Bad news: COVID-19 numbers are pretty meaningless

The number of cases, test positivity, hospitalizations and deaths being reported by local and state health departments, the mainstream media and the Centers for Disease Control and Prevention (CDC) have little meaning these days.

Case counts meant something very early on in the epidemic when each case reliably was associated with a certain risk of severe disease, hospitalization or death. Once there was a substantial number of people with immunity to severe outcomes due to recovery or vaccination those case counts became disconnected from expected outcomes.

While case counts were always underestimated due to the fraction of asymptomatic infection, now with the widespread availability of self-testing, reported cases greatly underestimate the number of circulating true cases even further.

Test positivity also used to be a reliable measure of the community burden of infection with increased positivity correlating with increased spread of infection and hospitalization. But because now many only seek medical testing to confirm a home-based positive test for employment sick-leave or other purposes, the frequency of those testing positive through medical testing is artificially high.

Remember the days of “flatten the curve,” the idea that we had to preserve hospital capacity through efforts to reduce the spread of infection? That curve was the number of hospital admissions due to COVID-19 and reasonably reflected the number of people admitted to hospitals severely ill with COVID-19. Now, however, due to the continued universal screening of all hospital admissions, a majority of reported COVID-19 hospitalizations are not hospitalized “for” COVID-19 but “with” COVID-19. 

I and several colleagues recently took a deep dive into COVID hospitalization at the Los Angeles County public hospital. Among 462 COVID hospitalizations, only 32 percent required oxygen, the most basic therapy for COVID-19, suggesting that the other 68 percent were admitted for reasons other than COVID-19.

Like testing results and hospitalizations, death counts are also an overestimate of deaths caused by COVID-19. There are various reasons why a positive COVID-19 test result might be added to a death certificate resulting in the reporting of COVID-19 related death in someone who only tested positive during their hospitalization because they were tested on admission.

In general, we rely on organizations like the CDC to conduct surveillance and monitoring of diseases in the United States. Over the decades the CDC has created and maintained robust systems for disease monitoring that include sentinel surveillance — where a select group of clinics or hospitals is intensively monitored for specific diseases, population-based surveys — where individuals are randomly sampled for the presence or absence of disease, and case reporting. Unfortunately, case reporting, which we are relying on now to understand COVID-19 in the United States is the weakest type of surveillance for an ongoing pandemic.

Moving forward, we must improve our sentinel hospital surveillance to include only those cases likely to be a true COVID-19 hospitalization. Many experts suggest that can be easily done by counting those cases that required oxygen therapy or specific COVID-19 treatment. Population-based surveys would be very useful, albeit expensive and time-consuming, but conducted in an ongoing statistically meaningful fashion could be very informative.

The bottom line is that we must understand and accept the limitations of current COVID-19 numbers. We should not be responding out of proportion to the severity of the epidemic. We should be focused on making sure vulnerable people have easy and timely access to effective treatment and investing in new vaccines that can truly prevent future infections.   

Jeffrey D. Klausner, MD, MPH, is a clinical professor of Medicine, Infectious Diseases, Population and Public Health Sciences at the Keck School of Medicine of the University of Southern California. He is a former U.S. Centers for Disease Prevention and Control medical officer and former San Francisco City and County deputy health officer.