The Blog

2020-

2020-12-18 b
THE COVID-CON I
"Results of this critical appraisal reveal information bias and selection bias in coronavirus mortality overestimation, most likely caused by misclassifying an influenza infection fatality rate as a case fatality rate."
+
"Unfortunately, even experts can make simple miscalculations that can lead to catastrophic results."
+
"The ethics of implementing fear-based public health campaigns needs to be reevaluated for the potential harm these strategies can cause."

Public Health Lessons Learned From Biases in Coronavirus Mortality Overestimation

Abstract

In testimony before US Congress on March 11, 2020, members of the House Oversight and Reform Committee were informed that estimated mortality for the novel coronavirus was 10-times higher than for seasonal influenza. Additional evidence, however, suggests the validity of this estimation could benefit from vetting for biases and miscalculations. The main objective of this article is to critically appraise the coronavirus mortality estimation presented to Congress. Informational texts from the World Health Organization and the Centers for Disease Control and Prevention are compared with coronavirus mortality calculations in Congressional testimony. Results of this critical appraisal reveal information bias and selection bias in coronavirus mortality overestimation, most likely caused by misclassifying an influenza infection fatality rate as a case fatality rate. Public health lessons learned for future infectious disease pandemics include: safeguarding against research biases that may underestimate or overestimate an associated risk of disease and mortality; reassessing the ethics of fear-based public health campaigns; and providing full public disclosure of adverse effects from severe mitigation measures to contain viral transmission.

[Introduction]
On September 23, 1998, the US National Aeronautics and Space Administration (NASA) permanently lost contact with the $125 million Mars Climate Orbiter. A simple miscalculation, failure to convert English measurements to metric measurements, doomed the Mars space mission. A later investigation found that backup quality assurance procedures were not in place at NASA to catch and correct this simple miscalculation. Fast forward 22 years to another crisis involving a US government agency: On March 11, 2020, the US Congress House Oversight and Reform Committee received information from the National Institute of Allergy and Infectious Diseases (NIAID) concerning the novel coronavirus, severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), and coronavirus-disease 2019 (COVID-19). Based on the data available at the time, Congress was informed that the estimated mortality rate for the coronavirus was 10-times higher than for seasonal influenza, which helped launch a campaign of social distancing, organizational and business lockdowns, and shelter-in-place orders.

Previous to the Congressional hearing, a less severe estimation of coronavirus mortality appeared in a February 28, 2020 editorial released by NIAID and the Centers for Disease Control and Prevention (CDC). Published online in the New England Journal of Medicine (NEJM.org), the editorial stated:

“…the overall clinical consequences of Covid-19 may ultimately be more akin to those of a severe seasonal influenza (which has a case fatality rate of approximately 0.1%).”

Almost as a parenthetical afterthought, the NEJM editorial inaccurately stated that 0.1% is the approximate case fatality rate of seasonal influenza. By contrast, the World Health Organization (WHO) reported that 0.1% or lower is the approximate influenza infection fatality rate,5 not the case fatality rate. To fully appreciate the significance of discrepancies in fatality rate usage by NIAID, the CDC, and the WHO, brief definitions of relevant epidemiological terms follow.

Case fatality rates (CFRs), infection fatality rates (IFRs), and mortality rates are used by epidemiologists to describe deaths during and after an infectious disease outbreak. The CDC defined a mortality rate as the frequency of deaths within a time period relative to the size of a well-defined population. Patients may be classified as having an influenza-like illness (ILI) such as COVID-19 according to standard criteria in a case definition. A CFR is defined as the proportion of deaths among confirmed cases of the disease. CFRs indicate the disease severity, while an IFR is defined as the proportion of deaths relative to the prevalence of infections within a population. IFRs are estimated following an outbreak, often based on representative samples of blood tests of the immune system in individuals exposed to a virus. Estimation of the IFR in COVID-19 is urgently needed to assess the scale of the coronavirus pandemic.

Because different types of fatality rates can vary widely, it is imperative to not confuse fatality rates with one another; else misleading calculations with significant consequences could result.

DISCUSSION

Confusion between CFRs and IFRs may seem trivial, and it is easy to overlook at first, but this confusion may have ultimately led to an unintentional miscalculation in coronavirus mortality estimation. IFRs from samples across the population include undiagnosed, asymptomatic, and mild infections, and are often lower compared with CFRs, which are based exclusively on relatively smaller groups of moderately to severely ill diagnosed cases at the beginning of an outbreak. Due to host defense mechanisms and autoimmunity provided by innate and adaptive immune responses, asymptomatic infections are often prevalent in influenza. With many asymptomatic infections already identified in COVID-19, it appears unlikely that the IFR in an ILI like COVID-19 would approximate the disease’s CFR. Presymptomatic infections can also lower the proportion of asymptomatic infections. For example, a CDC report found that asymptomatic individuals identified through reverse transcriptase-polymerase chain reaction (RT-PCR) testing developed symptoms a week later, and those individuals were re-classified as having been presymptomatic at the time of testing.

