Vaccine mandates for employees are emerging as the most feasible and effective policy to reduce the risk of disastrous COVID-19 outbreaks in the workplace.
Higher education and healthcare are leading the way in vaccine mandates, while other sectors are beginning to follow their lead.
State laws barring vaccine mandates will become increasingly unenforceable as they come into conflict with private employer initiatives to protect workers and consumers.
Concerns that verification of vaccination status is unworkable or that the Food and Drug Administration has approved COVID-19 vaccines only on an emergency basis will prove to be overblown.
While a clearly articulated, unified federal policy on vaccine mandates remains lacking, independent agencies such as the Equal Employment Opportunity Commission, the Occupational Safety and Health Administration, and the Centers for Disease Control have already begun to step into the void.
In the absence of clearly articulated federal and state policies, private employers have had to adopt their own vaccination policies to reduce the risk of a disastrous COVID-19 outbreak among returning workers. Higher education and healthcare have led the way in establishing vaccination mandates for students and employees, and there are strong indications that other sectors will follow suit. This trend is reinforced by the emergence of markets for vaccination-only services, as well as market-based approaches to vaccination verification. While workplace vaccine mandates have already proved feasible and effective, the alternative of mixing unmasked vaccinated with masked unvaccinated workers remains untested. State laws that prohibit vaccination mandates will prove unenforceable as they increasingly conflict with the policies of large private employers. Vaccine mandates are not only inevitable, but they are sound public policy. The federal government can play an important role in endorsing workplace mandates while protecting the most vulnerable and reducing disparities in disease burden.
We have (nearly) forgotten what we learned from the anti-smoking campaigns of the sixties and seventies.
On December 1, 1966, a lawyer in his 20s named John Banzhaf – later to become a professor at George Washington University Law School – petitioned WCBS-TV in New York City to provide free air time to respond to the pro-smoking views embedded in the cigarette commercials that the television station had been broadcasting. After two years of litigation, the United States Court of Appeals for the District of Columbia Circuit, in an opinion written by Chief Judge David L. Bazelon, found in Banzhaf’s favor. The Court upheld a ruling by the Federal Communications Commission requiring every TV station to broadcast one anti-smoking public service announcement during prime time for every three cigarette commercials it aired.
There followed a sustained period of prime-time anti-smoking ads, running night after night until the U.S. Congress finally banned all TV and radio advertising of cigarettes effective January 2, 1971. From 1968–1969, per capita cigarette consumption in the U.S. fell by 4.6 percent — a drop in smoking rates not seen since the lung cancer scare of December 1953, and not seen again until a major, nationwide increase in the federal cigarette excise tax in 1983.
Here, we reproduce what was arguably the most influential public service advertisement of the period, sponsored by the American Cancer Society, lasting all of eight seconds.
Combining Trust in Healthcare Providers with Peer Influence
The American Cancer Society public service ad worked because it combined two critical themes. First, the television-viewing public trusted healthcare providers – physicians, in particular – to offer reliable advice. Second, the salient reference to a large round number with lots of trailing zeros added the element of peer influence. The fusion of these two themes sent a clear message to many who doubted a connection between smoking and disease, or who felt they didn’t have the willpower to quit: All those doctors must know something.
So let’s bring back the nearly forgotten motif from a half-century ago. To that end, we first need to do some arithmetic. An estimated 84.5% of the nation’s 3.8 million registered nurses work in healthcare. There are 985,000 actively practicing physicians in the United States. A recent Kaiser Family Foundation survey, in the field during February 11 – March 7, 2021, reported that 68 percent of doctors and nurses in the United States had received at least one vaccine dose. (We do not include those expressing an intention to be vaccinated.) Putting all the numbers together, and recognizing that the survey data are already two months old, we end up with a conservative tally of at least 3 million vaccinated doctors and nurses.
So, here is our reincarnation of the American Cancer Society public service ad of 1969.
Social Desirability Bias: The Achilles Heel of COVID-19 Vaccine Surveys
In 1978, a decade after the prime-time, anti-smoking TV ads had begun to put the brakes on the 20th century tobacco juggernaut, Ken Warner, a young economist on the faculty of the University of Michigan School of Public Health in Ann Arbor, pointed out how survey-based estimates of cigarette smoking rates fell increasingly short of actual consumption based on tax receipts. With smoking rates then on the decline, Prof. Warner worried that survey respondents were ever more reluctant to tell the interviewer that they were engaging in a “socially undesirable” activity.
Even at the time, social desirability bias (SDB) had been a concern of survey experts. Rather than reveal their true intentions or beliefs, the problem was that people would tell the interviewer what they thought the interviewer wanted to hear. During the intervening 43 years, there has been an abundance of studies of the influence of SDB in surveys of such personal health-related behaviors as substance abuse and physical activity. Extensive reviews have been published. The problem has persisted even in apparently anonymous online surveys. In the modern version of SDB, survey researchers worry that respondents, ever wary of empty assurances of anonymity, will click on the button they want the online survey software to record.
The Kaiser Family Foundation vaccine monitor recently reported that the proportion of U.S. adults who said they’d already received one dose of the vaccine or intended to do so as soon as possible increased from 61 percent in March to 64 percent in April. Polling data from the Pew Research Center, in the field on February 21, indicated that 69 percent of U.S. adults either had already received one dose of the vaccine or were probably or definitely planning to get vaccinated. Yet the Center Center for Disease Control’s Vaccine Tracker reports that 58 percent of the U.S. adult population has actually had at least one vaccine dose and only 44 percent had been fully vaccinated. These data would suggest that there remains considerable excess demand for the COVID-19 vaccine. Yet there are signs that demand has slowed, and some states are returning allotted vaccine doses because of excess supply.
The concern here is not just that a particular survey at a particular point in time may have overstated vaccination intentions. The more knotty problem is that the increasing promotion of vaccination itself enhances the magnitude of the social desirability bias and thus exaggerates the apparent rise in vaccine acceptance. That’s exactly what Prof. Warner worried about in an entirely different context more than four decades ago. What made smoking increasingly undesirable way back then was the growing movement of the 1970s to restrict smoking in public places.
Continuum of Resistance to Behavior Change
As the daily pace of vaccination has decelerated in the United States, a new wave of commentary has become fashionable. The basic theme is that vaccine hesitancy is deeply ingrained in some people because of their distrust of the medical establishment or authority in general, or their core beliefs about personal freedom or the inviolability of the body, or their hardened sense of fatalism.
What we learned from the anti-smoking campaigns is that resistance to behavioral change lies along a continuum. At one end of the spectrum, tens of millions of smokers quit when they got the facts about tobacco and health, and tens of millions of people lined up to get their shot when they got the facts about the efficacy of the COVID-19 vaccines. At the other end, many smokers remained addicted and in denial, but came around when they could no longer bear the ostracism of having to smoke outside the office in the cold, or when they had to pay more for life insurance, or when they suddenly found themselves tethered to a heart monitor in the intensive care unit. Others who remain hesitant to get vaccinated will similarly come around when they find out – as inevitably will be the case – that proof of vaccination has become the access key to open the portal to life as we once knew it.
