for age btw and the Gallup data supports it. So, for more accurate (albeit less entertaining) left v right comparison: Republican males aged 20–40 living in the city vs Democratic males aged 20–40 living in the city.

Answer: Thanks for your contribution! I agree that univariate analysis is insufficient and I also agree that we must be cautious when drawing correlations based on overly simplistic analyses, let alone mistaking correlation for causation. I definitely would not want to be sharing data for entertainment, especially if it contributes to political polarization. So your contribution got me to thinking — have any researchers conducted more complex analyses to examine whether partisanship plays a role in behaviors like mask wearing or social distancing (including by controlling for numerous potentially confounding factors — like geography, age, sex, COVID burden, etc.)? As it turns out, they have!

Based on what I’ve read (see brief synthesis bulleted below) even after controlling for other factors, partisanship is significantly correlated with COVID-related behaviors. Nonetheless, as Dr. Sandro Galea, an epidemiologist and the dean of the Boston University School of Public Health who studies the politics of public health recently shared with FiveThirtyEight, “Nobody should ignore the fact that people on the political extremes are embracing polarizing positions on health behavior that should not be polarized, but I think the evidence we have indicates that most people have tried to be responsible and adopt the recommended behaviors, even at a time of immense polarization and confusion and discomfort.”

• This paper published earlier this month in Journal of Public Economics, Polarization and Public Health: Partisan Differences in Social Distancing during the Coronavirus Pandemic,”use[s] location data from a large sample of smartphones to show that areas with more Republicans engaged in less social distancing, controlling for other factors including public policies, population density, and local COVID cases and deaths. [Authors] then present new survey evidence of significant gaps at the individual level between Republicans and Democrats in self-reported social distancing, beliefs about personal COVID risk, and beliefs about the future severity of the pandemic.” The authors conclude, “Our empirical results show that partisan gaps in beliefs and behavior are real. GPS evidence reveals significant partisan gaps in actual social distancing behaviors. Survey evidence shows substantial gaps between Republicans and Democrats in beliefs about the severity of COVID-19 and the importance of social distancing. The raw partisan differences partly reflect the fact that Democrats are more likely to live in the dense, urban areas hardest hit by the crisis, and to be subject to policy restrictions — in other words, to face stronger individual incentives for social distancing. Even after controlling carefully.” [Note: this paper’s introduction section also gives a nice overview of other related research]
• This preprint publication (not peer reviewed) from three scientists at Syracuse University, UC Irvine, and Cornell, Partisanship, Health Behavior, and Policy Attitudesin the Early Stages of the COVID-19 Pandemic, presents analyses of survey data collected from 3,000 Americans during March 20–23. After controlling for age, marital status, race, income, education, geography (state), interest in the news, and frequency of news consumption, researchers found that “Republicans are less likely than Democrats to report responding with CDC-recommended behavior, and are less concerned about the pandemic, yet are more likely to support policies that restrict trade and movement across borders as a response to it. Democrats, by contrast, have responded by changing their personal health behaviors, and supporting policies that socialize the costs of testing and treatment. Partisanship is a more consistent predictor of behaviors, attitudes, and preferences than anything else that we measure.”

Q&A for 8/25:

#Face Masks #Protection

Question: I am currently in South Dakota where I would estimate that only about 1/3 of people are wearing masks indoors (at Walmart, etc). If I diligently wear my mask, how much protection do I get if they aren’t also wearing masks?

Answer: Thank you for wearing your mask and please keep doing so. Your mask wearing protects others, is likely to protect you, and helps to establish new social norms. As we’re all learning (many of us first hand), mask wearing varies dramatically by geography (Figure 1) and is highly correlated with political party affiliation. Indeed, a Gallup poll from July found that the majority of people across every demographic group reported wearing a mask in public very often or more frequently with one exception; “The one exception is Republicans, among whom a majority say they wear masks infrequently — either sometimes (18%), rarely (9%) or never (27%).”

When it comes to the protection masks afford, the consensus is solid that wearing a mask protects others by keeping respiratory droplets from spreading more widely (see Q&As of 6/29 and 6/22 #Face Masks). What’s less clear is whether and to what degree mask wearing protects the wearer. As discussed in our Q&A of 6/29, a meta-analysis published in Lancet found that wearing a mask dramatically reduced the risk of infection among health workers in hospital settings (absolute risk of infection 17·4% with no face mask vs. 3·1% with a face mask). It’s unclear, however, how applicable these risk reduction findings are to non-clinical settings. Like so much related to COVID, we need more research.

In the meantime, a group of medical doctors recently hypothesized in a perspective recently published in the Journal of General Internal Medicine that wearing a cloth mask also benefits the wearer by reducing the amount of virus the wearer is exposed to and thereby “leading to more mild and asymptomatic infection manifestations.” While the idea still needs to be tested, it is nonetheless another interesting avenue to explore and points to yet another reason why universal masking needs to be a key component of pandemic response.

Figure 1: Mask Wearing in the United States (snapshot of mid-July, from NY Times)

Q&A for 8/24:

#Excess Mortality

Note: I’m back from a fantastically relaxing vacation!

Question: A friend of mine questions the Covid death count, asserting that we do not know that each institution reporting their count can be relied upon to provide accurate numbers (i.e. some of the individuals they are reporting as having died of Covid may have, in fact, died of something else, but Covid is listed as the cause of death). He went on to say that years from now, when we look back, we will likely see that the total rate of death from all causes during the pandemic was pretty much on par with the total rate of death from all causes in prior years; now, people are dying from Covid as compared to prior years, when they were dying from other illnesses or health complications. Is there clear data that demonstrates the overall rate of death from all causes is significantly higher during the pandemic than it has been in recent years?

Answer: We talked about mortality reporting in our Q&A of 5/5 (#Coding Deaths), about COVID mortality as compared with influenza mortality in our Q&A of 5/22 (#Monitoring), and about case fatality rates in our Q&A of 6/10 (#CFR). These posts have touched on the issues of COVID death monitoring, and largely answer your friend’s first issue/question about reporting. None of these posts, however, answer your friend’s second question/assertion on all cause deaths, so let’s address it! The short answer is: To understand the impact of COVID-19, we need to calculate “excess mortality.” In doing so, we see that deaths are MUCH higher (already 200,000+ deaths higher) during the pandemic than in other recent times. Read on for more information…

Excess Mortality Measurement

OurWorldInData offers a nice overview of excess mortality and how it’s measured. It’s a simple concept: excess mortality is the number of deaths above what would have been expected under “normal” circumstances (excess mortality= observed number of deaths during crisis period — expected number of deaths during normal period).

Excess Mortality in the USA

The NY Times ran a fantastic article with loads of great data visualizations ~10 days ago, finding that “Across the United States, at least 200,000 more people have died than usual since March… This is about 60,000 higher than the number of deaths that have been directly linked to the coronavirus.” Meanwhile, earlier in July, JAMA published a couple of papers that also reveal large increases in excess mortality in the United States — one paper concluded “the number of deaths due to any cause increased by approximately 122 000 from March 1 to May 30, 2020, which is 28% higher than the reported number of COVID-19 deaths” while the other paper concluded, “Between March 1, 2020, and April 25, 2020, a total of 505,059 deaths were reported in the US; 87,001 (95% CI, 86,578–87,423) were excess deaths, of which 56,246 (65%) were attributed to COVID-19.” JAMA’s accompanying editorial concluded “The United States and other countries have been focused on understanding the mortality and morbidity that can be directly measured among individuals who have tested positive for COVID-19. However, the 2 new studies published today in JAMA and JAMA Internal Medicine suggest that up to one-third of excess deaths during the pandemic may occur in those who have not tested positive for COVID-19. Thus, these studies underscore the importance of continuing to measure excess deaths…” Finally, CDC offers regularly updated data on excess deaths, including data visualizations here. For ease of reference, I’ve included CDC’s excess death bar chart herein (Figure 1). Importantly, data reporting is slow, so the number of deaths in recent weeks is incomplete, which means that excess deaths will be higher than what is shown below. Nonetheless, since March the number of weekly deaths in the United States has been statistically significantly higher than would be expected in “normal” times. The data are clear and conclusive!

Figure 1. Excess Deaths in the USA (note: data for recent weeks are incomplete) (from CDC as updated 8/19/2020)

Q&A for 8/7:

#Immunity #Antibody Tests #PPV

Note: I will be on vacation from 8/8 through 8/22. I will not be writing daily Q&As while away [unless the mood strikes]. Please feel free to keep sending me your questions and know that I’ll answer them when I return, if not sooner.

Question: I just took an antibody test in DC! How much weight should I put on the results?

Answer: Short answer: the validity of the test results depends as there are several important issues at play, described below. In terms of the results, here’s my take:

• If you live in DC, had been sick with COVID-like symptoms earlier in the year, and you get a positive result back from the antibody test, the odds are pretty favorable that it’s a true positive;
• If you had not been sick with COVID-like symptoms earlier in the year and you get a positive result, the validity of the test is murkier;
• If you get a negative result, you can be assured that you do not have B-cell antibodies, but you won’t know anything about T-cell immunity; and
• We still don’t know how protective antibodies are and for how long, and even a positive antibody result is not a definitive positive, so please keep doing all the right public health practices — mask wearing, 6+ feet distance, hand washing, staying home when sick.

Here are the issues affecting result validity:

1. Your own health history: if you had an illness earlier this year that was accompanied by COVID-like symptoms (described in Q&A of 7/6 #Symptoms) then a positive test result is more likely to be a true positive.
2. The sensitivity/specificity of the test: Every diagnostic test has its own sensitivity/specificity (e.g. ability to detect presence/absence of disease; described in Q&A of 4/15). As CDC describes, since population prevalence of SARS-CoV-2 is still low, tests with higher specificity are preferred.
3. The population-level prevalence of the disease: The ability of a diagnostic test to accurately capture true positives/true negatives depends on the overall prevalence of the disease in the population. The lower the population-level prevalence, the higher the false positivity rate. I really like the Q&A of 4/15 on this topic, so if you’re curious check it out! For the US, CDC shares prevalence estimates here) and shared a write up on the first round of prevalence estimates in this paper published three weeks ago in JAMA, which concluded “it is likely that greater than 10 times more SARS-CoV-2 infections occurred than the number of reported COVID-19 cases.”
4. Lasting antibodies: Antibody tests look for the presence of B cells, which are the cells (lymphocytes) that identify invaders and basically mark them for other immune cells to destroy (like military intelligence) (for immune system refresher, see Q&A of 5/9; nice graphic from Science commentary in Figure 1). It seems that for some people who have recovered from COVID, their B cells quickly dissipate and may not be present in antibody tests. But B cells are only part of the immune response. Remember T cells? T-cell immune response could be long-lasting even if B cell response is not (see Q&A of 8/6). So no presence of B cells in the tests does not necessarily mean no presence of some underlying immunity. That said, COVID is still new and we have no idea how long immunity lasts or how protective antibodies are. So far, the fact that we have little evidence of re-infection is quite reassuring (more on that in Q&A of 7/23).

And diving into DC specifically, according to the Washington Post, DC is using 2 antibody tests — DiaSorin Liaison assay or the Abbott Alinity i serology test. The FDA shares sensitivity/specificity estimates here, and I’ve copied the Abbot table here for reference (Table 1). As the FDA estimates, the Abbot test has a positive predictive value (PPV) (i.e. probability that a positive is a real positive) of 84% when prevalence is 5%. Restated — there’s an 84% chance that the positive result is a true positive. If the population prevalence is <5%, the PPV will be much worse. If it is >5%, PPV will be much better. We don’t know what the population prevalence is in DC, but we could make a few guesses. DC has a population of ~700,000 and has 12,518 cases. If DC were finding all cases, the prevalence would be ~1.8%. But we know that we’ve been missing cases. If we extrapolate from CDC’s findings described above — that for every case we found in March-May, we missed 10 cases — then prevalence would be MUCH higher (getting closer to 20%). That is far too high given hospitalizations and the like in DC. Plus, according to the Washington Post, only 6% of antibody tests have been positive, and since there’s likely bias in who is getting an antibody test (people more likely than the general population to have been infected), the population prevalence would presumably be lower than that. All that to say, the FDA’s use of 5% prevalence to estimate positive predictive value is probably a pretty good choice for the DC-area.

Figure 1. Immune Response (from Science)

Table 1. Abbott Sensitivity/Specificity (from FDA)

Q&A for 8/6:

#Immunity #T-cells

Question: Is there any research as to whether people who have been infected with previous coronavirus-type infections in the past have some level of immunity to this strain?

Answer: Short answer== YES! One of the big outstanding questions is why the severity of SARS-CoV-2 infection varies so dramatically from person to person. We talked about this issue a little back in our Q&A of 5/9 (#Kids), which also provides a very quick synthesis of how the immune system works. And in our Q&A of 5/17 (#Immunity), we talked about emerging evidence that “pre-existing SARS-CoV-2−crossreactive T cell responses were observed in healthy donors, indicating some potential for pre-existing immunity in the human population.”Since then, we have more information. Yay for science! Research results published three weeks ago in Nature, “SARS-CoV-2-specific T cell immunity in cases of COVID-19 and SARS, and uninfected controls” and nicely synthesized in this Science commentary show three fascinating things:

1. T-cells appear to be key to lasting immunity; T-cell immune response persists among recovering COVID-19 patients. Those study participants who are recovering from SARS-CoV-2 (COVID-19) (n=36) all had T-cells (CD4 helper and CD8 killer) indicating ongoing immunity to the virus. As speculated in the Science commentary, “T-cell driven immunity is perhaps the way to reconcile the apparent paradox between (1) antibody responses that seem to be dropping week by week in convalescent patients but (2) few (if any) reliable reports of actual re-infection.”
2. Lasting immunity; T-cell immune response is long-lived among those recovered from SARS. Those who were infected with SARS-CoV-1 (e.g. SARS) have lasting immunity. 17 years(!) after SARS infection, all 23 study participants had a robust T-cell immune response. This bodes well for hopes of long-term immunity to SARS-CoV-2.
3. Cross-reactive immunity; T-cells from SARS and other zoonotic coronaviruses are cross-reactive to SARS-CoV-2. Study participants who had recovered from SARS and a number of study participants who were never exposed to SARS or COVID have T-cells that are reactive to SARS-CoV-2! Interestingly, these T-cells do not react to proteins found in typical human coronaviruses, but instead react to those found typically in animal-based coronaviruses. Again, as posited in the Science commentary, “That would argue that there has been past zoonotic coronavirus transmission in humans, unknown viruses that apparently did not lead to serious disease, which have provided some people with a level of T-cell based protection to the current pandemic.”

Q&A for 8/5:

#Mutations #Vaccinations

Question: In response to yesterday, Yikes! So, would the vaccines currently in development stages work against the G strain? Or only for the D strain?

Answer: One clarification from yesterday that’s related to your question and then the answer to your question. In short, we expect viruses to mutate. So far, the mutations of SARS-CoV-2 have been slow and not too dramatic, which means that vaccines in development now are more likely to be effective when rolled out [early next year, fingers crossed].

SARS-CoV-2 Strains

We’ve known about the G strain for some time (since March, which is a long time in the scheme of knowing things about SARS-CoV-2). And Vietnam has previously experienced cases from the G strain. As stated by the Minister of Health in late-July, and as the GISAID database confirms, Vietnam has previously recorded five strains (aka clades) among confirmed cases. Apparently, what’s come up in Vietnam is a sixth strain. Unfortunately, I haven’t found information about what this strain is. We do know — thanks to the GISAID database and this recent study published in Frontiers in Microbiology — that there are six major strains of SARS-CoV-2. Study authors report, “Currently, the G clade and its offspring, GH and GR, are the most common clades amongst the sequenced SARS-CoV-2 genomes, globally accounting for 74% of all world sequences. Specifically, the GR clade, carrying the combination of Spike D614G and Nucleocapsid RG203KR mutations, is currently the most common representative of the SARS-CoV-2 population worldwide. The original viral strain, represented by clade L, still accounts for 7% of the sequenced genomes, and the other derived clades S and V have similar frequencies in the global dataset.” So I’m not sure whether the Vietnam strain is one of these or a different strain. Check out Figures 1 and 2 to see the viral clades over time and by geography (note: O= others).

Vaccines

Good news is that scientists are still observing that SARS-CoV-2 is slow to mutate. Molecular biologist, Dr. Peter Thielen, recently described on NPR’s All Things Considered that “viruses circulating today look remarkably similar to the ones that appeared in China late last year…The targets for diagnostics and the targets for vaccine design still today remain the same as we would’ve designed them in January.” Similarly, biologist Dr. Benjamin Neuman recently told Healthline, “The virus is still so similar now to the initial sequence that there isn’t really much reason to think the differences will matter in terms of vaccine.”

Figure 1. SARS-CoV-2 Strains Over Time (from GISAID)

Figure 2. Recent Distribution of SARS-CoV-2 Strains by Country (from GISAID)

Q&A for 8/4:

#Mutation #Virulence #Vietnam

Question: What fresh hell is this?! Vietnam fights new COVID-19 strain with higher infection rate

Answer: For those of you who haven’t read recent news out of Vietnam, here’s a synopsis — after months of successfully containing the virus, Vietnam is in the midst of an outbreak that local researchers (e.g. CDC-Vietnam) and officials (e.g. Health Minister) believe is the result of a more virulent strain that has also recently been seen in Bangladesh, UK, and Ireland. So what do we know?

Mutations and Virulence/Severity

While it is relatively easy to recognize viral mutations, it is difficult to understand how they impact virulence and disease severity. We know, for example, that the viral genome has become dominated over time by a one-base mutation strain (G strain) (Figure 1). And thanks to this lab-based study published in early July in Cell, we now have evidence to suggest that the G strain is more virulent than the original D strain. Important note: evidence from this study does not support the notion that the G strain causes more severe outcomes. The strain identified in Vietnam is yet another genetic variant. We need more data — and we need to see the data Vietnam is using to make these determinations — in order to understand how this new mutation impacts virulence and severity.

More Background

On 29 July, Reuters reported that “Prime Minister Nguyen Xuan Phuc said the current wave of infections was different to a second wave Vietnam fought in March…” and on 2 August, Reuters reported that “Health Minister Nguyen Thanh Long said the strain of the virus detected in the new outbreak is a more contagious one… each infected person may infect about 5–6 people compared to 1.8–2.2 people in the previous period.” Vietnam has submitted the genetic code to the GISAID database for further study. Meanwhile, WHO/Vietnam pushed back on the Vietnamese government’s assessment, stating “Based on what is known now, the contagion of the virus and the severity of the disease it causes have not changed. There is no evidence that the mutations observed so far have led to an increase or decrease in the virulence or severity of the disease.” Finally, for some basic background on viral mutations, see our Q&A of 3/25 (#mutations) and more recently, this news report from mid-July published in Science.

Figure 1. Changing Genome (from Science)

Q&A for 8/3:

#Age Trends

Question: I’ve been reading about how cases are trending younger in lots of different states. Is this true nationally?

Answer: I was reading that too, with Dr. Fauci recently having stated earlier in July that “the average age of people getting infected is now a decade and a half younger than it was a few months ago.” Still, I hadn’t been able to find the supporting data. CDC shows cumulative age distribution of cases here, but because these aren’t time trend data, they don’t help us answer the question. A few days ago, CDC released its updated case surveillance public use dataset. And my friend, Dr. YJ Choi (find some of her other COVID-related work here) ran the numbers to see national trends. Sure enough, cases across the country are getting younger (Figure 1).

Figure 1. Weekly reported cases by age (%)

Q&A for 8/2:

#Cross-country differences #Deaths

Question: Follow up question based on yesterday’s graphic. It seems that “western democracies” have a much worse per capita death rate (New Zealand doesn’t count because they are a far away island that was able to close themselves off). What do we think is causing that? Is it that more developing or authoritarian countries either can’t or won’t share accurate death counts? Or do we think the rate of chronic health conditions are higher in richer countries thereby increasing the number of vulnerable people?

Answer: When it comes to variations in per capita deaths around the world, there are so many factors at play. This Q&A of 6/10 (#CFR) offers a good overview (pat myself on the back) of reasons case fatality rates may differ, and these reasons are also applicable to your question on per capita deaths. I’ll elaborate a bit more on differences between high and low-income countries. The reasons bulleted below help describe why deaths are or at least appear to be lower in low-income countries. [Note: I recognize that your question was about governance type rather than income level. I’m focusing here on income level because it’s a bit easier to parse. There is some research linking income level with governance type, but that relationship is still debatable.]

• Limited data in low-income countries. Very few low-income countries have functioning civil and vital registration systems, which means that in order to track deaths and cause of death, we must rely on surveys and indirect estimation. This makes it harder to track deaths in real time and oftentimes means that we have only a hazy understanding of the number of deaths and cause of death. It also means that countries with poorer reporting systems *appear* to be weathering the pandemic better, if we’re only basing judgement on statistics. It can also mean that some countries *appear* to be faring worse. For example, earlier in the pandemic, Belgium had one of the highest per capita death rates in the world, but this cross-country difference was largely attributed to Belgium’s comprehensive reporting system.
• Lack of transparency. We discussed a few data transparency issues related to China and Brazil in our Q&A of 6/26 (#Cross-country comparisons). We also discussed some data transparency issues in the United States in our Q&A of 7/15 (#Surveillance) [there are many more beyond that, will save for a different post]. Leaders concerned with maintaining power may be more inclined to keep information that paints them in a poor light away from the public. This would seem to be more a hallmark of authoritarianism, but even “western democracies” have had their share of challenges when it comes to COVID and data transparency. Back in May, for example, the government of the UK was refusing to disclose data from nursing homes aka “individual care homes.”
• Delayed introduction of COVID-19 in low-income countries. As shown in Figure 1 from this paper published a few days ago in Lancet Infectious Diseases, which describes how COVID traveled around the world during the pre-pandemic period, “During the first 11 weeks of the COVID-19 outbreak outside mainland China, cases were detected in half of all countries and locations globally, with an acceleration in case detection during weeks 9–11 of the outbreak. Although most countries and locations in the WHO European, Eastern Mediterranean, and South-East Asian regions reported confirmed cases by the time WHO characterised the outbreak as a pandemic, only a third of countries in the Americas and African WHO regions had reported cases, suggesting delayed introduction, delayed detection, or both.”
• Population age distribution. We know that COVID has a more severe impact on people of older ages. As a result, in countries with larger youth populations (generally low-income countries) we’d expect to see fewer per capita deaths.

Figure 1. Countries and Locations with Confirmed Cases of COVID-19 (12/31/19–03/10/20) (from Lancet Infectious Diseases)

Q&A for 8/1:

#Sweden

Question: I am not sure if you addressed this already but could you share your take on the Swedish example? How were they able to stay relatively open as a society? Let’s stipulate that the Swedes take their social responsibilities seriously and wear masks and social distance as they are being asked. Let’s also agree that access to medical care for everybody might be better in Sweden and that the Swedish society is overall healthier. Are there any other factors in play here? What can we learn from the Swedish experience if anything?

Answer: Sweden decided not to lock down, following the advice of its state epidemiologist, Dr. Anders Tegnell. It’s been interesting to follow the trajectory of Sweden over the months since it bucked the response of its Scandinavian neighbors and much of Europe. In my opinion, this BBC article published last week does a nice job laying out the current results of Sweden’s response.

• Sweden has one of the highest numbers of per capita deaths in the whole world, even higher than the United States and Brazil. (Figure 1)
• Sweden’s economy is contracting. According to the BBC article referenced above, “ various forecasts predict the Swedish economy will still shrink by about 5% this year. That’s less than other countries hit hard by Covid-19 such as Italy, Spain and the UK, but still similar to the rest of Scandinavia. Sweden’s unemployment rate of 9% remains the highest in the Nordics, up from 7.1% in March.”
• Sweden is still far, far from herd immunity. According to Swedish Public Health Agency research released in mid-June, about 6% of the population is known to have antibodies.
• Compared with its neighbors, Norway and Finland, both of which more aggressively responded to COVID, Sweden is faring worse in terms of health and economics.
• For trends in cases, test positivity, and deaths in Sweden and other countries, check out ourworldindata.org.

At this stage, it seems to me that the Swedish experience is a cautionary tale. That said, given the “marathon vs sprint” approach Sweden has taken, it’s possible that longer-term benefits are still to come. Even so, for those families who have already lost a loved one — the high per capita death rate means it’s a lot of families — I doubt that optimism for the future holds much weight in light of the tremendous loss.

Figure 1. Mapping Per Capita Deaths (from OurWorldinData.org)

Q&A for 7/31:

#Mortality #Age #Life Expectancy

Question: Trump said yesterday that the average age of death was 78, so I was trying to verify that number since we hear so much about younger people dying that I figured he was exaggerating data again, and I came across this chart from the CDC: https://www.cdc.gov/nchs/nvss/vsrr/covid_weekly/index.htm. Is this right? It looks like death rates for all age groups was really high in April, but has fallen substantially for all age levels to almost zero for all but the oldest age groups, and they are now in the less than 100 range. My two questions are: is this data right or is there some delay in reporting or something? Or have drs and hospitals figured out how to prevent death in the vast majority of patients?

Answer: So much to unpack here! The demographer in me loves this question! Let’s dig in (but not too deep since I’m trying to keep this relatively short). First, when it comes to average age of death, Trump was not exaggerating (though it’s more complex). Second, CDC’s NCHS data are always several weeks behind, so the data from the last several weeks is not representative of the true picture (ignore it). That said, there is a modest change in the proportion of deaths by age (Table 1) with people ages 45–64 more recently accounting for a larger proportion of deaths. Third, doctors and hospitals have improved their patient care thanks to experience and evidence, which seems to have improved patient outcomes though data are still limited.

Average Age of Death Varies by Race and Ethnicity

Based on this report published a couple of weeks ago in CDC’s Morbidity and Mortality Weekly Report (MMWR), the average age of death varies dramatically by race and ethnicity. Using data compiled as of 18 May, the authors find “median age was 71 years (IQR = 59–81 years) among Hispanic decedents, 72 years (IQR = 62–81 years) among all nonwhite, non-Hispanic decedents, and 81 years (IQR = 71–88 years) among white decedents. The percentages of Hispanic (34.9%) and nonwhite (29.5%) decedents who were aged <65 years were more than twice those of white decedents (13.2%)… Among the decedents during February 12–April 24, 2020, for whom supplementary information was provided, 9,997 (93.9%) resided in New York City, New Jersey, or the state of Washington, three areas with early widespread circulation of SARS-CoV-2; the median age among decedents in these three jurisdictions was 75 years, (IQR = 64–84 years). The median age among decedents residing in the other 13 jurisdictions was similar (78 years, [IQR = 68–85 years]).” [Note: IQR means inter-quartile range and it’s a measure of the spread of the data — for example, among Hispanic decedents, the middle 50% of all deaths were among people between the ages 62–81.]

CDC’s National Center for Health Statistics (NCHS) Data Are Always Several Weeks Behind

We talked about mortality reporting in our Q&A of 5/5 (#Coding Deaths), which also describes how death certificate data captured by CDC is always several weeks delayed. This means that the sharp decline in deaths over the last several weeks that you observed in NCHS’s “Provisional Death Counts” chart is due to data reporting lag rather than sharp decline in deaths. That said, deaths have fallen dramatically since their peaks in April. Unfortunately, deaths are back on the rise (Figure 1).

Patient Care has Improved

We know so much more every day and as a result, patient care is improving as are patient outcomes. We talked about some of this evidence in our Q&A of 5/31 (#Outcomes) The ongoing challenge (beyond the virus itself) is that in the context of sustained, widespread transmission, local health systems may become overloaded and in that case, patient care and patient outcomes will suffer.

One Other Note of Interest — Life Expectancy

1) This not yet peer-reviewed article by a world renowned demographer estimates that due to COVID, “In the US, life expectancy is projected to decline this year by more (−.68), [a reduction greater] than the worst year of the HIV epidemic, or the worst three years of the opioid crisis, and to reach its lowest level since 2008. Substantially larger reductions, exceeding two years, are projected for Ecuador, Chile, New York, New Jersey and Peru.”

Table 1. Proportion of Weekly Deaths by Age Group (data from NCHS)

Figure 1. Daily COVID Deaths in the USA (data from covidtracking.com)

Q&A for 7/30:

#School Reopening #Criteria #Teachers #R0

Question: I was just reading in Politico that the president of the American Federation of Teachers is warning that teacher strikes may be coming to protest unsafe working conditions if schools reopen in states where infections are surging. According to the article, “The union, which represents 1.7 million educators in the United States, adopted a resolution this month that says schools should only open in places where the average daily community infection rate among those tested for the coronavirus is below 5 percent and the transmission rate is below 1 percent.” Which states meet this criteria and where can I find these state-by-state data?

Answer: These criteria are written slightly differently than I usually see them. Put another way, the first indicator, “average daily community infection rate among those tested for the coronavirus is below 5 percent,” is “average test positivity rate <5%” and here I would expect the average to be the 7-day rolling average. We can get these data directly from covidtracking.com or we can use this data visualization tool from Johns Hopkins, which pulls data from covidtracking.com. When it comes to the second indicator, “transmission rate is below 1 percent,” this is the reproduction rate — R0 — and it is a measure of the average number of people an infectious person infects. Anything above 1 (not 1% as stated in Politico) is an indication that the virus spread is growing rather than contracting. R0 is a relatively simple concept, but it is challenging to measure and requires mathematical modeling, as described by this 2019 paper from Emerging Infectious Disease. RT.live (a “public service” site run by the cofounders of Instagram) uses case count and test positivity rate data from covidtracking.com to model R0 by state (Figure 1). For ease of reference, I made a table that shows which states meet the test positivity rate and R0 criteria (Table 1). As of July 29, only 6 states meet the American Federation of Teachers’ criteria for reopening — Delaware, Maine, Michigan, New York, Ohio, and Vermont.

Figure 1. Reproduction Rate by State (from RT.Live)

Table 1. Reopening by State

Q&A for 7/29:

#Waves

Question: What is the definition of a second wave? How do we differentiate wave 2 from wave 1? E.g., Does Louisiana currently have the second wave? Or, is it just first-wave-not-properly-flattened-and-now-bursting? It’s happening in Japan and Austria too. Of course their peak is so low…

Answer: I was wondering about this too, so thanks for asking! As I understand it, there is no official epidemiological definition of a “wave.” According to WHO’s “Pandemic Influenza Preparedness and Response: A WHO Guidance Document (2009)”, a second wave is simply the resurgence of the virus after a post-peak period — e.g. resurgence in cases and/or deaths in the time after pandemic activity appeared to have been decreasing. Most of us in public health hear “waves” and think of the course of deaths in the UK (and world, which followed a similar pattern) due to the 1918 flu (Figure 1). When it comes to COVID-19, the United States still hasn’t gotten over its first wave. Instead, we had a wave of cases that never crested. Cases reached a peak, plateaued there, and then grew again (Figure 2). If we were to zoom in on different states, however, we’d see that some are in the post-peak period and are over the first wave (like New York, Figure 3) while others are still in their first wave (like Tennessee, Figure 4). When it comes to a state like Louisiana, I think it fits the criteria of being in a second wave — it was in a post-peak period and then experienced a dramatic increase in cases (check out charts here). Finally, if you’re interested in reading another article on this topic, this one from The Conversation shares an epidemiologist’s description of waves.

Figure 1. Three waves of 1918 Flu, Deaths in the UK (from Emerging Infectious Diseases)

Figure 2. Cases in the USA (excluding New York State) — first wave plateaued then rose (data from covidtracking.com)

Figure 3. Cases in New York — wave 1 over

Figure 4. Cases in Tennessee — first wave still rising

Q&A for 7/28:

#Private School

Question: Following up on yesterday’s Q&A: your thoughts about private/parochial schools vis a vis public schools?

Answer: In comparison with public schools, private schools have fewer schools, a smaller student body, and fewer challenges (like busing) to take into consideration. They also generally have greater financial resources to put in place the safeguards CDC recommends (e.g. cohorting students, improving ventilation, expanding cleaning and disinfecting, conducting daily health checks, implementing staggered schedules, etc). For these reasons, private/parochial schools may have a greater degree of flexibility and more feasible options available to them for in-person learning as compared with public schools. That said, members of private/parochial schools are still members of their broader community. And in the context of widespread community transmission, private/parochial students and their families are at risk, especially if they are congregating in groups and indoors. While social, racial, and economic inequality are highlighted and amplified by COVID-19, it’s nevertheless still the case that all people — regardless of social standing, race, class — are at risk of acquiring, transmitting, and suffering from the virus, even those who attend private/parochial school.

Q&A for 7/27:

#School Reopening #Thresholds #Benchmarks

Question: Is there a threshold for community transmission that schools should consider for closing/reopening?

Answer: CDC released its updated school return guidance late last week. I’ve copied a few key portions below for ease of reference (Table 1). CDC suggests school closing only in the context of:

• substantial, uncontrolled [community] transmission — “If there is substantial, uncontrolled transmission, schools should work closely with local health officials to make decisions on whether to maintain school operations.”; and
• someone in the school tests positive — “If someone within the school community (e.g., student, teacher, staff) tested positive for SARS-CoV-2, assessing the level of risk is important to determine if, when, and for how long part or all of a school should be closed.”

With regard to “substantial, uncontrolled transmission,” CDC describes levels of community transmission here (Table 2), but these levels do not include specific numeric thresholds. Within its guidance, CDC suggests that school systems use “indicators such as healthcare capacity (e.g., staffing, ICU bed occupancy), changes in newly identified COVID-19 cases, and percentage of people testing positive for SARS-CoV-2 infections in the community might be useful to determine whether to maintain or modify school operations.” This is moderately helpful, but for me, the challenge is that without numeric thresholds, it’s much harder for us to understand school return decisions; the criteria are not transparent. For example, what does “substantial, controlled transmission” look like? Would a given leader define their community’s substantial transmission levels “controlled” or “uncontrolled” and how would that description be weighted based on political, economic, health, and welfare concerns (among others)?

If we were to consider as numeric thresholds/benchmarks the indicators CDC suggested for phased reopening (Table 3), then indicators of “sustained, controlled transmission” would include downward trajectory of new cases; downward trajectory of % tests positive; tests positivity at </=20%; and tests returned within 4 days. Other groups, like Resolve to Save Lives and the Pandemic Response Network have drafted pandemic severity/alert criteria (see here). Interestingly, they suggest that in-person learning be curtailed in the context of both “substantial, uncontrolled transmission” and “substantial, controlled transmission” (e..g “high alert” and “moderate alert” respectively). With regard to this component of the debate — whether to have in-person learning at “substantial, controlled transmission” — I don’t know which is the right answer.

Table 1. CDC Guidelines for Making Decisions about School Operations

Table 2. Levels of Community Transmission (from CDC)

Table 3. Gating Criteria for Phased Reopening (from CDC)

Q&A for 7/26:

#Motto #Mascot #Meme #Japan #Amabie

Question: Got any good news?

Answer: Happy Sunday! Here are two pieces of news that I really enjoyed, but I’m a little late to the party in sharing:

1. Earlier this month, Japan’s Fuji-Q Highland amusement park reopened with a new request for visitors riding roller coasters — “Please scream inside your heart.” This request is to limit the virus transmission since screaming and loud talking may increase respiratory droplet spread. And just like that, a meme was born along with calls to make this the new motto for 2020 (Figure 1).
2. In addition to giving us our 2020 motto, Japan has also given us our 2020 mascot, Amabie’ (pronounced a-ma-bee-ay) (Figures 2 and 3). As described in this recent article from JAMA, “Amabié — A Japanese Symbol of the COVID-19 Pandemic,” Amabie’ is “a legendary mermaid-like creature who is said to emerge from the sea to prophesize good harvests and epidemics.” She originally gained attention in 1846 after the local newspaper shared the story and sketch (Figure 2) from an unnamed officer who reportedly went to the seaside to explore a strange light. What he found there was a mermaid-like creature who told him, “I live in the sea. My name is Amabié. Good harvest will continue for six years. At the same time disease will spread. Draw me and show me to the people as soon as possible,” before submerging. According to the JAMA article, which is a really enjoyable read, Amabie’ has been trending in Japan as a popular symbol of the COVID-19 pandemic since mid-March. In response, “Japan’s Ministry of Health, Labor, and Welfare has leveraged the character’s popularity to raise awareness about the importance of staying home and physical distancing for prevention of infection, in line with other efforts to recruit manga and anime characters to advance public health messaging… Amabié is also a feature character in the official national coronavirus contact tracing app released June 19, 2020.”

Figure 1. Please Scream Inside Your Heart (from Twitter)

Figure 2. Original Drawing of Amabie’ , 1846 (from JAMA)

Figure 3. Recent Depictions of Amabie’ (from JAMA)

Q&A for 7/25:

#DC #Trends

Question: I read that two days ago, DC implemented a mandatory mask order that requires everyone ages 3+ who is not vigorously exercising or eating/drinking to wear a mask outside the home. I thought DC had been doing alright — didn’t you do a Q&A on that recently?

Answer: We talked about the status of DC in our Q&A of 7/14 and I did write then that “DC is still looking pretty good.” At that time, cases had only slightly risen and test positivity was still very low. In the days that followed, new cases in DC continued to increase alongside an increase in test positivity. As Figure 1 shows, cases are on the rise (yellow bars) — the highest they’ve been since early June. These rising cases are also occurring in the context of rising test positivity rates (blue line of line chart), which means that the case increase is not due to an increase in tests (green bars). Yes, hospitalizations are flat (orange bars) and deaths are low (red bars), but we know these are lagging indicators. Right now, the increasing cases alongside increasing test positivity rate is cause for raising alarm.

As we continually see with COVID, If not tamped down quickly, things can quickly spin out of control. In my opinion, DC is doing the right thing by raising alarm and asking people to carry out an intervention (mask wearing) we know to be protective. It’s likely that DC will need to do even more — roll back the phased reopening — if cases and test positivity continue to rise.

Figure 1. COVID in Washington, DC (data from covidtracking.com)

Q&A for 7/24:

#Children #Influenza

Question: Saw this on Politico Nightly: There are 317,711 reported Covid-19 cases in children and teenagers across the U.S., according to the COVKID Project, led by a team of epidemiologists who track cases in children and teens. There have been 77 reported deaths, and 805 intensive care hospitalizations in children 17 and under. How do these numbers compare in deaths and hospitalizations to other childhood illnesses like flu, infections, and RSV?

Answer: Thanks for asking! We talked some about childhood COVID/influenza comparisons in our Q&A of 5/23. Two months have since passed (both shockingly quickly and slowly because time is increasingly difficult to track these days) and we have more information.

1. First, data show that children are still far less likely to have complications/severe COVID as compared with adults. Indeed, there seems to be a linear relationship between COVID severity and age — each year of age is associated with increased risk of severe COVID outcomes, including death (Figures 1 and 2).
2. Second, data show that children are not impervious to the virus, as illustrated by the COVKID Project data you shared in your question. While rare, death and other severe outcomes are coming to the fore, including multisystem inflammatory syndrome (described in Q&A of 5/15).
3. Third, the latest evidence we have is from a comprehensive contact tracing study out of South Korea (published earlier this month in Emerging Infectious Diseases), which shows that households with index cases among children ages 0–9 have the fewest transmission events while households with index cases among children ages 10–19 have the highest level of transmission events, transmitting just as much as adults. Young children seem less likely to infect others while older children seem at least as likely as adults to infect others (Table 1).

Now to get more directly into your question. Table 2 shows the 10 leading causes of death among adolescents and children in the United States in 2016 (from this paper published by New England Journal of Medicine). As you can see, 20,360 children (defined in this study as people ages 1–19 years) died in 2016. Of these deaths, 60% were due to injury-related causes. Cancer and congenital abnormalities were the next leading causes of non-injury related deaths (~9% and 5% of all deaths, respectively).

Turning specifically to influenza, we can use CDC reporting to see that a total of 185 pediatric deaths (ages 0–18) were reported for the 2019–20 flu season (Figure 3). You can find flu hospitalization rates by age here and covid hospitalization rates by age here. The 77 COVID-related deaths reported in children and adolescents in the US since March are deaths over a 4 month period. If the numbers remained constant over a 12-month period, we would see more pediatric deaths due to COVID (231) than influenza deaths (185), but fewer deaths than an ailment like heart disease (599 deaths in 2016).

Figure 1. Case Fatality Rates by Age (from ourworldindata)

Figure 2. Case Fatality Rates by Older Age over Time in Maryland (from Dr. YJ Choi)

Table 1. Children ages 0–9 appear to be the least likely to pass coronavirus on to other household members. (from “Contact tracing during coronavirus disease outbreak, South Korea”)

Table 2. Ten Leading Cases of Child and Adolescent Death in the United States, 2016 (from this 2018 article in NEJM)

Figure 3. Influenza Deaths among Children (from CDC)

Q&A for 7/23:

#Reinfection #Reactivation

Question: Still finding myself confused about test results / transmission ability….If someone had the virus, tested positive, then tested positive months later, is that person able to transmit?

Answer: We talked about the issue of potential reinfection in our Q&A of Q&A of 5/19. Short answer is: We still aren’t sure but most folks who test positive after apparent recovery are not contagious. Now for more detail…

When it comes to people testing positive for the virus after apparent recovery, we have three hypotheses (described more in Q&A of 5/19):

1. They’ve become reinfected;
2. They are experiencing reactivation of dormant virus; and
3. The second round or positive results are false positive.

When I wrote the post 2 months ago, it was looking like hypothesis 3 was the most likely. Since then, we have a few additional case reports of reinfection/reactivation. For example, this paper published last month in the Journal of Infection describes reinfection/reactivation among 11 patients in France while this paper published earlier this month in the Journal of the American Geriatrics Society also describes reinfection or reactivation among 3 older women hospitalized in France. In both of these papers, reactivation was the preferred hypothesis for what was causing these patients to exhibit symptoms after apparent recovery — that is, their bodies had not fully cleared the virus and it made a resurgence. Thus far, the evidence still suggests that hypothesis 3 is most likely/common, followed by hypothesis 2. We still have no definitive data on anyone becoming reinfected (hypothesis 1).

I share all of this because it all has bearing on whether we would expect a person who tests positive after apparent recovery to be contagious. And thus far, the evidence points to two things: 1) since most people who test positive a second time are likely to fall into hypothesis bucket 3 — test is a false positive — they are not infected and cannot transmit the virus to others; and 2) for those (seemingly rare) folks who are exhibiting symptoms again and test positive a second time, we should consider them as possibly contagious and ask that they again self-isolate; health providers will also treat them as COVID patients in terms of protocols and the like.

Finally, I think this paper from Journal of Medical Virology offers a nice summation of the evidence (as of early June). And just yesterday the NY Times published this report that synthesizes current thinking on reinfection. (It’s like the Times read your question!)

Q&A for 7/22:

#Vaccination #Safety

Question: I, like many others, was deeply encouraged and hopeful to learn that the COVID-19 vaccine being developed by Oxford in the UK shows some promise. Not to be a downer, but I feel the need to ask: Especially under an accelerated timeline such as the researchers are working under, how do you test a vaccine to see if there could be any long-term effects down the line? Is there a way to see if a vaccine in 10+ years could lead to increased risks of cancer or other problems, without actually waiting that long? Or are there just some things that we know a vaccine is not capable of doing?

Answer: It’s better to address vaccination concerns early and often, so thanks for asking this question. I’d like to think that we can find the balance with expedited vaccine development and vaccine safety and I know that myriad groups are trying to do just that. Last month, Scientific American published this opinion piece that describes concerns about “telescoping the [vaccine] testing timelines” and back in March, Nature published this viewpoint expressing similar concerns. Meanwhile, vaccine development needs to be expedited to help prevent more infections and save lives in this bleak pandemic we’re suffering through!

How do we monitor vaccine safety?

We talked about the vaccine development cycle in our Q&A of 5/13. Throughout the cycle, scientists are monitoring not only whether the vaccine works to prevent infection, but also whether the vaccine leads to any adverse outcomes. This monitoring spans early phase animal studies, through clinical development and human trials, and for vaccines that make it to widespread use (e.g. those deemed both safe and effective), the monitoring continues for quality control and adverse event reporting in the general population. CDC offers this great history of vaccine concerns, as well as this helpful resource describing common vaccine questions and concerns. In short, the vaccines we currently use are very safe! No vaccine is linked to cancer.

Concerns about COVID-19 vaccine roll-out

In the context of expedited vaccine development for COVID-19, safety remains paramount. And concerns that groups are rushing vaccines to widespread use without enough safety data are valid. For example, Reuters reported last month that “China’s Central Military Commission approved the use of a [CanSino-developed] vaccine by the military on June 25 for a period of one year.” The vaccine is still in human trials, but is being rolled out to China’s military (est. 2 million individuals)! In a similar vein, here in the United States, we had the massive roll-out of a proposed COVID treatment — hydroxychloroquine — before its safety was well understood. As data came in, FDA cautioned against hydroxychloroquine use and revoked its emergency use authorization because the drug was found to cause serious heart problems (big adverse event) and no benefit for hospitalized patients (no effectiveness). Would the US similarly roll-out a vaccine before we had enough safety and effectiveness data? Would other countries?

