Counting the Bodies
No U.S. President wants to be held responsible for the deaths of fellow Americans. This, in large measure, explains why since the conclusion of the Persian Gulf War news coverage of dead bodies of U.S. armed forces personnel being returned to this country has been prohibited. Similarly, even though President Trump has made it clear that no number of dead bodies will stand in his way to reinvigorate the nation’s economy, he nevertheless appears concerned that the mounting toll of deaths from the COVID-19 pandemic could impair his re-election chances. That explains why he has begun to assert that the official death count from the virus (which now stands at over 96,000) is overstated.
The potential adverse impact of a high number of Covid-19 deaths could easily be decisive in the coming election, especially since by the end of this month the number of those deaths will equal, if not exceed, all of the deaths suffered by U.S. personnel in all of the wars fought since World War II, including the wars in Korea, Vietnam, Afghanistan and Iraq. That also explains why the President has embarked upon a campaign to blame those deaths on the Chinese government (which suppressed the extent of the threat posed by the coronavirus outbreak) and the World Health Organization (which failed to press China for more information about the virus and was slow to declare a pandemic). Even so, the U.S. death toll from just the first wave of the virus is now projected by the Institute for Health Metrics and Evaluation (“IHME”) to reach 143,000. Until the latter part of April, the Trump administration’s actions had been guided by the IHME projections; however, that changed when the President decided that he no longer wished to continue the nation’s social distancing restrictions through the month of May, a principal underlying assumption of those projections.
Part of the difficulty faced by the President in blaming China and the W.H.O. is that their actions affected all nations, not just the U.S.; and the number of COVID-19 deaths in this country far exceeds those reported in every other country. In fact, the percentage of individuals dying from the virus in the United States is already more than one quarter of all those reported throughout the world and, based upon the current IHME projections, that could increase to one-third of those deaths by the end of this year. The U.S. death toll from the virus looks particularly bad in relation to certain southeast Asian countries such as Japan, South Korea, Australia, New Zealand and Singapore, all of which lie relatively close to China where the virus originated. Their national death tolls currently stand at 744, 263, 99, 21 and 22, respectively.
In Europe, deaths from the virus were much greater than those experienced in southeast Asia; and it is not clear why that has been the case. One possible reason is that nations in southeast Asia had previously been hard hit by earlier strains of corona virus, like SARS and MERS, and were alert to the potential danger posed by COVID-19. Still, the U.S. Center for Disease Control (or “CDC”) has long been considered the world’s premiere public health agency so that it does not seem plausible that the nations of Southeast Asia actually had a better understanding of the danger about to befall them. Another possible explanation is that the virus that attacked the European countries was slightly different from the one that attacked the nations of Southeast Asia. There has been some speculation that the European strain of the virus, which made its way to New York and spread across this country, was more virulent than the one that moved through Southeast Asia. The strain that hit Southeast Asia was transported to the States of California and Washington, both of which recorded early cases of the virus. This theory could explain why their death tolls have been relatively low compared to those experienced in New York, New Jersey, Massachusetts, Connecticut, Pennsylvania, Michigan and Illinois. That explanation, however, has been rejected by epidemiologists who have examined both strains of the virus. While finding some differences in the two, they have found no evidence that the European strain is either more lethal or more readily transferable. Thus, it is difficult to explain the high death toll experienced in the U.S. on the basis of either the actions of China and/or the W.H.O. or the variations in the two strains of the virus.
Perhaps the best argument that the administration can muster in defense of the high death toll experienced in this country is that the population of U.S. far exceeds that of every other country on the earth except for China and India which makes it unfair to simply consider the total number of deaths without examining the virus’ death tolls on a per capita basis. Even on a per capita basis, however, the U.S. efforts in restraining the impact of the virus is far from exemplary. Set forth below is a table comparing the populations and death statistics (as compiled by Johns Hopkins University of Medicine) from each of Canada, Mexico and the eight European countries with the highest reported death tolls from the virus. (I use the term “reported death toll as I have no confidence that the number of deaths reported by the Chinese government has any relationship to reality).