Vaccination and seropositivity. The hyper-vaccinated Upper East and Upper West Sides. African American vaccination rates lag those of Spanish speakers.
In A Deeper Look at Vaccination Inequity in Los Angeles County, we delved into the inequities of the COVID-19 vaccine rollout in the largest county in the United States. Here, we head eastward to the Big Apple. This time around, we go beyond concerns of fairness and begin to worry about vaccine hesitancy.
Vaccination and Seropositivity
The map on the left-hand side of Figure 1 shows the percentage of the adult population that has so far received at least one dose of a COVID-19 vaccine. The map on the right shows the corresponding percentage of adults who have so far had a positive antibody test for SARS-CoV-2. The two maps look like negatives of each other.
As we noted in A Deeper Look, many people who’ve already come down with COVID-19 don’t think they need to be vaccinated. (Consider Sen. Rand Paul’s provocative tweet of November 17. ) While there is now a strong medical case that those already infected should get vaccinated as well – In fact, we’re doing just that at our own community health center. – the perception that one is naturally protected is nonetheless a critical determinant of vaccine demand.
Prior studies of equity in vaccinationrollout – and, in particular, studies of disparities in racial and ethnic vaccination rates – have ignored this important negative relationship between seropositivity and vaccination rates.
Map comparisons may be illuminating, but they’re hard to work with. So let’s convert the negative relationship for New York City into a two-way plot.
Figure 2 plots the percentage of adults who have had at least one COVID-19 vaccination against the percentage of adults who have had a positive antibody test. As in A Deeper Look, this two-way plot allows us to quantitatively test the relationship between past COVID-19 infection and the demand for vaccination.
In contrast to our two-way plot for Los Angeles County, here we rely on New York City’s database of antibody tests for prior COVID-19 infection. The estimated slope of the fitted line is –0.71, which means that, in the absence of any other confounding factors, 71 out of 100 previously infected individuals have so far chosen to deter vaccination.
Vaccination itself may eventually cause some currently marketed antibody tests to turn positive. If so, the trend seen in Figure 2 would understate the actual deterrent effect of prior infection on vaccine demand, so that the real number would be greater than 71 out of 100. Still, this bias is likely to be small, as the vast majority of antibody tests were done before the U.S. vaccination rollout.
In Figure 2, we’ve drawn attention to a few specific Zip Code Tabulation Areas (ZCTAs). At the upper left, there is a cluster of nine ZCTAs located in the Upper East Side and Upper West Side, affluent areas bordering Central Park. At the extreme right, we see Corona in Queens (11368), where 48 percent of adults so far tested have had detectable antibodies against coronavirus, followed by Borough Park in Brooklyn (11219), where the seropositivity rate has reached 46 percent. Tracking the outliers from right to left, we see East Elmhurst (11369), Jackson Heights (11372) and Elmurst (11373) in Queens. Along with Corona, these contiguous ZCTAs formed the Queens-Elmhurst hotspot that emerged in late March 2020 in the early days of New York City’s epidemic.
Upper East Side – Upper West Side
Figure 2 should not be interpreted to mean that prior SARS-CoV-2 infection is the only determinant of the likelihood of vaccination. It’s just that we need to take prior infection into account when we analyze other potential determinants.
To that end, Figure 3 shows the influence of household income. This new figure is the same as Figure 2 except for the color coding. The darker-colored points represent ZCTAs with a median household income of at least $67,800 (the median of the medians), while the lighter-color points show the ZCTAs with a median household income below that threshold. The data on income and other demographics come from the Census Bureau’s American Community Survey.
If higher household income in fact enhances vaccination rates – even after accounting for prior infection – then we would expect to see the darker-colored points situated above the line and lighter-colored points situated below the line.
Figure 3 shows us that income indeed matters . Its influence, however, is concentrated in the cluster of outlier points in the upper left, all of which correspond to ZCTAs on the affluent Upper East Side and Upper West Side, where median family incomes exceed $120,000 annually. To be sure,there is an obvious relationship between income and the likelihood of having ever been infected. However, in contrast to our findings for Los Angeles County in A Deeper Look, there is no income gradient in vaccination rates at any given level antibody-test positivity.
Figure 4 shows the corresponding plot to gauge the influence of age. Again, we’ve copied over Figure 2, but this time the darker color-code points correspond to those ZCTAs with a higher proportion of senior citizens. The figure confirms that age matters, even when we take prior infection into account. Like Figure 3, the influence of age is concentrated in those ZCTAs with the lowest prior infection rates, especially the cluster located on the Upper East Side and Upper West Side.
Black and African Americans
Figure 5 once again reproduces our basic Figure 2, but this time we’ve broken down the data points by the proportion of individuals in each ZCTA who describe themselves as Black or African American. In contrast to Figures 3 and 4, we’ve shaded the data points in three colors. The lightest shade represents 84 ZCTAs with less than 10.3 percent (the median across all ZCTAs), while the two darker shades identify the 83 ZCTAs with a proportion of Black or African Americans above that cutoff. The very darkest shade represents 47 ZCTAs that are further distinguished as being significantly below the fitted line.
Figure 5 shows a pattern quite different from those in Figures 3 and 4. While ZCTAs with a higher household income or a higher proportion of elderly residents have higher vaccination rates, the opposite trend is observed for the proportion of Black and African Americans. Moreover, the lower vaccination rates among ZCTAs with more Black and African Americans is seen for a wide range of antibody-test positivity.
Figure 6 uses this color scheme to identify the corresponding ZCTAs in our map of New York City. As above, the two darker shades identify those neighborhoods with a higher proportion of Black or African American residents. What is most striking is that the very darkest shaded ZCTAs tend to be grouped into clusters of spatially contiguous zones.
Figure 7 repeats the same exercise for the proportion of persons in each ZCTA who describe themselves as Spanish speakers. In contrast to the case of Black and African Americans in Figure 6, a higher proportion of Spanish speakers is associated with a higher vaccination rate. This is particularly the case for the ZCTAs that made up the Queens-Elmhurst hotspot described in Figure 2.
To construct Figure 7, we measured the proportion of individuals in each community describing themselves as Spanish speakers. We saw virtually the same pattern when we instead measured the proportion of individuals in each ZCTA who described themselves as Latino or Hispanic. The only difference was that the median proportion of self-described Latino/Hispanic persons was 19.5 percent.
Figure 8 shows the results of our statistical breakdown of the influence of all five factors on the likelihood of vaccination to date in New York City. The basic idea is to estimate the effect of each individual factor while holding the remaining four factors constant. The details, which mirror the approach we took for Los Angeles County in A Deeper Look, have been relegated to the Technical Notes below.
A low seropositivity rate (that is, fewer than 26.4 of adults having tested antibody-positive for coronavirus) was the most important factor. The proportion of adults vaccinated was 6.4 percentage points higher in those ZCTAs with a low seropositivity rate. The next two most important factors were a high median household income (4.6 percentage points) and a high proportion of seniors (4.3 percentage points). A high proportion of Spanish speakers raised the likelihood of vaccination by 2.9 percentage points, while a high proportion of Black/African Americans lowered the likelihood of vaccination by 2.2 percentage points.
