Nearly 60% of U.S. Population at Substantially Greater Risk of Death by Accident, than from COVID-19

Updated: Nov 19, 2020

In late April and early may, New York State performed a large (15,000 person) seroprevalence survey to attempt to identify the real spread of COVID-19 in that state. This was done because officials there were aware that case counts, especially in the early days of the epidemic, drastically underestimated the level of spread of the novel coronavirus in the city. New York City has done an exemplary job at providing publicly available day-by-day totals of deaths, hospitalizations, and observed cases--including logging the proportion of deaths of people with and without co-morbidities. Taking these two data-sets together, one can learn a significant amount about the actual mortality by age group, including the mortality rate for the population without co-morbidities.

The results of this analysis, which I will go into in great detail below, paint a very different picture than that painted by the aggregated data. To wit, the results show that for people without co-morbidities of all ages--fully 60% of the population--their likelihood of dying from COVID-19 is lower than their risk of dying from accidental death--like a car accident or drowning. For people under 45, their risk of dying in a by accidental death such as a car accident is 10x higher, for children, near infinitely higher. Beyond that, when removing the nursing home population, the risk for people with mild severity of co-morbidities who are not in nursing homes, or about to be in nursing homes is far, far lower than what the aggregated data show. Nationally, the nursing home population has sustained more than 40% of COVID-19 deaths, and in the northeast, rates of infection are likely to be well above 50% in nursing homes--explaining the astronomical death tolls in these states. Continue to read, and I will do my best to put all of this into a perspective that is digestible, if not perhaps easily digestible.

Table 1 below shows the raw results of over-laying anti-body, or seroprevalence studies, in the NY City area with death data for the same period.

Table 1: Actual Cases, Hospitalizations, Deaths and Mortality as of 5/1 in the NYC area based on on-going seroprevalence surveys and daily tracking of cases, deaths, and hospitalizations.


Seroprevalence surveys: Governor Cuomo's 6/15 press conference--used 5/1 seroprevalence, later dates did not provide granular age data. Age data, Governor Cuomo's 5/4 press conference. Death data for NYC for 5/1 here. Hospital data for NYC for 5/1 here. Case data for NYC for 5/1 here. New York City population data, here. Screenshots for each of the datasets above are provided at the end of this article.

The table below above shows that for children, COVID-19 including those with co-morbidities, in New York City was 10x less deadly than flu (statistics below in Table 2, from the 2017-2018 season--which was particularly bad). It also shows that for all people between the ages of 18-44--including those with co-morbidities--COVID-19 was a little more than 3x more deadly than the flu (flu mortality rates are shown in table two below). This is relative to the number of 18-44 year olds who die from flu, which is 0.02% of those infected. The mortality rate for the same group for COVID-19 is only 0.073%. 3 times is a significant increase, but flu is not something people between 18 and 45 go around in fear of, so a 3x increase should not greatly impact their fear of death. After 45, it starts to get murkier, but as will be seen later in this post, much of that has to do with the ravaging of the nursing home population, the about-to-enter-nursing home population, and the fact that in NYC, lockdowns forced the least healthy portion of the population--poorer essential workers--to bear the immunological brunt of this disease.

Table 2 below shows the population-adjusted incidence of flu cases, hospitalizations and deaths in NYC in the 2018-2018 flu season (a bad flu year). I put this here now, simply so you can make some comparisons to the level of cases, etc. We will be returning to it later for further analysis. It is important to note that in 2017-2018, the flu vaccine was particularly ineffective, at only 40%, with only 40% of people taking it. Thus, this represents not quite an un-vaccinated population, but close. Multiply everything by 1.2 to get what this would look like in an un-vaccinated scenario.

Table 2: Population-adjusted incidence of symptomatic flu, hospitalizations & deaths in NYC during 2017-2018 flu season (most deadly in last decade).

Source: CDC, estimated influenza disease burden by age group.

Now, I'd like to move onto something that is very important. As I noted, NYC has done an exemplary job of separating out those people who died with co-morbidities and those who did not. In Figure 1 below, you see that of New York City's 13,000 deaths by May 1st, only 81 of those were in people who had no co-morbidities.