Considerations for further addressing vaccine safety concerns

In addition to the safety monitoring components of the vaccine development cycle, another avenue to monitor safety that will require increased investment is post-market vaccine safety studies, which StatNews described last week. And still another avenue is data sharing and transparency. Results on vaccine safety should be published in peer-reviewed journals, like this piece published two days ago in Lancet, and should be clearly communicated to the public using myriad avenues (e.g. not peer review publication!). Recognizing well-founded fears among some individuals and groups of being used as guinea pigs for research (see Q&A of 6/7 for a few historic examples of how scientists in the US abused Black people in the name of science), community engagement in vaccine research, roll-out, and monitoring is necessary. Since Black and brown people have been disproportionately harmed by the coronavirus, they may be considered a priority group to receive vaccination. On the flip side, in the context of limited safety data, rolling out a vaccine first among people of color could easily be viewed as another way in which people of color are devalued. Johns Hopkins Center for Health Security’s Working Group on Reading Populations for COVID-19 Vaccination recently issued a set of guidelines to help policy makers work through these challenging issues.

Q&A for 7/21:

#Pregnancy

Question: I have a question for you — not sure if it would be good for your whole list, but either way I figured you may be able to point me to further resources I should look into! My spouse and I have been weighing the pros and cons of trying to conceive/pregnancy. We put our plans on hold back in March, thinking probably we should minimize our interaction with the healthcare system, and try to mitigate unknown risk to myself or potentially adverse effects to the pregnancy/birth outcome. We’re now re-weighing this as more time passes. Anything to add to your previous answer on this topic from 5/17?

Answer: We are learning more and more as time progresses, but there’s still so much we don’t know, especially when it comes to pregnancy, birth, and the post-partum period. Are pregnant women at increased risk of contracting COVID? Of having more severe outcomes in the event they do become infected? Of passing the infection on in utero? At delivery? Through breastfeeding? Does infection increase the risk of miscarriage? Of low birth weight? Of developmental disabilities? Answer: we are still unsure. In part, these limitations in our current knowledge are because we are only ~6 months into the pandemic, which means that we simply don’t have longitudinal data from women who became pregnant during the pandemic, gave birth, and completed the post-partum period. What a challenging time of limited information we’re living through. Please reach out to your doctor to discuss. In the meantime, here are a few helpful resources and new research findings:

A couple of recent news articles that address this issue well (in my opinion):

NY Times, “Why We Still Don’t Know Enough About Covid-19 and Pregnancy” (July 10)

NPR, “Safe Pregnancy As COVID-19 Surges: What’s Best For Mom And Baby?” (July 17)

Are pregnant women at increased risk of contracting COVID?

Physiological and mechanical changes in pregnancy increase susceptibility to infections in general, and data from New York hospitals during the height of the State’s epidemic show that nearly 1 in 5 pregnant women who were tested at time of delivery were found to be infected with the majority being asymptomatic. That said, we do not have evidence as to whether pregnant women are more susceptible to SARS-CoV-2 infection as compared with non-pregnant women. More research is required.

Are pregnant women at increased risk of severe outcomes?

Data from a CDC report published in late June, “Characteristics of Women of Reproductive Age with Laboratory-Confirmed SARS-CoV-2 Infection by Pregnancy Status — United States, January 22–June 7, 2020” suggest that pregnant women may have more adverse outcomes — hospitalization, ICU admission, and mechanical ventilation — than non-pregnant women, though rates of recovery are similar between pregnant/non-pregnant women. This study had many limitations, including lots of missing data, so the findings are far from rock solid. More research is required.

Can the virus be passed on to the fetus/baby in utero?

Two new pieces of evidence help on this one. First, evidence from Italy shows that the virus can be passed on in utero through the placenta, as recently reported in Nature. Second, evidence from a recent NIH study shows that the placenta’s cells are missing the ACE2 receptors, which are the main receptors the virus uses to infect the cells. This means that in utero infection is likely to be rare. More research is required!

Does COVID infection increase risk of miscarriage or still birth?

We do not know. There are case reports (see this one from JAMA, for example) indicating that if the virus gets into the placenta, it can cause miscarriage. A small study out of Northwestern University published in American Journal of Clinical Pathology showed that among 16 women who were infected with SARS-CoV-2 while pregnant exhibited evidence of placental injury. More research is required!

Q&A for 7/20:

#New York #Trends

Question: We haven’t looked at New York’s data for a while. How’s New York looking?

Answer: The state is still looking pretty good (Figure 1).

• Tests have plateaued around 65K for the last couple of months (green bars)
• Test positivity rate remains very low at 1.2% (line chart) and have been <2% since the beginning of June.
• Cases have risen with reopening (yellow bars) and are the highest they’ve been in a month (Figure 2). We would expect to see cases rise to some degree with reopening. Coupled with low test positivity rates, things are still looking pretty good, though we must be ever vigilant!
• Hospitalizations are at their lowest level (orange bars)
• Deaths are also at their lowest (red bars)

Some of the trends may be easier to visualize looking at weekly data, which is presented in Figure 3.

Figure 1. New York Tests, Cases, Hospitalizations, and Deaths (data from covidtracking.com)

Figure 2. Weekly Cases are as high as they were the week of June 7th

Figure 3. Weekly Trends in Tests, Cases, and Hospitalizations

Q&A for 7/19:

#Face Masks #Face Shields

Question: Are see-through face shields as effective as face masks at preventing COVID spread?

Answer: We do not have much scientific evidence directly comparing the effectiveness of face shields versus face masks for preventing the spread of SARS-CoV-2. Based on what we do know, face masks — specifically N95s and surgical masks — offer more protection as compared with face shields. Unfortunately, we have limited data about homemade masks as compared with face shields. And it’s important to note that face shields have some unique benefits. If I had to sum up — 1) both are good, but N95 and surgical masks are better; 2) if you find that you are having trouble consistently and correctly wearing face masks, face shields are a next-best alternative; 3) since most of us are wearing homemade masks rather than N95s or surgical masks (we’re still trying to keep the supply for healthcare workers) we need to pay good attention to fit — evidence shows that mask fit plus tight fabric weave and multiple fabric layers are key to effectiveness; 4) we do not have enough information to compare homemade masks and face shields.

• Face masks are better at preventing airborne spread. Because face shields leave gaps/openings on the sides and bottom of your face, they are less effective at limiting airborne spread of the virus. A recent systematic review published in Lancet showed that while both face masks and face shields limit the spread of SARS-CoV-2, face masks are more protective. Additionally, a 2014 study published in the Journal of Occupational and Environmental Hygiene of the effectiveness of face shields at protecting healthcare workers from influenza infection concluded that, “face shields provide a useful adjunct to respiratory protection for workers caring for patients with respiratory infections. However, they cannot be used as a substitute for respiratory protection when it is needed.” Note: face masks in these papers are N95s or surgical masks.
• Face shields offer eye protection. There is some evidence that the coronavirus can be transmitted through the eye (see Q&A of 4/4), though this transmission route seems far less common. Still, the eye protection offered by face shields is a benefit of this type of personal protective equipment.
• Face shields may be more user friendly — easier to consistently and correctly use. Face shields make communication easier (you can see a person’s face). They may also be more comfortable to wear, which may keep users from touching their faces or inappropriately removing the shield. For more on these behavioral considerations, see this recent JAMA viewpoint article, this NY Times article, this People article, and this AARP Magazine article and this visualization from Business Insider (Figure 1).

Figure 1. Face Masks vs. Face Shields (visualization from Business Insider based on JAMA viewpoint)

Q&A for 7/18:

#Airborne Transmission #Office buildings #Apartments

Question: To follow up from yesterday’s Q&A: We know that lots of people in the same room and poor ventilation is bad, but what about someplace like a hotel or office building? Has there been evidence that a sick person on one floor could infect someone on another floor through the ventilation system?

Answer: There is currently only very weak evidence of airborne transmission across floors of an office building or apartment building. Bulleted below are a few examples, though it is by no means an exhaustive list. As I understand it, most experts believe that the data are not strong enough to definitely say that airborne transmission through ventilation systems is happening, but that the current level of evidence is compelling enough to warrant concern and to necessitate further study. The big debate is over what action to take in light of limited evidence. For further discussion, in addition to links shared in yesterday’s post, see the report of Canada’s Rapid Expert Consultation on Airborne Transmission, these two recent JAMA viewpoint articles, and this report from Nature. Should we implement additional safety control measures (examples are bulleted in yesterday’s post), which can be costly or wait until we know more? Should we be more cautious at greater financial cost or less cautious at possibly greater health/life cost?

• The experience of a call center outbreak in Seoul points to airborne spread — the initial case worked on the 10th floor and had no contact with workers on the 11th floor who subsequently were infected.
• The experience of Hong Mei House in Hong Kong points to airborne spread — the first two people in the building to contract SARS-CoV-2 were in the same vertical block of apartments, but 10 floors apart and some experts believed the likely culprit was a broken exhaust pipe.
• Meanwhile, last month another apartment complex in Hong Kong, the Sha Tin public housing estate, was evacuated as experts feared a new cluster of cases stemmed from the sewage system.
• Looking at a different, but similar virus, SARS-CoV (of 2003), we have stronger evidence from the Amoy Gardens housing complex that points to the spread coming from a”rising plume of contaminated warm air in the air shaft generated from a middle-level apartment unit.”

Q&A for 7/17:

#Airborne Transmission #Ventilation #Risk

Question: I am sure you have addressed indoor transmission however I have a friend with a specific concern. We would welcome your perspective. She works in a 15 story busy trade mart (in Dallas TX) and may be spending 10 hour days in her showroom for a week in late August. There is central air conditioning throughout the building with no outside air entering the building. She is a 66 year old who is following mask, shield and sanitizer guidelines as best possible but is wanting your opinion on how dangerous the AC system is in this setting.

Answer: Thanks for bringing up airborne transmission, which we haven’t really discussed much! I would be very uncomfortable with the situation you’ve outlined. First, Dallas is an extreme hot spot right now with cases and hospitalizations quickly rising. Unless something dramatic happens in short order, Dallas will not be safe in late August for a trade show or the like. Second, because no outdoor air can enter the building — you can’t even open a window — and the ventilation system is unlikely to be supplemented with airborne infection controls (described below; ask your building maintenance folks what they are doing), I personally think it’s not a safe set-up. Though the scientific debate still continues as to what constitutes sufficient evidence for airborne transmission, what we’ve seen so far leads me to believe that the type of indoor setting you’re describing is highly risky. Read on for more details and thanks for following all of the other public health rules — they are so important and you are doing yourself and your community a service!

Transmission Routes

We know that SARS-CoV-2 spreads from person-to-person mostly from respiratory droplets that are expelled when coughing, sneezing, and talking (hence the masking rule). These droplets generally have a range of only 6 feet and hang in the air for just a short time (hence the 6+ feet rule). There’s also some evidence supporting fomite transmission (e.g. transmission from contaminated surfaces, see Q&A for 3/12) (hence the hand washing rule). Another route of transmission is looking increasingly likely — airborne transmission — transmission from tiny droplets that can hang in the air for prolonged periods and accumulate in poorly ventilated places to such a degree that they spread infection to others.

Role of Airborne Transmission

Increasing evidence from super-spreader events (see Q&A of 6/3) as well as lab-based studies of the virus’s “aerosol fitness” show that airborne transmission is both theoretically possible and a plausible explanation for a number of super-spreader events. In response to mounting evidence, 239 scientists wrote to WHO last week asking WHO to revise its guidance. And on 9 July, WHO issued an updated brief on the transmission of SARS-CoV-2 concluding that further study is needed on “the role and extent of airborne transmission outside of healthcare facilities, and in particular in close settings with poor ventilation”. The scientific debate on the role of airborne transmission rages on.

New Public Health Rules if Airborne Transmission is True Threat (quoted from letter of 239 scientists)

• Provide sufficient and effective ventilation (supply clean outdoor air, minimize recirculating air) particularly in public buildings, workplace environments, schools, hospitals, and aged care homes.
• Supplement general ventilation with airborne infection controls such as local exhaust, high efficiency air filtration, and germicidal ultraviolet lights.
• Avoid overcrowding, particularly in public transport and public buildings.

Q&A for 7/16:

#Contact Tracing

Question: I’ve read a few reports — this one, for example — suggesting that contact tracing isn’t working in the United States. Is that true? Is the United States so focused on individualism that we just can’t do contact tracing?

Answer: Contact tracing is time-consuming, skills-based, cultural-competency based, and requires cooperation from the public. It doesn’t work well in contexts where cases are rapidly increasing or where testing is limited/delayed. It’s difficult. Contact tracing is also a necessary component of our response to COVID-19, and in spite of the difficulties, we’ve had many successes.

What are contact tracing successes?

• Contact tracing limits community transmission. By quickly identifying clusters and isolating individuals, contact tracing limits the ongoing spread of the virus. The state of Massachusetts was an early adopter of widespread contact tracing and has seen steady declines in cases, test positivity rate, hospitalizations and other key metrics.
• Contact tracing gives us key data to build our evidence base on transmission risks and protections! Because of contact tracing, we know much more about risk of infection, particularly as it relates to location — indoor restaurants, bars, crowded parties. Additionally, our increasing knowledge of the role of airborne transmission and evidence of protection afforded by mask wearing come from the data collected through contact tracing.
• Contact tracing allows hotspot identification for increased public health enforcement. Contact tracing highlights locations that require enhanced protections for employees and customers (think: meat and poultry plants, crowded bars). In knowing the location of the outbreak, officials can implement harm mitigation plans, as described in the NY Times by an Allegheny, PA official.

How does contact tracing work?

Once a person tests positive, a contact tracer gets in touch with them to inquire about where they have been and who they have been around (generally defined as being within 6 feet of for a period of 10–15 minutes or longer) during the presumed period of infection (e.g. last 14 days). The contact tracer then reaches out to everyone who was potentially exposed and asks them to self-isolate for 14 days after the exposure, monitor themselves for symptoms, and get tested. For any exposed person who has symptoms, the process starts all over again. CDC offers more details here and Partners in Health offers this helpful diagram (Figure 1).

What must be in place for contact tracing to work? (not a comprehensive list)

• Limited Community Spread. When cases are rapidly rising, it is exceptionally difficult to effectively contact trace. There are just too many infected individuals. Contact tracing is to be implemented at scale in the period when community transmission is largely under control and the goal is to keep outbreaks contained. We need extensive contact tracing in place as we start reopening. If the virus is spreading too quickly for contact tracers to handle all the contacts, then it’s considered time to go back towards shut-down, including re-entering earlier phases of re-opening.
• Widespread Testing. For contact tracing to work, we must first identify cases and those cases must be identified within a short timeframe. If an infected individual cannot get a test, they remain unknown to contact tracing efforts. Meanwhile, if an infected individual receives their test result many days after testing, the people exposed during their period of infectivity will have been told to self-isolate too late to slow/contain the spread.
• Trained, Skilled, Community-Base Workforce. Johns Hopkins offers a few suggested best practices for contact tracing, noting that contact tracers do not need to have special public health skills — like epidemiology or the like. Most importantly, contact tracers must be active listeners, curious, meticulous (detail-oriented), and from the community they are working with.
• Community Trust and Participation. Folks must be willing to respond when contacted and provide details to contact tracers. In some communities, building this trust can be exceptionally difficult. This is why the key workforce skills described above are so imperative. Also important is widespread communication to set expectations and build knowledge and acceptance. Community members also need support, which means that wrap-around programs — to help people who are self-isolating receive food and medication, for example — must also be established.

What about American Exceptionalism?

The idea of American exceptionalism in the context of contact tracing, is akin to exceptionalism in the context of masking. We know that masking is necessary for limiting the spread of the virus. And just because we’re having trouble getting all Americans to wear masks doesn’t mean that we abandon the strategy. If anything, it means we invest more in campaigns to get Americans to wear face masks! In fact, after the White House Coronavirus Task Force shared CDC’s recommendation that all Americans wear face coverings when outside their homes back in early April, mask wearing increased from 61% to 74%. More can and should be done — to promote mask wearing and effective contract tracing! We must invest the resources for states and localities to hire, train, and manage contact tracing workforces while we must also invest the resources in educating the public, educating businesses, and building new social norms that bolster support for contact tracing efforts (including, for example, collecting contact information of clients to share with contact tracers in the event of an infection). This is not an insurmountable problem! We can contain the virus!

Figure 1. Contact Tracing Public Communication

Q&A for 7/15:

#CDC #Surveillance

Question: Last night the NY Times reported, “ Trump Administration Strips C.D.C. of Control of Coronavirus Data.” This news was alarming to read at first, but I see that some of the data sites you refer to in yesterday’s Q&A get their figures directly from each state health department. What are your thoughts on this move from the administration? How concerned should we be?

Answer: I was also alarmed when I read this news. CDC’s mission to protect the public’s health. Surveillance, including collecting and sharing key health data, is a key component of achieving its mission. Stripping CDC of its role in hospital surveillance for COVID-19 *seems* punitive and politically motivated. Do I *know* if it is punitive or politically motivated? No, I don’t. Am I concerned? Absolutely. So much of our COVID response has been politicised. The idea of key data now being reported outside of traditional systems via the private sector concerns me (and others) — will the data be manipulated?; will the data be available and accessible?; who is accountable? The NY Times article does a nice job of describing these concerns. Read on for additional background and details not provided in the Times article.

Role of NHSN

Typically, more than 16,000 health facilities across all 50 states + DC and Puerto Rico report a bevvy of data to CDC’s National Healthcare Safety Network (NHSN), the nation’s most widely used healthcare-associated infection tracking system. On March 27, NHSN launched its COVID-19 tracking module for hospitals to report COVID-related data and on March 29, VP Pence sent a letter to hospital administrators asking them to participate in sharing daily reports with NHSN. (Note: in May, NHSN launched a tracking module for reporting by nursing homes.) The hospital data reported to NHSN include information on hospital capacity and patient load, healthcare worker staffing, and healthcare supply. The data come either directly from the hospitals, from vendors hospitals hire to do their reporting, or from States who are authorized to collect the data and reporting on behalf of all hospitals in their jurisdiction. Local data are made available in real-time to local health departments for monitoring the outbreak and health systems needs. The data are also used by CDC and HHS, and are synthesized into public-facing data dashboards.

Friction between the White House and CDC

Back in mid-May, The Washington Post reported, “Growing friction between White House and CDC hobbles pandemic response.” In that report, the Post described how CDC’s data was continually questioned and viewed as flawed with Dr. Birx stating, “There is nothing from the CDC that I can trust.” Note: we discussed CDC’s processes for reporting COVID deaths in our Q&A of 5/5 and there is a time delay in death reporting and the deaths reported are likely an undercount. I have seen reports of data quality and completeness issues with the nursing home data CDC collects, and there are still many data gaps in other types of data CDC collects (how long did it take us to get data on race(?!)) In my opinion, it is fair to criticize CDC. The odd thing that happened (in my opinion) was that rather than trying to work with CDC to improve its systems, the federal government invested in a new, duplicative system to collect hospital data run by a private company, TeleTracking.

Non-competitive award to TeleTracking

As the Post reported, the $10.2 million, 6-month contract to TeleTracking was awarded by the Secretary of Preparedness and Response at HHS (outside of CDC) non-competitively back on April 6 with the mandate that TeleTracking collect the same data NHSN is collecting! On April 10, Secretary Azar wrote hospital administrators again asking them to report daily information via the above described avenues + TeleTracking. And now, July 15, HHS has advised hospitals to stop reporting data to CDC altogether. As the Post back on May 15 summed up, “a senior administration official said Trump’s son-in-law and senior advisor, Jared Kushner, and many of his allies do not trust the government agencies and prefer to work with officials in the private sector.”

Q&A for 7/14:

#Data Sources #DC #Trends

Question: I feel like I don’t actually know how DC is doing at this time. Like I saw a map floating around Twitter that we are trending badly, but I don’t honestly know at this point how to substantiate things. Help!

Answer: The internets are so full of it — the good, the bad, the ugly. Here are a few websites that I recommend for following the data: 1) CovidTracking.com keeps updated state-by-state data on COVID testing, cases, hospitalizations, and deaths; 2) DC’s Coronavirus Webpage provides key DC data updated daily; 3) NY Times Coronavirus Tracker provides great visualizations, including interactive state-by-state maps; 4) Johns Hopkins COVID Tracker provides county-level case data and much more; and 5) Our World in Data provides lots of COVID data and visualizations for the US and other countries.

Turning to DC’s current trajectory, see Figure 1. As I see it, DC is still looking pretty good.

• Tests are increasing (green bars), and test positivity rate remains low (line chart).
• Cases are beginning to rise (yellow bars). In the context of increasing tests and reopening, we’d expect to see increases in cases. So far, the increase is manageable (in my opinion), especially since the test positivity rate is still so low and hospitalizations are also low.
• Currently hospitalized patients have plateaued around 90 for the last several days (orange bars; note: DC did not report hospitalizations earlier in the pandemic).
• DC has not witnessed any deaths for the last 4 days (red bars).

Figure 1. COVID in Washington, DC (data from covidtracking.com)

Q&A for 7/13:

#Tourists #Spread

Question: I was just reading this article in the NY Times, “In Texas Beach City, Out-of-Towners Drove In an Outbreak,” which my neighborhood listserv is ranting about. We live in a small beach town that had been relatively COVID-free… that is, until a few weeks ago when we started opening up. Folk here who run Airbnb are totally into the move forward, but our county test positivity rate has doubled in two weeks! What do you think about the issue of tourists and travelers spreading the virus?

Answer: That article on Corpus Christi was such a good, sad read. For those of you who didn’t read it, the summary is: Corpus Christi had gotten daily case counts down to zero; it became known as a safe place; tourists flocked to the town; meanwhile, the town was unprepared with no mask regulations, no processes to help people maintain 6 feet of distance, etc.; now it is a hot zone with cases growing rapidly; some residents blame outsiders and some blame themselves. So how is this story relevant to other towns? Here are my thoughts:

1. No place is special. No place is immune. If Town X hasn’t been hit, it’s not because Town X is special or immune. Town X has just been lucky. And luck runs out.
2. We are interconnected. One take-away could be to shut your town off from others. This would be an exceptionally difficult path to follow. First, I’m not sure that towns or counties have the legal authority to close themselves off. Even if a town were able to do so, they would be forgoing the revenue that tourists bring, which could prove too economically damaging. Such towns would also presumably need to keep their residents within the town borders — no one comes, no one goes. Back in the pandemic of 1918, Gunnison, Colorado did this self-imposed quarantine to keep the flu out, and it worked. But given our interconnectedness, I’m doubtful it would work today. A related, more moderate solution would be to have everyone who comes/goes self-quarantine for 14 days. Hawaii has taken this approach and kept daily new cases relatively low, though daily cases are now back at peak levels (~30 new cases/day) even with the quarantine measures in place. This would be challenging to implement at a local level — who will enforce it?
3. Public health regulations and practices are imperative. Towns and communities must practice public health interventions before embarking on reopening. Masks must be required of employees and customers. “No mask, no service.” When expecting lines, markers should be painted on the ground to keep people paced 6+ feet apart. Can markers be made on beaches? Add them! What does your community want to do about bars and places where people put their guard down? Consider keeping them closed. Will folks renting their houses agree to only rent to families (e..g no party houses)? Does your community have the ability to track individuals who have been exposed? If not, set it up now!
4. It’s easy to blame others — “outsiders” — for problems. But Corpus Christi’s problems were also of their own doing. The town was not prepared. It continued operating as if the pandemic didn’t exist. This is a recipe for disaster. Please use this story as a cautionary tale. Make changes to your community now to limit the spread.
5. Be prepared to shut things [back] down if case numbers [continue to] rise. To use an analogy — We are like dry brush. Little fires can quickly turn into out-of-control blazes.

Q&A for 7/12:

#Timeline #Progression

Question: Will you please remind me (us) of the expected time lag between cases, hospitalizations, and deaths?

Answer: Sure! We talked about illness progression in our Q&A of 3/29 and CDC offers a nice synthesis of the data here and here. For easy visualization, I made this timeline (Figure 1). In sum, we’d expect to see an increase in cases generally about 5 days after exposure, an increase in hospitalizations about 14 days after exposure, an increase in deaths about 20 days after exposure. So, let’s see how that compares with recent trends in the United States (Figure 2). As you can see, cases started increasing in early/mid-June. June 9th was the lowest 7-day rolling average (20,065 cases). Hospitalizations started rising in mid/late June. June 20th was the lowest 7-day rolling average of hospitalizations (28,201). And deaths started rising in early July. July 6 was the lowest 7-day rolling average of deaths (478). Put another way, cases started increasing on June 10th (est. Day 5); hospitalizations started rising on June 21st (Day 15); deaths started to rise on July 7th (Day 34). This timeline fits within expectations (sadly). Things are going to get worse before they get better.

Figure 1. Timeline of COVID Progression

Figure 2. Timeline of COVID Progression in the United States (data from covidtracking.com)

Q&A for 7/11:

#Doctor #Dentist #Risk

Question: What is the relative risk of going to a doctor or clinic for health care visits? A kid in our “pod” (two families that have merged our circles of isolation) recently went in to get routine immunizations, and we are wondering if we should un-merge from that family for a couple weeks. Likewise, my doctor asked me to come in and get blood work done, and I wonder (a) if doing that is worth the risk of contracting Covid, and (b) if I should isolate from my family and the other family in our “pod” afterwards. On the one hand, health care workers are highly trained and likely to be very diligent about masks, disinfection protocols, etc. But on the other, these interactions are indoors, typically in small confined rooms with limited air flow, in establishments where lots of people are coming and going, and without social distancing (you can’t socially distance from someone while examining them, giving them a shot, or drawing their blood). Arguably, health-care settings also attract more sick people than, say, bars or restaurants. We are in one of the areas of the country where cases are recently exploding. I’m not suggesting people avoid going in for preventative health care, but more wondering whether a period of self-isolation afterward is appropriate. On a scale from 1=”isolate at home” to 10=”attend a huge indoor rally”, how would you rate the risk of going to a clinic for routine health care?

Answer: We’re all grappling with this issue right now. Another friend wrote a similar question about going to the dentist and then there’s the Q&A from 7/9 about risk from being in a hospital waiting room. As with everything COVID-related, we’re still operating with incomplete information and our understanding of risk is limited as a result. When it comes to scaling risk of seeing the doctor or dentist, I suggest the following:

1. Read up on safety guidelines. CDC’s suggestions for infection control in dental offices are here and suggestions for pediatric healthcare providers is here. Additionally, the American Academy of Pediatrics offers practice management tips during COVID here.
2. Check whether your provider is following safety guidelines. After you’ve read up on infection control suggestions, reach out to your provider to understand what they are doing to protect themselves and their patients. Compare their practices with what you read. Do they align? Are they weaker/stronger? If they don’t meet the guidelines, consider finding another provider.
3. Check community transmission levels. You can find community transmission data from your local health department (e.g. Nashville’s is here). If the rate of transmission is greater than 1.0, then your community is seeing rapid spread. Other data points you could use are — 1) number of daily hospitalizations and currently hospitalized; 2) number of daily cases and percent of tests that are positive [note: increases in hospitalizations OR in cases + test positivity rate are both worrying signs.]

So, how would I then use this data? Three scenarios:

1. IF the provider is following safety guidelines and community transmission is declining, THEN your risk is quite low. On a scale of 1–10, I’d put it at a 3 (moderately low risk). I would go. And I wouldn’t be overly concerned if someone else in my pod had gone.
2. IF the provider is not following safety guidelines and community transmission is declining, THEN I would consider finding another provider. Even if community transmission were contained, I’d still want to know that my provider was doing their utmost to keep us all safe. And if my provider weren’t doing the needful, it would make me question their judgement pertaining to other elements of their healthcare practice/provision. I’m not sure how to rank this on a scale of 1–10; it really depends on how far your provider deviates from the safety guidelines.
3. IF the provider is following safety guidelines and community transmission is increasing, THEN I would: a) consider whether you can wait until community transmission is declining; b) if you decide that you need to go, then talk with your provider about how you can further mitigate risk, especially if you have underlying conditions/risk factors (perhaps you could be the first patient of the day, for example). On a scale of 1–10, I’d rank the risk at a 4 or 5 (moderate risk). And if I had a good experience with my visit (unlike our friend who had to go to the ER with a bunch of sick people the other day!), I’d probably keep hanging with my pod, but limit interactions to outdoor, socially distanced interactions only.

A few more things:

• When it comes to routine immunizations, I’m of the opinion that the risk of missing or delaying immunizations is far greater than the risk of COVID infection stemming from the visit. If your pediatrician recommends it, please do not delay!
• For a run-down of infection transmission in waiting room settings, see the Q&A from 7/9. And for a list of most common cluster transmission points, see the chart from yesterday’s Q&A.
• If you’re interested in seeing how other groups have ranked risk, the Texas Medical Association’s COVID-19 Task Force and Committee on Infectious Disease recently ranked a set of activities from low to high risk. Their list (Figure 1) largely aligns with another recent ranking from four Michigan doctors with a few exceptions — compared with the Michigan list, Texas list ranks movie theaters, airplanes and office buildings riskier while it ranks playgrounds as less risky. These lists are not the end-all-be-all, but since we still don’t have anything like it from our national public health authorities, it’s a start. That said, risk will be dependent on where you live in terms of what’s going on with community transmission, so please keep paying attention to your local situation to make informed decisions.

Figure 1. Activity Risk Ranking by Texas Medical Association

Q&A for 7/10:

#Outdoor Risk

Question: What do we know about transmission risk in outdoor settings? We’ve been doing a lot more of seeing our friends in socially distanced, outdoor spaces because we know it’s safer than indoors. Even so, how much are we chancing it?

Answer: The more we know, the more we can be confident that: a) outdoor settings are much, much safer than indoor settings; and b) outdoor settings do not erase risk. Even when socializing outdoors, we need to take precautions, including by keeping our distance (6+ feet). The good news is that — to my knowledge — we do not have any evidence of transmission in the context of outdoors + socially distant. If you’re abiding by these parameters (and staying home if unwell, washing hands too), I think your risk is very low.

We talked about indoor/outdoor risk in our Q&A of 5/20 and 5/12. Over the last two months, we don’t (to my knowledge) have much rigorous science on outdoor transmission risks a la the study out of China that we discussed in our Q&A of 5/12. What we have now is a lot of news reports indicating where clusters of infections originated. We also have this database of literature and media reports, “COVID-19 Settings of Transmission”, compiled by researchers at the London School of Hygiene and Tropical Medicine and further described in their article, “What settings have been linked to SARS-CoV-2 transmission clusters?” I just became aware of this open access database last night and I admit, I am really excited to access it (nerd!)! Mind you, a database like this has a lot of limitations; but in the context of limited and emerging data, it’s a helpful resource. So here’s our outdoor transmission risk update:

• Of the total 266 cluster events compiled as of 6/7, only 7 were fully outdoors — 4 cluster events at building sites in Singapore, 1 at a seafood market in China (discussed in Q&A of 5/12), 1 at a playground in Germany. and 1 from a running partner in Italy. Note: 20 events of the 266 were in indoor/outdoor spaces.
• The most common cluster settings captured in the database are: households, large shared accommodation settings (e.g. dormitory, lodge, group home), sports settings, religious settings (e.g. indoor church services), food processing plants, and elder care settings (Table 1 and Figure 1).
• There are several events not yet captured in the database that are relevant. Last week, the New York Times provided a nice synthesis, including reports of COVID transmission at a DC socialite’s backyard soiree. And at the beginning of July, the Times reported that recent protests had not caused a wave in infections, “In Minnesota, an initiative that targeted demonstrators found that 1.5 percent of them tested positive. In Massachusetts, fewer than 3 percent of protesters did.”

Table 1. Cluster Settings (from COVID-19 Settings of Transmission Database)

Figure 1. Case Clusters by Setting and Size (from ScienceNews)

Q&A for 7/9:

#Hospital #Waiting Room #Risk

Question: My toddler busted her chin getting out of the tub the other day, prompting us to visit the emergency room. A couple of stitches later, she’s fine and was happy with the whole experience because she got lollipops and juice from the doctor — an extra special treat since she never gets those at home. Anyway, while we were sitting in the emergency room, we were around people coughing, people there for COVID tests, and just a general sea of humanity. Yes, they were wearing masks, but still, it was very disconcerting. Do we need to worry about exposure to COVID? Do we need to self-isolate?

Answer: I’m glad baby girl is fine and took it all in stride. I’m sorry that you had to go through that experience . Honestly, I’m shocked that the emergency room has everyone sitting together. I would have hoped (and expected!) they would have people with COVID-like symptoms and people wanting a COVID test to wait in a different area! Indeed, if the emergency room were following CDC guidelines for infection control, they would be screening and triaging everyone coming into the facility and separating anyone with COVID symptoms. Since that didn’t happen, I think you may have to consider yourself exposed and self-isolate. :( Here’s more background on why I’m suggesting that you may need to self-isolate:

1. There are a few differences in how CDC and other public health groups define “close contact.” I’m quoting the close contact definition that New Jersey uses because it directly addresses hospital waiting rooms: “A close contact is defined as being within approximately 6 feet (2meters) of a COVID-19 case for a prolonged period of time (approximately 10 minutes or longer); close contact can occur while caring for, living with, visiting, or sharing a health care waiting area or room with a COVID-19 case OR Having direct contact with infectious secretions of a COVID-19 case (for example, being coughed on).” Based on the waiting room experience you described, it sounds like you meet the criteria of being a “close contact.” Of course, I was not there and I do not know what your experience truly was, so please use your own best judgement.
2. While I have not found any direct evidence of COVID-19 infection resulting from exposure in a hospital waiting room, we do have some reason for concern based on evidence from other diseases (though like most things in science, it’s not clear cut). Here are a few pieces of related evidence that unfortunately do not offer much clarity. From what I’m reading, I think that if everyone was wearing masks, sitting far apart, and not playing with common toys or reading from a shared pile of magazines or the like, your risk is quite low.

• This article from June 2020, “Minimizing intra-hospital transmission of COVID-19: the role of social distancing” describes Singapore General Hospital’s recent experience of caring for COVID patients. Of the 75 cases they cared for in the first months of the pandemic, 1 case was not initially isolated and potentially exposed 18 other patients and 8 health care workers. None of the exposed patients or health care workers became infected. The authors believe this happy outcome was the result of social distancing + mask wearing by the infected patient.
• This article from 2018, “Infection prevention and control in paediatric office settings”, offers an informative synthesis of the evidence:
• Measles has been transmitted in paediatricians’ offices.
• There are reports of transmission of tuberculosis from physicians to patients in paediatricians’ offices.
• Pertussis has been transmitted to healthcare workers in paediatric ambulatory settings.
• In-office spread of common viral and gastrointestinal infections has not been reported, but is nonetheless probable if offices don’t take precautions.
• One study showed no increased risk of influenza-like illness while another study reported an increased risk of influenza-like illness in a family member in the 2 weeks following a well-child visit.
• This 2010 modeling study, “Potential for airborne transmission of infection in the waiting areas of healthcare premises” concluded, “Under normal circumstances the risk of acquiring a TB infection during a visit to a hospital waiting area is minimal. Likewise the risks associated with the transmission of influenza, although an order of magnitude greater than those for TB, are relatively small. By comparison, the risks associated with measles are high.”
• This 2012 cohort study, “Extremely low risk for acquisition of a respiratory viral infection in the emergency room of a large pediatric hospital during the winter season” followed 615 children in Athens, Greece who visited the emergency room, finding that only 22 (3·6%) children developed at least one symptom compatible with a respiratory viral infection within 1–7 days after the visit, and only 3 children (0·5%) developed an influenza‐like illness.

Q&A for 7/8:

#Vaccination Effectiveness #Herd Immunity #Anti-vaccination sentiment

Question: Tony Fauci recently said that a vaccine is unlikely to offer herd immunity, which confused me. I thought we would get herd immunity with a vaccine. What’s up with that?

Answer: In an interview with CNN last week, Dr. Fauci described two issues that would preclude herd immunity — 1) Anti-vaccination sentiment, “there is a general anti-science, anti-authority, anti-vaccine feeling among some people in this country — an alarmingly large percentage of people, relatively speaking.” and 2) Vaccine efficacy, “The best we’ve ever done is measles, which is 97 to 98 percent effective. That would be wonderful if we get there. I don’t think we will. I would settle for [a] 70, 75% effective vaccine.” In short, because the vaccine is unlikely to be fully effective against SARS-CoV-2 and because a large proportion of the population is unlikely to get the vaccine, we’ll fall well short of achieving herd immunity and the virus will continue to circulate in our communities, threatening those who are unable to get vaccinated and those for whom vaccination did not confer protection.

The need for vaccination education programs.

As this article published in the BMJ describes, the anti-vaccination movement is of grave concern for our fight against COVID-19 and other diseases. For example, a study of 1000 people in New York over 24–26 April found that “only 59% of respondents said they would get a vaccine and only 53% would give it to their children.” The expectation that we’ll miss herd immunity because of anti-vaccination sentiment calls for sweeping social behavior change efforts to disabuse people of anti-vaccination bias. Dr. Fauci calls for just that. In the CNN interview, Dr. Fauci says, “We have a [vaccination education] program right now that’s going to be extensive in reaching out to the community…” Unfortunately, this is yet another area of our government’s response that appears to be disjointed and neglected. CNN reached out to HHS for more information on the vaccination education program and reported, “Michael Caputo, an HHS spokesman, did not confirm the existence of a vaccine education campaign, adding that “I’d hate to see CNN put out [a] wildly incorrect story.”

Defining vaccine efficacy.

In addition to hearing a lot about herd immunity (for a refresh, see Q&A of 5/1, #Herd Immunity) we’re also hearing a lot about vaccine efficacy. It’s another epidemiological term bandied about as if we all know what it means. Let me give a quick overview — vaccine effectiveness measures the reduction in disease among the vaccinated group in the real world (rather than clinical setting). For example, if a coronavirus vaccine is 75% effective, it means that we will see a 75% reduction in the number of coronavirus cases among those vaccinated. CDC offers a full definition here. In the space of partial effectiveness, even for folks who are vaccinated and still get sick, we have evidence from other vaccination studies — here, flu — that vaccination can reduce the severity of the illness.

Q&A for 7/7:

#Case demographics #Youth #Risk

Question: Given that so many young people (< 35) are coming down with positive tests, and are notoriously being careless about risk-taking and making everyone vulnerable, how can we impress upon them to abide by appropriate protocol? The youthful invincible attitude cannot work for COVID spread!

Answer: There has been a lot of media coverage of younger people taking risks. Honestly, I’m afraid it’s a bit unfair to younger people. Let’s look into the issue more deeply.

Are we seeing more cases among younger people? Yes, we are seeing more cases among people ages 18–49. As Figure 1 shows, we are seeing far more tests among young people (yellow bar) and an increasing proportion of tests coming back positive over the last 3 weeks (yellow line). More tests and increasing proportion of tests being positive== more cases. [Note: CDC’s data on age groups lumps ages a bit awkwardly, hence the 18–49, 50–64 distinction I’m using for this post]

Is the proportion of cases skewing more youthful over time? Yes. We are seeing an increasing proportion of positive cases among people ages 18–49 as compared with people ages 50–64. Analyzing CDC data, we see that the proportion of positive cases among people ages 18–49 compared with those 50–64 has steadily increased over the last 12 weeks — from 1.2 cases to 1 in early April to 3.7 cases to 1 in late June (data not shown). This has accompanied an increase in testing among people ages 18–49 compared with those ages 50–64–1.8 tests to 1 in early April to 2.4 tests to 1 in late June (data not shown). As these data reveal, the increase in the proportion of cases among those ages 18–49 is steeper than the increase in the proportion of tests. The proportion of cases is increasingly skewing more youthful.

Do we know why we are seeing more cases among younger people? More cases could be due to at least four factors: 1) more tests; 2) young people facing more exposure as they are more engaged in the workforce; 3) older people increasingly avoiding unnecessary risk; 4) younger people increasingly taking unnecessary risk. We know that earlier in the epidemic those with mild/moderate symptoms were asked to stay home and many of them did not receive tests. As testing has expanded, younger people are increasingly able to test. That said, the fact that the proportion of tests coming back positive has been increasing over the last several weeks means that more testing is not the full reason for more cases. When it comes to the other three factors, we just don’t have enough data to know to what degree each of these factors is influencing the changing demographics of the epidemic in the United States. We do know that younger people are more likely to have lower paying jobs, more likely to be in riskier, essential-worker-type jobs — health care, meat-packing plants, grocery stores, and factories — and less like to have sick time off compared with older people.

How do we talk about risk? Social behavior change is of utmost importance! People need education, but education is not sufficient. And in the context of ongoing risk, we have to use harm reduction strategies, which means helping people navigate risk rather than telling people to simply avoid it — very few people can hole up at home for the long-term, and this is especially true for younger people. Figure 2 presents a helpful infographic on social behavior change from a recent article published in Nature. This paper gives a great primer on social behavior change. Meanwhile, the Johns Hopkins Center for Communication Programs is leading the COVID Communications Network to help professionals with messaging, including risk communication.

Figure 1. Trends in COVID Tests and Test Positivity Rates by Age Group (from CDC, updated July 3rd)

Figure 2. Social Behavior Change Issues (from Nature)

Q&A for 7/6:

#Allergies #Symptoms

Question: I have a persistent itchy throat. I spend a lot of time outdoors and think my issue is allergies, but how can I tell the difference between allergies and COVID symptoms? What are common COVID symptoms?

Answer: We talked about this allergy/COVID symptom differentiation issue a long time ago — back in our Q&A of 3/14(!) — but we know a lot more now, and some of what I wrote back then is no longer correct, so let’s revisit. First, I’m not a doctor and if you have any health concerns, please talk with your doctor. Second, CDC recently updated its list of COVID symptoms based on new evidence (Table 1). As you’ll see on this list, there are some overlaps with symptoms of common allergies, but some differentiation. For example, if your throat is scratchy (rather than sore), that’s not a sign of COVID. In early May, the World Allergy Organization Journal published this statement paper from the European Forum for Research and Education in Allergy and Airway Diseases, which describes commonalities and major differences between COVID and allergy symptoms (Figure 1). Upshot is, itchy throat==allergies.

When you look at the list of COVID symptoms (Table 1), it may also be helpful to understand how common each of the listed symptoms is. Unfortunately, CDC does not provide such information, which could be because it’s still too early in our understanding to give these types of estimates. Nonetheless, because I’m interested, I pulled a few pieces of data from recent research. Table 2 shows frequency of symptoms among children and adolescents who participated in a European cohort study. Among children and adolescents, the most common symptoms were fever (65%) and upper respiratory tract infection (54%) (diagnosed based on congestion/runny nose, sore throat, ear ache. cough, sinusitis). CDC shares more information on pediatric clinical presentation here. For adults, the European CDC offers a nice synthesis of recent evidence, which I’ve copied here for ease of reference:

• An Observational study of 1,420 patients with mild or moderate disease indicated that the most common symptoms were headache (70%), loss of smell (70%), nasal obstruction (68%), cough (63%), fatigue (63%), muscle pain (63%), runny nose (60%), gustatory dysfunction (54%) and sore throat (53%), and fever (45%).
• The International Severe Acute Respiratory and Emerging Infections Consortium reported on 25,849 hospitalized cases of COVID-19 across a broad clinical spectrum. The five most common symptoms at admission were history of fever, shortness of breath, cough, fatigue/malaise, and confusion.
• An analysis of data from 4,203 patients mostly from China identified fever (81%), cough (58%), and difficulty breathing (24%) as the most common clinical symptoms.
• A study among 20,133 hospitalized patients from acute care hospitals in England, Wales and Scotland identified clustering of symptoms with three common clusters: one respiratory symptom cluster with cough, sputum, shortness of breath, and fever; a musculoskeletal symptom cluster with myalgia, joint pain, headache, and fatigue; a cluster of enteric symptoms with abdominal pain, vomiting, and diarrhea. 29% of patients presented with gastrointestinal symptoms on admission, mostly in association with respiratory symptoms, while 4% of patients presented with only gastrointestinal symptoms.
• A systematic review identified olfactory and gustatory dysfunction as common symptoms with a 53% pooled prevalence across 10 studies with a total sample size of 1,627 patients from North America, Europe and Asia.
• A pooled analysis of five studies with 817 patients found altered taste sensation was common 50% of COVID-19 patients.
• Analysis of self-reported symptoms from 7,104 individuals who tested positive and used the COVID Symptoms Study app found, “Loss of smell and taste, together with fever or cough, should now enable us to identify 87.5% of symptomatic COVID-19 cases, although this is likely to be less in the early phases of the infection.”