A Cautionary Note on Vaccine Hesitancy
It comes as no surprise that the rollout of COVID-19 vaccinations in New York City exhibits significant evidence of economic inequality. What is more unexpected is that the inequality is so highly concentrated in a small area bounding both sides of Central Park in Upper Manhattan.
But that is not the real take-home lesson of this deeper look at vaccination rates in New York City. Once we account of other factors influencing demand or access – prior rates of COVID-19, median household income and proportion of senior citizens– we find that Spanish-speaking individuals have a higher than expected vaccination rate, while Black and African American individuals have a lower than expected rate.
One partial explanation is that the city government’s prioritizationof vaccine distribution, in combination with private efforts, is already having an impact – especially in those Latino communities in Queens that were hardest hit by COVID-19 last spring. That may explain why the data points for East Elmhurst (11369), Jackson Heights (11372) and Elmurst (11373) in Queens are so far above the line in Figure 7.
But what about the finding that the data points for predominantly Black and African American communities remain below the line, as we saw in Figure 5? And why do the communities farthest below the line seem to be clustered, as we saw in the map of Figure 6?
Perhaps we’re simply observing a temporary state, a condition of disequilibrium. As vaccines become more plentiful, the divergence between the two groups may disappear. The deeper concern, however, is that we may be confronting a serious problem of persistent vaccine hesitancy.
Spatial Clustering and Peer Effects
As economists well know, the spatial clustering of ZCTAs in the map of Figure 6 has two competing interpretations. On the one hand, the individuals in these clusters may simply share things in common that have nothing to do with the demand for or access to COVID-19 vaccines. On the other hand, there is the distinct possibility of peer effects. That is, individuals in close social proximity learn from each other.
Much has already been written about the drivers of vaccine hesitancy among Black and African Americans. Our findings here suggest that the underlying beliefs driving vaccine hesitancy are learned, shared, and reinforced through highly local, personal interactions.
When a patient tells us, “I just don’t trust those vaccines,” he’s really saying, “We don’t trust those vaccines.” When another patient tells us, “I heard that people have died from the second dose,” she’s really saying, “My core influential group says that people have died from the second shot.”
If so, then our task as physicians and nurses is not simply to factually and dispassionately allay our individual patients’ concerns about vaccine safety and effectiveness.
We need to leave the comfort of our examining rooms and step out into the community.
As in A Deeper Look, we specified a multivariate statistical model of vaccination rates across 167 Zip Code Tabulation Areas (ZCTA), where each of the five explanatory variables was binary, that is, either low or high. For each explanatory variable, the cutoff between low and high was the median value for the sample. Table 1 below shows our detailed results.
Binary Explanatory Variable
Low Confirmed Seropositivity
High Median Household Income
High Proportion Aged 65+
High Proportion Spanish Speakers
High Prop. Black/African American
Table 1. Linear regression of Percent Vaccinated against five binary variables: Low Seropositivity (up to 26.4%); High Median Household Income (greater than $67,800); High Proportion Aged 65+ (greater than 26.2%); High Proportion Spanish Speakers (greater than 12.4%); and High Proportion of Black/African Americans (greater than 10.3%). The data set included 167 ZCTAs with adult population greater than 9,000. The regression was weighted by the number of adults tested for SARS-CoV-2 antibodies in each ZCTA. The two right-most columns show the bounds of the estimated 95% confidence intervals. The adjusted R-squared statistic was 0.53. The estimated difference in coefficients between the High Proportion of Spanish Speakers and the High Proportion of Black/African Americans, equal to 5.08 percent, was statistically significant at the level p = 0.0005.
The use of binary explanatory variables allowed us to compare the magnitudes of each of the variables’ estimated effects. A high proportion of Spanish speakers in a ZCTA was associated with a 2.92 percentage point increase in the likelihood of vaccination, while a high proportion of Black or African Americans was associated with a 2.16 percentage point decrease. The spread between these coefficients, equal to 5.08 percentage points, was statistically significant.
Table 2 shows the results of an alternative linear multivariate model where each of the explanatory variables is continuous. The estimated coefficient of Percent Seropositivity was –0.38 (95% CI –0.49 to –0.27). That is, once we took all our explanatory factors in account, the impact of prior SARS-CoV-2 infection on vaccination rates was reduced by about half.
Continuous Explanatory Variable
Median Household Income ($000)
Percent Aged 65+
Percent Spanish Speakers
Percent Black or African American
Table 2. Linear regression of Percent Vaccinated against five continuous variables. The adjusted R-squared statistic was 0.53.
The decades-long effort to produce a workable HIV vaccine has hardly been a waste of public and private resources. To the contrary, the scientific know-how acquired along the way has served as the critical foundation for the development of vaccines against the novel, pandemic SARS-CoV-2 virus. We retell the real-world story of HIV vaccine research – with all its false leads and missteps – in a way that sheds light on the current state of the art of antiviral vaccines. We find that HIV-related R&D had more than a general spillover effect. In fact, the repeated failures of HIV vaccine trials have served as a critical stimulus to the development of successful vaccine technologies today. We rebut the counterargument that HIV vaccine development has been no more than a blind alley, and that recently developed vaccines against COVID-19 are really descendants of successful vaccines against Ebola, MERS, SARS-CoV-1 and human papillomavirus. These successful vaccines likewise owe much to the vicissitudes of HIV vaccine development.
The SARS-CoV-1 outbreak in 2002-2003 and the continuing reintroduction of MERS-CoV on the Arabian Peninsula a decade later sent a clear message to the public health and scientific communities well before the December 2019 outbreak of SARS-CoV-2 in Wuhan, China. At any minute, a novel, lethal respiratory coronavirus could emerge with the potential for pandemic human-to-human spread. Piggybacking on the major advances in virology, immunology and molecular biology achieved during the scientific confrontation with HIV, researchers already knew that human coronaviruses were positive-sense RNA viruses with a key spike glycoprotein that protruded from viral envelope. They already knew that the spike glycoprotein could bind to viral receptors on host target cells. The spike glycoprotein of SARS-Cov-1, they knew as well, could bind a receptor called ACE2 in human lungs. They knew it was coming and, at least from the scientific point of view of vaccine development, they were ready.
There’s more to the story than a simple comparison of two maps. The sources of inequity are systemic.
The Los Angeles Times recently posted two maps in an attempt to demonstrate the striking inequities in the COVID-19 vaccine rollout in Los Angeles County, far and away the largest county by population in the United States. One map showed the geographic distribution of vaccines to date, while the other showed the corresponding distribution of recent COVID-19 cases throughout the county.
In Figure 1 above, we display our own rendering of these two maps. The distribution of vaccinations on the left certainly looks different from the distribution of recent cases on the right. In fact, except for the broad swath of the San Gabriel Mountains and Los Angeles National Forest that cut across the county, the two maps look like negatives of each other.