Figure 1: Confirmed COVID-19 Deaths in NYC as of 5/1


Because we know the incidence of co-morbidities such as obesity, we can back into the number of people who were infected who had no co-morbidities and identify the likely risk of death for each age group for those people without co-morbidities. Let's start first by looking at the major co-morbidities that contribute to death in each age cohort. The following table is taken from New York State, on August 15, 2020. While New York City tracks the number of deaths occurring with co-morbidities, they do not tell you which co-morbidities are most prevalent by age. That is why the table below is so useful.

Table 3: Top-10 Comorbidities Contributing to Death fro COVID-19 by Age Group in New York State, as of 8/15/2020

Looking at the table above, it becomes very clear just how important obesity is in terms of driving mortality for COVID-19. Hypertension, Diabetes, Hyperlipidemia (they are using this to mean obesity, I think), Coronary Artery Disease, Renal Disease, Atrial Fibrillation and Stroke--7 of 10--are tightly linked to obesity. Even amongst COPD, chronic obstructive pulmonary disease patients, 30% are obese. Thus, the only two co-morbidities that are not associated with obesity are dementia and cancer, both of which here may be more proxies for age than actual risk factors predisposing mortality.

In fact, looking at this, it seems that really what we are looking at is not just obesity, but people who have metabolic syndrome, obesity as well as a variety of other co-morbidities that are contributing to that syndrome, and pre-disposing them to Diabetes II and others of these obesity-linked co-morbidities--especially in the youngest cohorts. This is further supported by the fact that in the under-60 cohort diabetes is equally or more highly-represented than hypertension. Amongst diabetics, 54% are obese. Beyond that, renal disease is here represented at roughly 25% the rate of deaths as those with diabetes, exacyly the rate at which renal disease presents in Diabetes II patients, generally. Nationally, obesity accounts for roughly 70% of all hypertension incidence, and approximately half of all Diabetes II incidence, with being overweight contributing another 40% to its incidence. Thus, obesity looks to be a pretty good proxy for co-morbidities, particularly in the under-65 cohort, where these co-morbidities are closely linked to obesity, and where there are very few non-obesity-linked co-morbidities. While there is some prevalence of these co-morbidities outside the under-65 obese population, there is likewise 25% of the obese population that is considered "metabolically healthy."

While with younger cohorts we can use obesity as a proxy for people with co-morbidities, as people age, that becomes less true. Fortunately, there is no reason to guess at the number of people with serious co-morbidities by age--this has been exhaustively studied. Using that data, we can identify the percent of people with co-morbidities by age. Figure 2 below shows the percent of people with co-morbidities by age.

Figure 2: Comorbidities by age in newly diagnosed cancer patients


This data is taken from 27,000 newly diagnosed cancer patients, meaning that these people are if anything more likely to have comorbidities than a normal person. The list of co-morbidities that are considered here is more comprehensive than the list for COVID-19, thus again, this would be likely to show a worst-case scenario for comorbidity presence in the population by age. .

There are only two points where that is not true. The first is with asthma, but asthma's effect on COVID-19 is complicated. For about half of asthmatics, allergic asthmatics, COVID-19 appears to be protective. Other types of asthma it is at best inconclusive. Thus, for our purposes, identifying co-morbidities contributing to death, asthma cancels itself out. Particularly because 30% of asthma incidence is also obesity-linked, meaning that once again, obesity proves to be a good proxy for comorbidities. The other point where the inputs above differ, is related to obesity. Obesity in the study referenced above is defined as a BMI over 38, where obesity for COVID-19 is defined as over 30. Again, to make sure that we are getting the most conservative, i.e. the worst-case scenario, if there is variance between obesity rates, and co-morbidity rates in an age group, we will choose whichever is larger. Thus, for people under 44, we used the obesity rate as a proxy for all co-morbidities, given that the study rate of people with all co-morbidities was lower for that cohort, and for the three older group, we used the study rate, which goes as high as 85% co-morbidity rate for people over 75.