Table 1. COVID Symptoms (CDC)

• Fever or chills
• Cough
• Shortness of breath or difficulty breathing
• Fatigue
• Muscle or body aches
• Headache
• New loss of taste or smell
• Sore throat
• Congestion or runny nose
• Nausea or vomiting
• Diarrhea

Figure 1. Similarities and Differences between COVID and Allergy Symptoms (from EUFOREA Statement)

Table 2. Frequency of Symptoms in Children and Adolescents (Lancet)

Q&A for 7/5:

#Masks #WHO #Medical vs. Surgical #FDA

Question: I was just reading that WHO recommends people ages 60+ and people with underlying comorbidities should wear a “medical mask” in settings where physical distancing cannot be achieved. My parents fit that criteria, so I wanted to buy them some medical masks to supplement the cloth masks they are currently using. But then I got confused — is there a difference between medical masks and surgical masks? Is any mask that looks like a doctor’s mask (you know the blue pleated ones) a “medical mask”? With so many options not to choose from, I’m struggling to know the difference!

Answer: Interesting question that caused me to revisit the WHO Guidance and FDA approvals, thanks! Longer answer follows, but in short:

• Medical masks and surgical masks are different! Unlike medical masks, surgical masks must meet more regulatory requirements regarding fluid penetration and flammability requirements. (Note: I do not believe that they necessarily *look* different)
• Please avoid purchasing surgical masks as there are still healthcare personal protective equipment supply issues and we should save the surgical masks (and N95s) for health care personnel.
• When it comes to shopping for medical masks, the main thing to pay attention to is ASTM or EN-rating. If the mask meets either ASTM F2100 or EN 14683 ratings, it has performance characteristics recommended by WHO.

Now the details. First up, Figure 1 is the WHO definition of “medical mask.” Here, to be called a “medical mask,” specific performance characteristics must be met. When it come to performance characteristics, WHO specifies ASTM F2100, which is the American Society for Testing and Materials (ASTM) set of specifications for “classifications, performance requirements, and test methods for the materials used in the construction of medical face masks”, and EN 14683, which is the European Standard. Importantly, the WHO definition makes “fluid penetration resistance” optional, which means that ASTM F2100 Level 1 masks (as opposed to Level 2 or 3) are appropriate for your parents to use (Figure 2).

In the United States, all medical devices are regulated by the FDA, which *typically* means that in order to be designated a “medical” or “surgical” mask, the mask must receive FDA approval. Due to the COVID public health emergency, we are in *atypical* times and FDA has therefore eased regulatory approvals for face masks. Under the FDA’s emergency use authorization (EUA), “face masks for a medical purpose that are not intended to provide liquid barrier protection” can be manufactured and marketed without FDA notification and do not have to comply with certain regulatory requirements. Meanwhile, FDA calls those face masks that are intended to provide fluid penetration resistance “surgical masks.” Surgical masks must meet additional ASTM standards and flammability requirements, though under the EUA, surgical masks can also be manufactured and marketed without FDA notification and do not have to comply with certain regulatory requirements. All face masks that are being manufactured and marketed under the EUA must “clearly and conspicuously state that the product has not been FDA cleared or approved, the product has been authorized by FDA under an EUA…” I raise this in the context of shopping — if you’re on Amazon and see this type of statement, it’s not necessarily an indictment on the product being sold, rather the product is new and following EUA guidelines.

Note: CDC offers this infographic on surgical mask and N95 mask differentiation and FDA offers this description (Figure 3).

Figure 1. WHO “Medical Mask” Definition

Figure 2. ASTM Levels of Mask Protection (from Cardinal Health Face Mask Selection Guide)

Figure 3. FDA Surgical Mask Description (from FDA)

Q&A for 7/4:

#Asymptomatic Transmission

Happy 4th of July!

Question: In follow up to yesterday’s Q&A, one piece that I’m still at a loss around: if she‘s infected but asymptomatic, then regardless of a 14 day quarantine without symptoms, she can still pass the virus along, right? That’s why I was thinking quarantine plus a test would be more confirmation that she’s not infected. Do you know of any research on detecting infection in asymptomatic carriers? (I imagine this sort of data would be hard to get!)

Answer: Good question! And you’re right, because of the threat of asymptomatic transmission, self-isolation + test would be better confirmation than self-isolation alone. Here’s an update on what we currently know about asymptomatic transmission (not enough, but more than last week!):

What proportion of infected individuals remain asymptomatic?

CDC estimates that 35% of infected individuals are asymptomatic (range: 20%-50%). The proportion could very well depend on the characteristics of the population (like age, underlying conditions, previous exposure to other coronaviruses). For example,in the nursing home study we discussed in our Q&A of 4/3, only 13% of those infected (3 of 23 individuals) remained asymptomatic. Meanwhile, a study published by Nature just 4 days ago presented fascinating results from testing nearly everyone in a small Italian town at the beginning of lockdown and two weeks after lock down. The scientists found “42.5% (95% CI 31.5–54.6%) of the confirmed SARS-CoV-2 infections detected across the two surveys were asymptomatic (i.e. did not have symptoms at the time of swab testing and did not develop symptoms afterwards).” This paper from mid-June, COVID-19: asymptomatic carrier transmission is an underestimated problem, and this paper published in early June, “Prevalence of Asymptomatic SARS-CoV-2 Infection” both provide helpful syntheses of evidence to date.

Is the virus able to spread as easily from asymptomatic individuals as symptomatic individuals?

CDC estimates that asymptomatic individuals are as infectious as symptomatic individuals (range: half as infectious to as infectious). In line with CDC’s estimate, this literature review published by Journal of Infection last week found that viral loads are similar between symptomatic and asymptomatic individuals.

What is the duration of infectivity among asymptomatic individuals?

Our current best guess at how long asymptomatic people may be infectious is 10 days after testing positive. WHO recommends asymptomatic individuals leave self-isolation/quarantine 10 days after testing positive. This recommendation aligns with findings from studies that measure the duration of virus detection in asymptomatic individuals. For example, this study of 24 individuals with asymptomatic infection in Nanjing found the median duration of infectivity as measured by time from positive test two consecutive negative tests was 9.5 days while this study from Jiangsu Province found the median duration to be 7 days.

For more background, see Q&As of 6/30, 6/9. 5/28, 4/18, and 4/3.

Q&A for 7/3:

#Self isolation #False Negative

Question: In mid August, our 30 yo nanny is travelling first to Arizona (ack!) to visit her parents, and then to Cancun with friends for vacation. We want to do our best to make sure she’s COVID-free before returning to work. Starting the day she returns home, we are planning to have her quarantine for a week, to see if she develops any symptoms. I thought it might also be wise to have her get a COVID test, to see if perhaps she was indeed infected, but is asymptomatic. My questions are: 1. Is having her stay away for 7 days post-return reasonable, or should we wait 10 or even 14? 2.If she gets a test, how soon can she get it and it still be useful? For example, going off our 7 day quarantine timeframe, if she gets the test on day 4 of being home, and we expect to get results 3 days later (in time for her to hopefully return to work on the 8th day), is there utility in getting a test that early? Would it show COVID infection if indeed there was one?

Answer: Any travel internationally necessitates 14-day self isolation upon return, per CDC (Figure 1). This 14 day period stems from data that show COVID-19 symptoms generally begin 5 days after exposure (range 2–14 days) (further described in Q&A of 3/29 and more recently this paper from Annals of Internal Medicine, which found that 97.5% of people who develop symptoms will do so within 11.5 days). In answer to your first question, your nanny needs to stay away from you and others for the full 14 days.

When it comes to your second question on how soon COVID tests can detect the virus, this is still an area of scientific exploration (read: we don’t fully know). For the most common swab tests (e.g. PCR tests), we can be confident that a positive result is a true positive (for more on PCR tests, see Q&A of 3/9; for more on sensitivity/specificity, see Q&A of 4/15), but we can’t be so confident in a negative result. This overview from Harvard Medical School offers a synthesis of findings from this study on false negative rates published in mid- May by Annals of Internal Medicine:

“If you get the nasal/throat swab or saliva test, you will get a false negative test result:

• 100% of the time on the day you are exposed to the virus. (There are so few viral particles in your nose or saliva so soon after infection that the test cannot detect them.)
• About 40% of the time if you are tested four days after exposure to the virus.
• About 20% of the time if you develop symptoms and are tested three days after those symptoms started.”

And when it comes to how long a person must wait for results to be returned, this really depends on lab capacity and test type. Generally, test results are available in 2–5 days.

To sum up, if your nanny goes the full 14 days without symptoms, she doesn’t need to get tested. If she chooses to skip her planned international travel and your family cannot be without her for 2 weeks, one happy(ish) medium could be to do as you suggest — ask her to self-isolate after returning from Arizona, get tested on the 4th day, and return to work when she receives her negative test result, assuming she remains without symptoms (this would likely be 7–9 days after her return).

Figure 1. Returning from International Travel

Q&A for 7/2:

#Social Behavior Change #Communication #Masks

Question: Without fail, every time I go to the grocery store, there are employees wearing masks incorrectly (under the nose or around the neck!) or not wearing them at all. Today, the employee supervising the self checkout approached me wearing the mask over her mouth and under her nose. When I pointed this out, diplomatically I thought, she failed to change her behavior and looked quite vexed with me. My experiences challenging the incorrect wearing of the mask rarely result in behavior change. My conclusion is it’s a failure in training and oversight on the part of grocery store management but I have no data to back this up. How do we get this behavior to change or do we just all order groceries for pick-up or delivery?

Answer: I don’t have an easy solution, but please don’t give up on people yet! Mask wearing is new and we have a long way to go in terms of people adopting the behavior, let alone adopting correct and consistent use. And there’s plenty of blame to go around when it comes to the lack of a national policy, lack of comprehensive public health communication campaigns, and the politicization of mask wearing (among other challenges). Behavior change is hard and interpersonal communication about adopting a given behavior is extra hard. We know that education is important, but not enough — as behavioral scientists recently wrote in Nature on the topic, persuasion, behavior modeling, and more will be critical. Here are a few tips for interpersonal communication, which align with points made in this recent ReWire article and The Atlantic article:

• Communicate risk in simple terms. For example, “Keep you and me safe. Wear a mask over your nose and mouth.” (for another example see, Figure 1)
• Do not shame, guilt, or judge people into compliance. It doesn’t work and can make things worse.
• Be honest. Empathy is powerful. Don’t pretend that wearing a mask is comfortable or that wearing a mask doesn’t have drawbacks — like impeding facial cues for communication.
• Model the behavior. Wear your mask and wear it correctly. (for example, Figure 2)

Figure 1. Messaging from Greyhound

Figure 2. How to Wear a Mask (from Africa CDC)

Q&A for 7/1:

#Travel #Airplanes

Question: How do you feel about flying? Is it safer than driving? Less safe? How about for children? Would you fly with your children right now to a place that has a downward trend? I hear things about the air actually being cleaner on airplanes and then I hear the opposite?

Answer: We talked about flying in our Q&A of 5/16, but it’s worth a revisit since there’s a bit more information now and we’re all still trying to figure out how to live our lives during pandemic times.

1. How do you feel about flying?: When it comes to my feelings about flying, if I can avoid it, I will — but I feel that way about most activities that put me into contact with groups of people.
2. Is it safer than driving? Less safe?: When it comes to flying vs. driving, I think it really depends on a host of factors that depend on your given circumstances. If you can make a long drive within a day and get to your destination, I recommend taking that route. You’ll be exposed to far fewer people, which means it will be safer than flying. But, if we’re talking about a cross-country trip with lots of stops, the calculus might be different as you’ll have more opportunities for exposure (and likely some mental health challenges if traveling with small kids).
3. How about for children?: when it comes to traveling with kiddos, the good news is that data continues to accumulate that the coronavirus tends to be mild in children. For example, just last week Lancet published a study, COVID-19 in children and adolescents in Europe: a multinational, multicentre cohort study, which confirms previous reports from China that COVID-19 is generally mild in children and infants. That said, we do see instances of severe disease in children — it’s far more rare than in adults, but it does happen. So if you are traveling with kids, be sure to be extra vigilant — cleaning spaces, washing hands, keeping distance, wearing masks, avoiding touching stuff and faces.
4. Would you fly with your children right now to a place that has a downward trend?: I don’t have reason to fly anywhere these days, so I’m avoiding it. But I did take a road trip down to Nashville a few weeks ago and I plan to take another road trip to a Georgia beach in August, both with my little boy. We talked about the risk/reward calculation in our Q&A of 5/16 and I think it still holds today. And when it comes to flying, here’s a bit more updated information:

• Dr. Erin Bromage (who wrote a helpful blog post in May about risk that we discussed in Q&A of 5/12) has another helpful blog post he updated in early June, “Flying in the Age of COVID” that offers a great overview and advice based on his knowledge and experience flying with his family to Australia in March.
• As I understand it and as CDC reports, the risk of flying is largely from the rigmarole surrounding flying — checking in, security, lining up to get on the plane, etc — and not the flight itself.
• In fact, BBC reported a few days ago, “Although many people might think that sitting in a confined space for long periods would inevitably spread infections, the chief engineer at aerospace giant Airbus insists that is not the case. Jean-Brice Dumont argues that the way modern aircraft are designed means that the air is intrinsically very clean. “Every two to three minutes, mathematically, all the air is renewed,” he says. “That means 20 to 30 times per hour, the air around you is completely renewed.” Put simply, air is collected from outside the aircraft, normally through the engine, and mixed with recycled air from the cabin. The recycled air, which is reused in part to keep temperature and humidity at the correct levels, is passed through HEPA (high-efficiency particulate air) filters that are similar to those used in hospitals.” Figure 1 from How It Works shows how air circulation in planes works.

Figure 1. Air Circulation in Airplanes (from How It Works)

Q&A for 6/30:

#Transmission #Asymptomatic

Question: There’s one thing I don’t have a good handle on….. isn’t it possible that ANYONE could transmit the virus, regardless of their history (testing positive, negative, symptomatic, asymptotic, had it but recovered, etc.)? In other words, is everyone a potential transmitter at any given time?

Answer: One of the great challenges of COVID is that viral shedding begins before people exhibit symptoms, which means that people can be spreading the infection before they know they are sick with it (see Q&A of 6/9). A new systematic review on SARS-CoV-2 viral load and infectivity published in the Journal of Infection yesterday found “little to no difference in viral load between pre-symptomatic, asymptomatic and symptomatic patients.” Furthermore, the study authors conclude that “viral load of SARS-CoV-2 peaks around symptom onset or a few days thereafter, and becomes undetectable about two weeks after symptom onset…however the duration of infectivity remains uncertain.” The ability to unknowingly spread the virus makes common public health interventions — like telling people to stay home when sick or monitoring an individual’s temperature as they enter establishments — less effective at slowing the spread of the virus. Because of pre-/asymptomatic transmission risk, it is generally safest if each of us considers ourselves a potential risk to others and takes action to protect others — it makes wearing a face mask and keeping at least 6 feet apart all the more important and relevant. A couple of additional notes —

• Thus far, we have no evidence that people who have recovered from COVID can transmit the virus, even people who go on to test positive after having recovered (see Q&A from 5/19). We’re still learning as we go, so this is not definitive, but it is nonetheless positive. On this front, it’s important to do our utmost to avoid stigmatizing those who have been sick with COVID.
• When it comes to those who test positive for antibodies without having previously tested positive for COVID, we do not know whether: a) their antibody test result represents a true positive or false positive (see Q&A of 4/15); b) the antibodies they exhibit offer protection from the coronavirus; c) they can be (re)infected with coronavirus (see Q&A of 4/25). These folks should assume that they are at risk of becoming infected and infecting others.
• Status is not fixed. Just because a person tested negative several weeks ago does not mean that they remain negative. It is imperative that everyone, even those who have tested negative, practice behaviors that reduce risk, including and especially wearing a face mask, keeping 6+ feet apart, avoiding gatherings (especially indoors), washing hands, etc.

Q&A for 6/29:

#Risk Reduction #Face masks

Question: Leaders across the spectrum have finally gotten on board with recommending the wearing of masks, but I fear that people think that as long as they wear a mask they can do whatever they want. Isn’t social distancing, especially indoors, more important than wearing masks to curtail the spread?

Answer: Social distancing and mask wearing are both extremely important for curbing virus spread. We should be doing both when we are out in the world in the presence of others. Social distancing does not negate the need for masks and masks do not negate the need for social distancing! While neither of these interventions make the risk of infection nil, if we want to keep ourselves and each other safe, we should be practicing both. We should also be helping to ensure that health providers have necessary personal protective equipment. Read on for the data…

A meta-analysis, “Physical distancing, face masks, and eye protection to prevent person-to-person transmission of SARS-CoV-2 and COVID-19”, published in Lancet earlier this month found that proximity increased risk of infection (absolute risk 12·8% with shorter distance vs 2·6% with further distance, risk difference −10·2%, 95% CI −11·5 to −7·5). Additionally, the strength of association was larger with increasing distance and remained regardless of face mask type worn. Study authors found an even higher risk reduction related to the use of face masks (absolute risk of infection 17·4% with no face mask vs. 3·1% with face mask, risk difference −14·3%, 95% CI −15·9 to −10·7) though most of the face mask studies were among health workers in hospital settings. Generally in non-healthcare settings, it is unclear to what degree face masks protect the wearer. What is clear in non-healthcare settings is that face masks keep the wearer’s germs from more widely spreading, thereby reducing the risk of infection for others (for more, see Q&A of 6/22).

Finally, the fear your raise reminds me of earlier conversations about condoms and HIV — there was a time when people were concerned that in teaching people about the relationship between condoms and reduced HIV acquisition risk and in widely distributing condoms, we would be encouraging more risk sexual behavior. Evidence shows that such programs do not increase risky sexual behavior and can even promote delayed sexual initiation among youth (among other positive findings). When it comes to COVID, I think the more comprehensive risk reduction communication we see, the more risk reduction behaviors we’ll see.

Q&A for 6/28:

#Transmission Seasonality

Question: Clearly, summertime conditions (heat, humidity, people spending more time outdoors) do not reduce coronavirus spread in any meaningful way (in contrast to seasonal flu, and to the other coronaviruses that cause common colds). Could this mean, conversely, that winter conditions won’t make Covid much worse? Many folks had been expecting Covid to “lay low” over the summer and explode in fall and winter, but maybe that’s not how this disease works? Maybe it has almost no seasonal component?

Answer: While I hope that the pandemic won’t be worse in the fall and winter, I don’t think that will be the case. It’s bad now and it’s likely to be worse then.

There has been a lot of debate, both political and scientific, about whether COVID would lessen or even “burn off” in the summer. We last talked about this in our Q&A of 3/22 and I still find informative this Harvard professor’s blog on the topic, which I cited back then. Because the virus is relatively easy to transmit and because we have no/limited immune protection, experts generally believe that the virus will overcome any effect warm/humid weather may have (Vox summarizes here). As experts pulled together by the National Academies for a rapid consultation on the matter wrote in their report of 4/7, “In summary, although experimental studies show a relationship between higher temperatures and humidity levels, and reduced survival of SARS-CoV-2 in the laboratory, there are many other factors besides environmental temperature, humidity, and survival of the virus outside of the host, that influence and determine transmission rates among humans in the “real world.”” This brings us to the issue of fall and winter transmission. Unfortunately, fall and winter bring conditions besides environmental temperature and humidity that make SARS-CoV-2 easier to spread — including more time indoors, more time in close quarters, and more illness (like flu and the common cold) that will both tax our health systems and our immune systems.

Finally, here’s where I could be wrong with my assessment up top. JAMA published an article earlier this month, “Temperature, Humidity, and Latitude Analysis to Estimate Potential Spread and Seasonality of COVID-19,” which concluded, “Using weather modeling, it may be possible to estimate the regions most likely to be at higher risk of substantial community spread of COVID-19 in the upcoming weeks and months, allowing for a concentration of public health efforts on surveillance and containment.” These findings indicate that environmental factors make some geographies more/less likely to be hit hard by the pandemic at different points in the year. But these findings do not prove a direct causation and have big limitations — including no incorporation of myriad factors into the model, including public health interventions or population density. So at this stage, I still wouldn’t put much weight on them and I stand by my assessment. The pandemic is likely to be worse in the fall and winter.

Q&A for 6/27:

#Testing

Question: The President recently stated at a rally in Oklahoma, “When you do testing to that extent, you’re going to find more people. You’re going to find more cases. So I said to my people, ‘Slow the testing down, please.’” Do we have any evidence of testing slowing in the United States?

Answer: These remarks are concerning, regardless of whether they were made in jest or they do not reflect an actual policy shift. COVID-19 has killed more than 125,000 Americans and counting. Downplaying the case estimates in the United States by claiming they are merely the result of increased testing only serves to downplay the threat and severity of the epidemic. These remarks are DANGEROUS, regardless of whether they equate to an actual slow-down in testing. Included herein is some background on the remarks, and a more direct answer to your question.

Background: After the President’s remarks were widely reported, White House officials walked them back, with one official telling CBS news that they were made “in jest”, another telling CNN that they were “tongue-and-cheek”, and the Vice President telling Governors that they were merely a “passing observation.” Then, speaking with reporters on June 23rd, the President himself responded to a reporter’s question on the matter (see C-SPAN for full transcript):

Q: Mr. President, at that rally, when you said you asked your people to slow down testing, were you just kidding or do you have a plan to slow down testing?
A: I don’t kid….The reason we have more cases than other countries is because our testing is so much.

Additionally, as Politico reports, “In an interview with Scripps Networks, Trump did not deny asking his administration to curtail coronavirus testing and instead contended that “if we did slow it down, we wouldn’t show nearly as many cases.”” Given these statements, plus the ongoing concern about the economy, especially reopening in light of increasing cases, it would seem plausible that the President is interested in slowing testing down. So let’s look into whether there’s any evidence of such a slow down.

Evidence: First, we have the response of Dr. Fauci and three other government officials on the matter during Tuesday’s House hearing. In response to questions on testing, Dr. Fauci stated, “To my knowledge, none of us have ever been told to slow down on testing — that just is a fact. In fact, we will be doing more testing … not only testing to specifically identify people [and] identify, isolate, and contact-trace, but also much more surveillance.” That’s a positive. BUT…

On Wednesday, a day after the House hearing and the President’s “I don’t kid” remarks, the administration said that it would no longer directly fund 13 of its original coronavirus testing sites in five states — “funding and support for the sites in Illinois, New Jersey, Colorado and Pennsylvania as well as Texas would end June 30.” In response to concerns, Reuters reports that “U.S. Assistant Secretary for Health Admiral Brett Giroir said that he had spoken to leaders of the five states, who agreed “it was the appropriate time to transition” to other options. He said the states could use the more than $10 billion allocated last month to support testing to keep the sites open if they chose to.” So that’s concerning, but not tangible proof.

Let’s look at testing numbers themselves. As Figure 1 shows, testing in the United States (excluding NYS since it has conducted a huge proportion of tests over time) has been increasing over time, although the increase is at a slower pace in June as compared with May (see slope). So, we have evidence of a slow down of the testing ramp up… but not evidence of fewer tests themselves. We’ll need to keep monitoring the situation and holding our leaders accountable.

Figure 1. Daily Tests in the US (excluding New York) [data from covidtracking.com]

Q&A for 6/26:

#Testing #Cross-Country Comparisons

Question: Unlike Trump, I wouldn’t advocate for less testing, but the rise in cases does make me wonder what the US testing rate is compared to Europe’s average and China’s average and Brazil’s average.

Answer: When it comes to the sheer volume of tests, the United States has indeed conducted the most tests (nearly 30 million total) of any country in the world (Table 1). Because the US is the third most populous country in the world (behind China and India), one of the richest countries in the world (richest by total GDP, one of the richest by GDP/capita), and one of the most hard-hit countries in the world (most COVID cases; most COVID deaths), we’d expect the US to have conducted a high number of tests. This isn’t a “win” to brag about in my opinion. Let’s look at a few other comparisons: a) testing per capita and b) test positivity rate.

By population, the US testing rate is now higher than most countries, though it is not the highest and it has taken several months for the US to achieve this level (Figure 1). When it comes to test positivity rates, the 7-day rolling average of daily percentage of tests that are positive in the US is 6.1%, 6.8% if NYS is removed (data from covidtracking.com). This positivity level is growing and is higher as compared with Europe and Australia, but lower as compared with South America and parts of Asia (Figure 2). This means that there’s still room to further ramp testing up in the US — higher positivity rates indicate that a proportion of the infected population is likely still untested. On this front, WHO issued guidance stating that positivity rates should be <5% for at least 14 days (assuming surveillance is comprehensive) before governments begin relaxing social distancing measures.

Finally, when it comes to making comparisons with China and Brazil, we are limited by a lack of transparency. As you’ll see in the below tables and charts, China’s data does not appear. That said, China does claim to have conducted the most tests of any country in the world — 90 million per Xinhua, as reported in The Hill on Monday — but experts question the country’s case and death counts. Brazil has also had its share of data transparency issues. Earlier this month, the Supreme Court of Brazil issued an order for the ministry of health to release COVID data. Still, from what we do know, Brazil lags far behind in testing and has the highest test positivity rate of all countries in the world (~37%).

If you want to explore further, here’s another shout-out to Dr. YJ Choi, who keeps updated trend data on the latest comparisons of COVID across OECD countries.

Table 1. Cumulative Tests (from Our World in Data)

Figure 1. Country-level Trends in Daily Testing Rate per 1,000 People (from Our World in Data)

Figure 2. Percentage of Daily COVID Tests that are Positive (from Our World in Data)

Q&A for 6/25:

#USA #Cases

Question: I’ve been reading about how many more cases are popping up across the United States. Is it really as bad as it sounds?

Answer: Uggghhh. Yes, it’s bad. COVID case numbers are rising in more than half of the states. And while the US accounts for only 4.3% of the world’s population, we have 25% of the world’s cases. Over the course of April and May, thanks to social distancing and non-medical interventions, the United States had succeeded in flattening the curve and had even begun to bend the curve (Figure 1). Unfortunately, our progress in bending the curve stalled in late May. Since early June, we’ve continued to see a sharp increase in the number of daily cases (Figure 1). It’s important to note, because the initial epicenter in the US was New York, the overall picture of the United States is skewed with New York’s experience. If we take New York out of the picture, we see that the US as a whole merely flattened the curve over April and May; we never really got to bending it (Figure 2a). And since early June, cases are back on the pre-shutdown trajectory (Figure 2b). It’s looking like we gave ourselves so much time to get our testing, tracing, and health facility response together at huge personal, societal, and economic cost just to land back where we started. A few more notes:

• The picture across the US varies dramatically by location (Figure 3) with the Northeast on a sharp downward trajectory and the South and West accounting for the large increases.
• When it comes to tracking cases over time, we do indeed have more testing (good!) (Figure 4a). But… the percentage of tests coming back positive is also increasing (bad!) (Figure 4b) .
• States don’t uniformly report hospitalization data, so we don’t have that piece to compare here. We can use CDC’s hospitalizations tracker, which does show hospitalizations on a downward slope (good!), but was last updated June 13th (bad!), and we know that hospitalizations lag cases. It’s possible that because we are doing more testing, we are capturing more mild and moderate cases as compared with earlier in the epidemic, so the case:hospitalization ratios won’t be as high as what we saw in New York (good!). For a reminder of timing of illness onset to hospitalization, see Q&A of 3/29.
• In spite of limited hospitalization data, CDC’s ensemble forecast of deaths, updated yesterday, suggest that “the number of new deaths over the next four weeks in Arizona, Arkansas, California, Florida, Hawaii, Missouri, Nevada, North Carolina, Oklahoma, South Carolina, Tennessee, Texas, and Utah will likely exceed the number reported over the last four weeks.” (Figure 5)

Figure 1. Daily COVID Cases in the United States (data from covidtracking.com)

Figures 2a and 2b. Daily COVID Cases in the United States Excluding New York (data from covidtracking.com)

Figure 3. Regional Differences in New Cases (chart from The COVID Tracking Project)

Figures 4a and 4b. Tests and Percent of Tests that are Positive in the USA, Excluding New York (data from covidtracking.com)

Figure 5. CDC’s Ensemble Forecast of Total Deaths suggests 130K-150K Deaths by July 18

Q&A for 6/24:

#Polio #MMR #Vaccine

Question: I was just reading this article published by the American Society of Microbiology. Could recent vaccination for other diseases help protect against COVID?

Answer: Ohh…. I’m glad to revisit this hypothesis now that there’s more information! In short, we still don’t know, but it’s seeming more plausible that live attenuated vaccines could offer some protection against COVID. We talked about BCG vaccination offering potential COVID protection a while back (see Q&A of 4/27). Since then, there have been several other reports, including an article on oral polio vaccine (OPV) and the article you cite on measles, mumps, rubella vaccine (MMR), about the plausibility of other live attenuated vaccines offering protection. Note: to my knowledge, scientists are arguing that existing live attenuated vaccines can serve as stop-gap measures in the context of no SARS-CoV-2 vaccine; everyone is still pushing for a SARS-CoV-2 vaccine!

As a reminder, live attenuated vaccines are those that use a weakened (attenuated) form of the virus to elicit an immune response. Because these types of vaccines create such a strong immune response, the protection they offer typically lasts a lifetime. What’s also fascinating about these types of vaccines (in my opinion) is that they not only allow a person’s body to develop antibodies against the specific virus, they also prompt an innate immune response that helps protect against other viruses! This article from Science published earlier this month, “Can existing live vaccines prevent COVID-19?”, does an amazing job laying out the evidence regarding the relationship between live attenuated vaccines, the innate immune response (for a refresher on the innate immune system, see our Q&A of 5/9), and long-term outcomes (e.g. reduced mortality). Scientists aren’t sure of the exact mechanisms by which live attenuated vaccines offer non-specific protection, but the long and short of it is that they seem to boost the immune system so it is better primed to quickly and effectively respond to virus invaders.

There are a number of BCG vaccine studies in the works now. I just checked clinicaltrials.gov and see that there’s also an MMR trial in the works out of the University of Cairo in Egypt and one OPV trial in the works out of the Bandim Health Project in Guinea-Bissau. The authors of the article published by the American Society of Microbiology state that they are also planning an MMR trial out of New Orleans.

If any live attenuated vaccine offers protection against COVID, we’d want to use the vaccine that is in the most abundant supply and has the fewest side effects. The authors of the article published in Science that I referenced earlier make a strong case for oral polio vaccine over BCG, arguing that OPV is safer, cheaper, in more abundant supply, and more likely to induce a common innate immunity reaction since both polioviruses and coronaviruses are positive-strand RNA viruses. Between OPV and MMR, I’m not sure which would be preferred. Finally, for a nice synthesis of SARS-CoV-2 vaccines, NPR published this piece earlier today.

Image 1. Oral Polio Vaccine, Haiti 2010 (image from USAID)

Q&A for 6/23:

#Blood Type

Question: I’ve been reading that some blood types are more at risk for contracting a severe version of the virus than others, was wondering if you could touch on this with more data? What I’m seeing is that people with Blood Type A are the most at risk and people with Blood Type O are the least.

Answer: I’m glad you asked. I had read some initial news reports about this finding a few weeks ago when the paper was a pre-print and then I set it aside to wait for the paper to be peer-reviewed. And now it has been! Published a few days ago, findings suggest that people with blood type A have higher odds of severe disease (45% increase in odds) than people with other blood types; people with blood type O have lower odds of severe disease (35% decrease in odds) compared with people of all other blood types. Let me give you an overview of the study and then some context regarding the blood type finding. One thing I want to note is that the difference in severity isn’t due to the form of virus contracted (e.g. a “more severe version”), but rather in how the body responds to the virus itself (e.g. a “more severe outcome”). So far, we do not have solid evidence about COVID-19 mutations causing more/less severe outcomes. Finally, if you have Blood Type O, do not rejoice — Lower odds does not mean no risk! Similarly, if you have Blood Type A, do not despair — Higher odds does not mean severe disease is a given!

Research overview: The new research, conducted by a group of scientists called the Severe COVID-19 GWAS Group (here, GWAS is Genome-Wide Association Study), was published in the New England Journal of Medicine a few days ago. Basically, in the search to understand why some people suffer more severe outcomes from the virus, this group of scientists turned to a relatively new study form — genome-wide association study. The scientists used the GWAS methodology to search people’s genes for small variations, called single nucleotide polymorphisms or SNPs (pronounced “snips”), that occur more frequently in people with a given disease compared to people without the disease. Here, the disease is severe COVID, defined as respiratory failure that necessitates use of oxygen therapy or ventilator support. The group compared the genomes of 1,980 study participants in Italy and Spain who had severe COVID with the genomes of 2,381 control participants in Italy and Spain whose infection status was unknown. The researchers found two gene clusters associated with severe COVID: Cluster 1 is involved in the regulation of the immune response, especially in the lungs, and Cluster 2 is involved in determining the ABO blood group type. When the scientists recognized Cluster 2’s relationship to blood type, they analyzed outcomes by blood type, controlling for age and sex, and found “a higher risk among persons with blood group A than among patients with other blood groups (odds ratio, 1.45; 95% CI, 1.20 to 1.75; P=1.48×10−4) and a protective effect for blood group O as compared with the other blood groups (odds ratio, 0.65; 95% CI, 0.53 to 0.79; P=1.06×10−5).” The authors further posit in the discussion section that blood group A may be more susceptible to acquiring the virus, but this is not a finding directly borne out in the data they present, so I wouldn’t put much weight in that component of the paper. As far as GWAS studies go, this one was quick and relatively small. Most (all?) scientists are waiting on more data before embracing the finding that blood type impacts disease severity. Indeed, the NIH is currently funding a set of researchers to use GWAS to study gene variants among 5,000 COVID patients in the US and Canada. That said, there is a body of science that supports the link between blood type and severity of other forms of disease, including other coronaviruses. And recent research out of Wuhan, “Association between ABO blood groups and risk of SARS-CoV2 pneumonia”, also supports the blood type findings. And if you want to explore blood type differences further, the Australian Academy of Science offers a really nice, easy to read synthesis (Figure 1).

Figure 1. Blood Type Differences (from: Australian Academy of Science)

Q&A for 6/22:

#Face Masks #Barriers #Pandemic Culture

I’m back!

Question: Are folks in Nashville (Tennessee) taking COVID seriously? How does the pandemic culture compare with the DC-area?

Answer: Pandemic culture between the DC area and Nashville area is vastly different in my estimation. From what I saw in Nashville, very few people wear masks. Shops don’t require customers to wear them and people just don’t seem to use them. Tourists are abundant, walking in large, unmasked groups, standing in big, undistanced lines, eating outside on crowded decks, having rooftop parties. Given how quickly cases are rising in Tennessee (Figure 1), I found the more “free wheeling” culture rather disheartening. As a reminder to us all, let me take a moment to reflect on why mask wearing is important:

Mask wearing over the nose and mouth keeps respiratory droplets from spreading more widely, thereby decreasing the ability of the virus to spread from the infected person to others. Because the virus can spread during the pre-symptomatic period (see Q&A of 6/9), it is important to wear a mask even if you are not experiencing illness. A growing body of scientific evidence supports face masks as a key intervention for reducing the risk of viral spread and slowing the spread of SARS-CoV-2. For example, a meta-analysis of 172 observational studies published in the Lancet earlier this month found that “Face mask use could result in a large reduction in risk of infection.” For a nice round-up of the additional accumulating evidence, check out this report that NPR published yesterday, “Yes, wearing masks helps. Here’s Why.”

All that said, I get why mask wearing is an issue — we have knowledge, attitude, and behavioral barriers to mask wearing (examples of which are in the bullets below). And if we are going to overcome reticence to mask wearing, we are going to have to tackle these barriers. In good news, many of the issues are similar to other public health challenges — like condom wearing for HIV/STD prevention and seat-belt wearing for injury prevention — and we can use our knowledge and solutions from these challenges to inform our approach to mask wearing.

• Knowledge: Public health communication hasn’t been clear or consistent. Initial CDC and WHO guidance was to only wear a mask if you had symptoms and then changed after a/pre-symptomatic transmission was recognized (for more on this see Q&A of 3/31) — People may not know that asymptomatic transmission is a real threat and therefore think mask wearing is unnecessary unless they feel sick; some scientists have argued that masks can increase face touching and thereby put users at greater risk (note: as more data has come out, this argument has been diminished); leadership on mask wearing has been limited — for example, the President shares guidance to wear masks but does not himself wear a mask.
• Attitudes: Trust in experts has waned; people may fear that if they wear a mask, they will be perceived as sick, weak, or fearful; for some, masks have been turned into a political symbol; for some, they have also been turned into a symbol of emasculation.
• Behaviors: Masks aren’t that comfortable to wear; they are not readily accessible (you typically have to buy them or make them, which takes time, money, and effort); because it’s a new behavior, there need to be ample triggers to remind individuals of the new behavior.

Figure 1. Cases in Tennessee are Rising (data from covidtracking.com)

Q&A for 6/12:

#Herd Immunity #Seroconversion

Note: I’ll be out of town on vacation next week, so Q&A will be paused until 6/22 (unless I can’t resist and decide to answer a question here or there!)

Question: Could the declines in new cases and deaths in Europe be due to herd immunity?

Answer: Many European countries have seen steep decreases in new cases and deaths over the last several weeks (Figure 1 and Figure 2 from Our World in Data). Two studies published yesterday in The Lancet show that these declines are not because of herd immunity. Thus far, the scientific consensus is that no country has seen infection rates sufficient to prevent a second wave. For a refresher on herd immunity, see our Q&A of 5/1. Now on to the results of those two studies:

1. Seroprevalence of anti-SARS-CoV-2 IgG antibodies in Geneva, Switzerland (SEROCoV-POP): a population-based study: This study presents results of a population-based household survey in Geneva, Switzerland that used weekly serosurveys to measure temporal changes in antibody presence among household members ages 5 and up. 2,766 individuals from 1,339 households participated between April 6th and May 9th. Based on survey results, the authors estimated that by late-May, only 10.8% of the Geneva population had been infected (95%CI 8.2%-13.9% — e.g. we can be 95% confident that the true estimate falls between 8.2% and 13.9%). Antibodies were less likely to be present among children ages 5–9 and adults ages 65+ compared with those 10–64 years of age. The authors conclude, “These results suggest that most of the population of Geneva remained uninfected during this wave of the pandemic, despite the high prevalence of COVID-19 in the region (5,000 reported clinical cases over <2·5 months in the population of half a million people). Assuming that the presence of IgG antibodies is associated with immunity, these results highlight that the epidemic is far from coming to an end by means of fewer susceptible people in the population.”
2. Have deaths from COVID-19 in Europe plateaued due to herd immunity?: This paper explores whether declines in COVID-19 deaths in Europe can be attributed to lockdowns, social distancing, and other intervention OR to achievement of herd immunity. Using European CDC data on cumulative deaths, timing of lockdowns, and other relevant seroprevalence surveys, the authors find that herd immunity is NOT the reason for the plateau in deaths. Their reasons include:

• Different levels of mortality at time of plateau: “Under herd immunity, the cumulative mortality rate due to COVID-19 per million of the population would be expected to plateau at roughly the same level in different countries (assuming similar basic reproduction numbers). This is not what the data show. For example, in Germany, the Netherlands, and Italy, all countries with good quality health care and testing capacity, the difference in mortality is several fold, with Germany at 95 deaths per million population, the Netherlands at 332 deaths per million population, and Italy at 525 deaths per million population (as of May 17, 2020).”
• Relationship between lockdown timing and mortality: “Countries that went into lockdown earlier experienced fewer deaths in the following 6 week period. This trend is therefore inconsistent with the herd immunity explanation; however, it is exactly what one would expect under the explanation that lockdowns are curtailing transmission and deaths, making them most effective when pre-lockdown transmission is low.”
• Low levels of seroprevalence and high infection mortality rate: “A strong and consistent relationship exists between the prevalence of antibodies to SARS-CoV-2 and mortality from COVID-19 in European populations, consistent with an infection fatality rate of 0·5–1·0%…which is many times higher than seasonal influenza (<0·1%)” This matters because if herd immunity had been reached through a large proportion of the population being infected, then we would expect to see a) a much higher seroprevalence rate; b) a much lower infection fatality rate; and c) very similar seroprevalence rates across the countries.

Figure 1. Country-Level Trends in Daily New Cases

Figure 2. Trends in Daily Confirmed Deaths in European Countries

Q&A for 6/11:

#Recovery

Question: Well this opinion in the NY Times, “Our Next Crisis Will Be Caring for Survivors of Covid-19”, is really depressing. And my 9/11 health friend has been pointing this out to me for weeks — that survival doesn’t mean “all better”. Is anyone collecting data on how many of those that “survive” are permanently injured?

Answer: So true; Survival doesn’t mean “all better.” Over the last month or two, there have been more and more news reports on this issue, especially the issue of long-term challenges experienced by patients coming off of ventilators and out of the ICU. This review article, recently published in Science of the Total Environment (a journal I was not previously familiar with), discusses the need for longitudinal cohort studies to understand the needs of COVID-19 survivors. It also offers a helpful visualization of the ways in which COVID infection could have long-term physical ramifications (Figure 1). Meanwhile, guidance published by the UK’s National Health Service last week, “After-care needs of in-patients recovering from COVID-19”, provides the best synthesis I’ve seen of the ongoing issues and needs of patients who survive COVID-19. For example, the guidance notes that post-intensive care syndrome (PICS), “an amalgamation of persistent physical, cognitive and psychological impairments, [has been found to] present in 56% of patients at 12 months following prolonged ventilation.” Check out the guidance for more evidence-informed detail!

Finally, I think you’ll be heartened to know that there are a number of longitudinal cohort studies being planned and currently recruiting study participants to understand long-term effects of COVID among people who recover. Here are a few longitudinal cohort studies in the works:

• A Longitudinal Study of COVID-19 Sequelae and Immunity: NIH-planned study that will begin recruitment this month with the aim of enrolling 900 participants who have recovered from COVID-19 to “characterize the clinical sequelae of acute infection, characterize the immune response to the virus, and follow the evolution of the immune response over time and determine the extent to which natural immunity is protective against re-infection.”
• Longitudinal COVID-19 Cohort Study: University of Vermont and Johns Hopkins Univeristy-planned study that began recruitment last month with the aim of enrolling 225 participants with COVID-19 to “longitudinally follow survivors for 12 months, to investigate short-term and longer-term inflammatory/immunologic and clinical outcomes during this pandemic.”
• Quality of Life and Physical Performance After Novel Coronavirus Infection: Kantonsspital Winterthur (Switzerland)-planned study that began recruitment last month with the aim of enrolling 60 participants with COVID-19 to “observe the long-term health-related quality of life (HRQOL) and physical performance in individuals hospitalized due to a COVID-19 infection.”
• Mapping Organ Health Following COVID-19 Disease Due to SARS-CoV-2 Infection: Mayo Clinic (UK) and Gemini (UK)-planned study that just began recruitment with the aim of enrolling 507 participants with COVID-19 who are recovered or recovering to “measure the prevalence of organ volume changes and damage in lungs, heart, kidney, liver, pancreas, spleen as assessed by MRI among those having recovered, or recovering”
• Longitudinal Population-based Observational Study of COVID-19 in the UK Population: Queen Mary University of London-led study that should be in the recruitment phase now with the aim of enrolling 12,000 participants who are UK residents age 16+ to “1. Determine risk factors for incident COVID-19 and for adverse outcomes of COVID-19 in the UK population; 2. Characterise the natural history of COVID-19 in the UK population; 3. Evaluate the impact of COVID-19 on the physical and mental health of the UK population; 4. Provide a resource from which to identify potential participants for future clinical trials, and to use data collected as comparison or control data for trial participants who have been randomised to receive one or more interventions.”

Figure 1. Potential Biological Impacts of COVID-19 (image from this paper)

Q&A for 6/10:

#Case Fatality #CFR

Question: You may have already answered this, but I haven’t been able to find it. What have you found to be the current fatality rate? I heard a really high number of 5.9% which I thought was much higher than reality. Could you also explain how it’s calculated?

Answer: I have talked about the case fatality rate (CFR, also called case fatality ratio) many times before, but I don’t believe I’ve ever described how it’s calculated and why it fluctuates. Let me remedy that gap! As you’ll see, that high CFR you heard was not off base (unfortunately). What I hope you’ll take away from this description is that COVID-19 is serious and far more serious than other conditions that it has frequently been compared with (e.g. flu).

CFR is a measure of disease severity. In essence, it is the proportion of people with a given condition who die from that condition. To calculate CFR, the numerator is deaths from the condition and the denominator is all people with the condition (deaths / cases) (Figure 1). Put to practice, Figure 2 shows the current CFR calculation and trends for the United States based on data from covidtracking.com. As of 6/9, the COVID-19 CFR for the United States is 5.4%. Now, if we were to use data from the Johns Hopkins Covid Dashboard, the current CFR in the US would be slightly different — 5.7%. This difference (5.4% vs. 5.7%) is due to slightly different numbers of cases and deaths compiled by these two aggregation services. As you can see in the line chart presented in Figure 2, CFR is not stagnant. And it’s important to recognize that CFR will continue to change based on multiple factors, including:

Data Completeness:

• Testing: As more mild cases are identified through widespread testing, we should expect CFR to decrease.
• Death Reporting and data completeness: For more on this, see Q&A from 5/5, but in essence, as we do a better job at capturing COVID deaths in hospital, long-term care, and home-based settings, we may see an increase in CFR. And for more on issues of data completeness, see Q&A of 5/22.