While the U.S. looks much better on the basis of its death toll viewed in relationship to its total population, for the nation with the greatest resources and the most advanced medical technology, its record in fighting the virus is still not one that merits applause. This is because the European countries listed above were attacked by the virus before it reached the U.S. Thus, the U.S. should have learned from their experience. In addition, the foregoing death tolls experienced by the European countries represent close to the maximum numbers of deaths that they are projected to experience this year; whereas the U.S. has only experienced a little over 60% of the COVID-19 deaths that it is expected to incur this year. The importance of this distinction can be seen in a comparison of the percentage of virus cases that have led to the patient’s death. For the most part, one would expect little variation from one country to the next. That, however, has not been the case. In the U.S., the percentage of confirm cases of the virus leading to death has been just under 6%, whereas in Western Europe the percentage of confirmed cases leading to death has been between 12% and 16%. Similarly, in those states that were first exposed to the virus (like New York, New Jersey and Connecticut), the percentage of deaths range between 7% and 9% of total confirmed cases as compared to those states that the virus did not begin to reach until a few weeks later (such as Tennessee, Nebraska, Arkansas and Indiana) where the percentages of deaths to confirmed cases range from 1% to 3%. Thus, the only appropriate way to compare a nation’s efforts to combat the virus is to examine its results measured from the day that it experienced its first death or its 10th death, and not on any arbitrary single date for all countries. For example, you might compare the results of each country on the 100th day following the date it recorded its tenth death from the virus. For this reason, simply examining death statistics on a per capita basis proves little.
Perhaps anticipating that a strategy of blaming China and the W.H.O. might not suffice, the President has nominally placed upon the nation’s governors the decision as to when their states should restart their economies. In this way he can blame them for the large death toll that we will have experienced by election day. While any defense is better than no defense, even this argument is not likely to enjoy much credibility. That’s because the President has been leading the chorus demanding that the economy be restarted, and his administration has taken a number of actions both compelling the states to proceed prematurely and impeding their ability to proceed in a manner that will minimize further deaths. These actions include repeated encouragement for them to proceed, initiating protests against their restrictions on non-essential business activities, opposing further fiscal aid compelling governors to restart their state’s economies in order to relieve increasing budgetary deficits, and touting the imminent arrival of vaccines that will prevent further spread of the virus. Thus, convincing the American public that the administration has not been at fault may turn out to be a tough sale.
More importantly, the principal reason for the nation’s high death toll will not be the premature reopening of its economy, but rather the long delay in taking any action to combat the spread of the virus. A Columbia University team of disease modelers investigated what would have been the nation’s death toll as of May 3rd had the administration acted earlier to prevent the spread of the virus. The team’s finding was that had the nation begun to lockdown non-essential businesses and mandated social distancing one week before it did, the number of deaths occurring before May 3rd (65,307) could have been reduced by 36,000; and had it done so two weeks earlier, the number of deaths could have been reduced by 54,000 (or a staggering 83%). Thus, by the time the administration chose to have the states decide when to restart their economies, the nation’s ultimate death toll had already become preordained.
This brings us back to the President’s current campaign to cast doubt of the death toll itself. You might think of it as Chapter Two of his book in which the first chapter is devoted to his contention that it wasn’t the Russians who interfered in the 2016. As might be expected, this theme has been picked up by Fox News and on-line conservative media. The defense, however, has gone beyond simply contending that the death numbers are overstated. There have been real efforts made to actually understate the number of deaths attributable to COVID-19.
While not well publicized, the CDC tracks confirmed cases of the virus and resulting deaths and posts that data on its website. It currently reflects approximately 50,000 less confirmed cases and approximately 3,000 fewer deaths than are reported on the Worldometer website and approximately 25,000 fewer confirmed cases and 2,000 fewer deaths than are reported on the Johns Hopkins website. Both of these private website collect their data directly from state and local health departments which raises the question why there would be such differences. It might just be a question of when the data is retrieved as the differences essentially represent a single day’s activity. These differences might also be explained by the CDC’s taking a time to review the data before posting it to its website. Even so, the relatively minor differences between the CDC’s postings and those of the other websites do not materially alter the conclusion that the handling of the virus in this country has been an unmitigated disaster.
The administration’s efforts to cast doubt on the accuracy of the virus’ death toll has not exactly experienced smooth sailing as Dr. Fauci, along with scores of other epidemiologists, has already publicly announced that the actual death toll for the virus is likely to be much greater than the published figures. This, in large measure, explains why the Trump administration is refusing to allow Dr. Fauci to testify before a House Committee. If I were a conspiracy theorist, I might be asserting that the administration had caused the good doctor to come in contact with someone who had tested positive for the virus, resulting in his being quarantined.
Dr. Fauci’s statement about the accuracy of the COVID-19 death toll, however, was quite mild compared to the reality. In fact, there is very good reason to believe that the actual death toll is already over 110,000. This is based on an analysis of what epidemiologists call “excess deaths.” It is widely accepted that the number of people who have been infected by the virus greatly exceeds the number of reported cases, which now stands at over 1.6 million in the U.S. In fact, epidemiologists believe the actual number of infected person in the U.S. is probably five to ten times that number. That’s why the number of reported infected persons is labeled “confirmed cases” which simply means the number of persons who have actually tested “positive” for the virus. Similarly, the number of deaths attributed to COVID-19 essentially represents the number of persons who died that had tested “positive” for the disease or have been medically diagnosed as having the disease based upon their symptoms. The problem is that in February, March and early April there were very few test kits available so that many people who died were simply never tested for the disease, which is not an uncommon occurrence. This is particularly true as many virus victims never made it to a hospital where the cause of their death could be determined.