Using these data, I calculated the proportion of infected people likely to have co-morbidities, and subtracted that from the total number of infected people, based on seroprevalence. That allowed me to come up with the rough number of people with zero co-morbidities infected in each age group. From there, I divided the number of deaths amongst people with no co-morbidities in each age cohort by the number of infected without co-morbidities to get to the mortality rate for each age cohort amongst people without co-morbidities.

Table 4: COVID-19 mortality rate by age for NYC residents without co-morbidities, relative to accidental death rates and suicides.

Sources: Seroprevalence & morbidity and mortality data as cited above. U.S. adult obesity rate by age, children's obesity. Annual deaths in NYC by cause of death. Co-morbidities by age.

Looking at Table 4 above, one can see that the estimated mortality rate for people without co-morbidities is vanishingly small, and in most cases, substantially lower than their risk of accidental death--which is predominantly driven by car accidents (no pun intended). It is also interesting to compare the number of suicides that occur in NYC each year, realizing that for healthy people, this risk of suicide is unquestionably being elevated due to our response to COVID-19. The CDC recently reported that 11% of all adults have contemplated suicide since the start of COVID, and more than 25% in the 18-24 range.

The reason that this is so important touches on the whole "herd immunity" question. Not only is the risk of death for people with co-morbidities low, the number of people without co-morbidities is very high--almost 60% in this conservative estimate (thank god for those kids :) The under-65 population without co-morbidities has a significantly lower risk of death due to accidental death or suicide than dying from COVID-19, and they account for nearly a million--almost 60%--of the estimated COVID-infected in NYC. The rates of people without co-morbidities over 65 get quite low, dropping to less than 15% above 75, but even in these populations, their risk of death from COVID is less than their risk of death due to accidental death. This gives much credence to the idea that just because someone is elderly, does not mean they or their doctors should give up if they get COVID. There are very important public health policy implications from all of this, that I will discuss at the end of the post.

Now, if you are obese, you might be freaking out right now. However, mild obesity (BMI of 30-35) is only associated with a 10% increased likelihood of death, moderate obesity (35-40) a 30% increased likelihood, and morbidly obese (40+), a 60% increase. Those are big numbers, but they are big numbers on top of small numbers, so while there is reason for concern, in most cases, obesity alone does not seem likely to bring you above your risk of dying from accidental death. Indeed, the numbers above surely give credence to the idea that it is not so much one thing that predisposes a person to death from COVID, but a host of factors piling one on top of another. For instance, while 54% of people with Type II diabetes are obese, only 12% of obese people have type II diabetes (Type II diabetes has a prevalence of 8.8% nationally)--and again, there are those 25% of obese people who are "metabolically healthy." Use this tool to see what your BMI is.

One final note before we begin to look at the elderly. The data in table four above, as I noted, is taken from the reported death and co-morbidity data in NYC as of 4/30/20--shown again figure 3 below. You will note that for all ages except those under 18, there are some deaths for which co-morbidities are unknown. Most of the patients for whom this data is unknown, are almost certainly the people who have the worst co-morbidities, as the majority of these patients would be people who were admitted to the hospital, then died before the hospital could get complete medical information about them. But let's for argument's sake assume that every one of those patients with unknown co-morbidities had zero co-morbidities. What then would be the risk for people without co-morbidities? For children, the risk would stay the same--0. For people from 18-44, the risk would increase by a factor of 6, so 0.018%, still half the risk of death due to accidental death. For people from 45-64, the mortality rate would jump to 0.165%--3x the risk of accidental death. However, since some portion of this group--those most likely to die from COVID-19--is about to enter a nursing home, for this group, this assumption,the idea that all of the people for whom co-morbidity data is unavailable have zero co-morbidities, is the least likely to be accurate. Indeed, it is surely not an accident that as the age of the deceased cohort approaches the age when some portion of that group might be in a nursing home, the likelihood of having data on co-morbidities is lower. This is precisely because these are likely the patients who were dumped to die at the hospital. Let's dive into that, next.

Figure 3: Deaths in NYC by age and co-morbidity as of 05/01/2020

I noted above that the population-based antibody testing performed by New York State was performed in grocery stores. There are literally zero people in nursing homes who would have been in that sample. And yet, nationally, 40% of all deaths from COVID-19, according to CMS records are amongst the nursing home population (at 1.4 million, this is 0.5% of the total population of the U.S.). This means, that when looking at the seroprevalence results in NYC vs. deaths, we are making an apples-to-oranges comparison.