Health Systems and Medical Interventions:

• Health System Capacity/Readiness: In settings where hospitals are overwhelmed, as we saw in Italy, the CFR increases because health workers are less equipped to provide high quality care for their patients.
• Effective Treatment: As more treatments become available and as health professionals become more adept at caring for COVID patients (discussed in Q&A of 5/31), we should expect to see the CFR decrease.
• Vaccination: As vaccinations become available and used widely, we should expect to see fewer cases and fewer deaths. Among those vaccinated who still become infected (perhaps because the vaccine is not fully effective) we should expect to see fewer severe outcomes, which would reduce CFR.

Population Dynamics:

• Age structure: Because COVID severity varies dramatically by age, in general, the older the population, the higher the CFR. For more on age-specific COVID mortality, see Q&A of 5/23.
• Underlying Conditions: We know that risk of more severe outcomes, including death, is associated with underlying conditions, including hypertension, obesity, diabetes, and more. For more on underlying conditions, see Q&A of 4/14 and Q&A of 6/8. For populations that have more underlying conditions, we’d expect to see higher CFR.

Social/Structural Inequities:

• Racism: In populations that are marginalized, we see higher CFR. For a discussion of racism and health, see Q&A of 6/7 and Q&A of 6/1.
• Poverty: In populations that are more impoverished, we see higher CFR (see this recent research, for example).

Non-medical Interventions:

• Social Distancing, Mask Wearing, etc.: The more widespread social distancing, mask wearing, hand washing, and the like, the lower we’d expect the CFR to be. This is because populations more vulnerable to severe outcomes are better protected from exposure and because health systems are also better protected from becoming overwhelmed.

Note: This list is not exhaustive. I’m sure in writing this morning that I’ve unintentionally omitted some other factors. But as you can see, there are lots of factors that influence CFR. No doubt, CFR is an imperfect measure, but it does give us insight into disease severity, and that’s exactly what it’s meant to measure!

Figure 1. CFR Calculation (from CDC website)

Figure 2. COVID-19 Case Fatality Rate of 6/9

Calculation:

105,981 deaths / 1,970,596 cases = 0.05378

0.05378 x 100 = 5.4% (rounded to nearest tenth)

Q&A for 6/9:

#Asymptomatic Transmission

Question: I was very confused when I read yesterday that WHO is suggesting that asymptomatic transmission is rare. I thought we had evidence that asymptomatic spread was a major concern. Is there new research or guidance on this topic?

Answer: Yesterday during a press conference and in response to a reporter’s question WHO’s Dr. Maria Van Kerkov made the statement that asymptomatic transmission is “very rare.” I was very confused and slightly shocked by this statement and the media reports about it. It flies in the face of what we currently know (okay, that can happen in science), and was said without presenting any data to back it up (not okay, that’s bad public health communication). In short, please disregard what WHO said on this topic yesterday. And now for a more detailed response:

A snapshot of what we currently know: There is a growing body of evidence that asymptomatic spread — or at least presymptomatic spread — is of real concern. Again, asymptomatic means a person is infected but never develops symptoms (or symptoms of any note) while presymptomatic means that they are infected but haven’t yet experienced symptoms. This study published recently in Nature found that viral loads are very high early stages of the infection. Meanwhile, a study still awaiting peer review estimates that the proportion of transmission that was attributed to the presymptomatic period was 48% in Singapore and 62% in Tianjin, China. CDC offers this synthesis of the early evidence here, stating “Recent epidemiologic, virologic, and modeling reports support the possibility of SARS-CoV-2 transmission from persons who are presymptomatic or asymptomatic.”

Challenges in public health communication: In response to the confusion, this morning WHO hosted a live Q&A to address the concerns. And the upshot is, WHO screwed up its communication yesterday. No guidance has changed. Dr. Maria Van Kerkov admitted error in using the words “very rare” and the new data she was basing her statement on are not robust enough to make such a statement or generalization. She clarified that yesterday’s statement was made with regard to people who are truly asymptomatic rather than presymptomatic. And she stated, “Some estimates of around 40 percent of transmission may be due to asymptomatic, but those are from models, and so I didn’t include that in my answer yesterday but wanted to make sure that I covered that here.” For further discussion on this issue, health expert Andy Slavitt offers an interesting take in his twitter thread.

Finally, our knowledge is not static and as we learn more, we adjust and even change our guidelines based on our evolving understanding. There’s still so much we don’t know and it’s quite possible, even expected, that we will have some major shifts in our understanding, including our understanding of asymptomatic cases and transmission. But as of June 9th, we don’t have any evidence that changes our understanding of the lay of the land pre-WHO press conference of June 8th. Please keep wearing your mask and please keep physical distance… please keep doing all the things to help keep us protected!

Q&A for 6/8:

#Diabetes

Question: I have heard recently that people with diabetes are at especially higher risk of mortality with COVID. Do you know of anything that distinguishes between type of diabetes, BMI, age, general health?

Answer: So far, we have no evidence that people with diabetes are more likely to contract SARS-CoV-2. That’s good news. The bad news is that once someone with diabetes contracts COVID, their risks of severe outcomes so far appear much higher as compared with someone without diabetes. Preliminary findings reported by CDC align with findings from China and Italy in showing that people with diabetes are more likely to experience severe outcomes and even death from COVID. The mechanism by which people with diabetes become more susceptible to severe outcomes is still unknown and a recent review of available literature on diabetes and COVID-19 published in the Journal of Clinical Hypertension highlights the many gaps in the current data and our understanding, concluding “the influence of diabetes on severity and outcome in COVID‐19 patients is not clear because of large gap in evidence.” Here, the authors cite the problem of unaccounted confounders — type of antidiabetic therapy used, other diabetes-associated conditions like obesity and age, how well diabetes is controlled, etc. More evidence is needed.

In this vein, the British National Health Service just released estimates of severe outcomes among patients hospitalized with COVID-19. These data, as reported by Diabetes UK, show that “People with type 1 diabetes were found to be 3.5 times more likely to die, and people with type 2 diabetes twice as likely to die, than people who don’t have diabetes when in hospital with coronavirus.” The full NHS paper on diabetes and COVID-19 is here. This paper also found that the presence of other factors — obesity, and age — further increased risk of death among people with diabetes.

Finally, the American Diabetes Association offers an FAQ on diabetes and coronvirus that you may find helpful too.

Q&A for 6/7:

#Race #Racism #Biology #Sociology #Inequality

Question: I saw on a House hearing on COVID on Thursday, a House rep said that there may be a possible “genetic component” for racial disparities in the COVID epidemic. The witness, a black woman M.D., pushed back by responding “There is no genetic basis for race…race is a social construct”. I won’t ask you to take on the second part (unless you’d like to), but I feel like it may be helpful if you could maybe explain to folks what Dr. Uche Blackstock said when she said that “There is no genetic basis for race”. Aren’t the features that make a person their race embedded in their DNA? Can you talk about how that’s a dangerous assumption in a public health setting? [Sidenote: I personally agree with Dr Blackstock, but some people may not have heard this line of argument before…]

Answer: Dr. Blackstock’s words are true. There is no genetic basis for race; race is a social construct. The history of race science is a long, painful, and ugly one. And like so many other issues, its roots are in slavery. Here’s a very brief (too brief) synthesis of race science and answer to your question.

Historians show that the concept of race stems from the Atlantic slave trade, which began in 1441. At that time, Portugal played a leading role in the Atlantic Slave Trade and in response, Portuguese leaders, chroniclers, and the Catholic Church placed Africans in a new category of enslaveable people — a group of “barbaric savages” who deserved “civilizing.” This categorization served the economic interests of the enslavers and assuaged any moral concern they may have had about the horrors they were inflicting on fellow human beings. A well-articulated language of racial inferiority based on skin color was birthed and spread throughout Europe and her colonies. Over the centuries, enslavers looked for ways to confirm their beliefs in their own superiority and the in the inferiority of darker-skinned “others”. This desire for confirmation (confirmation in the economic and social structures the enslavers developed at the expense of the enslaved) birthed race science.

Race science used to just be “science.” Its focus is on identifying biologic differences between the races and using these differences to create hierarchies. NYU provides this important timeline of race science. And FacingHistory.org summarizes, “Prominent scientists from many countries built upon each other’s conclusions…

• Carolus Linnaeus, an eighteenth-century Swedish naturalist, was among the first scientists to sort and categorize human beings. He regarded humanity as a species within the animal kingdom and divided the species into four varieties: European, American, Asiatic, and African.
• Petrus Camper, an eighteenth-century Dutch professor of anatomy, believed that the ancient Greeks had come closer than other people to human perfection. He used Greek statues to establish standards of beauty and ranked human faces by how closely they resembled his ideal.
• Johann Friedrich Blumenbach, a German scientist, coined the term Caucasian in 1795 “to describe the variety of mankind that originated on the southern slopes of Mount Caucasus” along Europe’s eastern border. He claimed it was the “original” race and therefore the most “beautiful.”
• Samuel George Morton, an American anthropologist, theorized in the mid-1800s that intelligence is linked to brain size. After measuring a vast number of skulls from around the world, he concluded that whites have larger skulls than other races and are therefore “superior.””

Dr. Samuel Morton was one of the most prominent race scientists, especially influential in the United States. His work became widely known and cited as evidence in support of ongoing repression and debasement of black and brown people. When Dr. Morton died in 1851, the South Carolina-based Charleston Medical Journal praised him for “giving to the negro his true position as an inferior race.” And Morton’s legacy, like the legacy of slavery, remains with us today. The horrific Tuskegee Syphillis experiments, built on racist beliefs about black bodies, ended as recently as 1972! North Carolina repealed its compulsory sterilization law in 2003! As Faulker wrote, “The past is never dead. It’s not even past.”

The proof? A member of the U.S. Congress identified biology as one of the factors driving the higher rates of COVID-19 mortality among Black Americans. On Thursday! And that Congressional leader is not alone. Recent research shows that the majority of white medical residents believe that black people have higher pain tolerance and that black skin is thicker than white skin. These beliefs then influence the medical treatment people of color receive. These pervasive beliefs stem from our racist history and present.

Now, someone reading this post might ask — what about some differences that are associated with race, like sickle cell trait? Isn’t that a racial genetic difference? Answer: No. It’s a genetic difference based not on skin color, but on geography — where humans have lived in places with high degrees of malaria, they have adapted this trait to help mitigate malaria’s impact. Angela Saini, science journalist and author of Superior: The Return of Race Science gives a nice synthesis of this race/geography conflation in this NPR interview. She also provides a deep dive into race science in her book, showing that current-day science still has hallmarks of race science. As Ramin Skibba writes for Smithsonian magazine, “Saini cites an example of a 2017 study claiming that race and biology indicate that the airways of asthmatic black Americans become more inflamed than those of asthmatic white Americans. Black Americans do suffer more from asthma than whites do, but they’re also affected more by environmental hazards like air pollution from highways and factories as well as disparities in access to high-quality healthcare. These many forms of inequality and structural racism — which sociologists have documented for decades — were swept under the rug in favor of a race variable that led to findings that could be easily misinterpreted.”

How does this all factor into our COVID response? I want to be clear, race is indeed exceptionally important to examine when it comes to health outcomes and disparities. As Dr. Jennifer Tsai wrote in Scientific American several years ago (pre-COVID), “The existence of racial disparities in pain management is an issue of racial difference. Black patients really are getting less pain medication, and yes, because of their race. But this has nothing to do with genetic susceptibility. Such racial logic fuels stereotypes that feed inequity….Race is enhanced as a descriptor when it is mobilized as a marker of potential risks drawn from external inequities and assumptions, rather than as a risk factor that is innately responsible for poorer health outcomes.” For COVID, stark differences in mortality by race are markers of stark inequalities by race.

So much more — books and books — to be read and written on this topic. In summary, there are no biological differences by race. Race is a made up concept used by those individuals, institutions, and systems that have profited from the economic and social systems born from slavery in order to justify their actions and themselves.

Q&A for 6/6:

#Sex #Risk #Harm Reduction

Question: Is there any information about transmission risk and sexual behavior?

Answer: Thanks for asking! We don’t have direct evidence about the risk of virus transmission and sexual activity. Here’s what we do know and what doctors and public health experts suggest.

In terms of what we know. First, because SARS-CoV-2 is spread through respiratory droplets and may spread from contaminated surfaces, any in-person activity carries risk, including and especially sex. Second, it’s still unclear whether the virus can be transmitted through semen, vaginal fluids, fecal matter, or urine. Two small studies of infected people found no evidence of the virus in semen or vaginal secretions. Another small study of 38 infected men detected the virus in 6 patients (16%). SARS-CoV-2 has also been detected in urine and anal swabs as well as fecal matter. Now, virus detection does not equate to transmission risk, but until it is proven otherwise, it would be prudent to consider exposure to semen, urine, and feces as potential avenues for transmission.

When it comes to advice from doctors and public health experts, there is widespread agreement that abstinence-only approach is not sustainable for most folks. A recently published comment in the Annals of Internal Medicine, “Sexual Health in the SARS-CoV-2 Era”. offers a lot of advice to clinicians counseling their patients. I’ve copied their table below, which synthesizes current recommendations for safer sex (Table 1). The DC Health Department also offers a bulleted list of suggestions (Table 2) and the NYC Department of Health offers a fact sheet.

Table 1. Sexual Approach and Recommendations (ordered by increasing risk)

Table 2. Sex During the COVID-19 Public Health Emergency (DC DoH)

Q&A for 6/5:

#Hydroxychloroquine

Question: I see hydroxychloroquine in the news a lot these days. Would you please share a synthesis?

Answer: Sure! So, first reason for all the news is that there was a big observational study about hydroxychloroquine’s role in COVID-19 patient treatment recently published in The Lancet and met with lots of press. Scientists quickly raised alarms about the data used in the analysis — a data set that was heretofore unknown, owned by a private company, and with far too many inconsistencies and implausible data to be true. Based on this alarm and at the request of the study authors who stated,”we can no longer vouch for the veracity of the primary data sources,” the study was retracted yesterday. In the wake of this storm, the New England Medical Journal on Wednesday published results of a randomized control trial by Boulware et al. examining whether hydroxycholorquine can serve as effective post-exposure prophylaxis (e.g. prevent COVID-19 infection among those exposed). Researchers found that hydroxycholoquine did not protect exposed individuals from getting COVID-19 (detailed in the below paragraph). That said, the study had many limitations, as discussed by the accompanying NEMJ editorial, which concludes that “The results reported by Boulware et al. are more provocative than definitive, suggesting that the potential prevention benefits of hydroxychloroquine remain to be determined.” With 212 studies of COVID-19 and hydroxycholoquine listed on clinicaltrials.gov (as of today), we should soon have more data for decision-making!

“The incidence of new illness compatible with Covid-19 did not differ significantly between participants receiving hydroxychloroquine (49 of 414 [11.8%]) and those receiving placebo (58 of 407 [14.3%]); the absolute difference was −2.4 percentage points (95% confidence interval, −7.0 to 2.2; P=0.35). Side effects were more common with hydroxychloroquine than with placebo (40.1% vs. 16.8%), but no serious adverse reactions were reported.”

Q&A for 6/4:

#Protesting #BLM

Question: I really want to join the protests, but I’m also really afraid of contracting the coronavirus and spreading it to others. I’ll admit, I’m also afraid of the police. Is there any way to safely protest?

Answer: Thank you for asking this question. I have the same desire and fears. And I am feeling so much admiration and gratitude for the heroes who are out there protesting. If you do choose to go protest, here are some recommendations based on my own understanding of virus transmission and also advice several public health experts have shared in the last few days, including in yesterday’s Washington Post article.

Protesting during the Pandemic:

• Masks: Wear your face mask to limit the spread to others in case you are asymptomatic.
• Hand sanitizer: Bring hand sanitizer and use it.
• Eye protection: Bring shatterproof googles and wear them too. If they get foggy, you can use your own spit to clean them. As discussed in Q&A of 4/4, goggles may offer protection from COVID. They will definitely help protect your eyes from potential tear gas exposure.
• Distance: Keep your distance from others as best as possible. You know the 6+ feet rule, but in the context of a protest this guidance seems extra challenging. At the very least, stay with your group and minimize contacts with others.
• Location: Stay outside and keep moving. Try and stay upwind.
• Yelling: Try to avoid yelling as it can spread respiratory droplets further. Bring drums, noisemakers, and carry signs instead.
• Feeling unwell?: Please don’t go.

Other Protesting Suggestions (non-COVID-related):

• Know your rights. ACLU shares a great list its website. For ease of reference, I’ve copied below ACLU’s list.
• Bring essentials:
• Backpack
• Water and snacks — Keep hydrated and use bottles with squirt tops are preferred so that you can use them to rinse your eyes out and skin off if exposed to tear gas. Note: water and other saline eye rinse solutions are recommended treatment (use water, not milk!).
• Cash
• Identification + emergency contact information
• Your phone and charger
• Medications
• Extra change of clothes in a plastic bag
• Hat for sun protection
• Avoid wearing make-up, sunscreen, or other oil-based lotions that could trap tear gas.
• Amnesty International offers a synthesis of protest recommendations.
• Showing Up for Racial Justice offers “Guidelines for White People Showing Up at Black-Led and People-of-Color-Led Protests”

And if you decide not to protest in the streets, you can still be engaged! For example, 26 Ways to Be in the Struggle beyond the Streets offers 26 suggestions.

ACLU Know Your Protest Rights

Q&A for 6/3:

#Super-spreading #Missouri

Question: I saw that there was one person at that lake of the Ozarks party that was clearly sick while he was there and began to show symptoms the day after. Do we know if anyone has fallen ill as a result of being exposed during that event? I’m super curious if the outside air/heat/pool made the virus harder to catch. And if there is no news on that event, do we have an reports of people at beaches falling ill 3–5 days afterwards and therefore were potentially exposed while at the beach?

Answer: I read that headline too, but I didn’t read further. Your question prompted me to do so, thanks! For those of you who didn’t read the story, the upshot is that a party-goer from Boone County, Missouri hung out in multiple, crowded bars on May 23–24 in Lake of the Ozarks (Camden County), MO. As the Camden County Department of Health reported on May 29th, the person “arrived here on Saturday [5/22] and developed illness on Sunday, so was likely incubating illness and possibly infectious at the time of the visit.”

In terms of transmission, it’s hard to answer your question because so many of the party-goers were from so many different places and there’s no solid contact tracing effort happening (to my knowledge). The local paper, KY3 just reported results from a cell phone tracking company, which uses anonymized cell phone location data to see where people disperse to from large gatherings, including Lake of the Ozarks during Memorial Day weekend. KY3 reports that “The data shows after the weekend was over, those blue dots, or cell phones, spread out all across the midwest. St. Louis, Kansas City, and Omaha, Nebraska saw a lot of activity…about 14 states that got some version of traffic.” All that said, to really know whether folks became infected from the party-goer, we’d need solid contact tracing data, which we don’t have.

Still, we could perhaps glean something from the data we do have, so let’s take a look. When it comes to Missouri as a whole, cases are on the rise. Yesterday, Missouri had the highest number of daily cases reported since the beginning of May (Figure 1). When it comes to Boone County, where the party-goer was from, Missouri is reporting a 25% increase(!) in cases over the last 7 days (increase from a total of 123 cases to 154 cases). When it comes to Camden County, where Lake of the Ozarks is, Missouri is reporting only 37 total cases, with only 1 new case in the last 14 days. Other places that party-goers seemed to have dispersed to — Omaha (rising cases), St. Louis (falling cases), and Kansas City (rising cases) — are seeing varying trends in the average number of reported cases over the past two weeks according to the NY Times county-level data aggregator. But trying to see how the virus may have spread by looking at these county-level data is really not that informative, especially since the places listed have quite large populations. Plus, Omaha, St. Louis and Kansas City were already hotspots. It is heartening to know that Camden County has not seen an increase in cases, while disheartening to see the massive increase in Boone Country.

When it comes to Spring Breakers, I read a number of news reports of students testing positive after partying at the beach during Spring Break, but I haven’t seen anything that showed the events were super-spreader events. So far, super-spreader events seem to happen in populous, indoor places — nightclubs in South Korea, conference in Boston, large dinner party in Connecticut, church events in Arkansas, choir practice in Washington, restaurant in China. One epidemiologist also noted that these crowded places also happen to be loud and speculated about the relationship between talking loudly/shouting/singing and increased spread. As I’ve said the old aphorism before, absence of evidence is not evidence of absence. There’s still so much we don’t know about the virus.

Figure 1. Daily Cases in Missouri (data viz from NY Times)

Q&A for 6/2:

#Tennessee #TN

Question: How is Tennessee looking?

Answer: Tennessee (my home state) has been lucky so far in having avoided any major spike in cases and deaths. But the trends of the last few days are not as promising as the trends of earlier May, particularly with regard to daily cases and hospitalizations (Figure 1).

• Daily deaths from COVID-19 were the highest in early April (58 deaths the week of April 5th) and have since remained relatively flat at around 35 deaths/week except the week of May 10th, which saw 53 deaths.
• Daily hospitalizations peaked in early May and fell through May 24th. Since then, however, hospitalizations have been on the rise. Currently, Tennessee is seeing an average of 25 people/day hospitalized.
• Daily cases have also been on the rise since mid-May with a current average of 421 new cases identified each day.
• The rise in cases is occurring alongside a decrease in daily tests. The percentage of tests that are positive currently averages 5.6% and has been increasing since a low of 3.2% on May 10th (data not shown).

Figure 1. Daily Deaths, Hospitalizations, Cases, and Tests in Tennessee

Q&A for 6/1:

#Racism #Disparities

Question: Let’s talk about racism and coronavirus.

Answer: Racism is a public health crisis. As Johns Hopkins Bloomberg School of Public Health’s Dean MacKenzie and Professor Cooper wrote last night,

“…Law enforcement violence is a public health issue. It is just one dimension of racism as a present and deadly force in our society. As shocking as these high profile examples are, they represent the tip of the iceberg of persistent racial inequities that constitute a crisis for public health. African American babies die before their first birthday at more than twice the rate of white newborns. African American women die at more than twice the rate of other women during pregnancy and childbirth. African American adults suffer far higher rates of hypertension, diabetes, and other serious chronic illnesses. The life expectancy of African Americans is 3.5 years shorter than for white Americans. The roots of these and other mortal disparities run deep to the structural and institutional racism that shapes policing, housing, transportation, education, and health. The COVID-19 pandemic has reexposed the consequences of this legacy. With less secure housing, less stable access to food, greater reliance on crowded public transit, more low-wage work without adequate protection, and less access to health care, many predominantly African American communities are suffering staggering losses. African Americans are nearly twice as likely to die from COVID-19 compared to others in the U.S. population.”

COVID-19 highlights and amplifies existing disparities resulting from ongoing systematic and institutional racism. Here are some data that show the stark racial divides in this country. Figure 1 is from the COVID Racial Data Tracker. It shows that Black people are dying at a much higher rate (nearly 2x) than would be expected. This finding is mirrored in recent research, which looked at county-level disparities in COVID cases and deaths, finding “significantly higher rates of COVID-19 diagnoses and deaths in disproportionately black counties compared to other counties.” Figure 2 was shared by CDC on Friday and presents hospitalization rates by race/ethnicity. As you can see, the age-adjusted hospitalization rates for American Indian, Alaska Native, and Non-Hispanic Black Americans are 4.5x higher than the rate of Non-Hispanic White Americans. The hospitalization rate of Hispanic Americans is 3.5x that of Non-Hispanic White Americans.

As scholar, writer, and Director of the Antiracist Research and Policy Center at American University, Ibram Kendi, wrote in The Atlantic last month,”…To explain the disparities in the mortality rate, too many politicians and commentators are noting that black people have more underlying medical conditions but, crucially, they’re not explaining why. Or they blame the choices made by black people, or poverty, or obesity — but not racism…. Black people are not to blame for racial disparities. Racism is to blame.” The solution? Actively being anti-racist and fighting for anti-racist policies at institutional and structural levels.

Figure 1. Black Americans are Dying at a Rate 2x higher than their Population Share.

Figure 2. COVID-19 hospitalization rates are 4.5x higher among American Indian, Alaska Native, and Non-Hispanic Black Americans as compared with Non-Hispanic White Americans.

Q&A for 5/31:

#Outcomes #Waiting

Question: I was just thinking that if I were to contract the coronavirus, I’d rather contract it later rather than sooner. I’ve heard more people say that they’d just rather get it over with, but I was thinking that it’d be better to wait because we’d have more science and doctors/hospitals would have more experience treating COVID patients. The science and experience would offer a better chance of having a better outcome. What do you think?

Answer: I agree. We know more and more each day and our knowledge and experience will improve patient treatment and outcomes. Please continue to try to avoid contracting the virus; the longer we can avoid SARS-CoV-2, the better! Here’s one interesting nugget of data from a prospective cohort study of 5,279 patients a a large academic medical center in New York City and Long Island (NY Langone) recently published in The BMJ that supports this position — while the risk of hospital admission remained constant during the study period (March 1st through April 8th), the risk of critical illness decreased. The authors state, “Our institution was stretched but not overwhelmed by the epidemic and did not experience important equipment or treatment shortages. The improvement in outcomes over time (in the setting of a functioning health system) raises the possibility that familiarity with the disease, ongoing iteration of protocols and practices in response to observed outcomes, and initiation of new treatments might improve outcomes even in the absence of vaccination or regimens known to be effective.”

The BMJ editors elaborated on this finding in an accompanying editorial, stating “Perhaps the most intriguing finding from the New York cohort was that risk of critical illness declined progressively over the study period, with a suggestion of declining mortality as well, without changes in risk of hospital admission. Several potential explanations worthy of future investigation include the influence of strain in hospital capacity on quality of care, allocation of resources, and disposition decisions in the emergency department; changes in care delivery over time, such as proning in awake, non-intubated patients to avoid intubation or better adherence to lung protective mechanical ventilation strategies; and changes in targeted therapy that might be beneficial (remdesivir and anticoagulation) or harmful (hydroxychloroquine).”

Q&A for 5/30:

#Swimming #Oceans #Pools

Question: Can I get the virus from swimming in a pool or the ocean?

Answer: The journal, Water Resources, recently published a research paper that addresses this issue, “Coronavirus in water environments: Occurrence, persistence and concentration methods.” I’ve copied their overarching findings in the below paragraph, but in short, coronaviruses can’t survive in treated water — swimming pools are not reservoirs of virus and you cannot get the virus from swimming in a pool, hot tub or the like. When it comes to other bodies of water that aren’t treated — lakes and oceans for example — we have less solid data. Even so, there seems to be general consensus among experts that the dilution effect in lakes and oceans would be so much that it would be basically impossible to get infected this way. So far, there are no known cases of transmission from lakes, oceans, and other bodies of freshwater. Based on the data we do have, it seems that water temperature dictates how infectious the virus remains with infectivity declining more rapidly in warmer water (temps 73°f and higher). Upshot is, the water part of swimming does not carry virus transmission risk. Rather, it’s the people part — if you are at a beach or in a pool with others, then the transmission risk comes from respiratory spread. If you do choose to go swimming, avoid being among crowds of people, keep 6+ feet of distance, and as possible, stay upwind!

“The data available suggest that: i) CoV seems to have a low stability in the environment and is very sensitive to oxidants, like chlorine; ii) CoV appears to be inactivated significantly faster in water than non-enveloped human enteric viruses with known waterborne transmission; iii) temperature is an important factor influencing viral survival (the titer of infectious virus declines more rapidly at 23°C-25 °C than at 4 °C); iv) there is no current evidence that human coronaviruses are present in surface or ground waters or are transmitted through contaminated drinking-water; v) further research is needed to adapt to enveloped viruses the methods commonly used for sampling and concentration of enteric, non enveloped viruses from water environments.”

Q&A for 5/29:

#Risk #Family #Visits

Question: Okay so I really *really* want to go home and see my family. If my partner and I stay in an AirBnB and visit in my parents backyard, 6 ft apart, is that the safest way to do this? We would drive their in our own car. Like, if I’m wearing a mask and gloves, am I allowed to hug my Mom? I feel like there’s no information on how to do any of this safely and it’s impossible to find helpful guidance.

Answer: Same! We are in the risk mitigation phase of the response and knowledge and information on this front is difficult to sift through. The first question I would ask — does your family want to see you? Everyone is operating with different concerns and risks; some families may not want visitors. Figure that question out at the get-go. Great! Now, let’s say you’ve had that conversation and your family does want to have you come home for a visit. What’s the best way to do so to minimize risk?

What you’ve described in your email is probably the best set of actions to minimize risk== drive to your destination (see Q&A of 5/16 for more advice on that piece), keep all family interactions outdoors and at a distance (see Q&A of 5/21 and 5/12 for more on indoor/outdoor risk), and as always use your mask and wash your hands. If you can, you might also want to consider getting tested before you travel just to have piece of mind that you’re not asymptomatic. But not hugging your mom sounds bleak. So, I’d suggest that you ask another set of questions to think about tiered risk — what are everyone’s underlying risk factors and how protected thus far has everyone been from exposure? The more risk factors, the more precautions I’d take.

When it comes to underlying risk, I’m talking age, weight, health issues like cardiovascular disease, autoimmune disorders, etc. When it comes to exposure, I’m talking rates of community-level transmission and types of activities individuals have participated in (ex: hike in the woods== lower risk; singing with choir== higher risk). If you or your family are in higher risk categories, then I would indeed keep interactions distanced and in outdoor spaces. If you and your family are in lower risk categories, then I would have conversations with them about what, if any, additional risk everyone is willing to accept. Hugs? Staying in the same house together? Eating at the dining room table? Maybe your mom, dad, partner, and you decide to stay together as a unit. Now you need to keep transparent conversations about activities that you are and are not comfortable with. Does your dad want to resume his spin class at the gym? Other risky activities? You’ve decided to operate as a cluster, so you have to protect each other and minimize risky activities. Set expectations. Do your best as individuals and as a unit to avoid situations that expose you to many people. The Atlantic published a nice synthesis article touching on some of these points a couple of days ago if you want to read more.

Finally, I think you’ll see a lot of guidance about avoiding seeing grandparents. This will definitely minimize risk. In my opinion, the guidance also becomes increasingly untenable as time goes by. My grandparents are in their 90s. We don’t have much time left together in the scheme of things. They have decided that they would rather take the risk in seeing us than forgo the joys of being together. So even though they are in high risk category of individual, we’re going for a visit. Yes, it’s a risk. But it’s one that we’re all willing to take and we’ll do our very best to minimize the threat.

Q&A for 5/28:

#Asymptomatic

Question: Do we know any more about characteristics and health outcomes of people remain asymptomatic? Do we know why some people remain asymptomatic?

Answer: Mostly, these questions are still outstanding, and scientists are trying to answer them. When it comes to your first question, there was an interesting piece of research published on this topic yesterday in JAMA, “Comparison of Clinical Characteristics of Patients with Asymptomatic vs Symptomatic Coronavirus Disease 2019 in Wuhan, China.” I’ll give you a snapshot of its findings.

In short, 78 individuals from 26 cluster cases were enrolled in the study based on the following criteria: 1) they had been exposed to SARS-CoV-2 at the Hunan seafood market or had close contact with a patient who had been hospitalized with confirmed COVID-19; 2) they were confirmed to have SARS-CoV-2 infection. Those 78 individuals were all hospitalized in same place, and provided the same care by the same health workers. 33 patients (42%) were asymptomatic, while 45 patients (58%) were symptomatic. Symptoms were monitored daily and patients were tested for the virus every 24–48 hours. Patients also received chest scans. I’ve copied the article’s Table below, which presents findings. What you see is that compared with symptomatic patients, asymptomatic patients were statistically significantly (p<0.05) more likely to be young and female; less likely to have pre-existing liver injury; more likely to have quicker recovery times and a shorter viral shedding period; less likely to have fluctuating test results; and less likely to have damage done to their immune systems.

Table. Differences between Symptomatic and Asymptomatic Patients

Q&A for 5/27:

#Nursing Homes #Mortality

Question: Follow-up question: How much “additional death” in nursing homes does this represent? How many years does one usually live in a nursing home and what is the “normal” rate of death for nursing homes? What percentage of the nursing home population dies on average every year?

Answer: Interesting question! In December, the New England Medical Journal published this short analysis, “Changes in Place of Death in the United States”. The authors find that in 2017, 21% of all deaths occurred in nursing homes (534,714 deaths in nursing homes in 2017). For more detail, see Figure 1, which I copied from the paper. Data for 2020 are limited when it comes to nursing home fatalities, both overall fatalities and those due to COVID-19. Based on what we do know, a much larger proportion of COVID-19 deaths is occurring in nursing homes (35% of all deaths) as compared with what we could consider to be the baseline (21% of all deaths). Finally, when it comes to overall mortality in nursing homes, a quick pubmed search provided a couple of articles that followed two different longitudinal cohorts of nursing home residents, one in Norway and the other in Iceland, both of which showed median survival of approximately 2.5 years (27 months and 31 months respectively). Of course, median survival depends dramatically on underlying health, age of admission, availability of home care services, and more. I’m not sure how well these data translate into the US context. In that vein, I did find some tables on nursing home residents for 2016 in the United States thanks to CDC (Figure 2). IF we were to assume that the data shared in the NEMJ article on nursing home deaths and the data shared from CDC on nursing home residents were comparable, it would seem that about 30% of nursing home and residential care residents die each year (534,714 deaths / 1,924,300 nursing and residential care residents). I’m not sure about this back of the envelope, however, as I’m not sure how residential care (e.g. assisted living) factors into the death estimates and I don’t like comparing data from two different sources.

Figure 1. Changes in Places of Death in the US

Figure 2. Nursing Home Residents in 2016

Q&A for 5/26:

#Nursing Homes

Question: I had read that Florida had done a much better job than New York in protecting nursing home residents. Is that true?

Answer: As of May 22nd, Florida had recorded 1,032 deaths of nursing home and assisted living residents and workers. As of May 24th, New York had recorded 3,282 confirmed deaths and 2,698 presumed deaths for a total of 5,980 deaths among nursing home/assisted living residents and workers. Florida does not include presumed deaths in its totals, so it is perhaps more appropriate to compared confirmed deaths in both states. Either way it’s abundantly clear that New York has experienced far more deaths in nursing and assisted living facilities than Florida. But let’s look a little further.

First, let’s add a denominator to both numbers — number of nursing home residents in each state. According to the Kiaser Family Foundation, Florida has 72,741 nursing facility residents while New York has 101,518 residents. They are respectively the 6th and 1st largest nursing facility resident populations in the country. This would put the COVID-19 mortality rate among nursing facility residents at 1.4% in Florida and 3.2% in New York (or 5.9% if you include presumed deaths). Again, New York is fairing far worse. That said, it’s important to add one more layer of context — how much of a toll the virus has taken on each state. As of May 24th, New York had experienced 23,391 deaths while Florida had experienced 2,316 deaths (data from covidtracking.com). Per the US Census Bureau, New York has a population of 19.45 million and Florida has a population of 21.48 million. So, New York has seen 120 deaths per 100,000 residents while Florida has experienced 11 deaths per 100,000 residents. New York has been hit 10x harder than Florida. But New York’s nursing homes have been hit only 2 to 4 times harder than Florida’s (depending on which NY data you use). Either way you slice it, it’s terrible. But once we add the broader context, I’m not sure that Florida is a shining example of how to get it right. I’m also confident that there’s plenty to criticize when it comes to New York’s response.

Moving forward, there is much to be done to prevent COVID-19 in nursing homes and assisted living facilities. JAMA included a nice commentary on this subject just a few days ago and the American Geriatrics Society issued a brief to guide policy last month. The recommendations are quite straightforward and include having comprehensive testing policies and supplies, making PPE widely available, providing paid sick leave for workers, and more.

Q&A for 5/25:

#Maryland

Question: You showed some Maryland data a couple of weeks ago. Any positive updates?

Answer: We talked about Maryland in our Q&A of 5/11. With another two weeks under our belts, here are some updated charts (Figures 1 and 2) I made using data from covidtracking.com. The good news is that deaths have continued a downward trend and testing capacity has increased. The bad news is that the percent positive tests is still very high, the number of cases is increasing, and the number of hospitalizations has plateaued for a number of weeks.

As you’ll see in the green bar chart, Maryland has been slowly but surely increasing its testing capacity. Meanwhile, the proportion of tests that are positive is still high at 17% (rolling 7-day average), meaning that testing needs to be further expanded. The number of daily cases in Maryland showed a very small dip starting around May 7th, but then showed an increasing trend around May 19th (yellow bar chart). Since then, we have averaged about 1,000 new cases/day. Since flattening the hospitalization curve in mid-April (orange bar chart), hospitalizations had plateaued for about a month, hovering around 1,000 patients hospitalized per week (beginning week of April 12). The week of May 10, hospitalizations declined to 823 patients, but in the last week, they’ve popped back up to 1,070 patients. Meanwhile, deaths have been decreasing since the week of April 19th (red bar chart).

Figure 1. Maryland Deaths, Hospitalizations, Cases, and Tests

Figure 2. Weekly Deaths, Hospitalizations, Cases, and Tests in Maryland

Q&A for 5/24:

#Travel #Self-isolation #Self-quarantine

Question: A friend recently traveled by plane to California to spend several weeks with a family member who had been in and out of the hospital for care unrelated to COVID-19. Within days of returning, she invited me to a “socially distanced happy hour” in her yard. I declined. Should she be self-quarantining?

Answer: We’re all navigating this new terrain, and I’m trying not to be judgmental of the decisions other folks are making — except when I decide to be really judgmental (wear your mask at the grocery store!). CDC recommends that all international travelers self isolate at home for 14 days upon return. And a few states, notably Hawaii, require everyone arriving or returning to the state to self-isolate for 14 days. As to our area, there are no such restrictions or recommendations for arriving/returning domestic travelers. As to your neighbor, it would seem that she’s had some elevated risks — taking care of someone who has been in and out of the hospital, traveling/flying, and likely being in two hotspots (CA + DMV). Because of her elevated risk, it would seem prudent for her to be more restricted in her interactions with others. At the very least, your neighbor friend should disclose her recent activities to those she’s inviting for socially distanced socializing so they can make their own risk/reward calculations.

Q&A for 5/23:

#Deaths #Age-specific Mortality # Flu #Children

Question: Another follow-up: Has anyone compared the death or incidence rates for COVID and the flu in the under-60 population? I presume flu is more deadly for children.

Answer: You are right. And thank you for the follow-up question. Flu is more deadly in children than COVID-19. I pulled some data from CDC’s provisional death counts between Feb. 1 and May 16, 2020 to highlight these differences (Figure 1). Among those <15 years of age, flu has indeed caused more deaths. Meanwhile, for those ages 15+, COVID-19 has caused more death. This article published in Bloomberg earlier this month does a nice job simply estimating and describing differences in age-specific mortality between COVID-19 and influenza. And Our World in Data nicely summarizes what we know about risk of death from COVID.

Figure 1. COVID-19 Deaths and Influenza Deaths by Age

Q&A for 5/22:

#Spread #Severity #Monitoring

Question: Follow-up question: But if the only people being tested are those that show up at testing facilities with symptoms, and legit estimates are that the actual infection rates are 4–10 times higher with many people asymptomatic or just refusing testing, then how can we believe our current numbers are a useful gauge for the spread and severity of this illness? Even without antibody tests, our denominators are still wrong.

Answer: Our numerators and denominators are both incorrect — we’re likely missing deaths and missing cases. I totally agree that we need widespread testing — testing that includes folks who are concerned that they may have been exposed, of folks who have only mild symptoms, of folks who are sick but do not require hospitalization, etc. Without such widespread testing, we won’t be able to accomplish our test, trace, isolate public health goals to constrain the spread of the virus. When it comes to understanding the spread and severity of the illness, while the data we have now are incomplete, they are nonetheless informative. For monitoring disease spread, we absolutely need widespread testing and contact tracing (Health Force!), but in their absence, we can make informed judgments based on a) hospitalization rates (daily admissions and overall bed use); b) percent of tests that are positive; c) number of daily cases. For more on these metrics, see Q&A of 4/16. When it comes to severity, what we’ve seen around the world — from Wuhan to New York to the US writ large — is that among cases that we do know about, approximately 20% are serious and require hospitalization. We also know that COVID is far more serious than the flu. I bring this up because early on in the spread of the virus, many people likened it to the flu to assuage concerns (this happened in 1918 too, when people likened the flu to “the old-fashioned grip”)! For more on the severity of COVID compared with the seasonal flu, this viewpoint published in JAMA Internal Medicine last week is a fantastic read that I highly recommend. Here’s one excerpt:

“The demand on hospital resources during the COVID-19 crisis has not occurred before in the US, even during the worst of influenza seasons. Yet public officials continue to draw comparisons between seasonal influenza and SARS-CoV-2 mortality, often in an attempt to minimize the effects of the unfolding pandemic. The root of such incorrect comparisons may be a knowledge gap regarding how seasonal influenza and COVID-19 data are publicly reported. The CDC, like many similar disease control agencies around the world, presents seasonal influenza morbidity and mortality not as raw counts but as calculated estimates based on submitted International Classification of Diseases codes.2 Between 2013–2014 and 2018–2019, the reported yearly estimated influenza deaths ranged from 23 000 to 61 000.3 Over that same time period, however, the number of counted influenza deaths was between 3448 and 15 620 yearly.4 On average, the CDC estimates of deaths attributed to influenza were nearly 6 times greater than its reported counted numbers. Conversely, COVID-19 fatalities are at present being counted and reported directly, not estimated. As a result, the more valid comparison would be to compare weekly counts of COVID-19 deaths to weekly counts of seasonal influenza deaths.

During the week ending April 21, 2020, 15 455 COVID-19 counted deaths were reported in the US.5 The reported number of counted deaths from the previous week, ending April 14, was 14 478. By contrast, according to the CDC, counted deaths during the peak week of the influenza seasons from 2013–2014 to 2019–2020 ranged from 351 (2015–2016, week 11 of 2016) to 1626 (2017–2018, week 3 of 2018).6 The mean number of counted deaths during the peak week of influenza seasons from 2013–2020 was 752.4 (95% CI, 558.8–946.1).7 These statistics on counted deaths suggest that the number of COVID-19 deaths for the week ending April 21 was 9.5-fold to 44.1-fold greater than the peak week of counted influenza deaths during the past 7 influenza seasons in the US, with a 20.5-fold mean increase (95% CI, 16.3–27.7).”

Q&A for 5/21:

#Monitoring #Antibody Tests #Virginia

Question: Did you see that VA was being accused of “cooking the books” because they were including antibody test results in their denominators? The real question is why isn’t everyone else? We won’t have an idea of the average severity, the R0, or the death rates if we aren’t trying to find and include as many ill in the denominator as we can.

Answer: Dang, I did not see that. And when I went to the Veteran’s Administration COVID-19 website, I didn’t find evidence that they are including antibody test results in their denominators. But then I realized that you might be referring to Virginia… ah ha! There are indeed lots of stories on this issue, including this prominent article published in The Atlantic. How did I miss them!? Anyway, according the the latest — a CNN report published yesterday — Virginia health officials say they are no longer combining COVID-19 PCR diagnostic test results with antibody results. In my opinion, this change is a good thing.

To elaborate, I agree that we do want and need to be conducting antibody tests to understand how far the virus has spread and have a deeper understanding of its severity. I’d love to see a two-stage cluster design household survey to really get reliable population-level estimates. In this space, CDC did just announce a study that will test samples from blood donors in 25 cities for antibodies to understand population-level prevalence. When it comes to state-level reporting of cases, however, I think it’s best to avoid lumping antibody tests with PCR tests for several reasons:

1. Monitoring Incidence: New cases cannot be identified by antibody tests. We need widespread testing that identifies new cases (incidence) so we can track how the virus is spreading in real time.
2. Making Re-Opening Decisions: The National Guidelines (described in Q&A of 4/19) that Governors are using to varying degrees to determine re-opening include a metric — proportion of tests that are positive — that can be influenced by the inclusion of antibody tests. The Richmond Times reported last week, “With a higher number of tests in the state’s log due to antibody tests, Virginia’s positive rate was skewed down by a percentage point, from 15% to 14%.” The test positivity rate should be based on diagnostic tests; other data should not be added because they may obscure our understanding of the situation, especially when these data are being used to inform re-opening decisions.
3. Monitoring Testing Capacity: Lumping the two types of tests may present a rosier picture of a state’s testing capacity and infrastructure than is the truth. As we focus on a test, trace, isolate approach, it is key to understanding how testing capacity is changing over time, including how many diagnostic tests per capita are being conducted. We need to compare apples with apples.
4. Minimizing False Positives: Until the population-level prevalence is much higher, antibody tests will yield far more false positives, which will artificially inflate the number of cases (see our Q&A of 4/15 for a deep dive into the sensitivity/specificity challenge).

Q&A for 5/20:

#Restaurants #Outdoor Risk #Viral Load

Question: A friend and I had a lengthy conversation this morning about safety in restaurants from air circulation: Even if restaurants employ good social distance practices, aren’t we at-risk from air conditioning/air circulation within the restaurant? And, isn’t this analogous to the days when we had “no smoking” sections in restaurants but all were exposed to smoke anyway? Finally, I wonder about outdoor dining and exposure to infected patrons via air currents from wind?

