Flawed Models Show Why COVID-19 Policies Must Consider Total Mortality
James Enstrom /
Policymakers need to scrutinize their epidemiological models.
In response to the coronavirus pandemic, the federal government has been heavily influenced by the Institute of Health Metrics and Evaluation’s computer model, which has projected from 60,000 to 240,000 COVID-19 deaths in the U.S.
This epidemiological model is now being criticized as flawed and misleading as a source of public information and for government decision-making. Besides the institute’s model, all other COVID-19 models are grounded in important assumptions about which there is currently little knowledge.
>>> When can America reopen? The National Coronavirus Recovery Commission, a project of The Heritage Foundation, is gathering America’s top thinkers together to figure that out. Learn more here.
Approaches other than models are needed to properly understand this pandemic.
One approach that has not been explored in any detail is the examination of deaths from all causes in addition to deaths from COVID-19, the disease caused by the new coronavirus.
Two major databases that track COVID-19 cases and deaths in the U.S., but not total mortality, are WorldOMeters and the Johns Hopkins University Hub. These trackers show major variation in COVID-19 mortality risk.
For example, as of April 22, WorldOMeters showed 47,681 COVID-19 deaths in the U.S. (50 states and the District of Columbia) with a rate of 144 deaths per 1 million people.
However, note that 71% of the deaths have occurred in six high-risk states (New York, New Jersey, Connecticut, Massachusetts, Louisiana, and Michigan) with 17% of U.S. residents and a death rate of 624 per million.
Ten percent have occurred in five medium-risk states (Rhode Island, Pennsylvania, Illinois, Maryland, and Indiana) and the District of Columbia with 12% of U.S. residents and a death rate of 124 per million, and 19% have occurred in the remaining 39 low-risk states with 71% of U.S. residents and a death rate of 40 per million.
It remains to be determined, however, whether these COVID-19 deaths have actually increased the total U.S. deaths this year. The best data on both COVID-19 deaths and total deaths in the U.S. come from the Centers for Disease Control and Prevention‘s National Center for Health Statistics.
During the five weeks ending Feb. 1 through Feb. 29, the Centers for Disease Control and Prevention reported 282,084 total deaths, which were 96% of the expected deaths based on concurrent 2017-2019 deaths. During the five weeks ending March 7 to April 4, the CDC reported 273,798 total deaths, which were 96% of the expected deaths.
Of the 9,474 COVID-19 deaths reported during these 10 weeks, 78.5% were among people over age 65, 21.4% were between the ages of 25 and 64, and only 0.1% were ages newborn to 24 years.
Those death counts through the end of March are preliminary, but they do not indicate that the total number of deaths in 2020 is greater than the comparable number of deaths during each of the three prior years.
Once the number of COVID-19 deaths and total deaths during the entire month of April are known, it will be clear whether there has been an increase in the total number of U.S. deaths this year.
One reason there may not be an increase in total deaths is because some deaths are being classified as COVID-19 deaths even when COVID-19 is not the underlying cause.
Normally, mortality statistics are compiled in accordance with World Health Organization regulations specifying that each death be assigned an underlying cause based on the current 10th revision of the International Statistical Classification of Diseases (ICD-10).
However, the Centers for Disease Control and Prevention reports that COVID-19 deaths are being coded to ICD-10 code U07.1 when COVID-19 is reported as a cause that “contributed to” death on the death certificate, but is not necessarily the “underlying cause.” Also, some of those deaths do not have laboratory confirmation of COVID-19 infection.
Thus, it’s possible that the focus on COVID-19 deaths has resulted in a lower number of deaths from seasonal flu, pneumonia, and other causes, compared with the number that would normally occur this year.
The CDC has stated that the number of flu hospitalizations estimated for this season is lower than total hospitalization estimates for any season since the CDC began making these estimates.
Furthermore, it’s possible that the lethality of COVID-19 is no greater than that of the seasonal flu.
A new Stanford University survey indicates that the population prevalence of COVID-19 in Santa Clara County, California, ranges from 2.5% to 4.2% and that the number of infected persons is 50 to 85 times the number of confirmed COVID-19 cases.
This preliminary finding suggests that at most 0.1% of infected persons will die from COVID-19, comparable to the seasonal flu death rate. Several other new studies indicate similarly lower fatality rates for COVID-19.
Americans need clarity. The federal government response to the coronavirus pandemic should not be based on flawed models, but rather on a localized public health approach that focuses on the high-risk areas of the United States and also on the high-risk elderly and those with comorbid conditions.
The emphasis should be on changes in personal behavior, such as staying at home for work or school if ill, covering coughs or sneezes, hand-washing, and avoiding those with respiratory symptoms.
Above all, the pandemic and COVID-19 deaths must be put in proper perspective, given the unprecedented societal and economic disruption of the current national lockdown.