One of the great frustrations in trying to analyze COVID-19 data, is that the datasets are often not matched up--different dates, or different age brackets, different geographies, or different samples. But usually there's a way to find your way into a reasonable proxy. So that's what we're going to try to do here.

A recent analysis in the New York Times, broke down the portion of each state's COVID deaths that occurred in nursing homes. Somewhat surprisingly, New York State comes in 2nd to last, i.e. having the smallest portion of its deaths attributed to nursing homes--this despite having the second highest deaths/million population nationally, at 1691 deaths/million--behind only New Jersey at 1798.

Figure 4: Nursing home deaths in the United States as of 8/10/20

It does somewhat defy belief that New York state, with 20% of all deaths in the U.S. and only 6% of the U.S. population, and the second highest deaths/capita nationally would have the second lowest percentage of deaths from nursing homes. Honestly, I wouldn't really care about this discrepancy, but for this analysis, this data is necessary for us to be able to estimate the real relative risk for non-nursing home seniors, and senior-adjacent people. But, as with most things, when something don't make sense, if you look deeper into the data, the answers come out of the woodwork.

As it happens, New York State is the only state in the U.S. that reports only those nursing home deaths that occurred in nursing homes. That is, New York state doesn't count nursing home residents who went to the hospital and died there--as does every other state in the U.S.. The number of people who died IN nursing homes in New York state is 6616--with only 7036 cases reported in nursing homes--i.e. roughly an 85% mortality rate for confirmed coronavirus cases in New York nursing homes. But presumably some of of the nursing home patients in New York had the benefit of hospital care, even if it was only to be admitted and die a few hours later. Indeed, a recent investigation of New York nursing homes by the AP, showed that there were 21,000 empty beds, 13,000 more than would be expected. 13,000 of New York's 32,887 deaths would be 40%, or right in-line with the national average.

So let's use these two data points and assume that at any given time, roughly 40% of the cumulative deaths in New York state and New York City are nursing home deaths. Thus, of New York City's 13,106 deaths on 5/1/2020, roughly 5242 would have been in nursing homes. (I think this is a low estimate, given Delaware at 61% was the very lowest of NY's neighbors, all of whom all embraced the deadly policy of forcing or strongly encouraging nursing homes to take back recovering COVID patients--without proof of a negative test before doing so. But hey, we'll give NY the benefit of the doubt and say that they are inline with the national average, not their regional cohorts who followed their lead). We also have data from Kaiser Family Foundation, that tells us the number of nursing home residents per state. In New York State, in 2019 that number was 89,000. We know the average breakdown by age in nursing homes. And of course we know New York City's population relative to New York state's. With these three pieces of data, we can find out the New York City nursing home population in each of these age groups, subtract that number from the total population of that age group, and use that number as a proper basis for applying the seroprevalence data.

Table 5: COVID-19 Mortality by Age in NYC, and Nursing Home Residency, Compared with All-Cause Mortality in NYC

Sources: All-cause mortality population-adjusted for NYC from CDC data; Nursing home mortality. Nursing home expected population, population-adjusted for NYC. Nursing home age distribution, population adjusted for NYC nursing home population.

Table 5 above, shows estimates of the mortality rate for the 65-74 and 75+ populations outside of nursing homes, based on NY seroprevalence data, NYC death data, and estimated nursing home deaths. Table 5 also presents COVID-19 mortality rates by age group relative to all-cause mortality in each group. This comparison provides many interesting lines of inquiry.