Answer: So many of us are trying to figure out our “new normal.” When it comes to risk of infection, put simply it’s about how much of the virus you’re exposed to (viral load) and how long your exposure is. Most outdoor activities will be safer than indoor activities because the sun, wind, and sheer openness will cause the virus to dilute far more quickly, which will thereby reduce both amount of virus you could be exposed to and the amount of time you could be exposed. The New York Times has had a couple of articles in the last few days that have provided a nice overview of risks and considerations when it comes to outdoor activities and restaurant eating (both indoor and al fresco). And for a bit more on the research on indoor/outdoor risks, see our Q&A of 5/12. I would write more, but those Times articles really do a fantastic job! Now, when it comes to amount of virus and amount of time thresholds for transmission, we still don’t know for sure (that’s our common refrain!). In the absence of direct evidence, a team researchers recently wrote a paper in The BMJ on the role of viral load in home-based transmission dynamics, which gives a nice overview of the state of evidence. The researchers state, “Although the infecting dose from a combination of droplets and environmental contamination cannot be easily measured, high quality experiments under controlled conditions in animal models can provide indirect evidence. We are not aware of infecting dose experiments with animal models of covid-19, but animal models of other viral infections show that variation in the infecting dose determines how many animals get infected and how severe the illness is.” And when it comes to length of exposure, CDC says that while data are insufficient to precisely define “prolonged exposure,” “15 min of close exposure can be used as an operational definition.”

Q&A for 5/19:

#Reactivation #Reinfection #False Positives

Question: I was just reading about sailors on the USS Theodore Roosevelt testing positive again after having already recovered from COVID-19. What’s going on?

Answer: I read about that too! Cases like this have also been observed in Switzerland, South Korea, Japan, Italy, and China. We are still learning and don’t really have a solid answer on what’s going on. Here are three hypotheses as to what’s happening:

• Hypothesis A: People are becoming infected for a second time (reinfection). This idea is considered by most scientists to be the least plausible because a) the timing between negative →positive is too short; b) patients don’t seem to have other sources of exposure; c) immunity should last in the short-term.
• Hypothesis B: People are experiencing a reactivation of the virus. This idea is considered plausible. We discussed reactivation in our Q&A of 4/11. The basic idea is that the virus lays dormant for some time and then gets reactivated (perhaps due to stress or some other factor). One reason this holds water is that a number of the reactivated cases are folks who experience a second bout of illness symptoms. Many of the discussion sections of the papers I linked to above elaborate on this hypothesis and one scientist laid out some considerations in this recent Letter to the Editor. On the flip side, because SARS-CoV-2 does not infiltrate into the host cell’s nucleus (like hepB or HIV), the chances of it reactivating are very low.
• Hypothesis C: Test results are false positives. This idea is also considered plausible; perhaps the most plausible given current data. The idea is that PCR tests identify RNA virus fragments, including fragments that are inactivated/dead. Because the tests do not differentiate between living/dead virus, they could be picking up the remnants of the virus in recovered patients. Indeed, the South Korea CDC reported at the end of April that dead virus fragments were the likely cause of 263 people testing positive again days and even weeks after marking full recoveries. In even better news, the South Korea CDC found zero transmission from such patients to others. This finding helps assuage concerns about the possibility of recovered individuals transmitting the virus to others — no evidence supports this!

Q&A for 5/18:

#Texas #Reopening

Question: I heard Texas had its highest daily number of confirmed cases over the weekend. I know that Texas relaxed social distancing. Are we already seeing a second wave?

Answer: Texas did have its highest number of daily cases — 1,801 — this Saturday. But it’s important to note that this increase in cases has occurred alongside a ramp-up in testing. It’s also important to note that Texas started reopening on April 30th, which was before the state had met basic reopening criteria (see Q&A of 4/29). In my opinion, it would be inappropriate to call an increase in cases a “second wave.” Texas looks like it is still riding its first wave. To bolster that point, here are a few figures I made thanks to data from covidtracking.com. As you’ll see in the top chart, the proportion of tests that are positive had largely plateaued around 5%, and has more recently begun to decrease as Texas further expands testing. In the last week alone, Texas increased it’s testing by nearly 60% (see bottom right bar chart)! This additional testing is finding more cases and the daily number of cases in Texas continues to rise (see blue line in top chart). Texas is also seeing a modest increase in daily deaths (see chart at bottom left). Based on the testing/case data, Texas does not seem to meet the initial criteria for reopening (described in Q&A of 4/19).

Figure. Texas Tests, Cases, and Deaths

Q&A for 5/17:

#Immunity #Pregnancy #Good News

Question: I liked the good news round-up from (last week?). What’s the good news this week?

Answer: Yeah, that was only a week ago. It feels somehow much longer ago…. time is especially strange these days. Anyway, we do have some good news from the last few days, so let’s embrace it!

1. Vaccines and Lasting Immunity: Three days ago, Cell published a paper from a team of researchers across CA, NC, and NY (Mt. Sinai!) further examining adaptive immunity to SARS-CoV-2 to inform vaccine development. Researchers examined T cells — both T Helper cells (CD4) and T Killer cells (CD8) — of non-hospitalized people who recently recovered from COVID-19 and people who were never infected. Reminder: for an overly simplistic review of immune functions see Q&A of 5/9. This paper is not an easy one to read, so I greatly appreciated this Science blog that described the study and its results in more lay terms. In essence, the study found that (quotes are from the study itself):

• “CD4+ T cell and antibody responses were observed in all COVID-19 cases, and CD8+ T cell responses were observed in most.” Among those who have recovered, researchers observed a robust adaptive immune system response, which means that immunity is likely to be [relatively] lasting.
• “These data suggest that a candidate COVID-19 vaccine consisting only of SARS-CoV-2 spike would be capable of eliciting SARS-CoV-2−specific CD4+ T cell responses of similar representation to that of natural COVID-19 disease.” This means that vaccines targeted to the Spike-protein may indeed be quite effective! Additional data in the paper provides direction for further vaccine enhancement.
• “Pre-existing SARS-CoV-2−crossreactive T cell responses were observed in healthy donors, indicating some potential for pre-existing immunity in the human population.” Some exposure to other types of coronaviruses may be protective (cross-reactive immunity)! And this could explain why the severity of the disease varies so dramatically. More research is required, but this is a fascinating bit of information to add to the pile.
• “Clearly more studies are required, but the data here appear to predominantly represent a classical TH1 response to SARS-CoV-2.” This means that there was no evidence of an antibody-dependent-enhancement response, which would have been very bad. Note: antibody-dependent-enhancement response occurs in diseases like Dengue where the immunity a person builds form a previous bout with the disease (or vaccination) actually makes clinical symptoms worse and the risk of severe outcomes higher if the person is subsequently infected.

2. Pregnancy: Earlier this week, a pre-print (not yet peer reviewed) article was released from a group of researchers across the UK describing “on a population-basis, the risk factors, characteristics and outcomes of pregnant women hospitalised with SARS-CoV-2 in the UK…” Researchers used a national observation study to study risk factors between 427 women who were hospitalized with SARS-CoV-2 and 694 women in the comparison cohort. In good news, the researchers found that pregnant women hospitalized for SARS-CoV-2 were no more likely than the general population to require respiratory support. Additionally, researchers report that “the majority of [pregnant] women do not have severe illness and that transmission of infection to infants of infected mothers may occur but is uncommon.” [Now, there were some findings in this paper that definitely don’t fit in the good news category — like increased association between black and minority ethnicity and hospitalization with SARS-CoV-2 during pregnancy — but we’ll save those important findings for different Q&A].

Q&A for 5/16:

#Travel

Question: I was thinking of taking a trip to visit some loved ones. Obviously, I don’t want to unwittingly bring the virus with me. Should I go? If I do go, what would be the safest way to go?

Answer: First off, CDC still recommends that you try and stay home as much as possible, especially if your trip is not essential. That said, seeing loved ones could very well be considered essential! I won’t tell you whether you should go; only you can make that risk/reward calculation. But here’s the challenge in that last statement — how can you make a risk/reward calculation when we still have limited information and when the information we have is not well communicated (in my opinion)?!? If you do decide to go, here are a few recommendations:

• In General: When it comes to travel, I would avoid prolonged time with groups of people; the longer you are in the company of an infected person, the higher your risk of acquisition (discussed in Q&A of 5/12). Since avoiding prolonged time with groups of people limits risk, for me driving would be my preferred way to visit folks. But even that has drawbacks.
• Mode of Transport:
• Planes: NPR had a great story yesterday, “How Risky is It to Fly?” that really works through the pros/cons of flight and how to be the safest if you do choose to fly (avoid the aisle seat; bring hand sanitizer and use it; wipe down your space; wear a mask; try to avoid the toilet; wash your hands, don’t touch your face; try to create space between yourself and others)
• Trains and buses: We talked challenges of public transport in our Q&A of 3/26. CDC has guidelines for safety of rail transit operators and guidelines for safety of bus transit operators that seem relevant to passengers too. Here, the above suggestions apply. Also, avoid crowded train cars and buses and if you can, open a window! I should note that the transmission risks to bus drivers are high. There have been numerous reports of bus drivers suffering disproportionate levels of infection. In London, for example, as of the beginning of May 28 city bus drivers had died due to COVID-19. Please do your best to protect transit workers — among other things, keep your distance, wear a mask, don’t travel if you’re even a wee bit sick, don’t travel if you’ve recently been exposed to someone who is sick, and express gratitude.
• Cars: If you choose to take a road trip, there are also a number of things to do to minimize risk. For example, use disposable gloves to pump gas; minimize stops at public places; be extremely vigilant of proper hygiene if you have to use a public restroom. AARP offers a few helpful suggestions. And if you were doing a road trip-type vacation, Forbes offers some helpful tips.
• Lodging: Pay attention to hygiene standards in making your choice of where to stay. Do some research before you travel. While hotels may have higher hygiene standards, because they also have more shared space (elevators, lobbies, daily cleaning, etc.), in my opinion, they likely carrier higher risk than vacation rentals (here, I’m referring to single-occupancy dwellings rather than sharing a space in someone’s house/apartment).
• Communication: Talk with your loved ones about shared expectations of time together. Do they want visitors? Are they willing to have you into their home? Would you be comfortable staying with them? How will you work collectively to continue to minimize risks? In addition to all the other adaptations, this new time calls for new conversations.

Q&A for 5/15:

#Kawasaki #Children

Question: What’s the story with the new COVID-related syndrome in kids that I’ve started hearing about… Kawasaki?

Answer: There’s always another layer of this onion. A few weeks back, doctors in the UK reported seeing a small number of children (8 children) who had been infected with SARS-CoV-2 presenting with a “significant systemic inflammatory response” that was similar to Kawasaki disease (described below). None of these children had previously been symptomatic. Since then, researchers in Italy found a 30-fold increased incidence of Kawasaki-like disease — 10 children presented in 1 month in Bergamo province. And as of May 12, the New York State Department of Health had identified 102 patients presenting with Kawasaki-like conditions.

As a result of these findings, CDC issued a Health Advisory yesterday that describes case criteria of this “multisystem inflammatory syndrome” and requests “healthcare providers report suspected cases to public health authorities to better characterize this newly recognized condition in the pediatric population.” CDC’s Advisory also notes: “It is currently unknown if multisystem inflammatory syndrome is specific to children or if it also occurs in adults. There is limited information currently available about risk factors, pathogenesis, clinical course, and treatment for MIS-C.” In all of these instances, the number of children affected is quite small as a proportion of the whole. Nonetheless, these findings are clearly concerning. Silver lining is that knowledge of this condition is becoming more widespread, which means that the medical community can more quickly identify patients and appropriately respond.

*Kawasaki disease is quite rare — it is an acute illness accompanied by “fever, rash, swelling of the hands and feet, irritation and redness of the whites of the eyes, swollen lymph glands in the neck, and irritation and inflammation of the mouth, lips, and throat.” It’s of unknown cause and primarily affects children <5 years of age. Kawasaki is a leading cause of acquired heart disease in the United States. It’s symptoms can be confused with those of toxic shock syndrome.

Q&A for 5/14:

#1918 Pandemic #Lessons

Question: I don’t know much about the 1918 pandemic that I hear frequently mentioned these days. If it was such an epic tragedy, why isn’t it a bigger part of our collective knowledge? And now that folks are talking so much about it, is there anything we can learn from it that’s applicable to the COVID-19 pandemic?

Answer: I was wondering about this too! I didn’t know about the 1918 pandemic until I was in graduate school studying public health. I’ve always assumed that the brutal pandemic — with an estimated death toll of 50 million — was left out of our history books because it was so overshadowed by World War I. Earlier today, the New York Times published an article that tackles your first question so well. I highly recommend reading it! As it turns out, yes, the pandemic was overshadowed by the November 1918 Allies victory. But there’s more to it. Historian and author of “Pandemic 1918: Eyewitness Accounts” Catharine Arnold, describes that “Part of the problem was that dying from the flu was considered unmanly. To die in a firefight, that reflected well on your family. But to die in a hospital bed… that was difficult for loved ones to accept. There was a mass decision to forget.” Additionally, because leaders including then-President Wilson were eager to focus on and sustain the war effort, they rarely mentioned the virus. Even though Wilson nearly died from the flu during negotiations over the Treaty of Versailles, he never released any statement on the pandemic!

As to your second question, history repeats itself. There are many lessons to be learned from the 1918 pandemic. There was no sweeping national response to the 1918 pandemic. Rather, each US city took its own approach and the approaches varied dramatically. Researchers have used data from these different scenarios to study the impact of various interventions. In my relatively quick skim of relevant articles in pubmed, here are just a few lessons that stuck out:

1. Social distancing measures work in the short-term if rapidly implemented: Cities that responded quickly with multiple non-pharmaceutical interventions (e.g. social distancing measures including closing schools, public events, etc.) had lower fatality rates. As study authors write, “rapid implementation of multiple non-pharmaceutical interventions can significantly reduce influenza transmission…”
2. Cities will not implement social distancing measures indefinitely: Study authors also found that most cities stopped widespread social distancing measures within 6 weeks (range 2–8 weeks). As another researcher wrote in a fascinating paper on the lessons of the 1918 pandemic,”Perhaps the most important “lesson” taught by the pandemic was the realization that those measures that worked the best to control a highly infectious disease — bans on public gatherings, school closures, and strict quarantine and isolation — were precisely the ones most difficult to implement in a modern mass society. As an article in the July 5, 1919, Literary Digest summed it up, influenza’s spread “… was simple to understand, but difficult to control.”
3. Getting people to be socially compliant with restrictions cannot last indefinitely: As one scientist wrote, “it was possible to get reasonable compliance with precautionary measures for a while, but not indefinitely, even in the more obedient social climate that prevailed in 1918. San Francisco had demonstrations in which citizens defiantly tore off their own masks.”
4. Virus spread renews upon social distancing relaxation: Study findings also show that “viral spread was renewed upon relaxation of such measures…. … no city in our analysis experienced a second wave while its main battery of NPIs was in place. Second waves occurred only after the relaxation of interventions.”
5. Cities less impacted during the first wave are more impacted during second wave: “Cities that had low peaks during the first wave were at greater risk of a large second wave. Cities that had lower peak mortality rates during the first wave also tended to experience their second waves after a shorter interval of time, ≈6–8 weeks after the first peak vs. 10–14 weeks for cities with higher peak mortality rates.”
6. Cities that fair best implement both early and effective interventions and reintroduce these interventions when transmission again increases: A second set of researchers found that “ Cities that introduced measures early in their epidemics achieved moderate but significant reductions in overall mortality. Larger reductions in peak mortality were achieved by extending the epidemic for longer…. the cities that got closest to the theoretical maximum possible reduction in mortality were those that implemented both early and effective interventions throughout the first peak and then were able to reintroduce these when transmission again increased.”
7. Political calculations and economic worries will stymie public health communication: As another set of researchers wrote, “Unfortunately, in 1918, demands of World War I influenced the timing and messaging of public health precautions in some cities. Many public health officials resisted and delayed community mitigation measures under pressure from civil authorities who believed morale, and subsequently wartime productivity, could suffer.”
8. Telling the truth is paramount: As another scholar wrote in a far more in-depth review of the US 1918 pandemic communications, “The US response to the 1918 flu offers a case study of a communication strategy to avoid…The communication strategy of either reassurance or silence had its effect. Its effect was terror.” [Read this paper; it gave me chills.]
9. Accurate case and death reporting will be challenging: As described in this paper, such challenges in 1918 were be both political and technical. One example — “In the January 1919 American Journal of Public Health issue, the editor wrote that data in many cases were incomplete and confused as “so great was the pressure for action, that very few were able to devote any time to observation for the sake of the future.”
10. Those living or working in crowded conditions will be harder hit: This paper describes, “As is almost always the case with communicable diseases, poor, disadvantaged, and malnourished persons and those who lived in crowded conditions were at higher risk of death in 1918.”

1918 Public Health Service Poster from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2862334/

Q&A for 5/13:

#Vaccine Timeline

Question: I keep seeing that it will apparently take us 12–18 months to develop a vaccine. Where does that projection come from, and is it realistic?

Answer: As of 11 May, there are 110 vaccination candidates under evaluation around the world, 8 of which are in clinical development (Table 1). The world is moving at “pandemic speed” to find a vaccine/s that is both safe and effective. As you may recall, the 12–18 month timeline came into our collective consciousness on March 2nd, when Dr. Fauci told the President during a televised coronavirus meeting that a vaccine would be deployable “… at the earliest, a year to a year and a half, no matter how fast you go.” Many public health experts believe that the 12–18 month estimate is “ridiculously optimistic,” as Dr. Paul Offit, the co-inventor of the successful rotavirus vaccine recently put it. Nevertheless, according to an article published by Science yesterday, “…the organizers of a U.S. government push called Operation Warp Speed have little use for conventional wisdom. The project, vaguely described to date but likely to be formally announced by the White House in the coming days, will pick a diverse set of vaccine candidates and pour essentially limitless resources into unprecedented comparative studies in animals, fast-tracked human trials, and manufacturing. Eschewing international cooperation — and any vaccine candidates from China — it hopes to have 300 million doses by January 2021 of a proven product, reserved for Americans.”

So, where do these timeline estimates stem from? Let’s examine the vaccine development cycle. According to CDC, there are 6 steps in the cycle:

1. Exploratory stage: identify candidate antigens
2. Pre-clinical stage: use tissue/cell culture and animal studies
3. Clinical development: generally a 3 phase process with each phase introducing the vaccine to a bigger and wider pool of individuals (Phase 3== thousands)
4. Regulatory review and approval: FDA reviews and if it passes muster, approves
5. Manufacturing: Usually the realm of major drug manufacturers in the private sector
6. Quality control: ongoing monitoring and reporting of adverse events

Generally, it takes years for a vaccination to make it through these cycles and into widespread use. And the overwhelming majority of vaccine candidates fail. Two weeks ago, the New York Times had a fascinating description — with great visualizations — of how the process works, how long the process takes, and what it would mean to actually have a vaccine in that “warp speed” timeframe. For a more in-depth answer to your question, please read it! In the meantime, Figure 1 is a screenshot of one of its visualizations, which shows in blue bars the COVID-19 vaccination goal and in grey, the typical vaccination timeline. As you can see, the timeline we’re aiming for is ridiculously optimistic, but that doesn’t mean it’s impossible. Scientists around the world are racing, funding is pouring in, and demand is urgent.

Table 1. Candidate Vaccines under Clinical Investigation (WHO)

Figure 1. Goal vs. Typical Timeline for COVID-19 Vaccination Development (NY Times)

Q&A for 5/12:

#Risk #Indoor vs. Outdoor #Running

Question: My Facebook feed is littered with people admonishing others for running outside and I think it’s gone a bit too far. Can you please do a Q&A on whether people need to fear getting the virus from that runner who just ran past them?

Answer: This is a perfect example of why we need better and widespread health communication about the coronavirus. There’s so much information for folks to sift through, including disinformation, and there’s no one CDC resource or the like that describes the risks of transmission under various scenarios. The best thing I’ve read on risk is this blog post from Professor Bromage that he published a few days ago.

Professor Bromage reminds us that “Successful Infection = Exposure to Virus x Time.” He writes (among other valuable takeaways) that “Social distancing rules are really to protect you with brief exposures or outdoor exposures. In these situations there is not enough time to achieve the infectious viral load when you are standing 6 feet apart or where wind and the infinite outdoor space for viral dilution reduces viral load. The effects of sunlight, heat, and humidity on viral survival, all serve to minimize the risk to everyone when outside.” Please take a read through if you haven’t had the chance yet. It will become all the clearer that running outside as an individual is a very low risk activity.

As pathologist, Dr. Kasten, recently described it, “[Consider the difference in the risk between taking a stroll through the park or a climb up a steep cliff face] sure, you could slip, fall, strike your head, and die on that path in the park. Likewise, you could free-solo successfully to the top of El Capitan. But most of us would accept the risk of the stroll and not accept [the risk of] dangling from the cliff. Breathing in someone’s sneeze cloud, close by, without a mask — that’s the cliff face. Jogging several feet away, or getting the mail — that’s the park.”

And now here’s a bit more information about some of the studies showing risk of indoor vs. outdoor activity:

1. A study of transmission dynamics in a restaurant in China showed that “droplet transmission was prompted by air-conditioned ventilation.” Indoor conditions and air ventilation matters a lot when it comes to viral transmission risk.
2. Two pre-print (not yet peer reviewed) studies released last month showed that the preponderance of transmission occurs indoors.

• The first study identified outbreaks of 3 or more people in January/February in municipalities in China to assess environmental exposure types. Researchers found that of the 318 outbreaks identified (7,324 cases), only 1 occurred in an outdoor environment and involved two people. “A 27-year-old man had a conversation outdoors with an individual who had returned from Wuhan on 25 January and had the onset of symptoms on 1 February.”
• The second study examined 110 cases in Japan to identify transmission settings and found that “the odds that a primary case transmitted COVID-19 in a closed environment was 18.7 times greater compared to an open-air environment (95% confidence interval [CI]: 6.0, 57.9).” This study manuscript is quite weak — missing lots of key details — so I wouldn’t put much weight on it until after peer review, but in the meantime, the results do fit within the trend of what we’re seeing.

Q&A for 5/11:

#Maryland #Nursing Homes #Curve

Question: I was surprised to learn that Maryland has not yet “peaked” with regard to the first wave of COVID infections. What’s the story?

Answer: It’s true, while Maryland has flattened its case and death curves, the state has not yet passed the initial peak. As you’ll see, cases are still increasing (Figure 1), as are the proportions of cases that are positive (Figure 2) albeit both are increasing at a slower rate. I put these two figures together to show that we can’t blame the increase in cases on the increase in tests. Paradoxically, COVID-19 hospitalizations have begun to decline (Figure 3) even while deaths have plateaued (Figure 4). Meanwhile, Maryland has had a strong social distancing response over the last 1.5 months and is currently ranked among the top 5 states for social distancing. I find this all confusing as I would have expected Maryland to be on the other side of the initial wave given how aggressive Maryland has been with social distancing. Perhaps it’s due to clusters or hot spots? For example, 20% of Maryland’s cases are from nursing homes and 49% of Maryland’s deaths are from nursing homes. Uggghhh. Meanwhile, 279 workers at a poultry plant on the Eastern Shore tested positive for COVID-19 at the beginning of the month. These types of localized clusters could very well be fueling the increase in cases. Additionally, it could be that Maryland has a number of essential employees and/or people who cannot afford to miss work, so despite the broad shutdown, there’s still communal spread. And with all this said, the true answer is: I really don’t know.

Figure 1. Cases in Maryland [data from covidtracking.com]

Figure 2. Proportion of Cases that are Positive [data from covidtracking.com]

Figure 3. Daily Total Hospitalizations [data from MD Health Department, https://coronavirus.maryland.gov/]

Figure 4. Daily Deaths from COVID-19 in Maryland [data from covidtracking.com]

Q&A for 5/10:

#Llamas #Treatment #Antigen Test

Question: Got any good news?

Answer: Happy Mother’s Day!!! Here are three pieces of good news that have come out in the last 3-ish days:

1. Anitbodies from llamas may be great candidate for prophylaxis treatment!

• Llamas have special antibodies that are much smaller than typical human antibodies and easily manipulated to link or fuse with human antibodies. As it turns out, these llama antibodies are ideally suited for attaching to SARS-CoV-2 spiky protein and destroying the virus. Scientists published in vitro (lab-based petri-dish) results in Cell earlier in the week. If additional evidence accumulates, these llama antibodies could be used by health workers, elderly, and other high risk individuals as a prophylaxis treatment. Folks who receive the treatment would be immediately protect for a month or two before needed to receive another dose!

2. 3-drug antiretroviral drug combination may lesson the duration and severity of COVID-19!

• On Friday, scientists published results in Lancet of the effects of a 3-drug treatment combination (antiretroviral drugs interferon beta-1b, lopinavir–ritonavir, and ribavirin) on clinical outcomes of patients infected with COVID-19. The radomized control trial compared the outcomes of patients who received the triple drug combo with outcomes of patients who received only lopinavir–ritonavir. The researchers found that the 3-drug combination significantly reduced the duration of the disease (7 days compared with 12 days) as measured by negative COVID-19 test result (7 days [IQR 5–11]) than the control group (12 days [8–15]; hazard ratio 4·37 [95% CI 1·86–10·24], p=0·0010) and significantly reduced the duration of hospital stay!

3. A new antigen test just approved by FDA may make testing for COVID-19 faster, cheaper, and simpler!

• Late Friday night, FDA issued its first emergency use authorization for a COVID-19 antigen test. Antigen tests quickly detect fragments of proteins found on or within the virus by testing samples collected from the nasal cavity using swabs. This third type of test is different from existing PCR tests and antibody tests and has a unique role in the fight against the virus. The antigen test is designed as a rapid test for the virus — like the flu tests, your doctor can use these tests to provide results in minutes. Unlike PCR tests, however, the antigen tests do not have high degree of sensitivity. A positive with an antigen test is a positive, but a negative test result may be a false negative (for a refresh on sensitivity/specificity, see Q&A of 4/15). FDA therefore recommends that those who test negative also receive a PCR confirmatory test.

Q&A for 5/9:

#Kids #Immune Response

Question: Why aren’t children as severely affected by COVID-19? If it’s about an adaptable immune system, would someone whose immune system has been exposed to lots of new things for their life have a similarly adaptable immune system? Or does it just not work that way?

Answer: The short answer is that we still don’t know. Here’s a quick run-down of what we current know (to my knowledge!). First, preliminary evidence shows that children are are less likely to be symptomatic or to develop severe symptoms as compared with adults (yay!). The preliminary evidence is, however, mixed when it comes to three other important issues: 1) whether children are just as likely as adults to become infected SARS-Cov-2; 2) whether children are just as likely as adults to transmit the virus; and 3) why children are less likely to be symptomatic or develop severe symptoms. As you’ll so often read in scientific publications, more research is needed. Even so, we can partially answer your question by exploring the hypotheses scientists have about why children are less likely to be symptomatic or develop severe symptoms. At this stage of our understanding, it seems like children’s highly adaptive immune systems are protecting them. In this case, it’s actually their lack of exposure to other viruses and infections that’s helping them (their immune systems aren’t stuck in old ways of doing things)!

• Quick refresh on our immune system: Our immune system has two parts — innate (e.g. the response we’re born with) and adaptive (e.g. the response we learn as we come into contact with our world). Both parts work together to mount an immune response.
• The innate system is the first responder when your body encounters an invader (virus, bacteria). Immune cells called phagocytes engulf the invader and in essence eat it up (like pac-man)! These phagocyte cells (macrophages and neutrophils) are made in our bone marrow and are produced throughout our life. They move through our blood system, ready to destroy viruses and bacteria they come into contact with. But sometimes, a virus or bacteria can overwhelm our innate system. So our adaptive system kicks in!
• The adaptive system produces antibodies to protect your body from various invaders. These antibodies are developed by cells called B lymphocytes a few days after first exposure to a new virus or bacteria. The two kinds of lymphocytes are B lymphocytes and T lymphocytes. Lymphocytes begin in the bone marrow and either stay there and mature into B cells, or go to the thymus gland to mature into T cells. B lymphocytes identify invaders and basically mark them for other immune cells to destroy (like military intelligence). T lymphocytes are either Helper T-cells, which stimulate B-cells to make antibodies and help killer cells develop, or Killer T-cells, which directly kill the cells that B lymphocytes have identified.
• When it comes to children, here are the current hypotheses:
• Children have fewer environmental exposures (ex: smoking) and underlying conditions (ex: cardiovascular disease) and are therefore less likely to have severe outcomes: This is compelling, but insufficient. For example, this hypothesis does not explain why children are less likely to have symptoms.
• Children have fewer ACE2 cells that the virus can use to replicate itself (angiotensin-converting enzyme-2 (ACE2): Interesting idea, but so far “there is no evidence of a lower degree of expression or function of the SARS-CoV-2 receptor (namely ACE2) in children.”
• Children have stronger innate immune response: Scientists are exploring the hypothesis that the natural antibodies children make in abundance are better at effectively reacting to unknown invaders.
• Children have more adaptable adaptive immune response: Scientists are also exploring the hypothesis that the types of memory B cells that children develop are far more adaptable than the types of memory B cells of adults, especially the memory B cells of the elderly, which are very good at recognizing known invaders, but very bad at recognizing unknown invaders. “With ageing, malnutrition, immunosuppression, and co-morbid states, our immune system loses the ability to adapt to novelty.” These two innate and adaptive immune response factors may also be the reason why adults are more likely to mount an overly-aggressive immune response.
• Finally, some scientists earlier hypothesized that because children are more exposed to coronaviruses in their school and play environments, and therefore have an immune system more primed to respond to SARS-CoV-2. This hypothesis doesn’t seem to hold much water given that newborns are also less likely to have symptoms or severe outcomes but haven’t been exposed to other coronaviruses. Something else is going on, which makes me think that it’s the way that children’s immune systems are primed for novelty that is protecting them.

Q&A for 5/8:

#Masks #N95s #Ovens

Question: I’ve been reading on the twitters that we can disinfect our masks by putting them in the oven and/or microwave. Any truth to that?

Answer: When it comes to sterilizing homemade and cloth face masks, CDC recommends, “A washing machine should suffice…” When it comes to sterilizing N95 masks at home, however, a washing machine is not your best bet — cleaning products will degrade the filtration efficacy. Microwaving N95s is not a good idea either; the Society of American Gastrointestinal and Endoscopic Surgeons says, “At-home microwaving is not recommended because of variable power settings, and metal portions of the masks may catch fire.” A couple of studies have come out in recent days that provide more guidance on the best way to sanitize your N95 or surgical mask at home, and ovens seem like they might work (details and caveats follow).

A couple of studies have come out in recent days that provide more guidance on the best way to sanitize your N95 or surgical mask at home. The first such study disinfected masks by — placing the mask in a paper bag (like a lunch bag); putting the paper bag in an oven safe container; preheating the oven on to 170°; and “cooking” the mask for 45 minutes. Researchers tested how well the mask filtration worked after each decontamination cycle for 10 cycles and found that “no reduction in average filtration efficacy was observed.” A second study, published three days ago, reports “We found that heat (≤85 °C [185 °F]) under various humidities was the most promising, nondestructive method for the preservation of filtration properties in meltblown fabrics as well as N95-grade respirators.” So, heating in the oven seems like a promising approach for disinfecting N95 and surgical masks. BUT (because it seems like there’s always a but!), I’d also caution that the promising results of the two studies I cited are a shift from previous evidence that had shown that decontamination using dry heat caused “substantial filter degradation.” Two studies based on lab conditions do not a strong evidence-base make. Ideally, we’d want more evidence before making a recommendation.

*Reminder, we regular civilians are supposed to save the N95s for health workers, so while supplies are short, please refrain from purchasing N95s and keep using cloth face masks.

Q&A for 5/7:

#New York #Trends #USA

Question: It’s been a few weeks since we looked at trends in New York and the USA. What’s the latest?

Answer: You’re right, it has been almost 3 weeks since we looked at the trends (see Q&A of 4/20). When it comes to New York, things are indeed improving. When it comes to the United States, the trends are less positive. For ease of comparison and discussion, I’ll present side-by-side figures for New York, the United States, and the US excluding NYS, since NYS has born such a disproportionate burden of COVID-19 disease in the country. The upshot as I see it is: 1) New York continues to improve, but at a slower rate than we’d like; 2) The United States excluding New York is still seeing increases in daily cases and daily deaths. While the country has flattened the curve, we have not yet gotten to the downhill side of it.

• New Cases: New York continues its steady downward trajectory in new cases. The US as a whole shows modest declines in new cases, whereas the US excluding New York shows small increases in new cases. (Figure 1)
• Proportion of Tests that are Positive: New York also continues it’s downward trajectory in the proportion of daily tests that are positive, as does the US as a whole and the US excluding New York. (Figure 2)
• Hospitalizations and Deaths: Importantly, the total number of New Yorkers hospitalized each day is on the decline as is the total number of deaths (Figure 3). All good news! On the not-so-good news front, the decline in hospitalizations and deaths has been slow (see how the natural curves displayed in Figure 3 have long tails). I don’t have hospitalization data for the US, so the tables in Figure 3 are daily deaths. Like New York, daily deaths in the United States have declined, but when we exclude New York, we see that daily deaths in the United States are increasing.
• Case Fatality Rate: The crude case fatality rate in New York has remained stubbornly high at 6.1%, higher than the US as a whole and the US excluding New York. (Figure 4). While New York’s crude case fatality rate seems to have plateaued, the rate for the United States continues to increase.

Finally, if you want to see curves for your state (yes, you do!), check out YJ’s latest post, Is my state ready to reopen? Seeing is believing.

[Data from covidtracking.com]

Figure 1. Daily Cases in New York State are Rapidly Decreasing; in the USA are Slowly Decreasing; and in the USA excluding NYS are Slowly Increasing

Figure 2. Proportion of Tests that are Positive in New York State, USA, and USA excluding NYS is Declining

Figure 3. Daily Number of Hospitalizations and Deaths in New York is Decreasing; Daily Deaths are Slowly Decreasing the US; and Slowly Increasing in the US excluding New York

Figure 4. Crude Case Fatality Rate Remains High in New York; is Increasing in the USA and USA excluding New York

Q&A for 5/6:

#Men #Gender #Sex #Disparity

Question: Why does COVID-19 seem to disproportionately affect men compared with women?

Answer: Data from China, Europe, and the United States show that higher proportions of men who become infected with COVID-19 have serious complications, including death. Data from New York City highlight the stark differences (Figure 1) with nearly twice as many men dying from COVID-19 as compared with women. A number of social, behavioral and biological mechanisms are likely at play here. These factors influence other health conditions as well. Indeed, around the world, women live longer than men.

On the social and behavioral front, men may be less likely to access health care until more severe complications arise (ex: more women are getting tested than men for COVID-19 in the US); they may have riskier behavior (like smoking); and they may make riskier choices (like lagging behind women in following social distancing guidelines). On the biological front, men are more likely than women to suffer from cardio-vascular diseases, which are high risk for severe COVID-19 disease. Women may also have an innate biological advantage of XX chromosomes, which is associated with stronger immune response. There is MUCH that could be written here! If you want to learn more, I recommend that you read this NPR story from earlier in April and this 2018 Our World in Data story on overall longevity differences between men and women.

Figure 1. NYC COVID-19 Cases, Hospitalizations, and Deaths by Sex

Q&A for 5/5:

#Coding Deaths #Counting Deaths

Question: When they are counting deaths from COVID, do you know if they are also counting secondary infections and/or underlying contributing factors on death certificates? It would be interesting to know how many of the deaths would also be counted as antibiotic-resistant infections. And when death statistics are compiled, does one death with multiple causes listed get listed as 4 deaths when listing the causes? For instance, my father-in-law recently died and they listed four causes — head and neck cancer (the underlying issue of course), lung infection (the pneumonia), kidney failure (as a result of his body fighting the infection) and I think ultimately something having to do with his lungs being filled with fluid. So ultimately it was the fluid that killed him, but the others obviously contributed. So when they collect the statistics, does his death count as “pneumonia”? or “head-and-neck cancer”?

Answer: This is a great question, thanks for asking. And I’m sorry for your recent loss of your father-in-law.

A longer answer follows, but here are the short answers to your questions, 1) multiple causes are counted; 2) antibiotic-resistant infections are counted, but for COVID-19, I don’t believe the data are available yet; 3) underlying cause is what’s counted when you’re looking at most cause of death tables (e.g. head and neck cancer). Now for the longer answer…

Recording Deaths

A “medical certifier” (e.g. coroner or health professional) completes Section 2 of the death certificate (Figure 1). The medical certifier lists the “underlying cause of death,” which is defined by the WHO as “the disease or injury which initiated the train of events leading directly to death, or the circumstances of the accident or violence which produced the fatal injury,” as well as the other conditions that contribute to the death (up to 20 conditions can be listed). The conditions are listed in causal order with the underlying cause being the lowest on the list (like the order you shared in your question). Per CDC, in the United States “there is an average of three causes listed per certificate. Approximately 20 percent have only one cause of death and 45 percent have three or more causes.”

Coding Deaths

After the medical certifier completes the bottom portion of the death certificate, the funeral director typically completes the top portion. Thereafter, the certificate is sent to the State’s vital statistics office (or the like) and each State then sends these certificates on to CDC’s National Center of Health Statistics (NCHS). At NCHS, the data are entered into a specialized computer system and either the computer or a trained coder translates these conditions into International Classification of Diseases, Tenth Revision (ICD-10) codes. These codes are used worldwide to enable comparable cause-of-death mortality data collection and analysis. And yes, these codes can be very specific, including specific codes for antibiotic resistant infections.

Tabulating Deaths

When it comes to aggregate statistics, cause of death data is reported in two ways: 1) underlying cause and 2) multiple cause. You use both sets of statistics for different, complementary analyses. The “leading cause of death” tables we frequently see or “top 10 causes of death” are based on underlying cause data. We also need to focus on multiple cause analysis to understand disease interaction, and consider other interventions to reduce mortality. This paper from Epidemiology gives a nice description of both types of statistics.

Calculating COVID-19 Deaths

This process applies to COVID-19 too. CDC describes COVID-19 death recording here and death coding and tabulation here. While in most cases, COVID-19 will be the underlying condition, CDC’s counts of COVID-19 deaths may include those for which COVID was the contributing cause depending on the analysis. Additionally, the ICD-10 code for COVID-19 (U07.1) is applied not only to “confirmed” deaths, but also to those that are “presumed” and “probable” (e.g. without laboratory confirmation). A couple of relevant notes:

1) CDC also collects death counts from states on a daily basis. These counts, listed here, are different from the counts listed with NCHS, which are always several weeks behind because they are based on death certificates received and entered into NCHS’s system; and

2) Our World in Data offers a really nice description of COVID-19 mortality around the world, including great graphs (for example, see Figure 2). I highly recommend!

Figure 1. U.S. Standard Death Certificate

Figure 2. Daily Confirmed COVID-19 Deaths (from ourworldindata.org)

Q&A for 5/4:

#Models #Projections

Question: I’m reading so much these days about projected COVID-19 deaths, especially with this weekend’s widely reported statements made by Dr. Birx on “Fox News Sunday” that “our projections have always been between 100,000 and 240,000 American lives lost…” What are the models telling us and how have they changed over time?

Answer: We’ve talked a bit about modeling in our Q&As of 4/5 and 4/28. The CDC now has a helpful resource page describing the various models, which was just updated on 1 May. I’ve taken a couple of figures from CDC’s resource pages to help answer your question. [Related aside: The IHME model is not listed on the page, but it has been listed in previous estimates. I’m not sure why it’s missing now, but it could be that the conditions that IHME were assuming — stringent national social distancing measures maintain — are no longer in effect nationally.]

As you can see in Figure 1, the many models do vary to some degree, with several estimating that deaths will exceed 100,000 by late-May and others estimating that the rate of increase will slow with deaths in the range of 70,000. The ensemble estimate, which summarizes estimates from the individual models, suggests about 90,000 deaths by late-May (right side of Figure 1). Much of this variation across models is due to: a) the assumptions that the modelers use to make their estimates (ex: CU-20 estimates only a 20% reduction in social contacts while CU-40 estimates a 40% reduction); and b) the various scientific methodologies the models use to make the predictions. If you want to further explore national and state-level estimates, this tool allows you to do so (it’s pretty fascinating!).

When it comes to how these estimates have changed over time, we can look at previous forecasts and see. For example, Figure 2 shows the cumulative death estimates as of April 13th across multiple models. Eyeballing it, the CU-30 estimate for May 1 seems to track the best with the actual number of deaths as of May 1 (60K-65K deaths — exact number varies depending on which source you use). Furthermore, if you wanted to see how IHME estimates have changed over time, this tool makes it very easy to visualize. Check it out!

It’s to be expected that models will change over time as scientists learn more and as modelers adapt their assumptions based on new data and learning. As Ed Yong so eloquently wrote last week in The Atlantic, “If measuring the present is hard, predicting the future is even harder. The mathematical models that have guided the world’s pandemic responses have been often portrayed as crystal balls. That is not their purpose. They instead describe a range of possibilities, and help scientists and policy makers to simulate what might happen pending different courses of action. Models reveal many possible fates, and allow us to choose one. And while distant projections are necessarily blurry, the path ahead is not unknowable.”

Figure 1. April 27th Forecast of Cumulative COVID-19 Deaths Nationally over the Next Month (from CDC)

Figure 2. April 13th Forecast of Cumulative COVID-19 Deaths Nationally over the Next Month (from CDC)

Q&A for 5/3:

#Far-UVC #Ultraviolet Light #Disinfectant

Question: You talked about sunlight in a previous post, but I was just reading about a special type of ultraviolet light — Far-UVC rays (check out this Columbia University article), which seems promising. What do you think?

Answer: Super interesting! We talked about sunlight and Vitamin D in our Q&A of 4/24, but we didn’t delve into ultraviolet light. Most ultraviolet light damages our cells and does so exceptionally quickly. The BBC article I referenced on 4/24 does an excellent job laying out the issues, including describing the potential benefits of Far-UVC. In short, Far-UVC does seem promising, especially for preventing antibiotic-resistant infections! But, the science is still very limited with only lab-based petri-dish and animal studies having been conducted. More research would be needed — research that includes human subjects — before we would start using widely using Far-UVC. Here’s a bit more information on what we currently know about Far-UVC:

• It is less dangerous to handle than other forms of UVC but still lethal to viruses and bacteria
• Per this Nature article, “due to its strong absorbance in biological materials, far-UVC light cannot penetrate even the outer (non living) layers of human skin or eye; however, because bacteria and viruses are of micrometer or smaller dimensions, far-UVC can penetrate and inactivate them.”
• Does not damage human skin cells based on culture dish (in vitro) study (note: the same study exposed hairless mice to Far-UVC and also noted that it did not damage mice skin cells); and
• Has been shown to prevent mouse wounds from becoming infected with the drug-resistant bacteria and also been shown to kill areosolized H1N1 influenza virus.

Finally, a pre-print study (not yet peer-reviewed) was just released last week, which studied Far-UVC effects on two aerosolized (e.g. airborne) coronaviruses (not SARS-CoV-2). The authors concluded that “very low doses of far-UVC light efficiently kill airborne human coronaviruses carried by aerosols.” The authors further recommended exploring the use of Far-UCV light for disinfecting indoor public spaces, not just for COVID-19, but for other types of viruses and bacteria. This is an exciting area of innovation! I look forward to more human studies to ensure safety before we delve into widespread use.

Q&A for 5/2:

#Contact Tracing #Workforce Proposals

Question: In your 4/26 Q+A you addressed contact tracing and concluded that there will need to be a massive workforce expansion in order to carry it out. Can you tell us how the CDC will expand their workforce, qualifications required to work in this field going forward, and who pays for it all?