... A comparison of coronavirus and seasonal influenza CFRs may have been intended during Congressional testimony, but due to misclassifying an IFR as a CFR, the comparison turned out to be between an adjusted coronavirus CFR of 1% and an influenza IFR of 0.1%. Had the adjusted coronavirus mortality rate not been lowered from 3% to 1%, fatality comparisons of the coronavirus to the IFR of seasonal influenza would have increased from 10-times higher to 20- to 30-times higher. By then, epidemiologists might have been alerted to the possibility of a miscalculation in such an alarming estimation.

Quality Assurance

Most people rely on trusted public health experts from organizations like the CDC to disseminate vital information on infectious diseases.26 Unfortunately, even experts can make simple miscalculations that can lead to catastrophic results.

... Mitigation Measures

... Scientists also warned of public health decisions made without reliable data of infection prevalence within the population.45,48 Lacking valid input data due to insufficient testing for disease prevalence, statistical modeling methods often relied on speculative assumptions, producing fearful predictions of increased mortality, which have often proved unreliable. A systematic review found that most diagnostic and predictive models for COVID-19 lack rigor, have a high risk of selection bias, and are likely to have lower predictive performance in actual practice compared with optimistic reports published in the research literature.

... Fear and Collateral Damage

Psychological adverse effects, such as anxiety, anger, and posttraumatic stress, have been linked to restrictive public health mitigation measures due to isolation, frustration, financial loss, and fear of infection.

... Fear, in contrast to moral civic duty and political orientation, was shown to be a more powerful predictor of compliance with mitigating behavior in response to a viral pandemic, but with decreasing well-being and poorer decision-making. Studies have shown that fear impairs performance of cognitive tasks through debilitating anxiety and worry. Even if a threat ceases to exist, prolonged fearful avoidance of threats is maladaptive and restricts a return to normal social interaction and productivity.

... Exaggerated levels of fear were driven by sensationalist media coverage during the COVID-19 pandemic.And yet, while the public was ordered to lockdown, overall costs and benefits to society from severe mitigation measures had not been assessed. Fear of infection also prevented people from seeking needed health-care services in hospitals during the pandemic.The ethics of implementing fear-based public health campaigns needs to be reevaluated for the potential harm these strategies can cause.

... In addition, legal and ethical violations associated with mitigation of pandemic diseases were previously investigated by the Institute of Medicine in 2007. People should have the right to full disclosure of all information pertinent to adverse impacts of mitigation measures during a pandemic, including information on legal and constitutional human rights issues, and the public should be guaranteed a voice in a transparent process as authorities establish public health policy.

Last, severe mitigating measures during the COVID-19 pandemic caused considerable global social and economic disruption.71 Enforced lockdowns increased domestic violence, closed businesses and schools, laid off workers, restricted travel, affected capital markets, threatened the security of low-income families, and saddled governments with massive debt. Between February and April 2020, US unemployment rose from 3.5%, the lowest in 50 years, to 14.7%. A recession in the United States was also officially declared in June 2020 by the National Bureau of Economic Research, ending 128 months of historic economic expansion. Of relevance, economic downturns are associated with higher suicide rates compared with times of prosperity, and increased suicide risk may be associated with economic stress as a consequence of severe mitigation measures during a pandemic.Relapses and newly diagnosed cases of alcohol use disorder were also predicted to increase due to social isolation, and harmful drinking in China increased 2-fold following the COVID-19 outbreak.As a global natural experiment, psychological outcomes from restrictive interventions in the COVID-19 pandemic require further investigations.

... CONCLUSIONS

Sampling bias in coronavirus mortality calculations led to a 10-fold increased mortality overestimation in March 11, 2020, US Congressional testimony. This bias most likely followed from information bias due to misclassifying a seasonal influenza IFR as a CFR, evident in a NEJM.org editorial. Evidence from the WHO confirmed that the approximate CFR of the coronavirus is generally no higher than that of seasonal influenza. By early May 2020, mortality levels from COVID-19 were considerably below predicted overestimations, a result that the public attributed to successful mitigating measures to contain the spread of the novel coronavirus.

This article presented important public health lessons learned from the COVID-19 pandemic. Reliable safeguards are needed in epidemiological research to prevent seemingly minor miscalculations from developing into disasters. (read more)

______________________

Permission is hereby granted to any and all to copy and paste any entry on this page and convey it electronically along with its URL,
______________________