But that doesn’t tell the whole story.
People who’ve already had COVID-19 don’t think they need to be vaccinated.
There is, in fact, an entirely plausible, alternative explanation for the apparent negative relation between the rate of vaccination and the incidence of COVID-19 in a community. In particular, people who have already come down with COVID-19 do not perceive the need to be vaccinated.
We could debate the science underlying the need for past sufferers of COVID-19 to get vaccinated. On the one hand, recent evidence suggests that the natural immunity acquired from a SARS-CoV-2 infection is indeed lasting, at least for 6-8 months. On the other hand, a single dose of an mRNA vaccine appears to confer significantly moreimmune protection in a previously infected individual than in an infection-naive person.
But this isn’t about scientific need for a vaccination. It’s about the perceived need for a vaccination. As economists would put it, prior infection reduces the demand for vaccination.
A Two-Way Plot
In Figure 2 above, we’ve plotted the percentage vaccinated in each community against the cumulative number of confirmed COVID-19 cases in that community. We’ve restricted our plot to the 204 communities within Los Angeles County that have a population of 10,000 or more. (Technically, these communities are called countywide statistical areas, or CSAs.) The size of each data point corresponds to community’s population. Strictly speaking, the horizontal axis measures the cumulative proportion of individuals who ever had a confirmed case of COVID-19, while the map at the right in Figure 1 shows the geographic distribution of recently confirmed cases. Still, the two variables are highly correlated.
The regression line fitting the data points in Figure 2 has a slope of –0.85. This would mean that for every individual who has had a confirmed case of COVID-19 since the epidemic began, 0.85 individuals deferred vaccination.
There is plenty of evidence that the number of confirmed cases substantially understates the actual number of cases of COVID-19. However, a lot of those unreported cases have been asymptomatic people who still don’t know they were infected. When it comes to our analysis of the effect of past infection on the perceived need for immunization, those people don’t count.
To be conservative, let’s assume that the true number of symptomatic infections is 25% greater than the confirmed number. Then our slope would drop to –0.85 / 1.25 = –0.68. That would still imply that most of the negative correlation seen in the two-map comparison can simply be accounted for by deferrals among those who already got sick.
The outliers offer a clue.
Figure 2 has a number of notable outlier points. We’ve annotated the five most obvious ones. With the possible exception of the Century City neighborhood of Los Angeles, where you can still buy a condo for under $1 million, housing prices in the neighborhoods of Beverly Crest, Encino, and Tarzana, as well as the City of Beverly Hills are way, way through the roof. It would be a delicate understatement to say that these outliers are affluent communities.
That gives us a clue that, quite apart from the influence of prior SARS-CoV-2 infection, income has played a role in the distribution of vaccinations to date.
Taking Income into Account
In Figure 3 below, we’re redrawn our two-way plot, coloring the data points according to the median household income of each community. (We set the cutoff at $65,200, the median of the medians.) The higher income communities tend to be clustered at the left, where the cumulative incidence of COVID-19 infection is relatively low. But we can also see a vertical gradient, with lower income communities situated below higher income communities. This gradient implies that, for any level of cumulative COVID-19 infection, higher income still yields a higher vaccination rate.
What about the elderly?
During most of the rollout to date, persons aged 65 years or more were classified as high priority for vaccination. So, in Figure 4 below,we carried out the same exercise as in Figure 3, recoloring our basic two-way plot to distinguish communities on the basis of the percentage of persons aged 65 or more.
Comparing Figures 3 and 4, we learn that the higher-income communities in Los Angeles County also tend to have a higher proportion of elderly persons. (The correlation coefficient was 0.78.) This strong correlation creates a nontrivial problem in the interpretation of our cross-community data. We had the impression from Figure 3 that higher income communities have fared better. But that may merely be a consequence of their higher proportion of persons 65+
An Attempted Statistical Solution
It is arguable that, at this juncture, we’ve reached a logical impasse. We might be able to pick out two communities which differ only by income and not by the proportion of elderly persons or the percent infected. But that sort of comparison would not necessarily tell us much about the equity of the vaccine rollout in the county as a whole.
One way out is to formulate a statistical model of each community’s vaccination rate in relation to our three factors: the cumulative proportion of confirmed COVID-19 cases; the median household income; and the percentage of persons aged 65 or more. We’ve done this in Table 1 below.
Binary Explanatory Variable
Low Confirmed COVID-19
High Median Household Income
High Proportion Aged 65+
Table 1. Linear regression of Percent Vaccinated against binary three variables: Low Confirmed COVID-19 (up to 8.9%); High Median Household Income (greater than $65,200); and High Proportion Aged 65+ (greater than 13.1%). The data set included the 204 CSAs in Figures 2–4. The regression was weighted by each community’s population. The two right-most columns show the bounds of the estimated 95% confidence intervals. The adjusted R-squared statistic was 0.60.
In Table 1, we treated each of the three explanatory variables as binary, that is, as either high or low. That allowed us to compare the magnitudes of each of the variables’ estimated effects. The table tells us that a low proportion of confirmed COVID-19 cases is associated with a 3.64 percentage point increase in a community’s vaccination rate. That’s very close to the estimated impact of a high median income. A high proportion of elderly persons, however, had an impact that was significantly lower.
For more on our statistical analysis, see the Technical Notes below.
A Deeper Look
Our results in Table 1 give us a reasonable idea of the magnitudes of the effects involved. A community with a low percentage of confirmed COVID-19 cases, with a high median income and a high proportion of elderly persons would have a vaccination rate 3.64 + 3.48 + 1.90 = 9.02 percent greater than a community at the other extreme. That’s the height of the bar in Figure 5 below, which shows each of our three estimated effects stacked atop one another. With a median vaccination rate of 13.1% to date, a combined increase of 9.02% is no small effect.
Our results confirm that the simple comparison of two maps in Figure 1 gives a distorted picture of the extent of inequality in the vaccine rollout to date in Los Angeles County. Still, when we took a deeper look at the data, we confirmed that a community’s median household income indeed had a significant positive impact on its vaccination rate.
How Inequality Works
Not a workday goes by without our witnessing first-hand the enormous inequities in access and outcomes that have been baked, seared, engraved and branded into our so-called system of healthcare. That our system of allocating lifesaving vaccinations is likewise unfair comes as no surprise.
But what is it about our system that makes the distribution of COVID-19 vaccines unfair?
Let’s put aside anecdotal evidence that persons with privilege have been able to jump the queue, and talk instead about systemic factors.
First, there is the hidden administration charge. When a patient receives a COVID-19 vaccine, he or she incurs no out-of-pocket expense. But the healthcare provider who administers the vaccine can charge the patient’s insurer for the service. Medicare, in particular, pays $28.39 for the administration of a single-dose vaccine and $16.94 for the first dose and $28.39 for the second dose of a two-dose vaccine. California Medicaid appears to be paying the same rate. This hidden administration charge creates a strong incentive in favor of vaccinating insured individuals. And to the extent that private insurers pay at higher rates, it creates an even stronger incentive to vaccinate privately insured individuals.