Answer: This is a big topic everywhere you turn. And there are numerous proposals on how to go about doing it, including Senator Gillibrand and Senator Bennet’s proposal, Health Force. Most experts agree that contact tracing can be easily taught and the skills do not require any higher-level educational background — a high school education suffices. Currently, it’s up to States and localities to establish their own contact tracing programs. For example, Partners in Health is working with the State of Massachusetts to establish a state-wide COVID-19 contact tracing program. In terms of a national response, experts suggest the cost will be multiple billions of dollars. Here’s a quick run down of the various proposals for a national response that I’m aware of:

• Health Force: Senator Gillibrand and Senator Bennet’s proposal to address the dual health and economic crises facing the country by establishing a new, federally funded and locally-managed Health Force. Health Force members would be recruited from their communities to work in their communities on behalf of state and local public health organizations. Force members would be quickly trained and employed to conduct contact tracing. They would also fulfill other key COVID-19 related tasks, like assisting with testing, vaccination campaigns, health messaging, data collection and entry, food and medicine delivery to isolated individuals, health and social system navigators, and more. After the current crisis abates, Health Force members would continue to serve vulnerable communities, helping to bridge health and wellness gaps. The proposal would provide funding to all states and territories to hire hundreds of thousands of individuals across the country.
• Pandemic Response and Opportunity Through National Service Act: This is a companion bill to Health Force that is being led by Senator Coons and colleagues. The idea is to expand service opportunities to respond to the panoply of COVID-19-related needs — health and beyond — across the country, including contact tracing. This legislation would expand AmeriCorps positions from 75,000 to 150,000 in Year 1 and to 300,000 in Years 2 and 3.
• Containment Corps: Senator Warren and Rep. Levin have also rolled out a strategy to address the contact tracing workforce gap. This plan would call for CDC to: a) develop a national strategy “to hire, train, and deploy nationwide individuals to augment public health authorities’ capacity [to conduct contact tracing]”; b) provide grants to states and local public health agencies to hire, train, and deploy such members; and c) provide regular reporting on such members. It would also require the Department of Labor to provide funding to state and tribal workforce agencies to link unemployed with Corps jobs.
• U.S. Public Health Jobs Corps: This is Joe Biden’s plan to mobilize at least 100,000 Americans across the country who could, “serve in a variety of important functions, including ensuring contact tracing reaches every under-served community in America — ideally by members of those communities themselves…. U.S. Public Health Jobs Corps would become the permanent foundation for a stronger national community public health service that could eventually transition to fight the opioid epidemic and address other national public health priorities.” This idea is a bit of a combination of the above proposals.
• Association of State and Territorial Health Officials: ASTHO recommends that Congress help scale up “existing capacity at the state, local and territorial” levels, rather than “set up a system outside existing public health agency response.” The national organization offers suggestions in terms of workforce and funding needs.
• Bipartisan Group of 16 Public Health Leaders: Last week, a bipartisan group of public health leaders sent a letter to Congress requesting $12 billion to help expand the contact tracing workforce and $34 billion to support other elements of the public health response. As NPR reports, “The officials estimate the workforce needs to increase by 180,000 until a vaccine is on the market.”

Q&A for 5/1:

#Herd Immunity

Question: I hear a lot about herd immunity, but I’m not sure that I get it. Would you please explain?

Answer: According to this history of herd immunity published in 2011, the term “herd immunity” was coined by scientists back in the 1920s. The basic idea is that if enough individuals in a given population develop immunity to a virus (generally through immunization, sometimes by widespread infection) then the vulnerable individuals in that population will be protected too (Figure 1). Imagine a herd of water buffalo surrounding their calves to protect them from lions and now you’re imagining immunized individuals surrounding vulnerable individuals and protecting them from bad viruses (here’s the video — it has a happy ending)! I think you probably already had this part of the knowledge down, so let me get more into the weeds with it for a minute.

Generally, 70% to 90% of a population needs immunity to achieve herd immunity. The proportion required to achieve herd immunity varies based on four factors: 1) how contagious an infection is; 2) how much immunity is conferred by the vaccine (or previous illness); 3) the social networks within the population (e.g. how people mingle); 4) the distribution of the vaccine/immunity within the population (e.g. heterogeneity of the population vaccinated/infected). Based on current estimates of how contagious the coronavirus is, scientists estimate that at least 70% of the population needs to become immune in order for the remaining 30% to be protected.

If the United States were to try to achieve herd immunity before a vaccine comes to the fore, it would mean that 231 million of us would need to be infected. And this presupposes that the antibodies we develop as a result of infection will protect us from reinfection, which is still an area of scientific inquiry (see Q&A of 4/25). Data from serology testing in one big hot spot, New York City, suggest that ~21% of NYC residents have been infected. Despite the massively heavy toll the virus has taken on the City and her residents, NYC is still far from reaching herd immunity. Epidemiologists of Johns Hopkins recently discussed this very issue and made the compelling case in their very brief article that, “As infectious disease epidemiologists, we wish to state clearly that herd immunity against COVID-19 will not be achieved at a population level in 2020, barring a public health catastrophe.”

Figure 1. Virus transmission in Susceptible vs. Immune Populations (from this Vaccines paper)

Q&A for 4/30:

#Recovery

Question: You talked about potential long-term effects of COVID-19 disease among recovered individuals a while back and there wasn’t much information. How about now?

Answer: The post you’re referring is from 3/18, which feels like forever ago. But in the scheme of science and following cohorts of recovered patients over time, it’s not that long. According to the Johns Hopkins Covid Tracker, as of today we have slightly over 1 million individuals who have recovered from COVID-19… wow! Unfortunately, I still haven’t seen much data about the health of recovered COVID-19 patients. Here’s a quick run-down of what I think we know:

First off, most people who contract COVID-19 will have a mild case and the symptoms they are likely to experience in recovery are also likely to be mild (ex: lingering fatigue). Even so, “It takes anything up to six weeks to recover from this disease,” Dr. Mike Ryan, executive director of the World Health Organization’s Health Emergencies Program, said in a March press briefing, and “people who suffer very severe illness can take months to recover from the illness.” Indeed, the types of issues hospitalized patients experience in recovery are more serious. Recently published results form a longitudinal study of 90 patients hospitalized due to COVID-19 showed that “94% (66/70) of patients who were discharged from hospital at the end of the study still had mild to substantial residual lung abnormalities on their last CT scans.” And among those who survive after being on a ventilator, the longer-term health conditions are even more severe with individuals also facing muscle atrophy and weakness.

All that said, my original answer from 3/18 remains today — we’ll know much more as we have more data tracking the health status of individuals who recover. In the meantime, if you want to know more about current thoughts on the matter, this Science article from earlier this month does a nice job laying out the concerns.

Q&A for 4/29:

#TX #Reopening #School #Day Care

Question: Our Governor is starting to reopen our state (Texas) on April 30th and our school decided to reopen on May 6th. Should I send my kids to school? They have detailed the precautions they plan to take, but I’m honestly not sure what the right way to go is.

Answer: This is such a hard question that many of us will soon be facing if we aren’t already. I would’t feel comfortable giving you advice as only you know your own circumstances and can make your own risk/reward calculation. What I can tell you is how I might approach this when the time comes for Maryland. And since I am a big fan of evidence-informed decision-making, I’d start with the criteria laid out in the federal government’s “Guidelines: Opening Up America Again.” We discussed these Guidelines in our Q&A of 4/19. As a reminder, before considering reopening, Governors must see these criteria met:

1. A sustained downward trend in flu-like illnesses and covid-like cases over 14 days;
2. A sustained downward trajectory in covid-cases OR proportion of positive tests over 14 days;
3. Ability of hospitals to treat patients without crisis care; and
4. Robust testing program in place for at-risk healthcare workers, including emerging antibody testing.

For Texas, I cannot speak to criteria 3 and 4. But we can look at data for criteria 1 and 2. For criteria 1, using data from the Texas Department of Health, Influenza Surveillance Network, we see that visits due to influenza-like illness have been declining over the course of April (Figure 1). For criteria 2, using data from covidtracking.com we see that there has been a downward trend in the proportion of positive tests since April 20th (Figure 2). If we were to expect Texas to follow the Guidelines, and *assume* that criteria 3 and 4 were met, we’d expect Texas to start Phase 1 of reopening on May 4th. Opening of schools and day care centers is NOT, however, part of Phase 1 of reopening. Rather, schools and day care centers are supposed to reopen in Phase 2 (see reopening table in Q&A of 4/19). So for Texas, this would mean that schools and day care centers would reopen only after Texas witnesses a continuing downward trend from May 4th- May 28th. This brings us to end-May as a potential date for schools/day care center reopening. If it were me, I’d be more comfortable with this type of return than the earlier return your school is offering.

Figure 1. Influenza-like illness

Figure 2. Cumulative Percent Positive Tests Began Trending Down on April 20th

Q&A for 4/28:

#DC #IHME

Question: The IHME projection model for COVID — which you so diligently analyzed earlier this month — showed that DC’s deaths per day would peak around April 15/16 and then begin to decline. Has that been the case, and does that then make the CHIME model the winner here (for now)?

Answer: Thanks for remembering my earlier post (Q&A of 4/5)!

IHME’s estimate for DC’s peak deaths was off. Sadly, DC experienced it’s highest mortality yet last week, and this week is on track to be even higher (Table 1). Additionally, the proportion of tests that are positive continues to increase (>21%), albeit at a slower rate than earlier in the epidemic (Table 2). And the cumulative case fatality rate is also increasing (4.8%). If you want to compare these numbers to earlier in the month, please check out the Q&A from 4/12. Unlike New York (Table 3), DC is not yet on the other side of the curve.

Because it’s late and time for dinner, I’m going to take a pass on talking about modeling winners… for now!

[data from: covidtracking.com]

Table 1. Weekly COVID-19 Deaths in Washington, DC

Table 2. COVID-19 Deaths, Cases, and Tests in Washington, DC as of 4/28

Table 3. COVID-19 Deaths, Hospitalizations, and Tests in New York as of 4/28

Q&A for 4/27:

#Vaccination #TB

Question: I’ve heard some rumblings about TB vaccination offering protection against COVID-19. What do we know?

Answer: Bacille Calmette-Guérin vaccine (BCG) is the vaccination for tuberculosis (TB) and it is offered to newborns in countries with high burden of TB. Because the US thankfully does not have a high burden of TB, the BCG vaccine is not commonly given here. A couple of ecological studies recently came out highlighting that countries that have high BCG vaccination rates do not seem to have high COVID-19 infection or mortality rates. On the heels of theses studies, WHO issued some guidance on this topic, stating “There is no evidence that the BCG vaccine protects people against infection with COVID-19 virus.” For ease of reference, I’ve synthesized the main points of the WHO guidance here. In short, there’s still a LOT we don’t know, but scientists are actively working to fill our knowledge gaps. In the meantime, BCG vaccination should be reserved for newborns in countries with high TB burden as BCG vaccination itself is in relatively short supply globally.

• While there’s some experimental evidence from animal and human studies that BCG vaccine has some effects on the immune system, the clinical relevance of these effects is unknown.
• As of 11 April, there are three pre-print manuscripts (e.g. not yet peer reviewed) that show COVID-19 prevalence is lower in countries with widespread BCG use compared with others. These types of ecological studies are prone to bias and are quite weak. We cannot rely on these findings as evidence to inform programs or policies. Instead, such findings call for more rigorous research.
• Rigorous research is happening with two registered clinical trials actively recruiting to explore impact of BCG vaccination given to health workers involved in the care of COVID patients.
• From Lancet article cited in WHO guidance, “Due to problems with vaccine production and limited supplier options in some countries, the global availability and procurement of BCG has been a challenge since 2013.”
• WHO’s current recommendation is: “BCG vaccination prevents severe forms of tuberculosis in children and diversion of local supplies may result in neonates not being vaccinated, resulting in an increase of disease and deaths from tuberculosis. In the absence of evidence, WHO does not recommend BCG vaccination for the prevention of COVID-19. WHO continues to recommend neonatal BCG vaccination in countries or settings with a high incidence of tuberculosis.”

Q&A for 4/26:

#Contact Tracing

Question: I’ve been reading a lot about the need for contact tracers to begin reopening. Can you explain what exactly contact tracing is, how it works, and why it’s important?

Answer: Contact tracing is a key element of public health response to infectious disease. When we identify a new case (e.g. someone who is infected with a virus that can be spread person to person, here SARS-CoV-2) we want to next identify everyone who has been in contact with that infected person. We want to find these contacts in order to alert them of their possible risk, to help link them to care, and to help prevent them from spreading the virus to others. Generally, the three components of contact tracing are:

1. Contact identification (ask the infected person to identify all the people they’ve interacted with during the period the may have been infectious);
2. Contact listing (find and talk with the contacts as quickly as possible, inform them of their risks and educate them about the disease, and especially for those high risk contacts, request that they self-quarantine); and
3. Contact follow-up (regular check-ins to monitor for symptoms and test for infection).

In the United States, our health departments are typically set up to do contact tracing related to sexually transmitted infections. When it comes to contact tracing for COVID-19, CDC has a number of resources and offers these principles to groups across the country who are gearing up to expand contact tracing.

The contact tracing component of the response is so important because once we stop using the blunt hammer of extreme social distancing, we need to have a more nimble and precise response that allows us to quickly identify all cases (widespread testing) and all potential cases (widespread contact tracing), and keeps these folks from engaging with the wider population (isolation) to thereby limit the spread of the virus. This will require a huge effort, including massive workforce expansion, as well as good will and trust of the people.

Q&A for 4/25:

#Antibodies #Vaccines

Question: If having antibodies doesn’t confer protection, then would a vaccine even work?

Answer: Great question! First, let me provide some more information on antibodies and protection, then a direct answer to your question.

Yesterday, WHO came out with some guidance on “immunity passports,” stating that “As of 24 April 2020, no study has evaluated whether the presence of antibodies to SARS-CoV-2 confers immunity to subsequent infection by this virus in humans.” Some media outlets ran with this, stating “WHO warns you may catch coronavirus more than once.” This take is, unfortunately, a bit of sensationalism (in my opinion). As the old saying goes, absence of evidence is not evidence of absence. In the case of antibodies and ongoing protection, with just 4 months into this new virus and disease, we just don’t know. And we wouldn’t want to base huge policy prescriptions — like immunity passports — on a hunch (as discussed in Q&A of 4/13 and 4/15). To my knowledge, most health experts believe, based on historic precedent, that antibodies would offer some degree of protection. Earlier this month, Dr. Fauci spoke with JAMA’s editor, stating that “It’s a reasonable assumption that this virus is not changing very much. If we get infected now and it comes back next February or March we think this person is going to be protected.” But, with lots of folks’ bodies/immune systems reacting in what seems to be different ways to the virus — with an estimated 80% of cases being mild, some people being asymptomatic, and still others having a variety of serious symptoms and outcomes, it’s even harder to make an educated guess. Moreover, having antibodies is more than a yes/no dichotomy; it’s a continuum — and the more antibodies you have, the more likely you are to be better protected. Just what level of antibodies required to protect against reinfection, however, is unknown, and just how long such immunity would last is also unknown. For other types of coronaviruses, the immunity offered against reinfection wanes after a year or two (as discussed in Q&A of 3/19).

So, what does this all mean for vaccines? The National Academies rapid expert consultation of April 8th has the best quick read on the subject that I’ve come across:

“The duration of antibody response and acquired immunity to reinfection will be critical to understanding 1) how effective vaccination is likely to be; 2) how durable immunity is; 3) whether it is possible to achieve herd immunity against COVID-19; 4) how safe it is for people who are positive in a serology test to return to work. One key uncertainty arises from the fact that we are early in the outbreak and survivors from the first weeks of infection in China are, at most, only three months since recovery….”

And copied herein is the experts’ list of gaps in knowledge. The knowledge that we’ll continue to gather will influence vaccine design and our expectations of vaccine efficacy, including how frequently vaccine boosters may need to be given.

National Academies Expert Consultation:

Q&A for 4/24:

#Sunlight #Vitamin D

Question: So… sunlight kills the coronavirus?

Answer: A lot of health journalists have written about this topic already, so rather than taking up much air time here, I’ll refer you to this BBC report, which is fascinating! WHO also tackles sunlight and coronavirus in it’s myth-busters series (see Figure below).

WHO has guidelines about UV radiation and sun exposure too, and they largely warn against sunlight exposure. That said, we also know that Vitamin D is good for us, including for helping our immune system fight off viruses. It’s difficult to get enough Vitamin D from food alone, so we need some sun exposure. While we don’t have science about the role of Vitamin D as a treatment or prophylaxis for COVID-19, some doctors are recommending Vitamin D supplementation nonetheless. The reasons for this push include: a) Vitamin D is known to play a role in immune response to viral infections and b) Vitamin D supplementation is generally safe with few side effects (as long as folks don’t overdose on supplements), so the risks are minimal. With more of us sheltering in place and potentially spending less time outside, doctors want to be sure that we’re getting enough Vitamin D — deficiencies are bad for our bones and overall health. Speaking of spending time outside, you don’t need a study to tell you that sunlight and nature have powerful effects on our mood and mental health. But if you want a study, here are just a couple of journal articles on the matter. So, go on, get outside! You’ll feel better for it… even if it doesn’t kill the coronavirus.

Figure 1.

Q&A for 4/23:

#Blood #Systemic #Stroke

Question: What is the story on these new COVID-related health problems I’m reading about — frequent blood clotting issues and strokes among young people?

Answer: Friendly reminder: I’m not a doctor; while I do have public health training, I do not have medical training. With that important caveat out there, here’s the lay of the land as I currently understand it:

Now that medical professionals have had increased opportunity to treat COVID-19 patients, they are starting to see common themes in terms of illness progression. Those observations beget scientific studies. And in the last two weeks or so, we’re having increased scientific evidence highlighting just how common coagulation disorders in COVID-19 hospitalized patients are. When it comes to the blood clotting issues described in the very informative Washington Post article cited in your question, doctors and scientists are still speculating on why patients are experiencing them. There’s a lot we still don’t know about how our bodies work and our immune systems and inflammatory responses are especially mystifying. One of the hypotheses is that the viral-associated coagulation disorders are caused by the inflammatory response to viral illness, which can “lead to an imbalance of the pro- and anti-coagulant state during viral infections.” This imbalance can make blood unable to coagulate (ex: excessive bleeding as seen in Ebola; called thrombocytopenia) or can make blood too sticky (ex: excessive clotting as seen with COVID-19; called disseminated intravascular coagulopathy (DIC)). The COVID-19 disease is clearly more than just a disease of the respiratory system — it seems to be far more of systemic disease than we originally thought.

Turning our attention to strokes among young people, this is also very likely related to the blood clotting issues. Blood clots in the arteries or the brain — called arterial thrombosis — can cause stroke. And the COVID-19 coagulation disorders we’re seeing include arterial thrombosis. So, this could explain why doctors are seeing increased numbers of COVID-19 patients presenting with stroke. I haven’t read any science on the risk of stroke to SARS-CoV-2 infected young people, but since doctors are taking note, I imagine that we’ll soon see more science on the matter. There’s always something new to learn when it comes to a new, highly contagious virus.

Q&A for 4/22:

#Wastewater Epidemiology #Poop #Sewage

Question: I was just reading that sewage can be a good way to trace COVID outbreaks. Is it true?

Answer: Epidemiologists have long been tracing disease through poop, sewage, and waste water systems. Indeed, the early tracking of typhoid cases in the late 1800s revealed it to be spread via excrement and contaminated water. The field of “sewage epidemiology” aka “wastewater epidemiology” is, however, relatively new, having first been proposed (to my knowledge) in 2001 by scientists who hypothesized that mass spectrometry screening could be used to identify drug residues in waste water and trace this back to population drug use. Since then, several studies have been conducted and published, concluding that through the novel mass spectrometry approach, “patterns and trends of drug abuse in local communities can be promptly monitored.” More recently, scientists have suggested using this approach to detect viral outbreaks. And back in March, researchers from Biobot, MIT, Harvard University, and Brigham and Women’s Hospital in Boston launched a pro bono program to test community sewers. As described in a fun Popular Mechanics article from last month, they are “looking for SARS-CoV-2, the virus that causes COVID-19, in people’s poop.”

While I haven’t seen any data from Biobot yet, a pre-print publication was posted just a few days ago exploring SARS-CoV-2 in wastewater in Paris. This study shows that it is possible to track coronavirus in waste water. Using time-trend data, the researchers found that concentrations of the virus in wastewater followed trends in COVID-19 fatalities. Also published just a few days ago was a similar study out of Austrialia (this one is even peer-reviewed!) showing a strong relationship between viral concentrations in waste water and coronavirus cases in Southeast Queensland. Both studies demonstrate that wastewater epidemiology can be a valuable tool for better surveying viral circulation in the population (v. important in case of asymptomatic transmission and limited testing), and can even serve as an early warning system to monitor coronavirus entrance into a given population. Science and Nature both had good articles on this cool new area of science if you want to learn more.

Q&A for 4/21:

#Treatment #Hydroxycholoroquine

Question: The President has not been touting hydroxychloroquine as much these days. Do we know more about how well the drug is working as treatment?

Answer: We’ll know much more once the clinical trials rigorously testing hydroxychloroquine are complete. In the meantime, the most recent and complete data we have are not promising.

A week ago, results from a Brazilian study on high dose hydroxycholoroquinee for COVID-19 treatment were released (again, pre-print, not yet peer reviewed). The study examined whether clinical outcomes were improved with a higher dose of hydroxycholoroquine compared with a lower dose. The study was stopped early because of the adverse outcomes (much higher risk of death) associated with the higher dose. These findings show that hydroxycholoroquine in too high a dose can be extremely harmful. But the findings don’t tell us about the impact — positive or negative — of a lower dose. Cue new study.

Just today, researchers released a pre-print manuscript submitted to the New England Journal of Medicine that explores outcomes of 368 male COVID-19 patients in the VA system who either received A) hydroxycholoroquine; B) hydroxycholoroquine + azithromycin; C) neither. The study used retrospective VA system record review and adjusted for myriad confounding factors. In fact, because patients who received an intervention (e.g. hydroxycholoroquine +) were more likely to have severe disease (and therefore more likely to die), the researchers used propensity score matching to help ensure that the comparisons between the groups were fair. That’s maybe a wee too science-y. What I’m trying to say is that the researchers did a really good job to try and make sure that the comparisons across the three groups really were comparisons of the impact of hydroxycholoroquine rather than some other factor or set of factors. And they found that the risk of death was highest among the Group A (adjusted hazard ratio, 2.61; 95% CI, 1.10 to 6.17; P=0.03). This was a statistically significant finding (e.g. very, very small chance that it’s not due to random error). Hydroxycholoroquine was not associated with improved outcomes but instead, was associated with increased risk of adverse outcomes! Soapbox moment — it’s findings like these and science like this, which highlight the need for clinical trials and evidence(!) before widespread use of a given drug.

Q&A for 4/20:

#Reopen #Promising Trends

Question: I’ve been hearing a lot about how the United States has turned a corner, or how we can now “see the light at the end of the tunnel”. Of course, a lot of this talk is related to reopening. How realistic is that perspective?

Answer: I was wondering about that too. We know that New York State has turned the corner with the proportion of positive test results decreasing, the number of hospitalizations decreasing, and daily deaths decreasing (see Figure 1). These promising trends are indeed something to celebrate! It’s even more meaningful since NYS represents the biggest outbreak in our country and one of the biggest in the world. But we shouldn’t confuse New York’s success for the success of the country as a whole.

I went ahead and removed NYS data from the USA data on testing, cases, and deaths (datasource: The COVID Tracking Project). What we see in Figure 2 is that in the USA (excluding NYS), daily/weekly deaths continue to increase; the proportion of tests that are positive continues to rise; the number of tests per day has been flat for several weeks; and the crude mortality rate continues to clip up. As a whole, the United States is no where near achieving the “gate metrics” required to consider reopening (listed in Q&A of 4/19). That’s the pessimistic read. The optimistic read is that widespread social distancing measures across the country have indeed curbed the growth of the epidemic. For example, even though daily/weekly deaths are increasing, the rate of increase is much slower. Similarly, the proportion of tests that are positive is increasing at a much slower rate. Hospitalization data from CDC’s COVID-Net are even more promising with overall rates of hospitalizations decreasing (Figure 3). So all in all, I think we should be proud of what we’ve accomplished together during this disaster. And, we have a long way to go to come out the other side.

Figure 1. New York State’s Promising Trends

Figure 2. US Not-Quite-Promising Trends (excluding NYS data)

Figure 3. Weekly Rates of Hospitalizations across COVID-NET are Decreasing

Q&A for 4/19:

#Guidelines #Reopen

Question: The Administration recently issued “Guidelines: Opening Up America Again.” What do you make of them?

Answer: Thanks for asking this question as it forced me to read the new guidelines rather than the newspaper reports and twitter occasional snark I had been consuming! Since I imagine that most of you also haven’t read the guidelines, l’ve written up a synthesis (see bullets and Table 1 below). So quickly, what do I think? Well, I’m happy to see something rather than nothing! And it’s good to see that the plan is tied to metrics rather than random dates (Easter!). But, what’s fascinating and (in my opinion) depressing here is that NOTHING is the responsibility of the federal government in what is a national emergency. As someone who has grown up believing in the federal system, this abdication of responsibility is shocking. As you’ll see in the Guidelines, it’s each State’s responsibility to set up testing, contact tracing, and to assure that hospitals have necessary PPE. I’m not arguing that States shouldn’t have responsibility too, but where it the national partnership? What is the federal responsibility?

Next issue in these guidelines is that the testing strategy is still focused on symptomatic testing. We now know that transmission during the asymptomatic period is a huge issue (see Q&A of yesterday). So why are we still focusing testing on symptomatic individuals. We need more testing!! Third, and this has been talked about a lot in the last few days, these guidelines are not comprehensive. For example, what do Governors do if they are in Phase 1 and see a re-emergence? Do we move back to extreme social distancing? They are also lacking specifics. For example, what are schools and daycare centers supposed to do with kids with sniffles (as they always have)? Do they get sent home? Similarly, what degree of employer-based contact tracing is required? Finally, an ongoing issue, as our colleague Caitlin pointed out — how can we continue to expect people to stay home when sick or when possibly exposed when most folks in our country don’t even have sick leave?

Okay, I could say more, but I’ll stop there. If you want to see a set of guidelines that have more meat, check out Johns Hopkins Center for Health Security’s “Public Health Principles for aPhased Reopening During COVID-19:Guidance for Governors” that was just released.

Synthesis

• Reopening will be based on Governor’s decision, ideally in regional collaboration, and will be done in phases.
• Before considering reopening, Governors must see these criteria met:

1. A sustained downward trend in flu-like illnesses and covid-like cases over 14 days;
2. A sustained downward trajectory in covid-cases OR proportion of positive tests over 14 days;
3. Ability of hospitals to treat patients without crisis care; and
4. Robust testing program in place for at-risk healthcare workers, including emerging antibody testing.

• Before considering reopening, States must have in place:

1. Testing and Contact Tracing: Ability to quickly set up testing sites for symptomatic people and test such people; Ability to trace contacts of those who test positive; Use of sentinel surveillance sites to screen for asymptomatic infection and conduct contact tracing
2. Healthcare System Capacity: Ability to quickly and independently provide necessary PPE and medical equipment; Ability to surge ICU capacity
3. Have in place plans: to protect health and safety of workers in critical industries, of those living and working in high-risk facilities, of mass transit workers and users; to advise citizens of social distancing and face covering protocols; and to monitor conditions and take steps to limit and mitigate any rebound

• In all phases, individuals must: practice good hygiene and stay home if sick.
• In all phases, employers must: develop and implement appropriate policies (travel, social distancing, PPE, etc.); monitor workforce for symptoms and not allow symptomatic folks to come to work; develop and implement procedures for workforce contact tracing following positive test result of an employee.

Table 1. Reopening by Phase

Q&A for 4/18:

#Asymptomatic #Testing

Question: It seems like a lot of COVID-19 transmission may be from asymptomatic carriers. How do we know that we’re not already all infected? Wouldn’t we just be locked down for nothing? Why can’t we reopen?

Answer: First, you’re right on the transmission front. There is increasing data showing that the virus is highly transmissible during the initial, asymptomatic/pre-symptomatic period. We also have recent findings from testing on the USS Roosevelt that ~350 of 600 sailors who test positive did not have symptoms. Given these findings, a person would do well to ask whether the coronavirus is far more contagious and far less deadly than originally thought — with most folks serving as unknowing viral hosts and only a small proportion of folks having serious symptoms and outcomes. We’ll that’s where it’s important to have time-trend data — like a prospective or retrospective study. Here we’d want to follow people who are asymptomatic when tested positive and examine whether symptoms come later (and how much later). Happily, we do have a few such datasets now. And what they are showing is that about 70%-80% who originally test positive while asymptomatic go on to develop symptoms (for more data/discussion, see Q&A of 4/3). So, based on what we know now, it’s very unlikely that most of us have already unknowingly had the virus. We would have had some symptoms. We also would have seen far higher numbers of deaths across the country, especially in non-epicenter areas. So while we still don’t know the true R0 or case fatality of COVID-19, nothing that we currently know would suggest that we’ve all already been exposed — that we’re already safe — and that we can simply reopen.

And this brings us back to why it’s so important to practice social distancing until we have widespread testing — if folks who feel fine are walking around spreading the virus, then simply telling sick people to stay home is NOT going to stop the pandemic. It also brings us back to why widespread testing is so important — if transmission occurs during a/pre-symptomatic period, then waiting to test someone until they have symptoms is too late! Because our country has not yet brought widespread testing to the fore and does not have the requisite contract tracers, if we reopened now, we’d basically be back in the same position we were in during February. It’s exceptionally infuriating that these necessary components of the response — testing, contact tracing — remain unfilled, especially recognizing the toll social distancing is taking.

Q&A for 4/17:

#New York #Data

Question: What’s the New York data update of the week?

Answer: It’s been another week since we checked in on New York epidemic data (see Q&A of 4/10). If you’ve been following the news or watching Governor Cuomo’s daily press briefings, you’ll have a good idea of where things are. Still, it’s good to see some data and draw your own conclusions. So, here goes:

• Daily Number Hospitalized (Table 1) has plateaued (~17,000 — ~19,000 per day since April 7th) and the Daily Number of ICU Patients (Table 1) has also plateaued (~5,000 — ~5,200 per day since April 11th). For both measures, there is even a slight dip in the last two days, but it’s still too early to call it the dip a decreasing trend. This continues to be good news! Of course, there are some caveats to this good news (discussed below).
• Proportion of Hospitalized Patients in ICU (Table 2) is increasing from 23% to 28% over the last month. Additionally, the proportional increase has continued even as hospitalizations have plateaued. Now, the increase hasn’t been that dramatic, but these data (plus stories I’ve read in the NY Times) does make me wonder — is there a bias in terms of who is coming to the hospital now? Could some of the plateau in hospitalizations be the result of only sicker people getting hospitalized? This could be both because sicker people are trying to avoid going to the hospital and/or because EMTs are being more selective in who to take to the hospital. On the other hand, perhaps it only means that more people are staying alive and in the ICU longer, which is supported by Table 3. Maybe it’s a combination of the two? So many questions…
• Daily Hospital Admissions (Table 3) seem to be decreasing. If we’re not just curbing, but are really blunting the epidemic, this is exactly what we would want to see. Now, it’s still too early to tell if this is a legit trend, but it’s promising nonetheless!
• Number of New Cases remains high (Table 4). New York is still seeing a large number of positive cases everyday. Just yesterday, 8,505 cases were identified. representing 35% of everyone tested. Despite New York’s successes in conducting 20,000+ tests each day, it’s still far from adequate, as described in our Q&A of 4/10. Additionally, even though we know we need to be testing more people, the daily number of total tests completed remains relatively stagnant.
• Daily Proportion of Positive Cases among All Tested may be decreasing (Table 5). As Table 5 shows, while the overall proportion of positive cases among all tested remains exceptionally high at 40%, it does seem that we may have started seeing slightly lower proportions of positive cases in recent days, which is modestly positive. I mean — yesterday’s proportion of 35% positive is nothing to write home about, but I suppose I’m feeling optimistic and hopeful that we’re seeing a declining trend. I might feel differently tomorrow. Yesterday, The Atlantic published a good discussion on this metric if you want to learn more.
• Daily Deaths remain high (Table 6). With 630 deaths yesterday alone, the death toll in New York is staggering. It was >700/day for more than a week, so getting into the 600s is better… But damn, it’s terrible. And we know that it’s an undercount.
• Crude Fatality Rate is increasing (Table 7). I guess I’m leaving this email with bummer news. In addition to the high numbers of deaths, the NYS case fatality rate is also increasing. Now there are lots of reasons this could be. Clearly, based on Table 5, we’re missing a lot of cases in our counts. So, this would inflate case fatality rates. But, we also know that we’re not capturing all COVID-related deaths, which would deflate case fatality rates. If we’re able to really expand testing to the degree necessary, we should start to see case fatality decline (our denominator gets bigger).

Table 1. Daily Number Hospitalized and in ICU in NYS has Plateaued

Table 2. Proportion of Hospitalized Patients who are in ICU is Increasing

Table 3. Daily Hospital Admissions Are Declining

Table 4. Daily Number of New Cases Remains High; Number of Daily Tests has Plateaued

Table 5. Proportion of Tests that are Positive Remains Very High

Table 6. Number of Deaths is Dramatically Increasing

Table 7. Case Fatality Rate is Increasing

Q&A for 4/16:

#Reopening #Social Distancing

Question: What are the hard and fast requirements for reopening society and the economy? Is chatter about reopening premature? What will reopening look like across the US? What will have to be in place and what will be required of individuals?

Answer: So many good questions that so many people are trying to answer. I think we should be talking openly and honestly about all of these things and we should be working on evidence-informed plans for reopening now! Actual reopening now would be far too premature, but planning for it now is a necessity. To dive into your other questions, let’s first remember why we are social distancing, then let’s talk about what’s required to ease up on social distancing, and finally what reopening might look like.

Why Social Distance?: The big goals of social distancing are:

1. To curb and ideally stop community transmission (e.g. stop the spread of infection from unknown exposure);
2. To get our health systems prepared (e.g. needed supplies, equipment, and personnel in place);
3. To get our public health system working at massive scale (e.g. ready for widespread testing, contact tracing, suspected case isolation, surveillance)

How do we know when we can ease social distancing?: The short answer to this question is that we ease social distancing when the above three goals are achieved. Ideally, Goals 2 and 3 will be achieved at national scale while Goal 1 will be achieved on a community-by-community basis. If you’re curious about how we might measure success in achieving goals 1, 2, and 3, see the longer subsection below. Ultimately, we’ll be past social distancing when we have a safe and effective vaccine that is rolled out across the population. An effective and widely available treatment will also help towards reopening.

What might reopening look like?: The goal of reopening is to get us back to functioning while stopping community transmission from starting back up again. To do this, we need to: a) only start reopening after we’ve achieved the three goals listed above; 2) start reopening on a slow basis so that our public health systems can quickly identify new cases, identify all contacts of those new cases, and ensure that these folks are self isolating; 3) as things improve (we keep new cases and community transmission abated), we expand reopening. Note: moving from social distancing to expanded reopening will not happen on a national-level — we’re just too big — and will instead happen at state and local levels. Additionally, we’ll likely find ourselves in a dance — moving between expanding and constricting reopening and possibly even an/other period(s) of social distancing.

What does success for Goals 1, 2, and 3 look like? (a non-exhaustive list of several key indicators)

1. To assess whether we’ve curbed/stopped community transmission, we’d expect to see:

• Decrease in number of new cases: This would require wide-scale testing, which we do not have; In New York, where we do have much wider scale testing than elsewhere in the country, we are seeing a curb in community transmission with the number of new cases being quite similar each day.
• Decrease in the proportion of positive tests: If we keep testing criteria as stringent as it currently is in most places (based on CDC guidelines), then we might want to look at the proportion of tests that have positive results. We’d want to see that proportion decline over time; In New York, even as testing has expanded, we still see that about 40% of tests are positive.
• Decrease in number of new COVID hospitalizations and daily total of hospitalized COVID patients: In absence of testing data, we would look to hospitalization data; In New York, we’re seeing that social distancing successfully curbed growth in hospitalizations, but we’re still seeing ~2,000 new COVID hospitalizations a day and about 18,000 patients hospitalized each day, which is a plateau.
• Given the above, I’d argue that New York has curbed community transmission, but has not stopped community transmission.

2. To assess whether our hospitals are prepared, we’d expect to see:

• Supply chains effectively working with systems in place to address supply shortages (e.g. in context of limitation, systems effectively and transparently share resources) — Right time, right place, right quantity, right quality supply chain performance measures could be used. In absence of these measures, we’d base success on hospital and gubernatorial reports.

3. To assess whether our public health system is prepared, we’d expect to see:

• A higher per capita testing rate, representing widespread testing using more relaxed criteria than CDC’s current criteria — along the lines of South Korea or even Italy (see chart below; much of the rise in US per capita testing is thanks to New York). As I mentioned in our Q&A of 4/10, epidemiologists suggest that testing should be so widespread that only 1 of every 10 tests is positive.
• Contact tracing in place to find all known contacts of each positive case and ask those contacts to self isolate. This would be based on number of workers hired into these roles. Estimates of need range from 100,000 to >300,000.
• Existing surveillance systems providing real-time data on coronavirus incidence and spread.

see: https://ourworldindata.org/covid-testing

Q&A for 4/15:

#Sensitivity #Specificity #Antibody Tests

Question: The other day when you brought up the issues related to “immunity certification”, you mentioned a couple of “science” questions related to antibody testing — “What is the sensitivity and specificity of the tests? How comfortable are we with people being told they are immune and they actually aren’t?” Would you please elaborate for us non-public health people?

Answer: Oh, thanks for asking! Sensitivity and specificity are epidemiology terms that sound more complicated than they are. The concepts are about measuring the precision of a diagnostic test. Sensitivity is the ability of a diagnostic test to capture everyone who has the condition. Specificity is the ability of a test to identify everyone who does NOT have the condition. As sensitivity of a test increases, specificity decreases and vice versa — it’s a trade-off between the two. Copied below is a little table they use to teach us these concepts in epidemiology class (Table 1).

Now, just because a test has high specificity does not mean it’s a good test. You could have a totally bogus test tell *everyone* that they are positive for a disease and the test will be 100% sensitive — it captures everyone who actually has the disease, but it’s worthless because it tells *everyone* they have the disease. So, an acceptable test must be both reasonably sensitive and specific. But there’s yet another sticky wicket here. The lower the prevalence of a given condition, the higher the number of false positives we’ll find (see Table 2). In the hypothetical examples I’ve made up in Table 2, if 10% of a population of 1,000 actually has COVID antibodies, the antibody test will incorrectly identify 90 people as having antibodies (false positive). But if 50% of a population of 1,000 actually has COVID antibodies, only 50 people will be incorrectly identified as having antibodies. Since scientists still believe that a preponderance of our population has NOT been exposed to COVID and therefore does not have COVID antibodies, using antibody tests to issue “immunity certifications” could be exceptionally problematic as a number of people will mistakenly believe that they have markers of immunity that they actually do not have. Imagine all these people taking fewer health precautions, visiting nursing homes, and spreading disease. So many problems!

All that said, antibody testing can be extremely beneficial on the population-level rather than the individual-level. For example, we want to know how far the coronavirus has spread, so we need to know the proportion of people who have been infected. This is a very different use than the individual-level use that folks supporting “immunity certificates” are promoting. Finally, if you want to learn more, NPR had a really good write up on this published just today! Also, if this kind of stuff excites you and you want to learn more, check out Johns Hopkins free open course, Fundamentals of Epidemiology.

Table 1. Sensitivity and Specificity

Table 2. Put into Practice

Q&A for 4/14:

#Risk factors #Obesity

Question: What are the risk factors associated with severe COVID-19 outcomes and how does obesity fit into the picture?

Answer: As time goes by, we have more and more information about risk factors related to severe outcomes. Just a few days ago, CDC issued a report that provides US-specific information. This report examined the characteristics of 1,482 patients hospitalized with COVID-19 in 14 states during the month of March. Of the hospitalized patients, 75% were aged 50+ and 55% were male. Unfortunately, data on underlying conditions were available for only 178 (12%) of these patients. That’s a big limitation — we don’t know how representative these patients are among all 1,482 hospitalized. Nevertheless, among these 178 patients, 89% had one or more underlying conditions — the most common were hypertension (50%), obesity (48%), chronic lung disease (35%), diabetes (28%), and cardiovascular disease (28%). The authors state, “These findings suggest that older adults have elevated rates of COVID-19–associated hospitalization and the majority of persons hospitalized with COVID-19 have underlying medical conditions.” These US-based findings largely align with other findings out of China, with the exception that the China studies did not, to my knowledge, explore obesity.

On the obesity question, let’s turn to a study from France that was also published a few days ago. The study examined the relationship between obesity and the severe outcome — mechanical ventilation — among 124 consecutive COVID-19 patients admitted in intensive care in a single French center. Researchers found that obesity was strongly associated with the need for mechanical ventilation regardless of age, diabetes, and hypertension. ICU patients with BMI >35 kg/m2 (severe obesity) were 7x more likely to need mechanical ventilation compared with ICU patients with with BMI <25 kg/m2 (normal weight). The authors concluded, “Obesity is a risk factor for SARS-CoV-2 severity requiring increased attention to preventive measures in susceptible individuals.” As shown in both the CDC and French studies, initial data suggest that obesity is an independent risk factor for severe COVID-19 outcomes. But, given the small size of both studies, more research is needed, including research to understand the biological mechanisms behind the observed relationship.

On a related note, the NY Times article you cite in your question included interesting observations from NY doctors that placing patients on their sides may help with oxygenation and reduce the need for mechanical ventilation. It would be great to have more science to back up this observation. In the meantime, it does align with ways that people with COPD are taught to reduce shortness of breath (described in Q&A of 4/9).

Q&A for 4/13:

#Immunology Certificates #Antibody

Question: I’m hearing a lot of chatter about “immunology certificates” being a ticket to economic recovery. What’s the thinking on this front?

Answer: I’ve been hearing more about this too lately. The basic idea is — folks who have developed antibodies to SARS-CoV-2 are presumably immune to the virus. These immune folks are ideally suited for helping to care for the sick, for taking on essential positions that are at increased risk of exposure, and for helping to get our economy revving again. On first blush, it sounds pretty good. Dig a bit deeper and it becomes a worrisome and troubling road. There are societal, scientific, and ethical reasons to worry. Here are a few examples of each challenge, but this is not an exhaustive list:

On the societal front, creating a new “in” and “out” group is exceptionally problematic given human nature and existing inequities.

• For example, just yesterday, The New York Times published this fascinating opinion, “The Dangerous History of Immunoprivilege,” from scholar, Ms. Kathryn Olivarius, describing the history of Yellow Fever in New Orleans where “immunocapital”widened the economic divide. “Black people, with limited access to health care, were of course as scared of yellow fever as anyone else. But those enslaved people who’d acquired immunity increased their monetary value to their owners by up to 50 percent. In essence, black people’s immunity became white people’s capital.”
• Meanwhile, behavioral scientists have noted that social cohesion will suffer. Robert West, a professor of health psychology at University College London, told Wired, “There’s so much evidence on ‘in group’ and ‘out group’ work that, even when you set up arbitrary ‘in groups’ and ‘out groups’, people become quite tribal.” And Adam Oliver, a behavioral economist at the London School noted, “The whole approach might also undermine the message that we are all in this together, which is crucial if we are going to get through this relatively quickly.

On the science front, this article in STAT News lays out a number of the issues. Even if we decided the societal risks were worth it, we still have so much to learn before we know whether this proposed approach would even work!

• Does antibody presence confer COVID-19 immunity? Does immunity require a given level of antibody presence? How long does protection last? Is it protective even in high concentrations of SARS-CoV-2?
• How well do antibody tests perform at detecting COVID-19 antibodies? What is the sensitivity and specificity of the tests? How comfortable are we with people being told they are immune and they actually aren’t?
• How will we differentiate people with active infections who exhibit antibodies from those who are recovered and have antibodies? What do we make of recovery in light of reports of virus reactivation? (per Q&A of 4/11)

On the ethical front:

• How can testing be rolled out in an equitable way without further straining the health system? If we rely on folks who tested positive and recovered, then we are privileging those who were able to get tested. If we rely on antibody tests, will that focus take away resources from the existing testing, tracking, quarantine approach epidemiology recommends? Just today, Kaiser News reported that the town in Colorado that was experimenting with population-based antibody testing (mentioned in Q&A of 3/30) had to stop because the resource pull on existing laboratories in terms of PPE and supplies was just too much.
• What are the perverse incentives for tying economic engagement/activity to immune status? How many people will aim to infect themselves for a chance to work?
• What are the ramifications — moral, legal, etc. — of document falsification? Of falsely believing you are immune?

Q&A for 4/12:

#DC #TN #KY #MD #NY #USA

Question: I don’t suppose you’ve created a similar set of graphs for DC as you have for NY? Just interested in seeing where DC is in terms of the curve.

Answer: Alas, I do not have the same level of detail for DC as I have for NY. But thanks to your nudge, I did go ahead and create a few sets of graphs for DC, MD, TN, and KY (TN and KY because that’s where my parents live) using data from covidtracking.com. I’ve enclosed the tables below — along with tables for NY and USA — but haven’t included any exposition. If you have specific questions, let me know. And if are curious about Maryland, check out my friend’s analyses (Dr. YJ Choi) that offer far more insight than the charts herein.

DC Since I do not have daily hospitalization numbers and since testing has been relatively flat, I do not know where we are at in the curve.

Maryland

Tennessee

Kentucky

New York

USA

Q&A for 4/11:

#Reactivation #Infection #Transmission

Question: The South Korea CDC stated earlier this week that COVID-19 may “reactivate” in cured patients. How much of an issue do you think this is?