First of all, it must be highlighted that overall annual mortality in nursing homes is 50%--the average length of stay in a nursing home is 2.2 years, and since nursing home populations are relatively stable, it makes sense that 50% of residents would die each year--this is supported by CDC data that notes that 28% of all deaths (780,000) occur in nursing homes patients--again, 50% of that population of 1.5 million. The reason to point this out, is to underscore just how fragile this population is--and how exposed it has been due to broad, vs. targeted lockdowns. The general mortality rate for people aged 65-74 who live in nursing homes is 30x that for this population outside of nursing homes--i.e. 1.5% annual mortality rate for seniors between 65 and 74 who are not in a nursing home, and 50% for those in nursing homes. Something like this difference is also likely true for COVID-19, given that COVID-19 exacerbates all of the co-morbidities that are the main drivers for all-cause mortality--heart disease, COPD, diabetes, etc. When you remove nursing home resident deaths from NYC deaths, the mortality rate from COVID-19 in the 65+ population does not look that dissimilar to this group's all-cause mortality. That in itself is significant, because this group's all-cause mortality is driven by just those diseases that COVID-19 preys upon. Of further importance is that there is a population within this 65+ group who will/would be entering a nursing home this year. Those people's mortality would not be off that hyper-mortal group within nursing homes whose year-over-year mortality is 50%--31% the first year, and 70% the second year.

As per usual, I am not drawing attention to this to try and minimize the risks associated with COVID-19. There are two reasons that I want to point this out. First, is to offer some level of comfort to those who are over 65. Those of us who are young and healthy are able to draw a significant measure of comfort from the fact that we are unlikely to die of COVID. It is far easier to disaggregate the data for our cohort. However, given the ultra-high mortality of nursing home populations which is driving mortality figures for the 65+ cohort, the 65+ population that is not in or considering entering a nursing home, should also be able to take a measure of comfort. This group's likelihood of dying from COVID is not much higher than--and tightly linked to--their likelihood of dying from whatever previously existing conditions they already have. For example, the risk of death from COVID-19 for the 65-74 group that is not in nursing homes is 1.2-3%. Their risk of dying at all this year is 1.5%. That their risk of death due to COVID-19 is predominantly born by those who already have significant comorbidities is under-scored by the very low mortality rate due to COVID-19 in the 65-74 group without comorbidities.

The second reason the point this out, is to draw attention to the fact that by trying to protect everyone, we are in effect protecting no one. This, perversely, results in preferential attack of those who are most at-risk, as they are the least able to distance themselves, given their reliance on others.

What's more, it is not as though we don't, or didn't, have a good idea of who those people were and are. They are, predominantly, the same people who are at high-risk from all diseases--those in nursing homes, and those in the severe area of the figure below.

The counter-argument to what I am saying here is that "we can't possibly protect the nursing home and at-risk population if we don't do a lock-down." The truth is that lockdowns unquestionably did NOT protect these populations. The Northeast has proved this in spades. A recently released paper in JAMA showed that in May in a random sampling of nursing home residents 28% were positive for COVID-19. Let me make this very clear. This isn't an antibody sample, this is a sample between May 2 and May 19th of nursing home residents showing that 28% of them had an active infection. Data from CDC seroprevalence surveys from the same time period show that only 5% of the general population in Connecticut had been exposed. Again, the difference in these tests says volumes. Serology, or antibody tests, give a picture of who has been infected over all time. PCR tests--what was carried out in the nursing homes--tell you only who has an infection at given time--and they tend to miss about 30% of active infections. This means that the nursing home population is being exposed--at a minimum--at a 6x rate to the general population. I say at a minimum because again, that 28% was at a given point in time--certainly others within the nursing homes got the disease before and after.

Increasingly, it is looking like lockdowns, and all of our attempts to stop, or slow the virus, have resulted in accelerating its course through our most vulnerable populations. Based on the data above, it seems likely that while the 99.5% of the population that is least at risk for this disease is plodding along towards "herd immunity", the most at-risk 0.5%, our nursing home population, is acquiring it apace--and at a tremendous cost in lives. The same appears to be true of our poor and minority communities. In Figure 5 below, this is underscored by Queens showing nearly double the rate of infection of Manhattan--32% to 17%. And locally, far worse, such as this neighborhood in Queens, where 68% of whose residents have antibodies to the virus. Given that these communities are both a) more likely to live in multi-generational homes, and b) at higher risk for the kind of co-morbidities leading to death from COVID-19, this is particularly unacceptable.