Second, there is the web-based sign-up. It obviously makes sense to set up an online appointment system. But the initial, crash-prone website was overloaded with page after page of required information to be filled out. In fact, one of the hidden functions of the website was to require registrants to upload images of the front and back of their insurance cards to make sure that providers could bill for the administration fee. This sign-up system heavily favored those with Internet access, especially those with laptops who could simultaneously open up ten browser windows – each with the required information automatically inserted by auto-fill – until they finally got in.
Third, there are the automobile-based vaccination sites. At the risk of overstating the obvious, lower-income people who rely on public transportation simply can’t wait in line for hours in their cars until it’s their turn.
What’s more, community health centers, with their physicians, nurse practitioners and other providers standing ready to give personalized counseling, take full advantage of the trusted advisor model, which has been receiving increasing empirical support.
What if you heard that you should not get a mammogram right after your COVID-19 vaccine shot? Or what if you’re wondering whether the skin test for tuberculosis you had last week might interfere with your upcoming COVID-19 vaccine? Or what if you’re allergic to dust mites? Or what if you’re just plain suspicious of authoritative statements by experts who, despite appearing to speak English, don’t seem to be understandable?
Do you think a busy vaccinator at a mega-pod site will answer your questions? Or a busy pharmacist at one of the new sites to be rolled out in March 2021? With the best of intentions, they’ll suggest that you call your doctor for advice.
Statistical models may help summarize complicated relationships between multiple variables, but they may also entail hidden, unverifiable assumptions. One such assumption here is that the contributions of our three factors are independent of each other. Still, alternative statistical models that we tested did not show any significant interactions between our explanatory factors. Another is that our binary distinction between high and low levels of each variable obscures more complicated nonlinear relationships. Our tests of models with continuous explanatory variables did not appear to support this possibility.
One such alternative test was a linear model with continuous explanatory variables, rather than the binary explanatory variables shown in Table 1. As shown in Table 2 below, the estimated slope coefficient for the cumulative percent of confirmed COVID-19 cases was -0.37 (95% CI, –0.51 to –0.23), which was less than half of the slope in the bivariate plot of Figure 2. That is, once we took all our explanatory factors in account, the impact of prior SARS-CoV-2 infection on vaccination rates was attenuated considerably. The estimated coefficient of a community’s median household income was 0.077 (95% CI, 0.046 to 0.108). That is, a $10,000 increase in a community’s median household income was associated with an increase of 0.77 percentage points in the vaccination rate.
Continuous Explanatory Variable
Percent Confirmed COVID-19
Median Household Income ($000)
Percent Aged 65+
Table 2. Linear regression of Percent Vaccinated against three continuous variables: Cumulative Percent Confirmed COVID-19 ; Median Household Income (in $thousands); and Percent of the Population Aged 65+. The data set included the 204 CSAs in Figures 2–4. The regression was weighted by each community’s population. The two right-most columns show the bounds of the estimated 95% confidence intervals. The adjusted R-squared statistic was 0.68.
No significant long-term change in smartphone-tracking indicators can be detected since the Los Angeles Mayor’s order of December 2 or the California public health officer’s regional orders of December 3 and 6.
Figure 1 below shows the average time spent at home by smartphones located within the City of Los Angeles, within the remainder of Los Angeles County, and within Orange County before and after the issuance of city- and statewide stay-at-home orders at the beginning of December.
Mayor Eric Garcetti’s December 2 order applied only to the City of Los Angeles, which is contained within Los Angeles County. The state public health officer’s order of December 3, followed by a supplemental order on December 6, applied to the entire Southern California region, including all of Orange and Los Angeles counties. (Municipalities within Los Angeles County but outside the City of Los Angeles include the cities of Beverly Hills, West Hollywood, Pasadena, Long Beach and Santa Monica, to name a few.)
Figure 2 below shows the corresponding trends in the percentage of devices staying completely at home.
Let’s look first at the trends outside the City of Los Angeles. Comparing Sunday November 29 with the following Sunday December 6 in Figure 1, we see that the December 2 and 3 orders may have been associated with about a half-hour increase in average stay-at-home time. But this short-term effect appears to have been dissipated during the subsequent two weeks. What’s more, Figure 2 suggests that if there was indeed a decrease in social mobility in Orange County or the rest of Los Angeles County, it started before the orders were issued.
Examining the blue trend lines for the City of Los Angeles in both Figures 1 and 2, we find it even harder to discern an impact. Again, if there was indeed a small short-term effect after the orders were issued, it appears to be gone two weeks later.
These two graphs do not show impressive drops in social mobility after the city- and statewide orders. There may have been some minor short-term reductions in out-of-home movement, but they have not been persistent.
Why So Little Impact?
The December 2-6 orders did more than require people to stay at home. Mayor Garcetti’s order, for example, prohibited “all public and private gatherings of any number of people from more than one household,” except for outdoor faith-based services and outdoor protests while wearing a mask. Acting State Public Health Officer Eric Pan’s order, for example, provided that “all retailers may operate indoors at no more than 20% capacity.” Perhaps these accompanying provisions have had a favorable effect in reducing coronavirus transmission.
Still, Garcetti’s order was entitled Targeted Safer at Home Order. Right off the bat, the order reads, “Subject only to the exceptions outlined in this Order, all persons living within the City of Los Angeles are hereby ordered to remain in their homes.” Pan’s order was entitled Regional Stay At Home Order. Paragraph #2 reads, “All individuals living in the Region shall stay home or at their place of residence except as necessary to conduct activities associated with the operation, maintenance, or usage of critical infrastructure.” Yet the evidence is that residents of Los Angeles and Orange counties did not stay at home any more or less than did before these stay-at-home orders.
The reflex interpretation of these findings is that they are simply one more manifestation of COVID-19 burnout and pandemic fatigue. Our own view is that there is now a serious problem of signal versus noise.
There have been so many orders and revised orders and supplemental orders that it has become nearly impossible to ascertain what restrictions on mobility are actually in effect. If we could have performed a focused survey of public awareness or an analysis of social media content, we’d want to know how many Southern Californians were even aware that the new stay-at-home orders were in effect.
Whatever the interpretation, these findings reinforce a critical conclusion.
Our calculations are derived from the SafeGraph Social Distancing database, which follows the GPS pings of an anonymous panel of smartphones equipped with location-tracking software. Each mobile device is assigned a home or origin based on the census block group (CBG) where it commonly spends the night. All CBG codes in Orange County begin with the 6-character string 06059, while all CBG codes in Los Angeles County begin with the 6-character code 06037. For example, the CBG codes for the three census block groups within census tract 1011.10 in the City of Los Angeles would be: 060371011101, 060371011102, and 060371011103. We can thus use census tracts to distinguish CBGs belonging to the City of Los Angeles from other municipalities within Los Angeles County.