Answer: This issue of reactivation has come up a bit in the last few weeks. Back in our Q&A of 3/30, I shared an NPR story about four Wuhan residents who had recovered from COVID-19 and tested negative for the virus, but then tested positive several days later. There’s also a new article coming out of Zhongnan Hospital in Wuhan that explores of 55 patients who had been hospitalized with laboratory confirmed SARS-CoV-2 pneumonia. Of those patients, 5 (9%) who were discharged presented with SARS-CoV-2 reactivation. The authors state, “The reactivated patients included 1 asymptomatic patient and 4 symptomatic patients, which suggests the reactivation potential of asymptomatic or minimally symptomatic patients. The time from SARS-CoV-2 negative to positive ranged from 4 to 17 days, suggesting that recovered patients still may be virus carriers and require additional round of viral detection and isolation.” The authors did not find any clinical markers that were predictors of reactivation. The WHO confirmed today that it is investigating these reports, stating, “We are aware that some patients are PCR positive after they clinically recover, but we need systematic collection of samples from recovered patients to better understand how long they shed live virus.” Ugh! Yet another COVID-19 mystery to explore. I have no idea how big of an issue this could be, but the initial findings are concerning.

Finally, if you want to understand more about how viral reactivation works, here’s a scientific overview and here’s a quick synthesis — Basically, viruses can go through three different phases of replication — lytic (super productive replication), latent (dormant phase, which is uncommon for most human viruses), and persistent (slow replication that does little damage to host cells). Most human viruses replicate in lytic phase and a few viruses, like hepB are persistent. Herpes is the most common type of virus that has a latent phase. I am not aware of any research showing that coronaviruses have latent periods. I’m glad to know WHO and others are investigating!

Q&A for 4/10:

#New York #Deaths #Hospitalizations

Question: How’s our weekly New York check-in looking?

Answer: Glad you asked! We checked out New York’s data back in last week’s Q&A of 4/2 where we saw some glimmers of hope — the rate of hospitalization increase was slowing. We also saw some unfortunate trends — the daily numbers of deaths and the case fatality rate were steadily increasing. It’s one week later with another week of social distancing under our belts — about 3 weeks total of widespread social distancing in New York. So, let’s check back in with our three tracking indicators — daily # hospitalized patients, daily # icu patients, daily # deaths — and one more visualization for good measure.

First some good news. As you’ll see in Figure 1, the daily number of patients hospitalized and in ICU is growing at a much slower rate. You can really see the curves leveling off. This would be what we’d expect after three weeks of social distancing! Yay! The numbers hospitalized are also much lower than multiple models had predicted. Double yay! What we’re doing seems to be working!

Now some bad news. As you’ll see in Figure 2, the number of daily deaths is high — a dispiriting trend. To date, at least 7,844 New Yorkers have died due to COVID-19 — and this is all within about a 3 week period. Meanwhile, the case fatality rate is still increasing by the day (Figure 3). As of 4/10, the case fatality rate is: 4.6%. In other bad news, New York City is reporting much higher numbers of deaths at home — 258 deaths on Monday compared with a typical of 25 — that are likely attributable to COVID-19, but not yet captured in reporting data. This news also presents a challenge for understanding hospitalization numbers — are sick people too afraid to go to the hospital? If so, that could be one bad reason that we’re seeing hospitalization and ICU patient numbers leveling-off. That’s mere speculation and I’ll have to dig further into that issue for a different Q&A. The fact that deaths are still quite high while hospitalized patients and ICU patients level off is to be expected as deaths are a lagging indicator (e.g. you would only see deaths begin to level and drop several days to weeks after hospitalizations level and drop).

Because I don’t want to end with bad news, let me close with mixed news (because nothing can be straightforward with this damn virus!) The good news is that New York is now testing at a higher rate per capita than anywhere else in the United States, even more than South Korea, and the daily testing continues to increase. Yesterday, New York conducted 26,336 tests! The bad news is that 40% of tests are still coming back positive (Figure 4). This means that we are still far from conducting enough tests to identify everyone infected. Epidemiologists suggest that a rate of 1 positive test for every 10 tests administered represents a benchmark of a testing system that is able to capture all infections. This type of system is especially important for communities that are easing social distancing — such communities require extensive testing, contact tracing, and case quarantine in order to minimize the risk of community transmission.

Figure 1: Daily Number of Hospitalized and ICU Patients in New York State Is Increasing at a Slower Rate

Figure 2. Daily Deaths Due to COVID-19 in New York State Have Exceeded 700/day over the Last Four Days

Figure 3. Crude Fatality Rate in NYS is Increasing by the Day; 4.6% as of 4/10

Figure 4. NYS Testing Continues to Increase and Positive Cases Continue to Be a Large Proportion of All Tested (40%)

Q&A for 4/9:

#Shortness of Breath #Treatment #Movement

Question: Chris Cuomo said on his show the other night that if you get the COVID, you should not lie down for long periods of time because that allows fluid to collect in your lungs. That you need to get up and move around every few hours. Is that true? (makes a lot of sense to me since that’s what they told me after my c-section when I was retaining a ton of fluids) And if that is true, why aren’t the drs saying that to the public? There are a lot of people self-treating at home right now because their symptoms aren’t severe. (clip here and description here)

Answer: I’m not a doctor, as you all know! Please talk with your doctor if you or your family member becomes sick. Now that I’ve put that preamble out there, here’s the short answer: I haven’t found any research on the need to get up and move as a treatment for COVID-19. But I have seen evidence of how gravity can and should be used to help alleviate shortness of breath (aka dyspnea), which is a common symptom of COVID-19. Upshot is== science shows that movement can alleviate symptoms, but science does not show that movement will help with clinical outcomes; meanwhile for patients on ventilators in the ICU, science shows that lying in prone position appears to result in better clinical outcomes. Here’s the longer answer:

Most of the research I found is related to either a) best body position for those put on ventilator and b) body position for those with chronic lung conditions like COPD. When it comes to ventilator support, the preponderance of research supports the practice of keeping ventilated patients in prone position because it “can prevent ventilator-induced lung injury in acute respiratory distress syndrome (ARDS) patients receiving conventional mechanical ventilation and, hence, may have the potential to improve survival.” But that’s not our focus. What about body position and movement for folks who are self-treating for COVID-19?

Well, Salon interviewed a couple of pulmonary specialists on the topic of breathing exercises — but I think it applies to movement too — and their take-away is “The breathing exercises can help keep the lungs fully expanded as much as possible, but will it keep you from having worse symptoms? No, that’s really dependent on so many other intrinsic factors… There’s a lot of anxiety going around about all of this and taking the time to do mindful breathing certainly will help with that aspect and that may make someone feel better — even if physiologically it’s not making a big difference in the way the lung is working.” Anxiety is a big issue that Chris Cuomo and other COVID patients talk about, so in that space, taking some action, like breathing exercises and stretches, can really help. University of Maryland offers breathing exercises for COVID-19 patients, noting that “When you’re short of breath, anxiety can increase and make it even harder to breathe. These exercises can help you stay relaxed when you feel your symptoms escalate and even prevent shortness of breath from occurring in the first place.”

Another thing Chris Cuomo talks about is inability to sleep. Sleep is clearly very important for your body’s ability to mount an effective immune response to the virus. And if you aren’t sleeping because you can’t breathe well… that’s going to be a big problem. The Lung Institute offers a nice set of diagrams of various body positions to help patients who suffer from shortness of breath, which is one of the more concerning symptoms of COVID-19 too. I’ve copied one below for ease of reference. Additionally, COPD Support describes the benefits of various body positions to alleviate shortness of breath symptoms. If you can change your position to breathe better, you’ll likely be able to sleep better and this may help your body to be better equipped to fight the infection.

Q&A for 4/8:

#Testing #New York

Question: At the beginning of March, New York State was conducting 40–50 COVID-19 tests a day. By the beginning of April, NYS was conducting 20,000+ tests a day (see bar chart below). How did it manage to increase testing by such an enormous magnitude?

Answer: New York State has conducted the most tests per capita of any U.S. State and even more than South Korea. This success is attributable to the following factors:

1. Existing Capacity: New York State has strong laboratory capacity with 28 labs the New York Department of Health can call upon.
2. Leadership: New York State’s leaders were quick to recognize that testing would be a key component of the pandemic response and made strong calls for testing capacity early on in the epidemic.
3. Regulatory Approval: On 13 March, FDA authorized New York’s Department of Health to process thousands of coronavirus tests per day by contracting with private labs. This allowed the NY Department of Health to authorize its 28 labs to conduct tests without need to engage with FDA or get an Emergency Use Authorization. That same day, FDA also authorized automated testing.
4. Public-Private Collaboration: Meanwhile, the State of New York had been working on public-private partnerships with labs like BioReference with the aim of establishing drive-thru testing sites and increasing lab capacity. Once FDA approval was received, New York was able to quickly implement its plans.
5. Existing Capacity/Regulatory Approval/Innovation: Additionally, New York’s larger labs had the financial and technical capabilities to purchase and use automated testing equipment as it was approved by FDA. For example, Northwell adopted Hologic testing once it received FDA approval on 16 March, allowing it to test 1,150 tests/day.
6. Innovation: Finally, on 29 March, New York State’s Wadsworth Lab announced its new, less intrusive test for COVID-19, which allows testing to be done through a saliva sample and a self-administered short nasal swab in the presence of a health care professional, thereby reducing risk of exposure while increasing testing availability.

To curb the pandemic, testing must be wide-scale and population-based. New York’s testing began after community transmission had taken hold. Testing has thus far focused on identifying infection among those who exhibit illness. With 40% of New York’s tests coming back positive, more testing is needed to ensure that asymptomatic and pre-symptomatic infections are identified so that these individuals can self-quarantine, their contacts can be traced, and infection spread can be minimized.

Q&A for 4/7:

#Donations #Helping

Question: I get asked by people where to donate for COVID response. I have the international-focused groups to recommend, but do not have such recommendations for domestic organizations. Do you have any suggestions? How can individuals help?

Answer: Staying at home is one of the absolute best things we can do to help, but it feels (at least for me) very passive. So, let’s think about more of the active ways we can help, of which donating our time and money is of high priority! Here is a list of some ways to help, both financial and non-financial, though this list is in no way authoritative or complete. Please add more!

• Donate money: First, remember that many of the ills of the world that charities try to mitigate are amplified by the pandemic. The groups your regularly donate to still need your donations. If you’re able, please continue to donate. Meanwhile, some groups at the forefront of responding to the epidemic need additional financial assistance. These groups may not have been in your basket of charities, but you may want to consider adding them. They include (but aren’t limited to) groups that help:
• Feed people, like local food banks, Meals on Wheels, No Kid Hungry, Feeding America
• House people, like local shelters such as Coalition for the Homeless or Covenant House
• Support women and children who are experiencing violence, like local domestic violence organizations
• Support prisoners who are an especially vulnerable population, through bail fund relief or Parole Preparation Project
• Support undocumented immigrants who are an especially vulnerable population, through groups like the Nat’l Domestic Workers Alliance Coronavirus Care Fund
• Support artists who make our films/books/music/etc, through organizations listed here.
• Donate blood: Blood banks are facing severe shortages of blood. If you’re able, please consider donating.
• Donate supplies: A number of hospitals and other health organizations are accepting homemade personal protective equipment. Consider making masks or gowns. For example, Johns Hopkins is accepting homemade mask donations. Other groups that are still serving people need cleaning supplies.
• Donate space: If you belong to a house of worship, social club, or another organization with large space that is not being used, consider working with organization leaders to donate the space to expand community shelter or day facilities for the homeless while maintaining social distancing or even to expand hospital bed space.
• Donate expertise: In addition to donating medical expertise that is sorely needed, a number of other types of expertise are needed now too. For example, New Jersey needs programmers to help code an antiquated system. I have a friend donating her statistical expertise to track the epidemic (go, YJ!). Your expertise could be useful too, so please share it where possible!
• Provide emotional support: Consider writing letters to our elders or providing companionship. Or consider becoming a crisis support volunteer or remote mentor to a kid in need through groups like iCloudBe.

Q&A for 4/6:

#Tigers #Cats #Transmission

Question: This is so bizarre. If tigers can get it, is there a risk that domestic cats could be involved in spreading it? I had thought there was pretty high confidence that non-human animals are not playing a role in the pandemic at this point, but could that be wrong..?

Answer: This upsetting news adds yet another layer to the COVID-19 mystery and our collective anxiety. Plus, as my colleague Jessica points out, people may be extra devastated by the news having watched “Tiger King,” which is still #1 on Netflix. Frivolity aside, I had the same reaction as you, a real — whaaaaaa???!!! So, I turned to science to try to figure it out. And here’s what I’ve learned from my quick research. First off — in response to your question on cat->human transmission, we don’t know. But so far, it seems unlikely and please do NOT do anything rash like banish your cat! Second, we do know a few things, and here’s what I’ve quickly learned:

1. Cats can contract other types of coronaviruses, like feline infectious peritonitis. (see: Black’s Veterinary Dictionary)
2. During the SARS epidemic, it was found that domestic cats can contract SARS and can pass it along to other animals housed with them. This was in a laboratory setting, not in “real life” conditions.
3. More recently, again in a laboratory setting, it’s been shown that domestic cats can become infected with SARS-Cov-2 and can pass it between themselves. This pre-print was just released on the subject last week (it’s not yet peer reviewed and is quite small). The researchers exposed several domestic cats to high levels of the virus and put these cats in cages beside cats who were not exposed. The exposed cats subsequently became infected, as did the unexposed cats. Note: dogs were also included in this study and were found to have low susceptibility to infection.
4. Now we have this example of Bronx Zoo tigers who all exhibited COVID-19 respiratory symptoms similar to us humans. One of the tigers was tested for COVID-19 and was shown to be infected. The working hypothesis is that a zookeeper, who was asymptomatic at the time of working with the tigers, passed the infection to them. This is the first example of human->cat transmission that we have.
5. What we do NOT have is any evidence of cats spreading COVID-19 to humans. As folks like to say, an absence of evidence is not evidence of absence. But based on contact tracing done elsewhere, it seems very unlikely that domestic cats are spreading the virus to humans.
6. Cats can transmit other diseases to humans, but most all of those on the list are parasitic or bacterial. Cornell has a nice breakdown of cat->human transmission. To my knowledge, the only *viral* disease we have evidence of cat-> human transmission is rabies.
7. CDC has guidance on COVID-19 and pets. This guidance states, “It is still recommended that people sick with COVID-19 limit contact with animals until more information is known about the virus. This can help ensure both you and your animals stay healthy.”
8. Also, USDA just released a great Q&A on this topic that I recommend you check out.

Q&A for 4/5:

#Models #IHME #CHIME

Question: I’ve been hearing a lot about a couple of models decision-makers are using to forecast hospital utilization needs. I saw that UPenn’s CHIME tool shows that the apex for DC’s COVID cases would come around mid-June/July, but University of Washington’s IHME tool projects DC’s peak to be April 15th or so. I also saw that the DC government is favoring CHIME because it accounts for the fact that not everybody is going to practice social distancing. Would you discuss the pros/cons of each model or talk more broadly about what we should consider before taking these projections as gospel?

Answer: Dang, good question. First off, models are only as good as the assumptions and data that go into them. And in the context of COVID-19, we don’t have much data and our assumptions are still changing as we’re learning more about the new virus all the time. Threading this needle of highlighting health system gaps for program and policy decision-makers — as all of these models do — with reminding folks that the estimates are *not* to be taken as gospel — which none of these models should — is challenging to say the least. On this point, I recommend a couple of articles from The Atlantic and Fivethirtyeight.

All that said, I was curious about these models too, so I read the methodology of each (see here and here). I find the IHME model far more difficult to understand than the CHIME tool (note: I generally find IHME models difficult to understand…. and I was heartened to read that it’s not just me!) Enclosed is a short table I made describing the two models and highlighting some of their differences as I see them.

I’m not in a position to say which is better. But there are a few things that stand out. First, IHME’s model assumes that our social distancing measures will have the same impact on the spread of the virus (R0) as Wuhan’s drastic measures did. This is a lofty assumption and will likely downplay the risk of ongoing transmission, especially in states that have not implemented social distancing measures (I just read today that Georgia reopened its beaches!). By making this assumption, the model estimates that COVID-19 hospitalizations across the US will peak at some point in April. Given what we know of the slow roll-out of social distancing measures in large swaths of the country, especially the south, this estimate seems optimistic. Meanwhile, Susceptible, Infected, Recovered (SIR) models like the one used by CHIME are often limited by assuming homogeneous risk across a population, which can inflate the number of expected cases. The CHIME model partially overcomes this limitation by allowing users to modify the inputs based on local expectations/data. That said, the population susceptible assumption is still a challenge for this model, and may result in overestimates of health system needs.

Personally, I prefer the ability CHIME gives its users to modify the inputs/expectations, especially with regard to social distancing effectiveness. If you’re looking for estimates of deaths, however, of these two models, you can only get estimated deaths from IHME. And if you don’t have a good idea of the inputs for your region, you might prefer the IHME dashboard, which doesn’t require any user input. Finally, if you were wondering how CHIME compares to the oft-cited Imperial College model, check out this post CHIME modelers wrote on this very question (upshot== models preform similarly). Regardless of the model we choose to use, I think the main message is the same: COVID-19 cases will outstrip our hospital capacity, so we need to take measures now to prepare as best as possible for the worst and hope for the best.

Table 1. IHME and CHIME Comparison

Q&A for 4/4:

#Eyeglasses #PPE

Question: Now that we’re all supposed to be wearing masks out and about in the world, should we be wearing other PPE too?

Answer: Interesting question. My most recent Q&A on masks was on 3/31 and I talked about gloves on 3/26. In short, Masks== wear them; Gloves== it depends, but for activities like walking around the grocery store, experts don’t generally recommend them. So that brings us to eye wear. How about goggles? Or even eyeglasses for that matter?

First, let’s remember why we are wearing masks. It’s to prevent us from spreading the virus, not to keep us from contracting the virus. So does walking around with goggles or glasses limit spread of the virus? Scientists have been studying this very issue and so far, the virus does not appear to be spread via tears. That’s good. Wearing goggles will not likely help in terms of reducing our ability to spread the virus. So, no need to wear them for activities like going to the grocery store. But..

On the flip side, scientists also find that it is likely that the coronavirus can be contracted through the eye, which aligns with evidence that influenza can be contracted through the eye. Indeed, it was an ophthalmologist, Dr. Li Wenliang, who was the first to raise alarm about the new virus back in December and subsequently succumbed to the disease in February. Additionally, conjunctivitis seems to be a less common symptom of COVID-19 and other coronaviruses, which *could* suggest ocular transmission. Evidence is also mounting that the virus can both a) be spread through finer vapors — the very act of exhaling — not just larger respiratory droplets from coughing or sneezing; and b) may hang in the air of poorly ventilated spaces for several hours. I bring this up to say that if you have to be in close quarters for any reason — like riding the bus or caring for a sick family member — it would be beneficial to also protect yourself by wearing goggles.

When it comes to wearing eyeglasses as protection, I haven’t found any solid data on the issue. Ophthalmologists say that eyeglasses can offer a bit of a barrier — if someone sneezes, the glasses will help shield your eyes (I don’t think we need science for that one!). Plus, folks who wear eyeglasses have less reason to touch their eyes than those who wear contacts. For those reasons, they suggest switching from contacts to eyeglasses. But in terms of protection from airborne transmission that I described above, goggles would be a far better option since they offer full protection of the eyes rather than partial protection offered by glasses. For health providers, CDC recommends wearing a face shield or goggles. CDC does caution that goggles can fog, but there are lots of hacks to limit the issue of fogging — like using liquid soap! Finally, if you want to keep up-to-date on eye issues and COVID-19, the American Academy of Ophthalmology offers a great resource page on the topic.

Q&A for 4/3:

#Asymptomatic #Pre-symptomatic

Question: What’s the difference between asymptomatic and pre-symptomatic?

Answer: I’m glad you asked because I see the word “asymptomatic” used far more than “pre-symptomatic” when it comes to characterizing folks who test positive for COVID-19 but do not have symptoms. And there is an important difference. Asymptomatic folks are those who test positive for the virus but never exhibit or experience symptoms. Meanwhile, pre-symptomatic folks are folks who exhibit symptoms some time after they test positive. A== without Pre==before. Just today, WHO epidemiologist, Maria Van Kerkhove, said that data out of China suggest that 75% of patients originally listed as asymptomatic go on to develop symptoms. ProPublica reports that “This matches up with the CDC’s findings at the nursing facility in Washington. Of the 13 positive patients who initially reported no symptoms during testing,10 later developed symptoms.” For the CDC analysis itself, see this fascinating MMWR report that CDC just published today. All this to say, when Iceland, for example, finds that 50% of those tested via population-based testing report being asymptomatic, it does not mean that 50% of people who contract COVID-19 never have symptoms. Instead, what scientists are finding is that the overwhelming majority of people who contract COVID-19 go on to develop symptoms. But one of the extra-challenging components of COVID-19 is that there seems to be a pre-symptomatic phase of infection where a person does not feel sick, but is able to pass the virus on to others. The ProPublica article I cited above gives a nice run-down of the issues, so check it out if you want to know more!

Q&A for 4/2:

#Data #Monitoring #New York

Question: Last week you described measures we would observe to see if our social distancing was working. Now that a week has passed, what do those indicators show?

Answer: Thanks for remembering! In my opinion, it’s still too early to tell whether social distancing is working. But since we have more data now, let’s take a look! Back in the Q&A of 3/25, I described three indicators we should be tracking to understand how successful we’ve been in containing the epidemic. Those measures are:

• Number of patients hospitalized (daily, weekly, cumulative) [Update: I originally wrote admissions, but I think daily hospitalized patients is a better measure]
• Number of ICU patients (daily, weekly, cumulative); [Update: I originally wrote admissions, but I think daily ICU patients is a better measure]
• Number of deaths (daily, weekly, cumulative)

Thanks to Governor Cuomo’s daily press briefings, we have these data points for New York. First, looking at Figure 1, we see that the number of hospitalized patients, ICU patients, and deaths continues to increase on a daily basis and has increased unabated since reporting began earlier this month. That’s bad news, but it is not unexpected news.

Figure 1. Daily Number of New York Hospitalized Patients, ICU Patients and Deaths Continues to Increase.

Now look at Figure 2, which shows the rate of increase in hospitalizations. Good news! The growth rate (e.g. doubling time) has slowed since March 27th. This is an indication that social distancing measures — put into place state-wide on March 18th (educational facilities closed), March 20th (non-essential biz closed), and March 22nd (mandatory stay-home order) — seem to be working. If these trends hold, it means that we are slowing the pace of the epidemic. All that said, I’m still not sure how much I’d hang my hat on that assessment. Based on findings from China (described in Q&A of 3/29), we expect folks to know they are sick within about 5 days of infection and to seek medical treatment (hospitalized care) within 7–13 days of symptom onset. So, presumably, we’d start seeing growth rate in hospitalizations begin to slow about 15 days after social distancing begins. If we start the clock at March 18, then we wouldn’t expect to see growth rates in hospitalizations slow until on/about April 2nd, and that’s a pretty early estimate.

Figure 2. The rate of increase in hospitalizations has slowed; doubling time is now estimated to be 6 days rather than 3.

Now, if we look at New York’s deaths in a cumulative way (rather than daily as is shown in Figure 1 above), we see that the increase continues along an exponential trend — per Figure 3, look at that steep slope since late March! 😞 Institute for Health Metrics and Evaluation’s COVID-19 model estimates that New York won’t hit it’s peak daily number of deaths for another 8 days. And if we look at crude mortality rates over time, we find them getting worse (Figure 4), which is obviously not a good sign.

Figure 3. Cumulative COVID-19 Deaths in New York are Increasing at an Exponential Rate, totaling 2,372 as of April 2nd.

Figure 4. Crude Mortality Rate in New York State Seems to be Worsening. Today it is 2.6%.

Finally, looking at how New York data compare with the data coming out of China, per the WHO report, we find that clinical outcomes are quite similar. [Note: Given how limited testing is in other areas of the US, I wouldn’t try this analysis on a national-level. Since NY has been doing such a comparatively good job of expanded testing, it offers a good opportunity to check.] Among positive cases in New York, about 20% are hospitalized (as of April 2, 92.3K total positive, 19.5K total hospitalized), which aligns with what China experienced too. Of those who are hospitalized, about 25% are in intensive care. Put another way, about 5% of all cases in New York thus far have required intensive care. And as to mortality rates, we’re finding a crude rate of approximately 2%. All of this aligns with China’s data. But if the trend in crude death rate continues, we will be worse off than China — either a reflection of fewer tests conducted among folks with minor illness, a reflection of a faltering health health system, or a combination thereof.

Q&A for 4/1:

#Drugs #Treatment #Clinical Trials

Question: A number of drugs have recently been hyped by numerous groups, the President included. How much hope should we have for drugs like hydrochloroquine?

Answer: Scientists are racing to try and find effective treatments for COVID-19. There are so many different potential treatments being floated, it’s hard to keep track. WHO has identified the 4 most promising drugs — a) remdisivir, b) chloroquine and hydroxychloroquine, c) ritonavir/lopinavir, d) ritonavir/lopinavir and interferon-beta — and recently launched a global trial to test them. These drugs were chosen for WHO by a panel of scientists who assessed the evidence and selected the drugs that have the highest likelihood of working, the most safety data from previous use, and are the most likely to have supplies sufficient to treat lots of people if the clinical trial(s) show they work. I’ve included below information about each of those four drugs.

In times of crisis, it’s especially important to be hopeful. But right now, we just don’t have enough solid information to know which of these drugs, if any, will work. And as with most drugs, these drugs have side effects, which means that it’s imperative to talk with your doctor before making decisions about using the drugs and/or joining a clinical trial (if that’s what you were thinking!). Finally, every clinical trial includes a data safety monitoring board that periodically reviews data on patient outcomes to assess whether the trial can continue. Simplistically — if a given treatment shows big improvements, the trial will be prematurely stopped so everyone can get the treatment. And if a given treatment shows big detriments, the trial will be prematurely stopped and folks will stop getting the treatment.

• Remdisivir is an antiretroviral drug that is taken intravenously. It is supposed to work by inhibiting viral reproduction. And it has shown promise in animal studies for treatment of other coronaviruses — MERS and SARS. It was also tested in humans as an Ebola treatment. While remdisivir did not prove to be effective for treating Ebola, it did prove to be safe for humans to use. Back in February, NIH started a drug trial of remdisivir for treatment of COVID-19. China is also conducting clinical trials of the drug and other avenues of investigation are underway. NIH’s study is scheduled to be completed today, but data from the study won’t be available until likely later this month. Chemical and Engineering News had an informative story on this topic that you might be interested in reading. In short, while scientists are hopeful, they also caution that remdisivir will likely be most effective for patients who are in early stages of the illness progression and because most patients enrolled in the trials will be at later stages of illness, the effect may be minimized.
• Chloroquine and hydroxychloroquine are old malaria drugs that are currently used to treat lupus and rheumatoid arthritis. They are taken orally. Studies have shown that these drugs can block coronaviruses (SARS) from infecting cells based on in-vitro science (e.g. lab-based science in the test tube). This article in Scientific American gives a good overview of the drugs, which are currently being tested for treating and preventing COVID-19 in clinical trials. Results from a very small, non-randomized study of 20 people treated with hydroxychloroquine daily in France showed that those who were treated had less virus in their system at treatment day 6. Anecdotal reports from China have also been positive, but to my knowledge, no scientific data supporting the claims have been published under peer review. [note: yesterday, a pre-publication (e.g. not yet peer-reviewed) was presented, which did show promising results based on a small study of 62 patients] In absence of better science, we just don’t know. Plus, both drugs do have known side-effects and risks. So, until we know whether these drugs work, and in case you were thinking about it — do NOT take them as prophylaxis! And if you are sick with COVID-19, talk with your doctor about the pros/cons of use.
• Ritonavir/lopinavir are HIV antiretroviral drugs that are taken orally. Lopinavir is a protease inhibitor that prevents viral replication and ritonavir is another type of protease inhibitor that works to enhance lopinavir. Lopinavir was shown to be effective against MERS in lab and animal tests. Initial results from a COVID-19 treatment study in China were, however, disappointing as they showed no statistically significant difference between intervention and control arms. That said, the study was underpowered (not enough people for statistically significant differences to be identified) and the direction of the differences (ex: fewer deaths in intervention arm) did show that this drug combo is likely worth exploring in large trials, which the WHO is doing. If you’re interested, this editorial in New England Medical Journal gives a good overview of the China study. Like all drugs, these drugs also have side effects. In the China study, lopinavir–ritonavir treatment was stopped early in 13 patients (13.8%) because of adverse events.
• Ritonavir/lopinavir and interferon-beta are the drugs described in the bullet above plus interferon-beta, which is a polypeptide that has immunomodulatory properties (e.g. enhances/modifies response of immune system) and is used for treatment for multiple sclerosis. This drug combination is given intravenously and has been shown to improve pulmonary function in the context of MERS. Most of what I wrote in the above bullet applies here too. Some Chinese doctors have been using a Cuban interferon drug called Alpha 2a, which they are speaking highly of, but I do not believe any data has been provided in the scientific literature yet.

Q&A for 3/31:

#Face masks #N95s #Homemade

Question: I’ve been reading a lot lately about face masks. Should I be wearing one when I go out?

Answer: This issue has evolved so much over the last few weeks, including since I wrote about face masks in my Q&A of 3/20. So, what’s the current word? First, guidelines vary dramatically by country. Here in the United States, CDC’s guidance remains the same as I described it back on 3/20. However, CDC is now reviewing its guidance in light of increased evidence of substantial asymptomatic and pre-symptomatic spread. As a reminder, CDC’s current mask-wearing guidance advises ill people to wear masks, but does not advise well people to wear them. Such guidance is perplexing in the context of COVID-19 — when a person is asymptomatic or pre-symptomatic, they would incorrectly classify themselves as a “well person,” and would then be walking around without a mask, unknowingly spreading illness. Upshot here is== when it comes to protecting yourself, your family, and your community, there’s absolutely nothing to lose in wearing a homemade mask to go out into the world (except for maybe some side-eye at the grocery store, but in this crisis, who cares!?!) and potentially lots to gain in terms of reducing virus spread. Now, there’s one big caveat here — with N95 masks and surgical masks in short supply, we need to save those for the doctors and nurses. This means that we need to make and use homemade masks (more on that below).

Question: I want to help! What’s the best way to make masks for myself, my family, doctors and nurses? Can surgical masks I make help health workers even though they aren’t N95s?

Answer: Awesome! I love the can-do attitude. And with personal protective equipment in short supply, every bit we can do can help. Unfortunately, we don’t have guidelines from CDC or other federal health officials for how best to make masks for health workers and others. I have seen LOTS of directions online, like this DIY published in the New York Times today and this post from Vanderbilt University Medical Center, which also offers step-by-step instructions. Good news is — for the most part, surgical masks appear to be nearly identical to N95 masks when it comes to health worker protection against respiratory illness. More specifically, a literature review from Center for Evidence Based Medicine updated just yesterday states, “Standard surgical masks are as effective as respirator masks (e.g. N95, FFP2, FFP3) for preventing infection of healthcare workers in outbreaks of viral respiratory illnesses such as influenza. No head to head trial of these masks in COVID-19 has yet been published, and neither type of mask prevents all infection. Both types of mask need to be used in combination with other PPE measures. Respirator masks are recommended for protection during aerosol generating procedures (AGPs). Rapid reviews on wider PPE measures, and what counts as an AGP, are ongoing.” This means that the masks you make at home — if well made — can do the job! Now we just need CDC on the case to give us the best guidance for how to make the most effective masks!

Q&A for 3/30:

#Immunity #Antibody #Test

Question: My brother believes he had COVID-19 and got over it. Like so many others, he wasn’t able to get tested. Can he be tested now to find out if he had it? And if he did have it, would he now be immune to the virus?

Answer: Scientists have developed a new antibody test for COVID-19 that will allow for folks like your brother to know whether they were infected with SARS-CoV-2. The tests use an individual’s blood to identify whether their immune system came into contact with the virus. Specifically, the tests check to see if a person has immunological markers — antibodies to SARS-CoV-2 — in their blood. These antibodies are the proteins the immune system makes to neutralize the virus. Mount Sinai in New York has developed one such test. Other such tests are being developed and deployed in Asia and Europe and even a small town in Colorado. I do not know when the tests will be available for broad public use. The beauty of these tests is that they can:

• Give us a better picture of the COVID-19 infection rate;
• Identify individuals, especially health workers, who have already been exposed to the virus and are therefore most likely immune and able to safely go back to work. This is especially important for easing health workforce burdens.
• Help get the economy going as folks who exhibit immunity can presumably stop social distancing;
• Identify folks who can donate blood plasma for clinical trials of potential therapeutic treatments.

Generally, the presence of antibodies confers immunity. But we are still not sure how long immunity lasts and we are making that generalization based on our understanding of other infectious diseases. We are hopeful it’s the case for COVID-19 too. For more on this issue, please see Q&A of 3/19. What’s one of the more perplexing findings of late is that there is a small group of health workers from Wuhan who had COVID-19, recovered (felt better and tested negative) and are now testing positive again. NPR had a fascinating story on the subject over the weekend. I’m not sure what this means (if anything) for immunity. If nothing less, it highlights just how much we still don’t know.

Q&A for 3/29:

#Clinical Progression #Time to Hospitalization #Exercise

Question: What is the timing of the clinical course progression of COVID-19. How long between becoming infected, becoming sick, needing hospitalization, etc?

Answer: Like everything else COVID-related, we are still learning. Here’s a brief timeline based on what we currently know, which is limited (so this is likely to change as we learn more):

Average Time [all of these estimates are based on data from Wuhan, China]

• Infection to symptoms: 5 days (most symptoms appear between 4–7 days)
• Symptoms to doctor visit: generally 5–6 days
• Symptoms to hospitalization: 7 days, 9 days, 12.5 days (depends on data source/study population)
• Symptoms to ICU admission: 10 days (range: 6–12 days)

We know from data from Wuhan, China that the average time from infection to symptom onset is 5 days with the preponderance of symptoms occurring 4–7 days after infection. We also estimate — again, based on the first 425 confirmed cases in Wuhan — that time from illness onset to first doctor visit is about 5 days, and the average duration of illness onset to hospitalization rages generally from 9 to 13 days. Other data presented back in February from one hospital in Wuhan showed that the average time from symptom onset to hospitalization is 7 days (range: 4–8 days) and average time from symptom onset to ICU admission is 10 days (range: 6–12 days). A recent paper published in Lancet tracking the clinical progression of the first five patients diagnosed in Europe gives this nice visualization, which highlights how different the clinical course is by individual — each of these patients had very different background characteristics and outcomes.

Question: Will a history of running help reduce COVID-19 severity? Like, if I have enhanced lung capacity from exercise, will that help?

Answer: Exercise is good for you, so keep at it! When it comes to answering your question, we don’t have this type of data for COVID-19 specifically (no surprise there), but we can look at the relationship between exercise and respiratory disease mortality since it’s been studied in other contexts.

• First, there’s a pretty big body of evidence to support the positive relationship between exercise and immune function.
• Second, there’s been some scientific debate about the short-term effects of high-intensity exercise on the immune system. A review of the evidence published in 2018 shows that we do not need to worry — “rather than suppressing immunity, contemporary evidence shows that an acute bout of exercise improves immune surveillance.”
• Third, when we look more specifically at the relationship between exercise and another respiratory condition — pneumonia — we find some marginal benefit: a) “The risk of community-acquired pneumonia decreased with increasing physical activity among women” but that was not the case among men in the study; b) “Women in the highest quintile of walking were less likely to develop pneumonia compared to women who walked the least (multivariate adjusted RR=0.82; 95% CI, 0.69–0.98)” but there was no statistically significant difference among the other categories (this study only included women); c) “Higher doses of running and walking were associated with lower risk of respiratory disease, pneumonia, and aspiration pneumonia mortality in a dose-dependent manner, and the effects of running and walking appear equivalent.”

Q&A for 3/28:

#Treatment #Ibuprofen

Question: If you’re in the hospital diagnosed with coronavirus, what are they actually doing to help you? I’m not sure if there’s a cure or not, so do they just give you fluids and put you on a ventilator? Is there actually anything they can do?

Answer: If you have a mild case, without much trouble breathing or other complications, you should just stay home and manage yourself like you would with a bad cold. There are no treatments for COVID-19 (like tamiflu) and there is no cure. But, if you do have a more severe case, doctors and nurses can help! CDC provides clinical management and treatment guidelines, which describe some measures. This paper from the Society of Critical Care Medicine provides far more comprehensive guidance for clinicians. To my knowledge (again: I’m a lay person not a health care provider), a patient would only be put on a ventilator if they were in critical condition with severely compromised respiratory system (e.g. their lungs were not working). Before that point, patients would get additional oxygen to supplement in the context of diminished lung capacity. Patients would also be treated for fever (likely with acetaminophen/paracetamol). Other actions could be taken based on the patient’s condition. In terms of other drugs, you may have also heard that New York is starting widespread testing/treatment using unproven hydroxychloroquine and chloroquine. There’s no evidence that this drug has positive treatment effects and scientists are highly skeptical of the drug’s potential benefit. Meanwhile, folks who need the drugs — like those suffering from rheumatoid arthritis and lupus — are having trouble getting their medication because of the surge in demand. So if/when you hear about a new and amazing treatment, please be skeptical until we have the evidence to back the claim.

Question: I have seen many warnings from non-reputable sources that we should avoid Ibuprofen and Advil because COVID-19 thrives on it. I have seen a COVID study sheet for practitioners that states “NSAIDs- strongly recommend avoiding NSAID use as reports of NSAID use have preceded clinical deterioration in some patients with COVID-19+.” Is this true and where is the data/source backing this up?

Answer: In short, we do not have good science on the issue and in the absense of strong data, some reputable medical professionals and associations are recommending use of paracetamol (eg. Tylenol) to treat COVID-19. Now for the longer response: I saw a lot of that floating around on social media two weeks ago (which feels like a lifetime ago) too. A lot of it was prompted by France’s Minister of Health, who made the statement on social media on 14 March that “Taking anti-inflammatory drugs (ibuprofen, cortisone . . .) could be an aggravating factor for the infection. If you have a fever, take paracetamol.” The British Medical Journal published a description of the issue on 17 March, “Covid-19: ibuprofen should not be used for managing symptoms, say doctors and scientists” and a follow-up report on 23 March, “Covid-19: European drugs agency to review safety of ibuprofen.” As to guidance/suggestions from reputable sources:

• The European Medicines Agency stated on 18 March that “There is currently no scientific evidence establishing a link between ibuprofen and worsening of covid-19. EMA is monitoring the situation closely and will review any new information that becomes available on this issue in the context of the pandemic….”
• The Society of Critical Care Management offers this guideline — “The use of non-steroidal anti-inflammatory drugs to treat fever in patients with COVID-19 continues to be debated. Until more evidence is available, we suggest using acetaminophen/paracetamol to treat fever.”
• And a letter in Science offered this suggestion on 27 March, “…should patients with clinically complicated SARS-CoV-2 infections be administered NSAIDs as a treatment? No. There is no evidence of benefit. If such a patient were also to have poor kidney function, maintenance of renal blood flow becomes critically dependent on vasodilator prostaglandins… Such a situation might also predispose the patient to the gastrointestinal and cardiovascular complications of NSAIDs. However, until we have robust evidence, patients in chronic pain should continue to take their NSAIDs rather than turn to opiates. Given that the elderly appear to comprise the predominant at-risk group for severe COVID-19 at this time, an association between NSAIDs and the disease may merely reflect reverse causality — that is, infection makes you more susceptible to adverse effects of NSAIDs on the infection.”
• Finally, the Center for Evidence Based Medicine (CEBM) reminds patients and providers that fever can be good and does not always need to be treated. Fever is the body’s strategic response to infection and can inhibit viral reproduction. Other potential benefits of fevers are described by the CEBM in the link I shared. So, another thought is, as long as the fever isn’t too distressing, consider avoiding/reducing consumption of fever reducers.

Q&A for 3/27:

#Hospitalizations #Data Use #Testing

Question: So I was confused with this… “Cuomo said it was encouraging that hospitalizations were projected to double every 4.7 days on Tuesday, compared with Monday, when the number was doubling every 3.4 days, and Sunday, when the figure was every two days.” Not only is it rather early to see impact on hospitalization (as you said!), but also the numbers/measure themselves are unclear. Do you know the source?

Answer: I have also been confused by these pronouncements! I reached out to NY Department of Health yesterday to inquire about the data and see if we could access the data, but I haven’t heard back yet. In the meantime, I looked at the data from The COVID Tracking Project, which is an amazing set of state-level and national data on daily tests conducted, cases, hospitalizations, and deaths. It’s just what I wanted! There are some data quality issues, but I’ll work with what I’ve got! And here’s what I’ve got (see line chart).

Overly simplistically — Here, the change from Saturday to Sunday was +371, which is about 1/4 of Sunday’s total, meaning it would take about 4 days for the number of hospitalizations to double. But then between Sunday and Monday, the change was +661, which is about 40% of Saturday’s total, which would mean that hospitalizations would double within 2-3 days. Indeed, by Tuesday, hospitalizations had doubled since Saturday. But the hospitalization increase between Monday and Tuesday was smaller than Sunday to Monday and a smaller proportion of the overall daily hospitalizations (+599), which would mean that doubling time would be slower — hospitalizations would double in about 5 days. So, the Governor saw this as a positive. Now, my overly simplistic numbers are not those the Governor is working with, but you get the general idea. The outlook seemed worse on Monday than on Tuesday. Now if we check out hospitalizations from yesterday, it kind of blows the whole statement that things are getting better — that we are slowing the spread — out of the water. Upshot — I caution reading too much into this right now because we have so few data points. In my opinion, it is definitely too early for the Governor to make pronouncements about the impact of social distancing or the like on the trajectory of the epidemic in New York.

Question: What does it mean now that the United States has the most confirmed cases of coronavirus in the world? Is it because we’ve ramped up testing? Because we have a large population? Or because we’ve been ineffective in our response?

Answer: I will give the US credit that in the last week, we really have expanded testing dramatically. Yay! Check out this table I made also thanks to data compiled by The COVID Tracking Project. The expansion is impressive, and the New York Times has a great article today on the topic with way better and far more visualizations than mine. Check it out! Despite the impressive increase in testing, per capita, we’re still quite far behind (we are at just ~200 tests per 100,000 people compared with Italy at ~600 tests per 100,000 and South Korea at ~700 tests per 100,000). And even at our high (so far) of ~100,000 test a day, it’s still only 2/3 the level public health experts recommend — 150,000. Finally, I’d say that our high case count is a reflection of our response, which was too late and is still poorly coordinated among other challenges (described further in Q&A of 3/24). And if you’re interested, Jeffery Sachs wrote an opinion today on this very topic.

Q&A for 3/26:

#Reuse #Masks #Gloves #Transmission #TakeOut #Transport

Question: Can we re-use protective gear, such as rubber gloves and masks? Is there an effective way to sterilize them after use?

Answer: In the ideal world, we would not reuse personal protective gear, like gloves or masks. But, we’re not living in an ideal world, protective equipment is hard to come by for everyone, and health workers are especially at risk. So what do we do since we’re not living in an ideal world?

Gloves: CDC is adamant that if you are living with someone who has COVID-19, you must clean frequently and thoroughly, and wear disposable gloves while you clean and disinfect. When it comes to just walking around the world, health experts do NOT recommend wearing gloves. Gloves won’t stop you from touching your face and they will be washed far less frequently than your own hands. If you’re concerned just walking around the world, just remind yourself to touch less stuff, stop touching your face, keep washing your hands, and stay home as much as possible!

Masks: When it comes to facemasks, CDC still recommends that among lay people, only sick people wear masks to protect others (for more on this, see Q&A from 3/20). When it comes to health workers, CDC has guidance about reusing masks when supplies are low. I think these guidelines would apply to mask use among lay people too — if your supply is low, keep reusing unless the mask is soiled, damaged, or hard to breathe through.

Question: Do we have any insight as to if there are activities that are more likely to infect people than others? Like for example, do we know if people were more likely to have gotten sick going to the grocery store versus commuting to work? Based on how others got sick, am I more likely to become infected picking up takeout versus having food delivered, etc?

Answer: Short answer: We do not know. Longer answer: Initial data from China showed that the preponderance of cases (75%-85%) came from familial transmission — folks living together. So far, it seems that the respiratory virus is mostly transmitted through people who are in close contact (e.g. within 6 feet of each other) and through respiratory droplets from sneezes or coughs. Whatever your activities are, try to keep a safe distance from other folks and minimize face-to-face interaction.

Transportation: When it comes to public transport, try to minimize it too. I say that because there have been a few studies looking at transportation and influenza that have found a relationship. One observational study of London Underground use and flu spread showed a correlation between the two. A case-control study, also in London, showed that people who saw their doctors presenting with acute respiratory infection were almost six times as likely to have used public transport in the previous 5 days. And 2016 systematic review of transportation and flu spread showed increased risk of flu and other types of coronavirus acquisition from various transportation types, especially air and cruise ship travel (which is not surprising given what we know of COVID-19 spread on cruise ships.)

Food: Since I couldn’t find more science on the matter, I did find this nice HuffPo article that synthesizes some of the risks of takeout/delivery and offers advice from some experts. In good news, there’s no evidence that coronavirus is transmitted through food, and there’s another good article on that front with tips too from Live Science, here.