Figure 5: Seroprevalence in NYC by Borough


Given these two facts, 1) the fact that our approaches heretofore have resulted in the highest levels of infection in the parts of our population most likely to die and 2) that a significant portion of our population--possibly as high as 65%, but certainly no lower than 45%--has a 10-fold+ lower risk of dying of COVID than dying due to some accident, it seems like we must re-think our approach--and fast.

Incredibly, what we appear to have with COVID-19 is a scenario where you really can have your cake and eat it to. Allowing the young and healthy to go about their business, while maintaining a temporary, voluntary "social sorting" based on risk would reduce their own risk of death due to suicide, improve the economy, and make it easier to protect the at-risk population. Up until now, the at-risk population has been practically the sole target available to COVID-19. We could provide them with protection that heretofore has been sorely lacking, using a more coordinated, short-term approach to shield them.

How long does it take for COVID-19 to go through a population if left unfettered? We need only look at Bergamo Italy, where 57% of the population has antibodies to the disease. In Bergamo, the disease ripped through in a matter of weeks--this is why their hospitals were overwhelmed. We, too, could allow this disease to go through our population in a matter of weeks, and by using an intelligent "social sorting" approach, we could dramatically reduce the number of deaths. Nor would such an approach overwhelm our hospitals. NYC's hospitals were not overwhelmed, despite the disease being allowed reach almost exclusively the most at-risk. In NYC, the hospitalization rate for those under 65 (including those with co-morbidities), was 1.4%--about twice the hospitalization rate for this group for flu. But remember, in NYC, the people exposed to COVID-19 were those most likely to have complications resulting in hospitalization.

Obviously, one can't infect deliberately infect people--although I think it can be argued that thus far, encouraging those who can stay home to stay home is effectively deliberately infecting those who can't. But we'll leave that there... Anyway, while we would never deliberately infect people with this virus, we can provide people with clear, but nuanced information that allows each person to understand their risk relative to other activities, and that allows them to make their own choices, based on that information. If this were done in a coordinated fashion, where we also provided some means for those at-risk populations to really protect themselves, as the healthy and wealthy have been doing up until now, we might be able to significantly reduce the remaining deaths due to COVID-19.

American public health isn't a big fan of nuance. American public health officials don't believe the American public is capable of understanding a nuanced message. This is why they constantly give us blanket guidance--even when such guidance does nothing to improve outcomes (I'll save that for another post, but trust me, this isn't the first time that the expedient of an easy message has worsened health outcomes). This is why our public health officials insist on trying to scare healthy, young adults into believing that they are also at-risk--because our public health officials think we can't understand relative risk, and different risk profiles. That young adults have somewhat ignored public health officials' advice is reflective of the fact that despite the CDC refusing to disaggregate the data, these young adults have nonetheless been able to internalize a more nuanced understanding of their risk. How much more effective could we be if we were to take the 160,000 deaths we have and provide truly targeted, nuanced messages based on risk? How many more lives could be saved if we were to provide people in multi-generational homes where someone had to work, with ways to protect their at-risk family members? How many people with co-morbidities could we save if we told them their real risk based on the severity and concomitance of those co-morbidities, informing those who were truly at risk to prepare to "hang-tight" for a few more weeks, and that after that they could hug their grandkids? The answer is, we would save far more than we will save by continuing to obscure cohort-based relative risk.

There are groups for whom this disease is thousands or tens of thousands more risky than others. It is verges on criminal to obscure this acute risk of a small group of people by lumping their risk in with large groups that bear almost no risk. Nonetheless, our public health officials continue to shame young people for getting together, despite the fact that not being around other people is significantly increasing those young people's risk of death due to suicide. This is also despite the fact that these same smug public health officials have been the architects of policies that if judged on their results look more like efforts to reduce entitlement spending by killing off the old and the poor than efforts to save those constituents' lives.

Appendix 1--If I were in charge.. :)

If I were a public health official, these would the the policies that I would try to put in place. I would call it "Operation Hugs". Because that is one of the things I have missed more than anything else since this began, despite not being a "hugger."