Under such circumstances it is standard procedure to determine how many people in a given jurisdiction have historically died during a specified period. There is surprisingly little variation (usually less than a 5% deviation from a 20-year average) in the number of deaths in a widely populated jurisdiction like the City of New York. During a six-week period between mid-March and late April of this year the number of total deaths in New York City exceeded the average by roughly 30%, with the implication that these “excess deaths” less the number of deaths found to be attributed to COVID-19 were also likely caused by the virus. A similar analysis conducted in other areas of the country during the same period turned up additional “excess deaths” which aggregate over 35,000 deaths, the net effect of which would raise the death toll from the virus in this country by roughly 15,000.
Even though the vast majority of reported COVID-19 deaths have been confirmed by tests for the disease, there is nevertheless some room for doubt. Perhaps the main area of uncertainty regarding the cause of death is that in any given case there may have been multiple possible causes of the patient’s death, requiring a judgment call by the attending physician. This could even be true where the patient has tested positive for the virus. In such circumstances it must be remembered that COVID-19’s most vulnerable targets are the elderly (who generally have a multitude of physical ailments) and those with serious underlying medical conditions like diabetes and heart, respiratory or liver disease. This means that in a significant number of cases there may be a legitimate issue as to the cause of death; and, considering the onslaught of very sick patients flooding the hospitals during late March and April, there was little time for healthcare professionals to stop and do a thorough analysis as to which of a patient’s physical infirmities was actually responsible for his or her ultimate demise. Thus, there is some basis for the administration’s insistence that there could be over-counting. If you will recall, the Trump Administration’s Dr. Birx suggested that the Coronavirus Taskforce might be looking into how the cause of death was being determined. (You can always count on the lady with the large collection of scarfs to find a way to curry favor with the President.)
Aside from making an actual determination as to the cause of death, the reporting process is fraught with problems. That’s because each hospital and each jurisdiction (city, county or state) has its own way of collecting health data and some are far more timely and efficient in doing so than others. In fact, it is not uncommon for the cause of death to be determined several days after the patient has died.
Even assuming that state and local health department officials are dedicated civil servants trying to do their very best under trying circumstances, there is clear evidence that the process of collecting the data is not immune from political influence. We have already seen a few instances of this. For example, in order to make it look like the virus was coming under control so as to justify the lifting of restrictions on non-essential businesses, the State of Georgia decided that the new case data should not be based on the date that the individual had tested positive for the virus, but rather on the date the individual first experienced symptoms of the disease, usually several days before he or she had been tested. To further give the impression that progress was being made against the disease, that state’s Department of Health published a bar chart depicting the number of confirmed cases and deaths in its five largest counties going steadily downward. The only problem was that the dates on the x-axis of the chart had been shuffled out of order to give that impression. Governor Kemp apologized for what his Department of Heath characterized as a glitch in the computer program on which the chart had been composed. As the Atlanta Journal-Constitution put it, “Some of these errors could be forgiven as mistakes made during a chaotic time, but putting days in the wrong order, as the recently withdrawn chart did, makes no sense.”
Similarly, in the State of Florida which has also been under pressure from the Trump administration to restart its economy has been reported as not including within its death figures those of its seasonal residents who have died in the state. In addition, Florida has recently fired the technician who had developed the software the state uses to track the COVID-19 cases and deaths. The reason given for her termination was “insubordination.” That insubordination, it turns out, was her refusal to manually alter the data shown on the state’s COVID-19 dashboard.
The heavy hand of politics has also apparently invaded the testing data. The federal government has been heavily and rightfully criticized for the low amount of testing that has taken place. This criticism centers of the fact that the extent of a state’s testing for the virus is critical to the decision to when it should reopen its economy. It has now been reported that the CDC has been combining the number of viral tests (which detect the presence of the virus) with the number of antibody tests (which detect whether the individual has been exposed to the virus) to make the testing numbers appear larger. While both of these types of tests provide useful information (each of a different variety), when combined they obscure the resulting information that would otherwise be useful. Moreover, antibody tests are notoriously inaccurate. Thus, it is inconceivable that the epidemiologists at the CDC were not aware that combining the two varieties of tests provides misleading information. Yet, the CDC not only committed this scientific sin, but also apparently encouraged the states to do the same thing.
It is still far too early to predict how many Americans will die as a result of the virus by election day, but the likelihood is that the President will need a very strong economy to overcome the stench caused by all of the dead bodies.