Q&A for 3/25:

#Indicators #Success #Mutations

Question: What would the data have to look like to know if what we are doing is working? And what if we still don’t have enough tests?

Answer: I especially like this question — it’s a monitoring and evaluation question and M&E is my jam! So prepare for a long-ish answer.

First, let’s just quickly outline what we want to know:

1. What are we doing (outputs)? (ex: percentage of population that is practicing social distancing; percentage practicing correctly)
2. What are we seeing (outcomes)? (ex: daily number of new infections, hospitalization #s/rates, intensive care #s/rates, virus mutation rates)
3. Why does it matter (impact)? (mortality rates, reproduction rates, vaccine/treatment effectiveness)

Without either: a) widespread testing; or b) effective population-based surveillance, tracking changes in reproduction rates (R zero) and mortality rates will continue to be a challenge. So, instead of focusing on impact, we need to focus on the outcomes, where we have better data. Here, our indicators include:

• Number of hospital admissions (daily, weekly, cumulative)
• Number of ICU admissions (daily, weekly, cumulative)
• Number of deaths (daily, weekly, cumulative)

For each of these indicators, we’d want to understand how affected vulnerable communities are, so we’d also want to disaggregate the data at least by age (typically 5-year age bands), underlying health conditions (e.g. heart disease, lung disease), and health worker status. We’d also want to continue to track virus mutations to understand how it is changing. These data, which scientists share via the online platform, GISAID (more on that below), will help us to know how effective our vaccines may be.

Now that we’ve got our indicators, we also need to remind ourselves that our what we are doing now will only be made visible in the future… we have to get Back to the Future! (just watched all 3 movies with my 7-year-old, Great Scott!) Since it takes on average 5 days for an infected person to begin to show symptoms and since it takes an additional 5+ days for a person with symptoms to become seriously ill, the folks who are showing up to hospitals now became infected at least 10 days ago, but probably even longer than that. Since we just started social distancing on 3/16(ish), we wouldn’t expect to see a decrease in the number of hospital admissions now… we need to wait still a few more days, and even more days for ICU admissions, and more for death. Really crudely — 5 days to feel sick, +5 more days to feel really sick, +X days to be admitted to ICU, +X days to die. So, we should expect to see our outcome measures (above) continue to worsen in the immediate term. If our social distancing now is effective, we’d expect to start to see hospital admissions decline in about 2 weeks or so (back of the envelope guess). But that decline will be a decline not from the numbers we’re seeing today, rather from the higher numbers we’ll be seeing two weeks from today.

If you want to know more about influenza monitoring (which is pretty close to COVID-19 monitoring), check out CDC’s weekly influenza report. Australia also has a nice description of its influenza surveillance framework here.

Question: Do we know if there is any potential for the Coronavirus to mutate, and what are the signs that a mutation has occurred?

Answer: Yes, this coronavirus, like all viruses, has the potential to mutate. (see Q&A from 3/11 and 3/19). Scientists around the world continue to monitor mutations via genomic sequencing data that are shared on the GISAID platform. Good news is, SARS-CoV-2 seems to be mutating at a very slow rate. Yesterday, the Washington Post had a great article on this very topic. So far, scientists are seeing very few mutations, they have found no evidence of a specific virus “strain” to increased severity, and they think this is great news for having a vaccine that will work for a long time (like measles rather than flu).

Q&A for 3/24:

#Responsibility #State #Fed #PPE #Testing

Question: It shouldn’t be just the feds taking responsibility for testing and PPE. What role are/should States and hospital systems taking? Are they not doing enough? Were they caught off guard expecting the federal government to come to the rescue?

Answer: I have mixed feelings about assessing blame for failures while we’re still in the crisis. Part of me thinks — let’s just look forward and assess later. The other part of me thinks — if we don’t identify recent failings now, we won’t be positioned to overcome those failings as the crisis continues. So with the idea that exploring failings now is helpful to our response moving forward, here’s my starting answer:

Testing. States and localities continue to try to take responsibility for testing, but have been stymied in the initial response due to CDC’s flawed tests, FDA’s slow EUA determinations, and more (The Atlantic recently had a great overview of these testing challenges). More recently, CDC’s International Reagent Resource Center, which states and labs rely upon, is reportedly not keeping up with demand, further stymieing state/local testing capabilities.

Protective Equipment (PPE). I don’t have enough visibility into state-level resources. At the national-level, we’ve known for some time that PPE shortages will be a massive problem. WHO warned about PPE shortages becoming a global problem as early as 7 February. Given that PPE shortage is a national-level problem, it is first and foremost the Federal Government’s problem to address. Asking states to figure it out on their own diminishes the federal system, puts states in competition with one another, and limits our ability to effectively and efficiently respond. Furthermore, it flies in the face of federal responsibilities outlined in the National Strategy for Pandemic Influenza, further described below.

Pandemic roles/responsibilities. Preparedness and response is the responsibility of federal, state, and local governments with key roles for the private sector and individuals too. The roles of each of these groups vary. Back in 2005, in the wake of the H5N1 outbreak, the federal government developed our National Strategy for Pandemic Influenza. Elements of the strategy have since been updated by CDC and HHS, but for my own ease of response, I’m sticking with the 2005 list of response expectations. I’ve used the list to make this chart of domestic response expectations that includes my initial assessment of the initial response.

Q&A for 3/23:

#Social Distancing #Quarantine

Question: How long to we have to stay quarantined from everyone?

Answer: I wish I knew the answer. I don’t think anyone really knows. But we’re not operating in a total information vacuum, so I’m sharing a synthesis what I do know. One take-away — one round of social distancing will likely not be enough.

• 2 weeks: Back on March 16, the President issued “15 days to slow the epidemic,” which made it sound like we’d be spending 2 weeks isolating ourselves at home.
• 1 month or more: On Friday, Dr. Fauci made the following statement about social distancing during an interview on NBC’s Today show, “If you look at the trajectory of the curves of outbreaks and other areas, at least going to be several weeks. I cannot see that all of a sudden, next week or two weeks from now it’s going to be over. I don’t think there’s a chance of that. I think it’s going to be several weeks.”
• 2 months or more: We also have the experience of China to look at — where large swaths of the country were shut-down for 2 months — and we should continue to learn from other country’s experiences. For example, Hong Kong is again taking social distancing measures after having lifted them earlier in the month.
• 3 months or more: Finally, we have the Imperial College model-based estimates of virus spread under various scenarios, which suggested a strategy that includes 3-months of social distancing followed by easement which would then lead to periodic social distancing based on triggers — hospital ICU cases (see Figure below).

Among the many important aspects of these various scenarios is the knowledge that we cannot be practicing social distancing in absence of other key public health interventions. We absolutely have to have ongoing testing, surveillance, contact tracing, case quarantine, and more! The New York Times yesterday included a fantastic article on everything we need to be doing.

Does social distancing need to happen? YES! Can it go on forever? ABSOLUTELY NOT. Is it social distancing the solution? IT’S ONLY PART OF THE SOLUTION. Should you practice social distancing? YES, PLEASE.

Related rant: The fact that New York is again limiting testing because it is trying to save its PPE is heartbreaking and infuriating. The United States has been and continues to fail in this response. To ask people to practice social distancing — shut down their businesses, lose their livelihoods, miss out on the opportunity for an education — while the U.S. Federal Government does not do the needful in the rest of the public health response is reprehensible.

Q&A for 3/22:

A point-by-point response to Aaron Ginn’s post of 21 March [also shared on a stand-alone Medium post]

Question: A lot of folks are reading a post on Medium yesterday that shows that we’re not being data driven and we’re all overreacting to coronavirus. Did you read it? What do you think?

Answer: I did read that post on Medium, which has since been taken down because it is “under investigation or found in violation of Medium’s rules.” The author, Aaron Ginn, a Silicon Valley self-described “technologist” who has written for Breitbart, re-posted to a website called ZeroHedge. In short, this post was exceptionally problematic and I’m glad that Medium has taken it down because it is such a flawed reading of data that gives readers incorrect information and the wrong take-aways. Rather than just leaving it at that, I feel obliged to walk through some of the overarching problems and I am going to do so by going though each of the post’s subheadings. Here’s my peer review of Andrew Ginn’s post.

Ginn’s overarching theme — We are all overreacting to the coronavirus and our overreaction is worse than the virus itself.

Response: I am angry too, but for a different reason. The United States did not get testing rolled out quickly enough — both for testing suspected cases and for conducting ongoing population-based surveillance — to keep the virus from moving through our communities. And because we missed that boat, we are all now gravely suffering an economic catastrophe. Even though we were late, our current rush to contain the virus is warranted. Listen to the experts and not to this self-described “technologist.”

Response by subheading:

1) Total cases are the wrong metric: Fair point. But no one is pointing to total cases in isolation, even Johns Hopkins highlights total cases alongside total deaths and total recovered. And most of the news coming out of places like China is focused on new case counts per day. Furthermore, understanding the number of people a new virus has infected is not a “vanity data point” as the author points out. It’s our denominator for all other analyses! Want to talk about case mortality rates? First, we need to know the number of cases.

2) Time lapsing new cases gives us perspective: Yes! And as a reminder, whenever you’re looking at a graph, first look at the X and Y axis. Here, the X axis is days, in 5-day chunks. But the Y axis, it’s different — it’s a logarithmic scale. We use log scale when we’re dealing with orders of magnitude (e.g. the values at the higher end are much larger than the values at the lower end of the range). So, what do we see — the case count growing at exponential rate. What’s the take-away? Answer: A small outbreak can quickly become a massive outbreak.

3) On a per capita basis, we shouldn’t be panicking: Agree. Panic is not helpful. And yes, it’s good to put things into perspective, so I appreciate the per capita breakdown. But remember the above section? It means that individual risk can change rapidly.

4) COVID-19 is spreading, but probably not accelerating: Even if the rate of spread remains constant, as long as the rate is greater than 1, we will see acceleration in the daily numbers of people who contract the virus. But let’s look further at what the author posits. He states, “daily growth rates declined over time across all countries regardless of particular policy solutions, such as shutting the borders or social distancing.” And then shares data from China, South Korea, Italy, and Iran. These countries have all implemented stringent controls over time, so to state that growth rates naturally decline is a fallacy. There’s no counterfactual here to test against. Second, look at the title of the graph, it’s “daily percentage increase in #of reported COVID-19 cases.” Daily percentage increase in # of cases is NOT equivalent to daily growth rates. If we start with 40 cases on day 1 and have a 200% increase, it’s an increase of 80 people, bringing case count to 120. Now if those new 80 people infect another 160 people, our case count on day 3 will be 280, but that’s only a 43% increase (we had a 200% increase the day before!). Why am I bringing this up? To point out that the rate of spread is difficult to visualize in charts like this.

5) Watch the bell curve: I’m honestly not sure what this section of the paper is trying to convey. The author states that “CDC and WHO are optimizing virality and healthcare utilization, while ignoring economic shock to our system. Both organizations assume you are going to get infected, eventually, and it won’t be that bad.” Huh? Where is this massive assumption on behalf of the author coming from? And taken together, these sentences don’t make sense. The first sentence is that CDC and WHO are trying to save our health system and second is that CDC and WHO don’t think it will be that bad if a person gets infected. And when it comes to exponential growth of any given thing, when there are limiting factors (here, number of human hosts for the virus), yes, you will generally have a bell curve. So what?

6) A low probability of catching COVID-19: This all depends on how quickly we can control the pandemic. The wider the pandemic expands, the greater your probability of catching the coronavirus. Data from Wuhan (see Lancet study here) show that before strict measures were enforced, the daily reproduction number (R0) was 2.35 and declined to 1.05 after travel restrictions were introduced. Government policy and collective action of individuals can have an impact on spread of the virus. Indeed, the same WHO report the author cites to boost his dubious claim that “research show that COVID-19 doesn’t spread as easily as we first thought or the media had us believe” states:

“The COVID-19 virus is a new pathogen that is highly contagious, can spread quickly, and must be considered capable of causing enormous health, economic and societal impacts in any setting. It is not SARS and it is not influenza. Building scenarios and strategies only on the basis of well-known pathogens risks failing to exploit all possible measures to slow transmission of the COVID-19 virus, reduce disease and save lives.”

A couple more points — we know that when it comes to respiratory viruses, mass gatherings can contribute to case spread and we have evidence from various recent experiences to support this claim — see, for example, the connection between Boston COVID-19 outbreak and the Biogen Conference. And we know that with exponential growth, the risk to the population grows concurrently. A low risk today does not equate to a low risk next week.

7) Common transmission surfaces: Agree, this study cited shows that the virus can live on surfaces, but did not go so far to show whether the virus can be transmitted from these surfaces to individuals. The key take-away on this front is that we absolutely have to keep washing our hands and avoiding touching our faces and cleaning our spaces because fomite transmission is within the realm of possibilities and now we have more data showing that it’s plausible.

8) COVID-19 will likely “burn off” in the summer: We have no evidence to support this claim. We can hope it’s right, but let’s not incorrectly use data to support this hope. First, we do have a number of cases in the southern hemisphere now! Check out data from Australia, Brazil, Malaysia, Singapore, etc. Second, many countries in Africa do not have widespread testing — we shouldn’t confuse the lack of data with the lack of cases. Third, this Harvard Center for Communicable Disease scientist, Dr. Lipsitch, puts it best, when answering the question, “Will COVID-19 go away on its own in warmer weather?”. His answer, “probably not.” To understand his rationale and the science behind it, check out his post.

9) Children and teens aren’t at risk. Yes, the good news in this pandemic is that children and teens seem to fair much better than adults, especially older adults. Children and teens become infected at similar rates as adults, and it’s still unclear how much transmission from children to children or children to adults occurs. Now, there is a risk difference between babies, children <10 and teens. We have seen some adverse outcomes in babies, some death in teens and no deaths so far in children. So, risk of adverse outcomes increases with age. And risk to teens is much, much less than risk to adults, especially older adults. But claiming no risk is also incorrect. This recent paper in Pediatrics shows that there were severe and critical cases among children and teens infected with COVID-19 in China. So, let’s not make sweeping claims that are incorrect in an effort to assuage people’s fears.

10) A strong, but unknown viral effect: The degree of coronavirus spread will depend on the interventions we implement. We can change R0. Indeed, the change in R0 in China was not a result of natural shift; it was the result of China’s extremely strong efforts to curb the virus’s spread. What we’ve seen in places that are at the beginning of the epidemic and places that have not implemented strong measure to control the epidemic is a growth rate that is exponential — with each person on average infecting at least 2 others.

11) What about asymptomatic spread?: The author states that data on asymptomatic spread “is still unclear but increasingly unlikely.” It’s actually the reverse. Yes, asymptomatic spread is still unclear, but the data is pointing to it being increasingly likely. Nature had a good summation of the evidence to date in an article published on 3/20. This evidence further points to the need for social distancing measures to be in effect.

12) 93% of people who think they are positive are not: This is a misuse of data. First, getting a test does not mean a person thinks they are positive. Second, the testing strategies have varied dramatically by country and even over time in country, percent positive in the US now and percent positive in South Korea have no relationship without that context. Most importantly, widespread testing is NOT about understanding a person’s underlying risk of acquiring coronavirus. Rather, it’s about identifying who is infected so that we can track their close contacts, minimize spread, protect health workers who may come in contact with the infected, and identify treatment strategies, if possible.

13) 1% of cases will be severe: No. We are still understanding mortality rates connected with COVID-19 and using US data from mid-March, before the US was more widely testing and before most cases had been resolved is completely inappropriate. As I’m typing, the US has 26,747 cases and 340 deaths (per Hopkins), which yields a very crude case fatality rate of 1.2% and that’s not including all the other cases that are severe but don’t result in death. Moreover, if our health system falters as we’re seeing in Italy, we could have much higher case fatality rate (it’s 9% there right now!).

14) Declining fatality rate: Yes, as we capture more mild cases of COVID-19 through expanded testing, we can hope to see our fatality rate decline. And as we protect our health system so that it can adequately respond, we should also hope to see more people survive. If we do not do the needful, however, we will see increasing fatality rates, as Italy is seeing now. China had declining fatality rates because of strong interventions, not by some magic.

15) What should we do?: Stop spreading false narratives. Stop basing decisions on faulty read of the data. Follow the advice of public health experts. Ramp up testing NOW. Hold our leaders accountable for inaction earlier when it could have staved off this health and economic crisis. Bring economic relief to the millions of Americans suffering.

In case you’re wondering who I am and what credentials I bring to this peer review:

I am a trained public health professional. I received my Master in Health Science from Johns Hopkins Bloomberg School of Public Health, where I studied demography and epidemiology. For the last 13 years, I have worked as a Public Health Advisor first for the CDC and most recently for the U.S. Agency for International Development. I have written peer-reviewed publications published in journals including Bulletin of the WHO and Lancet. I serve on the Editorial Board of Global Health Science and Practice and have served as peer reviewer for a variety of public health journals. The views shared herein are my own and do not represent the views of the U.S. Government or any other entity with which I am affiliated.

Q&A for 3/21:

#Money #Transmission #Treatment

Question: Aren’t open cash transactions spreading the disease? I read COVID-19 can live for some amount of hours on cardboard, so I would think that paper money with all the bodily oils would harbor live virus cells for at least some period of time.

Answer: Oh yeah, paper money does seem problematic, especially in light of the results from a recent study that showed that the virus can live on cardboard for 24 hours (see Q&A from 3/12 for more information on that front). CBS actually had an informative story about this very issue last week, where a number of epidemiologists expressed concern about handling cash and the spread of COVID-19. I also looked up the science specifically about paper money and here are a few interesting nuggets. First, scientists have been concerned about money for at least 135 years! (Schaarschmidt J. Upon the occurrence of bacteria and minute algae on the surface of paper money. Nature. 1884;30:360.) Second, a number of studies do show that paper money can indeed harbor a variety of microorganisms, some of which can cause pneumonia and some of which are antibiotic resistant. A 2017 study that took DNA swabs of circulating $1 bills in New York found a variety of genetic materials on the bills, including living bacteria. That study’s authors concluded, “As a whole, these results confirm and deepen several previous studies suggesting that money harbors a diverse and viable microbial population.” To sum up — try to avoid paper money and if you can’t just keep washing your hands and avoid touching your face. You can’t get COVID-19 simply by touching the money. Rather, the money is the vector by which the virus can spread from a person’s hand to their mouth or nose (e.g. an orifice through which the virus can enter the body).

Question: What’s the story with new treatments for COVID-19 folks are discussing? I’ve been hearing about remdesivir, a Cuban interferon drug called Alfa 2B, the use of antibodies from recovered patients, and now an antimalarial drug, chloroquine.

Answer: There are no treatments available for COVID-19 yet. And as much as some folks in the Administration tout the availability of new and amazing drugs, it’s hype that’s not yet backed by science. And the danger of using drugs that aren’t tested is that you can actually make things worse for folks. Remember thalidomide? Despite that bleak news, scientists are actively working to gather evidence about the effects of various potential treatment regimens through clinical trials. And with that evidence, doctors will be able to ultimately provide tested and effective treatment for their patients suffering from the virus. For a list of potential treatment options, see here. And for a description of studies FDA and NIH are working on, see this FDA press release from 3/19. Currently, to my knowledge and based on yesterday’s call with FDA, doctors caring for COVID-19 patients can consider off-label use of chloroquine and can reach out to FDA about participation in the remdesivir trial.

Q&A for 3/20:

#FaceMask #PPE

Question: Why does the US not recommend wearing face masks?

Answer: After hearing CDC’s Dr. Nancy Messonnier state on multiple occasions that we should not be buying or using face masks if we are healthy, while simultaneously seeing so many people in China, Japan, and South Korea wearing masks, I became curious about this too! First off, CDC’s guidance is: 1) people who are well should not wear a face mask; 2) people who are sick should wear a face mask when in public or interacting with others in order to minimize virus spread and protect others; 3) health care providers should wear a face mask to protect their patients and protect themselves. So there’s the background. What’s the evidence? Most of the evidence is outlined in CDC’s 2017 report, Community Mitigation Guidelines to Prevent Pandemic Influenza and what’s shown in the report (recommendations from which are copied below) is that there is limited data on protective effects of mask wearing among healthy people. Couple that with the fact that we are in a situation where face masks are in high demand and supplies are limited — and it makes even more sense that CDC would caution well people from buying and using face masks. Indeed, CDC just issued Strategies to Optimize Personal Protective Equipment Supply because supply shortages are so critical. So, if you’re living in the US, it’s best to save the face masks for those who are sick and for our health care providers. Now, if you’re living in East Asia, cultural norms will likely mean that you wear a face mask regardless of your underlying health (South China Morning Post published a fascinating article a few days ago on East/West face mask cultural differences). And in response over the last month, China boosted its face mask production capacity by 450 percent!

Q&A for 3/19:

#Reinfection #Immunodeficient #Youth

Question: Can you catch coronavirus more than once or are you permanently immune like after measles or something?

Answer: Short answer: We aren’t sure. Longer answer: We currently know that COVID-19 can stay circulating in a person for several weeks after recovery (see Q&A from 3/9). Good news: Some epidemiologists think this may be beneficial for longer-term protection since it gives your body’s immune system a longer time to build effective antibodies; There was a really small study of Macaques and reinfection released on March 14 that has not yet been peer reviewed and is not necessarily applicable to humans, which shows that over a short time, Macaques were not able to be reinfected with COVID-19. Bad news: We also know that viruses mutate (see Q&A from 3/11), which may create a different strain that a body’s immune system will have to mount a new response to. And we know that a person stays immune to another type of coronavirus — the common cold — for only 1–2 years after getting it. Finally, scientists are also hypothesizing that there is a dose-response relationship with COVID-19 — the more a person is exposed, the more severe the disease — scientists observed this phenomena with SARS. In summation: It seems there may be degrees of “catching COVID-19” (bad news especially for health care workers), that once you catch it, the virus may not be fully gone even when you are recovered, and our longer-term immune response is still unknown.

Question: Would you explain the threat that covid-19 poses to immunodeficient communities? And how that can impact our society as whole as well? Personally, I have a lot of loved ones with other health conditions, everything from asthma, to Crohn’s disease to transplant recipients on immunosuppressants. A lot of them are “young” and otherwise “healthy” people, but it is still scary to think about what would happen if one of them became ill with covid-19. And from my understanding, people with other health conditions are no small percentage of our population.

Answer: Per most of my answers, we have limited data about COVID-19 since it’s new and our knowledge is still growing. But, we can explore this good question with existing information. First, let’s look at the proportion of adults who are immunocompromised in the United States. I’ve copied herein a table from a 2016 research letter in JAMA that explores prevalence of immunosuppression among US adults. What we see is that among those 18–39, the prevalence is 1.6%, which is the equivalent of about 1.5 million Americans ages 18–39. Based on existing knowledge, we would expect an immunocompromised person to have more serious complications of a respiratory disease like COVID-19 as they have a weakened immune system that just isn’t as well equipped to fight against the “intruder”. Research on other types of respiratory infections — here, influenza — shows that “Immunocompromised patients with influenza had more severe disease/complications, longer viral shedding, and more antiviral resistance while demonstrating less clinical symptoms and signs on clinical assessment.” So, not only would we hypothesize that immonocompromised individuals would have more serious complications of COVID-19, we might also hypothesize that they are at increased risk of transmitting COVID-19 to others. For more on how the immune system works, check out this good article from Bulletin of the Atomic Scientists. CDC has also recently issued COVID-19 guidelines for people living with Asthma and people living with HIV, and has offered guidance to healthcare providers on how to handle immunocompromised patients. In short, even if you are young, if you are immunocompromised, please take additional precaution to protect yourself from COVID-19.

Q&A for 3/18:

#Recovery #Sequelae #Mortality

Question: There have been concerns about the long term health effects of many of those who recover from COVID-19, such as lung scarring and liver damage. What do we know about the permanent or long term health effects from other similar infections?

Answer: It’s true, we still don’t know the long-term sequelae of COVID-19. Some very preliminary data from Hong Kong that you cite in your question show diminished lung capacity among those recovered, but we just don’t have enough information yet to know what (if any) the prolonged impacts of COVID-19 may be. Looking at other coronaviruses could give us a clue. With the SARS outbreak of 2003, various studies found that lung function was diminished over time among those who recovered. A more recent study published in Nature found that “The most severe sequelae after rehabilitation from SARS are femoral head necrosis [e.g. hip joint doesn’t get enough blood and dies] and pulmonary fibrosis [e.g. damaged lung tissue that makes it hard to breathe].” Another paper looking at long-term effects of MERS also found that “Lung fibrosis may develop in a substantial number of patients who have recovered from Middle East respiratory syndrome coronavirus (MERS-CoV).” We will know more as more folks recover.

Question: The numbers you presented yesterday seem to be worst-case scenario type numbers. What’s up?! First, tell me what the numbers would be not in worst-case scenario. Then, describe for me how these mortality numbers differ from underlying mortality estimates.

Answer: If you want expert opinion, please read the report that the Imperial College COVID-19 Response Team released yesterday. [It won’t make you feel any better.] For the estimates I presented yesterday, I used the early-stage case fatality rates by age-group estimates from the China CDC. I haven’t seen any other age-specific case fatality rates, but I do acknowledge that these represent one place and one time point for an evolving pandemic. The case mortality rates will vary based on a range of factors and, fingers-crossed, the rates will be lower than these early numbers from China. As to typical mortality in the US, CDC has data on age-specific mortality and I’ve copied the table for 2018 herein for comparative purposes. I also changed the assumptions of the data I shared yesterday such that age-specific mortality is 1/2 that of what was shown in China and 30% of people become infected (see chart below). In this better-case scenario, we still have 568K deaths, which is equivalent to 20% of all deaths in the US in 2018. These are all estimates and we can change the estimates not just by changing our assumptions, but also by changing our behavior. My original purpose in showing age-specific mortality yesterday was to remind us that COVID-19 is not just a disease of the elderly. Even though mortality rates are much lower among younger ages, the threats of inaction or too little action are threats to us all.

Q&A for 3/17:

#Mortality #Elderly #Youth

Question: This may sound callous, but are we really upending our children’s education, the employment of thousands upon thousands, and the health of our entire economy to save the elderly? Dying is the only guarantee of old age. And dying from a respiratory illness has always been one of the most common ways old people finally succumb.

Answer: I appreciate the honest question. And I ran some numbers (see chart below). What I want us all to remember is that while COVID-19 hits the elderly the hardest, it still brings morbidity and mortality to other age groups. If the epidemic goes unabated, we could expect to see 2.27 million deaths, of whom 859,000 deaths will be to folks <70. If you’re wondering how these numbers compare with other deaths, CDC lists mortality rates for US population, including by cause. In 2017, the US experienced 2.81 million deaths TOTAL, of which 55,672 were due to influenza and pneumonia. COVID-19 poses a MUCH greater risk to the whole population, not just the elderly. Moreover, if our health system breaks, we will see higher mortality from all other causes due to poorer health care leading to poorer health outcomes. If you don’t believe my numbers (again, all assumptions-based and you can change the assumptions), you may want to check data from disease modeling experts, who show that “In an unmitigated epidemic, we would predict.

Q&A for 3/16:

#Exercise #SocialDistancing #Normal

(reminder: these are my own takes; also, I’m not a medical provider, I’m just someone trained in public health who reads a lot of public health journals)

Question: Am I allowed to go outside and run to workout? Like is that bad form for social distancing?

Answer: YES, get outside and take a walk and go for a run! Enjoy nature! Be good to yourself physically. You mentally and emotionally need it! And don’t worry, you are not putting anyone at risk. Social distancing==> Nature integrating! [only caveat: no marathons as it’s too many people too close together]

Question: How long do we think humans and the economy can continue this? And if/when we go back to normal, won’t infections peak again?

Answer: Good, depressing question. Clearly, this thing not only has terrible health impacts, but unknown social impacts and devastating economic impacts. And the US has managed to bungle most of the key pieces of good public health practice — e.g. rapid testing of suspected cases, clear public health messaging from the outset, and specific requests made of individuals to take action to protect themselves and others. Singapore seems to be a model country in this respect, having learned lessons from previous outbreaks to squash the current one. I suppose we can look to other countries and historic examples to understand how long we can continue this and how we can maintain our sanity in the process. Coming together to sing Rihanna out our windows sounds like a good plan, even if it’s fake. Using our time at home to think deep thoughts and work out creative ideas a la Isaac Newton is a good way to keep social blues away. But obviously, our massive self-quarantine/social distancing response to COVID-19 has psychological ramifications. A recent paper in Lancet shares how to address psychological impact of quarantine, and I admit, it’s all basically stuff you’d already assume — good communication, keeping time in quarantine as short as possible, ensuring adequate supplies, etc. As to getting back to normal, yes, COVID-19 may come back. Right now, we’re giving ourselves time to prepare so that when life returns to normal and COVID-19 does come back, we can quickly detect it and respond before it has the opportunity to move through communities at exponential speed. We’re also giving ourselves time to develop effective treatments and vaccines. [And if you aren’t up-to-date on existing vaccines, get up to date! Go get your flu shot!] After things have returned to “normal” and COVID-19 does return, our systems will be far better prepared to handle it and fewer of us will be at risk. Now let’s get to work (from home) on economic stimulus!

Q&A for 3/15:

#Elders #Asymptomatic #BeingCautious #Models #Iceburg

Question: I’ve been out in the world the last few weeks mixing and mingling with folks who know folks who are currently under self-quarantine and/or sick or recovering from COVID-19 or the flu (hard to know since testing has been so limited). I’m supposed to go see my 96-year-old mom next weekend, but I’m so scared of being an asymptomatic carrier. What should I do?

Answer: I’m also really concerned about seeing my elders. Again, COVID-19 hits the elderly the hardest — we see increasing severity, including morbidity, associated with increasingly older ages. Obviously, we need to do our utmost to protect this vulnerable population. Since the coronavirus seems to be transmitted most via respiratory droplets, we shouldn’t expect to see much asymptomatic transmission. But that’s an assumption… this virus is new and there’s still so much we just don’t know. As a recent journal article from China states, “The question of “the degree to which presymptomatic or asymptomatic infections can transmit” is not fully understood. There is an urgent need to screen infected carriers in larger close contacts or in the general population, and assess their risk for transmission.” Yet another reason that we need to have widespread testing! CDC does have guidance here that recommends long-term care facilities and nursing homes, “Restrict all visitation except for certain compassionate care situations, such as end of life situations.” Given what we know coupled with what we don’t know, I think it’s better to be overly cautious than not cautious enough. We generally visit my 94-year-old father-in-law every weekend. We’re not going this weekend.

Question: On Thursday, Dr. Acton, the Director of the Ohio Department of Health, said “”We know now, just the fact of community spread, says that at least 1 percent, at the very least, 1 percent of our population is carrying this virus in Ohio today… We have 11.7 million people. So the math is over 100,000. So that just gives you a sense of how this virus spreads and is spreading quickly.” How did Ohio arrive at that estimate?

Answer: Since Thursday, Ohio has backtracked a bit from this estimate, describing it as a “guesstimate.” But we do have modelers working on these very issues. And the models are only as good as the inputs — the data and the modeler’s assumptions. Since we have such limited data in the US right now regarding scope of community spread (blerg! testing challenges have so hampered our response!) I really wouldn’t put much faith on estimates you hear like this. The main thing to keep in mind is that we still haven’t turned the light on, so we just can’t see the full scope of the outbreak in communities across the US. Public health leaders like Dr. Acton are generally seeking to convey the fact that the scope of infections is bigger than we know (tip of the iceberg, so to speak). All that said, here’s a fascinating set of assumptions that modelers are currently using to estimate infections — start with death assumptions. Assume case-fatality rate of 1 percent and that it takes 15 days for an infected person to die, then it means if you have 1 death in Ohio now, you probably had 100 people infected 15 days ago. Now assume that it takes 5 days for cases to double, then over the course of those 15 days, we’d expect to have 800 cases today. Dr. Edmonds from London School of Hygiene and Tropical Medicine gives a good explanation here.

Q&A for 3/14:

#Symptoms #Allergies #Testing #FalsePositive #FalseNegative

Question: I’ve been developing a bit of a cough and have a bit of a scratchy throat. I thought it might be allergies, but I’ve never had them before. Is there any way to differentiate between allergy symptoms and COVID-19?

Answer: Obviously, I’m not a doctor. If you’re concerned about your health, definitely talk to your doctor and always pay attention to your body. And if you feel sick, please stay home, call your doctor, and take action based on that conversation. All that preamble said, this is a great question and one that many of us are asking. Per Weather Channel, tree pollen in DC is HIGH right now. And if you’re not from the DC area, even if you’ve lived here for a couple of years, it’s likely that allergies will hit you harder. Here’s a great article on the topic. Now when it comes to differentiating allergies and COVID-19 symptoms, there are a few things that make the two very different. First, you will NOT have a fever with allergies, but you generally WILL have a fever with COVID-19. Second, you generally will have sneezing and a scratchy throat with allergies, but you will generally NOT have those symptoms with COVID-19. As with all my answers, we’re operating with limited data. We do know that clinical characteristics out of China among those hospitalized show that “the incidence of fever was 89.1%, the incidence of cough was 72.2%, and the incidence of muscle soreness or fatigue was 42.5%… Diarrhea, hemoptysis, headache, sore throat, shock, and other symptoms only occur in a small number of patients.”

Question: I saw there were reports that President Bolsonaro in Brazil tested positive for COVID-19 and then an hour later another test was apparently negative. Is there a way of knowing which one was the false one?

Answer: This question is both a political one and a science one. On the politics side, one has to think about the benefits and drawbacks of a country’s leader admitting to illness. Add that with the information flow on social media and challenges abound. I’m going to set aside the political piece now and focus on the science piece. It is possible to have a false positive (e.g. test says your positive but you really aren’t), just as it is possible to have a false negative (e.g. test says you’re negative, but you really have the virus). In the case of COVID-19, false-negatives are more likely. For false positive COVID-19 tests, it’s generally because of test contamination. This is lab-based error and is, to my knowledge, quite rare [see more about PCR tests in the Q&A from 3/9]. Meanwhile, a negative test result does not necessarily mean that the person is not infected with COVID-19. It means that the test did not pick up any of the virus in the sample, which could be because of a variety of factors. LabCorps has a good description of those factors here. Finally, if I were tested and learned that one test was positive and one negative, I’d assume I was positive, especially since I was presumably tested because I was likely exposed and/or exhibiting symptoms.

Q&A for 3/13:

#Sick #Self-Quarantine

Question: If you get sick with COVID-19, how long do you need to self-quarantine? How long do you stay contagious?

Answer: By now, we’ve all heard that if we are exposed to coronavirus, we should self-quarantine for 14 days. So far, research shows that COVID-19 symptoms generally begin 5 days after exposure (range 2–14 days). If you don’t have symptoms within 14 days, you’re good to go back into the world. But what if you start to become sick within those 14 days? How long do you need to avoid seeing your parents or grandparents? How long do you need to self-quarantine then? These are questions scientists are still grappling with. We think that people are most contagious when they are most symptomatic (the sickest). Research published in the Lancet two days ago shows that median duration of viral shedding among individuals who were hospitalized due to COVID-19 was 20 days in survivors (range: 8–37 days) with one survivor shedding 37 days after COVID-19 onset. Again, we have limited data and these data are among those who were more severely impacted by COVID-19. My advice — If you’re sick, use your best judgement. Based on current data, if you’re still coughing, you’re likely still able to spread the virus. So, as long as you’re coughing, stay away from others!

Q&A for 3/12:

#Surfaces #Air #Cardboard #Plastic

Question: How long can coronavirus live on surfaces?

Answer: A new study was released yesterday exploring how long SARS-CoV-2 can live on surfaces and in the air. The study is not yet peer reviewed, so we should view the results with some skepticism. That said, the study found that “viable virus could be detected in aerosols [e.g. the air] up to 3 hours post aerosolization, up to 4 hours on copper, up to 24 hours on cardboard and up to 2–3 days on plastic and stainless steel.” Even if we have skepticism with this study, another recent study of other types of coronaviruses like SARS and MERS also show that it can stay viable on surfaces for extended periods of time. These results indicate that air and surface transmission of the virus is plausible and remind us of the vulnerability of virus spread in hospital settings, households, and group gatherings. Please be sure to use good cleaning products (bleach!) and keep washing your hands! Try not to touch your face (so difficult).

Q&A for 3/11:

#Worry #Risk #BluntTheCurve #Terminology #VirusMutations

Question: Why are we all so worried about COVID-19?

Answer: First off, you’re probably <60 years of age and you’re probably in pretty good health. You’re probably thinking that you personally, do not need to worry about yourself — yeah, you’ll probably get COVID-19 within the next year or so and yeah, being sick sucks. But being sick isn’t the end of the world! So, why the freak out? Is it warranted?

· Yes, the widespread action — of the right kind — is warranted. (examples of the right kind of action are here) [note: freaking out and buying a bunch of face masks from Amazon is the wrong kind of action]

· First, we need to protect the elderly, especially those with underlying conditions like heart or lung disease. Data from China reveal that folks age 80+ who get COVID-19 are at very high risk of adverse outcomes, including death — 15 of every 100 individuals age 80+ who contract COVID-19 die. In a situation like this, where we have no vaccination and no treatment, it is imperative that we try to minimize the elderly’s exposure to COVID-19. This means that we also need to try and avoid getting sick. And if we are getting sick, we need to stay home and avoid interactions with others as much as possible.

· Second, we need to protect our health system. Data from China also show that about 20% of all COVID-19 cases are serious and require medical treatment, including hospitalization. Our health system will be overrun if we all become sick at the same(ish) time. This will lead to terrible health outcomes not just for those infected with COVID-19, but for those who need to access the health system for everything else! Death rates sky rocket when health systems aren’t functioning. This is why social distancing measures are so imperative. We need to “blunt the curve” (see graphic below) so that even if all of us do end up becoming sick from COVID-19, we’re doing it at staggered times and not overwhelming the health system.

· We need to be concerned and we need to be taking proactive action to help our most vulnerable and to help our health system respond.

· Finally, unlike flu, we have no treatments and no vaccine for COVID-19 an no herd immunity.

Question: What’s the difference in terminology between SARS-CoV-2 and COVID-19?

Answer: SARS-CoV-2 is the virus and COVID-19 is the disease. Think of it like HIV/AIDS. HIV is the virus, AIDS is the disease.

Question: What’s the story with virus mutations?

Answer: [providing a synthesis of a good article from Science] Like all viruses, SARS-CoV-2 evolves over time through random mutations, only some of which are caught and corrected by the virus’s error correction machinery. SARS-CoV-2 accumulates an average of about one to two mutations per month, which is a slower rate than the flu. Scientists have been sharing SARS-CoV-2 genome sequences on the online platform GISAID and studying these sequences to a) answer the basic question as to which pathology is causing the disease (e.g. identify a novel coronavirus); and b) analyze how the virus is changing over time and place. The first piece is obviously critical. This second piece is also very important, but it can lead to over-interpretation of results. On the positive side, genome sequencing led scientists to correctly conclude that the new coronavirus had been circulating in Washington State for quite some time before being detected. On the flip side, it has led to over-interpretation — like a scientist concluding that the outbreak in Italy came from cases in Munich using a small dataset that was insufficient to make such a claim.

Q&A for 3/10:

#Italy #SouthKorea #SelfQuarantine #Testing

Question: Italy and South Korea both have similar case counts, but South Korea’s mortality rate is much lower. What’s driving the difference? Similarly, Japan has twice the population size of Italy, but far fewer cases and deaths. Why have these countries suffered so differently?

Answer: The true answer to this question will make a fascinating public health dissertation. I’m going to focus on the Italy/South Korea issue here since Japan has a more localized epidemic than either of the other two countries. As of March 10th:

· Italy has 10,149 confirmed cases, of whom 631 have died which gives a crude death rate of 6.2% — far higher than the WHO’s estimate (3.4%) and far higher than the mortality rate of South Korea, which has 7,513 confirmed cases with 54 deaths (0.7%)

· Italy also has an aging population with half of the population over age 47. Indeed, nearly a quarter of Italy’s population is age 65+, whereas South Korea’s population is younger with only 16% of the population age 65+ (note: for the United States, 17% of our population is age 65+). [if you like population data, see here]

· Because COVID-19 hits the elderly the hardest, we would expect to see higher mortality in countries with larger population of older folks.

· Based on the experience in China, we also saw much higher mortality in Wuhan, the epicenter of the outbreak. This is likely a result of an overloaded health system that was unable to adequately take care of those in need. We are witnessing an overloaded health care system in Northern Italy, which is also a reason why we are seeing increased mortality.

· There are probably myriad other factors, including socio-cultural factors, smoking prevalence, the speed of the initial public health response, and more.

Question: If I a person is self-quarantined, what does this mean for their interactions with others in their household? Are they also required to self-quarantine?

Answer: CDC has a great list of 10 things you can do to manage your health at home if you fear you have COVID-19. As much as possible, try to avoid your family members! If you do have to self-quarantine, here’s HHS guidance. While not a given, it’s likely your household members will also need to follow self-quarantine guidelines; by the time you are identified as being at risk, you’ve probably already been in close proximity with those living in your household.

Question: What’s the status of high–throughput tests? Rapid diagnostic tests?

Answer: High-throughput tests (also called automated tests) are for use in a diagnostic system that can process 1,000 tests in 24 hours. Yesterday, HHS announced it was partnering with Hologic, Inc through BARDA to develop such a system for COVID-19 testing. Per the announcement, “BARDA and Hologic expect that necessary development will be completed in a matter of weeks which then would allow the U.S. Food and Drug Administration (FDA) to consider granting Emergency Use Authorization (EUA) for the diagnostic test.” With regard to rapid diagnostic tests, several companies are working on development, which will allow doctors, nurses, or even individuals themselves to conduct COVID-19 tests (depending on the test type). Here’s one example. Generally, rapid diagnostic tests result in higher rates of false-positives than PCR-based tests, which is an important distinction early in an outbreak when we are working to find every case and their contacts (e.g. we do not want to waste resources investigating non-cases). At this stage, however, rapid diagnostic test kits in harder hit communities would be welcome. The Gates Foundation is working on home-based test kits for pilot in Seattle.

Q&A for 3/09:

#Mortality #Comorbidity #Youth #Testing #PCR #RDT

Question: I’ve seen a lot of news stories state that people with preexisting respiratory conditions, such as asthma, are at a higher risk of death from COVID-19. But those news stories rarely, if ever, mention if the severity of the diagnosis is important. For example, if someone, even a young person, has a mild respiratory condition, do they have a significantly higher risk of dying from COVID-19? Or is that person only at a significantly higher risk of death if their respiratory condition is so severe that they consistently require heavy medication or hospitalization?

Answer: We still have limited data, so we’re operating with uncertainty. While we are still unsure of the true case fatality rate of COVID-19, WHO recently cited case mortality rate of 3.4%.

· COVID-19 is the most dangerous to those age 60+, with increasing age associated with increased risk of adverse outcomes, including death. Data from China show that the mortality rate among those age 80+ is 14.8%. Note: As people age, they are also more likely to have underlying conditions that put them at greater risk — conditions like heart disease, diabetes, and lung disease. Data from China also show that co-morbidities can increase risk of death by up to 2.5x. Again, because data are limited, we don’t have much understanding of how serious underlying health conditions among young people influence their disease progression.

· So far, we know that young people are at very low risk of dying of COVID-19, regardless of underlying conditions. Indeed, young age seems to be protective. Data from China show that people under age 18 made up only 2.4% of all reported cases and only one person age 10–19 died. So far in South Korea, no one under age 30 has died. In Japan, no one under age 50 has died.

· Finally, when we hear about underlying conditions (co-morbidities), we’re talking serious underlying conditions rather than mild conditions.

Question: If a person tests positive for COVID-19, are they actually positive? What about someone whose test results are negative?

Answer: All of the coronavirus tests currently being used start with a technique called polymerase chain reaction (PCR), which can detect tiny amounts of a virus’s genetic material. This means that tests for COVID-19 are highly sensitive. Generally, a positive result really is a positive result. [minor caveat here is that tests should be confirmed as testing quality issues like contamination can lead to false positives] There are, however, several challenges with PCR testing:

· Because PCR tests pick up the virus’s genetic material, we still don’t know how long a person will test positive after they have recovered from COVID-19. Recent case reports from China showed that several folks who had recovered from coronavirus were still testing positive 5–13 days after hospital discharge. We need more information to know whether there are any implications for ongoing transmission and/or what to make of a positive test result among those who have recovered from illness or never showed any signs of illness.

· Even though PCR tests can pick up tiny amounts of genetic material, it still requires the virus to be replicating in the body for some time. This means that a person can test negative while simply being in the early stages of the coronavirus infection. Here, a negative result may not mean that the person is actually negative. This is why retesting is important for those who have been exposed to the virus. Resting has been regularly conducted among those quarantined.

· PCR requires more skills and is therefore less widely available than other types of tests. For example, PCR testing is more complex than rapid diagnostic testing that your doctor can conduct for the flu. This means that testing is lab-based rather than bedside-based. PCR testing is much more sensitive than rapid flu tests, which is important in places where coronavirus is not yet endemic; we do not want to see many false positives when we’re actively trying to identify all cases and their contacts!