For each calendar day from November 1 through December 22 and each CBG in Los Angeles and Orange counties, we extracted a record showing the total number of devices (device_count), the number of devices staying completely at home (completely_home_device_count), and the mean in-home dwell time (mean_home_dwell_time). For each calendar date and each of the three major geographic subdivisions (City of Los Angeles, rest of Los Angeles County, Orange County), we computed the overall mean time staying at home (Figure 1) and the overall percentage of devices staying completely at home (Figure 2).
Los Angeles County is fast becoming the epicenter of the pandemic in the United States. ⦿ Multi-generational households remain the principal pathway for coronavirus transmission. ⦿ Community health centers will be critical to getting through the hard winter to come.
Go Where the Virus Is.
The map below captures the main message of this article: Go where the virus is.
The rose-shaded areas are the Los Angeles County communities where the per-capita incidence of newly confirmed COVID-19 cases grew the fastest during the three weeks ending December 17. While these communities together contain half of the population of the county, their residents now make up nearly two-thirds of all newly diagnosed infections.
The black circles mark the locations of 353 member clinics of the Community Clinic Association of Los Angeles County (CCALAC). These community health centers are uniquely situated to confront the epidemic surge that is already overwhelming the county’s acute-care hospital capacity.
Actually, these health centers don’t have to move an inch to go where the virus is. They’re already there.
The Winter Surge Has Arrived.
Figure 2 below updates the weekly incidence of newly diagnosed COVID-19 cases in Los Angeles County, running from the week starting March 1 through the week starting December 6, 2020. When we last studied the Los Angeles County epidemic, we could see the emergence of a fourth phase during September. Now, the Phase IV surge is obvious. The incidence of newly confirmed COVID-19 cases per 100,000 during the week of December 6 is nearly four-fold the peak incidence seen during the week of July 12.
Los Angeles Mayor Eric Garcetti’s recent stay-at-home order may indeed retard the growth of new cases in the weeks to come. But it will not change the fundamentals of SARS-CoV-2 transmission. Herd immunity from mass vaccination is likely to be several months off.
There is, to put it graphically, no easy way to extinguish the virus-flames that will rage during the winter months to come.
Multi-Generational Household Transmission
Figure 3 compares two maps of Los Angeles County. Each map is broken down into countywide statistical areas (CSAs), a hybrid geographic classification of independent municipalities such as the City of Beverly Hills, neighborhoods of Los Angeles such as Hollywood, and unincorporated places such as Hacienda Heights.
On the left, the CSAs are color-coded according to the number of newly confirmed COVID-19 cases per 100,000 recorded during November 27 – December 17, 2000. On the right, the same CSAs are coded according to the proportion of households that we’ve identified as at risk for multi-generational transmission. As explained in this detailed report (which, for those who care, remains under peer review), we classified a household as at risk for multi-generational transmission if it had at least four persons, at least one person 18–34 years of age and another person was at least 50 years of age.
The two maps in Figure 3 show striking a concordance. Those communities with the highest prevalence of at-risk households, as depicted on the right, had the highest incidence of recent COVID-19 infection, as shown on the left. While we’ve previously noted the concordance with cumulative incidence through September 19 and through November 26, the comparison here is with the number of newly incident cases during the last three weeks of Phase IV.
In short, transmission within multi-generational households continues to dominate the Los Angeles County pandemic.
For those who don’t see the straight visual comparison of two maps as convincing, the technical section below offers a more rigorous demonstration.
The Data Fit With What We See on the Battlefield.
There isn’t the remotest doubt at this juncture that the numbers of confirmed cases reported through voluntary testing substantially understate the actual numbers of incident SARS-CoV-2 infections. Asymptomatic persons, we now know, are responsible for at least 40-45 percent of cases and play a dominant role in disease transmission. Still, the official public statistics fit with what we’re now seeing, as they say, en el campo de batalla.
The clinical histories, one by one, are remarkably the same. One person – usually an asymptomatic or pre-symptomatic young adult – has unknowingly imported the infection into the household. By the time the first member of the household displays any symptoms, all of them – children, adolescents, parents, aunts, uncles, grandparents – have already been infected.
In one household, a woman in her 40’s, having just come down with body aches, fever, and loss of smell, isolated her daughter and her mother in one bedroom and her son and father in another bedroom. The 23-year-old son, who turned out to be patient zero in this particular family, infected his 83-year-old grandfather roommate, who naturally became the focus of considerable clinical attention. (I have intentionally altered the ages of family members.)
The supreme irony of this and so many other cases is that nearly everyone in the household was taking special care to avoid contagion. The parents would make limited trips to the market, always wearing masks. Many grandparents would never go out at all.
Are the young adults who unintentionally import their infections into their multi-generational families simply being careless or ignorant? Not exactly. A young adult who turned out to be patient zero in his own family always wore his mask at work. When he went jogging with his buddies after work, they kept their distance. After one run, they all went for some cold refreshments.
One of the friends ordered hot tea. He just had allergies, he said.
It’s all in the math. Multi-generational households at risk for viral transmission make up 13.8 percent of all households in Los Angeles County. With an estimated 3.3 million households in the county, we’re talking about 455,000 multi-generation households where an asymptomatic or mild SARS-CoV-2 infection in a younger household member would put older household members at significant risk.
Let’s round off the size of the average multi-generation household to just 5 members. During the last three weeks, well over 1 percent of these households were infected. And let’s say the typical super-spreader event generates 100 cases, at least during the first round of transmission. The rest is arithmetic.
During the past three weeks, propagation of the virus through multi-generational households in Los Angeles County has been the equivalent of 200 so-called super-spreader events.
Community Health Centers Are the Key.
The Google Map below is a navigable version of the screenshot shown at the beginning of this article. It is worth repeating that the hot spots with the highest incidence of new infection are precisely the places where multi-generational families are most prevalent.
What’s more, the members of those multi-generational families at highest risk for transmission are, by and large, already patients of nearby community health centers.
Here Are Some of the Things Health Centers Could Do.
Here is an abbreviated list of the things that these community health centers could do – if only they were immediately given adequate resources.
Community health centers need to be equipped to perform high-volume SARS-CoV-2 testing of their patients, with a focus on bringing in all household members at the same time.
It is widely acknowledged that testing needs to be more proactive. We can’t continue to sit back and wait for symptomatic people to come through the door. Workplaces and educational institutions have engaged in routine testing of asymptomatic people. Community health centers need to be the next locus of routine testing. Health centers know who are their high-risk patients. They are ideally positioned to bring in the entire families of these high-risk patients for routine testing.
Community health centers need to be equipped with telemedicine technology, so that providers can rapidly respond to patients’ inquiries and concerns.
Surveys give us only an incomplete picture of the public’s knowledge of such basic issues as how COVID-19 is spread. The plain fact is that patients have an abundance of unanswered questions that apply to their specific circumstances.
Intelligent, useful answers to these questions require detailed knowledge of the incubation period of the virus, the time profile of viral shedding before and after symptoms appear, the likelihood of false-positive or false-negative tests, and the emergence of serious complications in the second critical week . Meaningful answers require clinical knowledge of whether COVID-19 can cause diarrhea, earache, body rashes, back pain. and anxiety due to a low oxygen level. Healthcare providers at community health centers are ideally positioned to answer these questions.