  1. Convey that our efforts here-to-fore, while well-intentioned have done little to protect the most at-risk in our communities, and that in many cases they may have put them at greater risk. Convey that for this reason it is necessary that we make a change in our public health policy that is cognizant of the fact that while we can't stop this disease, we can do a better job protecting the most at-risk members of our society by focusing all of our efforts on that population. This should be coupled with clear guidance about relative risk by age and co-morbidity, with the caveat that no one is at zero risk, but that this approach allows each individual person to take control of their lives and make their own decisions, no longer relying on the false protections of the government. Point out that while we would never encourage someone to actively seek to get the virus, that there are many who rightly perceive their risk of serious consequences from contact with the virus to be low, and indeed that many of those same people may be at significantly elevated risk of adverse consequences--including, but not limited to, suicide--due to our attempts to stop the virus.

  2. Communicate real risk based on age and co-morbidities, putting numbers into a context that normal people can understand.

  3. Provide detailed information about co-morbidities related to deaths by age group, and by race within age groups.

  4. Clearly delineate between risk-of-death of elderly who are in nursing homes, versus the elderly who are not.

  5. Provide vouchers or some other tool to allow people who have at-risk family members in their homes to move, or move their at-risk members, to a safer, less-exposed location for a short period of time.

  6. Establish a benchmark for the number of weekly deaths and hospitalizations that constitutes a "flattened curve." I would propose the high point of the 2017-2018 flu season for each state--most states are currently below this. If this threshold is exceeded, re-introduce caps on gatherings or large events only.

  7. Specify a set date within the next 4-6 weeks when all restrictions on movement, association, mask requirements etc, would be lifted. This would allow all people, especially those at-risk, or those living with at-risk people to prepare during the ostensibly-safer status quo.

  8. As part of this preparation period, inform low-risk people of the need to make plans to keep clear of higher risk people in their lives once their societies reopen, and to make plans to provide mental, material, and emotional support for those people.

  9. Set-up a register for at-risk people without family support to get support from their communities and local governments.

  10. Provide guidance on the most effective masks for protecting people FROM the virus, and how to wear them to maximize protection.

  11. Mass-produce and distribute said masks to those people who identify themselves as being at high risk.

  12. Once opening date is reached, allow people and businesses to make decisions on what they chose to do to limit their exposure, and their clients. Some businesses would cater to people with lower risk profiles, other to those with higher risk profiles. Others would create socially-sorted hours, or locales. Presumably those businesses (like grocery stores) who served all people, would create policies that made their most at-risk customers feel safe--or they would not be patronized by those groups. If not, provide a government incentive for essential stores to create a particularly hygienic environment for at-risk people. It must be made clear that just as world governments have been unable to protect people, businesses also are not going to be able to do this, and that the onus of prevention remains on the individual, laying special emphasis on the fact that our prior efforts have resulted in higher deaths amongst the most vulnerable, and we must adjust course in order to prevent that from continuing to happen.

  13. Organize some federal program to provide financial support to businesses that would enable them to allow high risk workers to stay out of work and continue to be paid at some level for 4-6 weeks while the disease continues the rest of its trajectory. Provide guidance to businesses on which workers ought to qualify for said designation. Strongly discourage usage by low-risk groups.

  14. Organize services to support at-risk people, and discourage those services from being used by low-risk groups--i.e. even utilizing commercials grocery deliver services should be discouraged for low-risk people during the initial four weeks of "perstroika," to ensure that these services are available for higher risk people.

  15. Encourage mask-wearing in mixed non-grocery, non-transit environments amongst at-risk only, as means of conveying to no-low-risk that they need to keep their distance and be solicitous of those people wearing masks to protect themselves.

  16. Provide guidance for states on performing community seroprevalence, including breakdown by age, zip code, race, and nursing home status.

  17. Provide guidance for states on how to conduct population-based T-cell testing to understand prior population exposure and immunity to COVID-19, as well as those people with cross-reactive T-cells that are not specific to COVID-19, but which in many cases appear to be protective in up to 50% of the population who has not had COVID-19.

  18. Encourage un-restricted play amongst children who are at low-risk, ensuring that higher risk children (who are extremely few) are allowed to stay back, but with the promise of being able to re-enter the fray in only a few short weeks.

Appendix 2, more sources:

Figure 1


Figure 2

Figure 3: Disaggregated co-morbidity prevalence by age

323 views0 comments