Healthcare providers at community health centers are uniquely positioned to respond rapidly to patients who are at risk of deterioration.
If any reader has the remotest doubt where people in the United States turn when they feel really sick from COVID-19, here is the answer. They call 911.
And this is by no means an abuse of the public safety system. In a December 16 news release announcing that Los Angeles County now had more than 5,000 COVID-19 hospitalizations, the Department of Public Health stated in plain English: “If you are having difficulty breathing, go to an emergency room or call 911.” A December 22 news release announcing another 1,000 COVID-19-related deaths in the past two weeks stated in plain Spanish: “Si tiene dificultad para respirar, vaya a la sala de emergencias o llame al 911.”
Healthcare providers in community health centers are ideally positioned to address patients’ concerns. Who else is going to tell a COVID-19-stricken patient who can hardly eat anything whether she should still take her blood pressure or her diabetes pills? Who else is going to intelligently advise a patient when chest tightness becomes so severe that it’s time to go to the ER? The Department of Public Health has urged providers: “Do not send patients to emergency departments unless absolutely medically necessary.” Physicians, nurse practitioners, physicians’ assistants, and nurses know their patients. They can respond to the challenge if given the resources to do so.
Community health centers could distribute pulse oximeters to their patients and instruct them on their use.
The pulse oximeter is now widely acknowledged as a critical tool in the outpatient evaluation of a patient acutely ill with COVID-19. Community health centers are ideally situated to distribute these devices to patients most in need. New York City is distributing oximeters to healthcare providers, and Los Angeles County health centers should get them, too. In abundance.
Community health centers are well situated to provide timely outpatient treatment to high-risk COVID-19 patients.
There is growing evidence that monoclonal antibody infusions need to be given as early as possible to high-risk patients during the initial viremic phase of their illness. Recent controlled trials suggest that these medications may be ineffective if we wait until the patient is so ill as to need hospitalization. The same may be true for the antiviral drug remdesivir.
We need to identify high-risk patients at the onset of symptoms and establish outpatient facilities where trained personnel can administer these therapies. When it comes to Los Angeles County, community health centers could serve as the ideal locus for this new model of care. It is not difficult to imagine a health center-based nurse administering an infusion to a 70-year-old diabetic with a history heart and kidney problems within a couple of days of symptoms.
What About Social Distancing Measures?
How does this fit in with the now-enormous literature on the effectiveness of the social distancing measures? No one disputes that a great many of the emergency measures enacted by governments worldwide have substantially reduced the propagation of the virus. But the time has come to realize that these now-conventional measures – stay-at-home orders, closures of bars, restaurants, gyms, hair salons, movie theaters, restrictions on large gatherings – are no longer enough. And what’s worse, their enforcement may represent a misallocation of limited resources.
Los Angeles County authorities need to start thinking beyond their standard public health toolbox. They need to adapt their policies to the unique facts on the ground.
If public authorities in Los Angeles County diverted substantial additional resources to community health centers, we’d still have a real chance at putting out the virus-flames.
Technical Details on Multi-Generational Household Transmission
Figure 4 offers another way visualize the critical role of intra-household transmission in the Los Angeles epidemic. We’ve graphed the number of newly recorded cases per 100,000 during November 17 – December 17 against the corresponding number of new COVID-19 cases recorded during October 17 – November 26, 2000. Each data point represents a CSA. The larger the circle, the higher proportion of at-risk multi-generational households in the community.
The linear relationship in the figure is striking. It tells us that those communities recording the most new cases at the start of the recent surge continue to serve as foci of new COVID-19 cases as the surge continues. And the fact that the data points at the upper right are the fattest, while those at the lower left are the skinniest, confirms that CSAs with lots of multi-generational households continue to drive the epidemic.
Figure 5 below displays an updated version of the graphs that we showed in our October 17 and November 28 posts. The graph relates the recent incidence of COVID-19 infection on the vertical axis to the prevalence of at-risk households across some 300 CSAs in Los Angeles County, as measured on the horizontal axis. The graph tells us that a 10-percentage point increase in the prevalence of at-risk households is associated with a 51-percent increase in COVID-19 diagnoses. This relationship has not let up since the days of Phase II, when the epidemic began to concentrate in specific communities.
The TPR is uninformative in an epidemic dominated by asymptomatic transmission. And it doesn’t give us a clue what to do in Philadelphia during the hard winter ahead.
During the ongoing COVID-19 epidemic, public health practitioners and policymakers have increasingly relied on the test positivity rate (TPR) to decide whether to impose or relax constraints on social mobility and how much to expand testing capacity. Yet there is a genuine question whether the TPR – which basically gauges the number of positive cases as a percent of all persons tested – may be steering us off course.
During an epidemic where testing for infection is non-random and voluntary, the TPR may indeed tell us how well our testing program is identifying more severe, symptomatic cases. But in the current COVID-19 epidemic where asymptomatic persons are responsible for at least 40-45 percent of cases and play a dominant role in disease transmission, the TPR may not tell us how well we’re identifying infectious individuals.
When Testing Is Mandatory and Random
When testing is mandatory and random – for example, when the University of Wisconsin-Madison required all on-campus students to be tested regularly during an outbreak – the TPR may indeed accurately gauge the infection rate. If, say, one percent of all persons randomly tested during the past week come out positive, then we can reasonably estimate the current incidence rate per week to be about one percent. To be sure, there can still be false positives and false negatives, and there can be retesting of the same individuals. And we certainly don’t want to mix tests for active infection with antibody tests. While these concerns are valid, they are not fundamental.
When Testing is Voluntary and Non-Random
When testing is voluntary and non-random, however, it is widely acknowledged that the sickest, most symptomatic individuals will queue up for testing first. This behavioral observation has led to the conclusion that as testing capacity is expanded, the TPR will decline. If only the individuals in the first row in Figure 1 are tested, the TPR will be 40 percent (that is, 4 out of 10). If we expanded out capacity to test both rows, the TPR will be down to 30 percent (that is, 6 out of 20). This straightforward syllogism seems to imply that a low TPR is a favorable sign that our testing program “is casting a wide enough net.” Most authorities, including the World Health Organization (WHO), put the cutoff for a sufficiently low TPR at 5 percent, but some analysts peg the cutoff at 3 percent. Whatever cutoff is designated, the TPR is to be interpreted as “a measure of whether we’re doing enough testing.”
The Catch: A Large (and Growing) Pool of Asymptomatic Infected Persons
But there’s a catch. In the current COVID-19 epidemic, a lower TPR means our testing program is adequately identifying the sick people who are voluntarily queuing up to be tested. A further expansion of voluntary testing capacity may indeed lower the TPR, but it will not necessarily identify asymptomatic infectious individuals because they’re not even getting in line.
In Figure 2, we’ve included another group of 20 such asymptomatic people. Eight of them are positive and they don’t know it. If we really could test everyone, the TPR would be back up to 35 percent (that is, 14 out of 40).
In more technical terms, the critical assumption of a declining marginal yield to expanded testing breaks down when there is a significant pool of asymptomatic infected individuals with an extremely low propensity to get tested voluntarily. And as we’ll see shortly, the continued growth of this pool of asymptomatic infected individuals is the principal challenge underlying the hard winter months ahead.
Some Real Data from Philadelphia
Some real data from the city of Philadelphia in the United States, recently downloaded from its online dashboard, will clarify the point. At issue here is not whether Philadelphia’s current caseload is representative of other cities, but whether the problems in interpreting Philadelphia’s TPR are typical.
Figure 3 shows the weekly incidence of test-positive COVID-19 cases per 100,000 population in the city. To facilitate the presentation, we have partitioned the graph into four phases. (We’ve similarly partitioned the COVID-19 epidemic in Los Angeles County into four phases.)
In Phase I, from the week starting March 9 through the week starting March 30, 2020, highlighted in burgundy red, cases were rising rapidly from 3 to 178 per 100,000, equivalent to an average doubling time of 4.7 days. In Phase II, from the week starting April 6 through the week starting May 25, highlighted in green, the case incidence rate fell as emergency social distancing measures enacted in mid-March took effect.
During Phase III, from the week starting June 1 through the week starting September 14, highlighted in orange, the case incidence rate remained stable at about 50 per 100,000. Since then, with the arrival of Phase IV, highlighted in navy blue, the case incidence rate has topped 400 per 100, rising with an average doubling time of about 18 days. While reporting delays temporarily render the data for the last week in November unreliable, the case incidence has surely continued to rise.
The One-Case-per-Thousand-per-Week Threshold
Figure 3 includes an additional, horizontal dashed line at the level of 100 per 100,000 cases per week, equivalent to 1 case per 1,000 per week. This line crosses the epidemic curve during the week of March 23 (Phase I), the weeks of May 11 and May 18 (Phase II), and the week of October 12 (Phase IV). At each crossing point, the case incidence rate was the same, but the underlying dynamics of the city’s epidemic were different. Crossing the threshold from above (Phase II) is not the same as crossing from below, and crossing the threshold during the early surge of cases (Phase I) is not the same as a re-crossing when previously relaxed social distancing measures have proved insufficient (Phase IV).
Dynamic Pueyo Plot
Figure 4 plots the case incidence rate against the TPR, once again calculated on a weekly basis. The data points have been color-coded to correspond to the four phases identified in Figure 3. With the exception of a few points tightly clustered together in Phase III, the corresponding week is noted beside each point. (For visual clarity, the data point for the week of March 9 has been omitted.) The area of each point (not the diameter) is proportional to the total number of tests performed during each week, ranging from 2,147 tests during the week of March 16 to 60,980 tests during the week of November 16.
We’re calling Figure 4 a dynamic Pueyo plot because Tomas Pueyo appears to be the first analyst to plot case incidence against TPR. Pueyo’s version of the plot, however, was a static, cross-sectional comparison of different countries at a particular point in time, and not a dynamic rendering of the course of the epidemic in a single locality.
The Critical Question
The critical question is: What, precisely, does the additional information on the TPR in Figure 4 add to our knowledge of the dynamics of the disease propagation, the adequacy of testing, or the most appropriate epidemic-control measures in Philadelphia during the hard winter ahead?
Figure 4 tells us that the first time the epidemic curve crossed the 100-cases-per-100,000 threshold during the week of March 23, the TPR was 28 percent, far above the WHO’s acceptable level of 5 percent. The second time, during the weeks of May 11 and May 18, the TPR was 12 percent, still beyond the acceptable range. The implication is that the observed decline in case incidence during Phase II in the weeks after the mid-March declaration of emergency may have been an artifact of inadequate testing. Such a conclusion would defy the now-enormous literature on the effectiveness of the social distancing measures put into effect during Phase II.
The fact that the orange data points in Figure 4 are situated at or near the 5-percent TPR threshold would seem to imply that Philadelphia had adequate testing during Phase III. But if that conclusion has any substantive meaning at all, then it must apply as well to the first four weeks of Phase IV, when the number of cases per week more than doubled. Does this mean that no additional aggressive testing measures were required?
It is now evident that during the week of September 21, Philadelphia was on the verge of a new and more deadly resurgence of infections, yet its TPR of 2.5 percent was telling us everything was just fine.
The raw number of positives indeed obscures the number of negatives or the number tested. And there isn’t the remotest doubt that current counts of positive test results substantially understate the actual incidence of infection. Still, when it comes to assessing the dynamic state of the epidemic in Philadelphia or anywhere else, the number tested may be a more misleading denominator than the standard population count.
One might argue that the Philadelphia data are entirely consistent with the TPR paradigm. The expansion in testing seen in the last two months – as indicated by the enlarging blue data points in Figure 4 – was accompanied by an increase in the TPR. This observation, some analysts may contend, precisely illustrates the value of the TPR in identifying a dangerous outbreak. The problem with this logic is that both the number of positive tests and the number of total tests are endogenous variables. Without a natural experiment, we can’t tell whether new positive cases were pushing total testing, or whether expanded testing was pulling the number of case counts. Another analyst might observe, “The percent positive will be high if the number of positive tests is too high, or if the number of total tests is too low.” But that’s just having it both ways.
The Hard Winter Ahead
The first equation of the classic SIR model of epidemics teaches us that the incidence of new cases is a product of two basic factors: (1) the number of contagious individuals and (2) the rate at which these contagious individuals transmit their infections to other susceptible individuals.
During Phase I of the epidemic in Philadelphia, Los Angeles, New York City, San Antonio,Broward County, Florida, and countless other metropolitan areas throughout the world, the surge in cases was driven by factor #2. At the very start, after all, we had no social distancing measures, so that a relatively small number of contagious persons could propagate their infections widely.
During this Phase IV of the epidemic – which appears to be continuing into the cold, hard winter to come – the new surge in cases is being driven by factor #1. The growing number of asymptomatic individuals has now overwhelmed what social distancing measures we have put in place. Despite marginal improvements in testing capacity of the order of magnitude seen in Philadelphia, we haven’t come anywhere near the level of testing we need.
Without routine, repeated testing of asymptomatic carriers through widespread rapid testing on a far more massive scale than we’ve seen so far, we’ll continue to be “operating in the dark,” no matter what the test positivity rate is.
Comments and Responses
In response to the thoughtful comment below by Prof. Raphael Thomadsen. As an indicator of the adequacy of testing, the TPR appears to have little or no informative value. When it comes to gauging the severity of the epidemic, the TPR might serve as a proxy for the case incidence rate. But that just begs the question: What does the graph of TPR in Figure 4 tell us that we haven’t already learned from the graph of the case incidence rate in Figure 3? As to whether there were different test regimes, take a look at the graph below. There was clearly a different regime during Phase I, when the CDC was still struggling to come up with a reliable PCR test for coronavirus infection. After that, the total number of tests has climbed steadily at about 5 percent per week.