For more than a year, it has emerged that many hospitalizations officially classified as due to COVID-19 are instead of patients without symptoms of COVID who are admitted for other reasons but who also test positive. Since almost everyone is still routinely swabbed upon admission to hospital (although the largest infection control organization has recommended against it), many patients with other conditions also receive a positive test result, especially during ongoing Omicron flare-ups, thus exaggerating the number of hospitalizations tabulated as caused by COVID-19. UCLA researchers who examined data from the Los Angeles County public hospital found that more than two-thirds of official hospitalizations for COVID-19 since January 2022 were actually “with” rather than “for” the illness.
A rigorous assessment from Massachusetts determined that a comparable proportion of COVID hospitalizations were actually incidental to the coronavirus. An attending physician at Emory Decatur Hospital (and former president of the Georgia Chapter of the Infectious Diseases Society) quoted by Dr Wen estimates that around 90% of patients diagnosed with COVID at his hospital are now being treated for another illness. Wen also cited the Tufts Hospital epidemiologist, who similarly observes that recently the proportion of patients hospitalized with COVID-19 has been as low as 10% of the number of people who would have the disease. All of this is entirely consistent with the reality that by March 2022 over 95% of people had already been infected or vaccinated or generally both, and the resulting robust population immunity, combined with the less virulent nature of Omicron, causes much less serious results.
Growing recognition of the overcount of COVID-19 hospitalizations has caused some local authorities as well as the CDC to try to better estimate true levels. Misclassified hospitalizations obviously suggest that there have also been misclassified deaths, but a parallel acknowledgment that undoubtedly many official COVID-19 deaths are also due to people dying with rather than from the coronavirus just beginning to emerge. CDC guidelines still state that any death from (any) illness occurring within 30 days of a positive test result is automatically classified as due to COVID-19. Therefore, if the current prevalence in the population is, say, 3% (towards the lower end of typical levels during major outbreaks like this), then the background prevalence among people admitted to hospitals for d ‘other reasons – and also among those that end up dying – would also be around 3%. Considering that approximately 9,200 total deaths occur daily in the United States, then in this hypothetical scenario, some 275 deaths attributed to COVID (or approximately two-thirds of the official daily tally) would in fact have been due to other causes.
Milwaukee County’s former chief medical examiner has conducted a careful review of some 4,000 reported COVID-19 deaths during the pandemic there. His research found nearly half had no connection to COVID or, in some cases, only a ‘marginal’ association, such as terminal cancer patients whose death may have been hastened. days or weeks after catching the disease. An analysis of LA County and national data collected during the most recent waves of the highly contagious (but significantly less lethal) variants of Omicron suggests that deaths from COVID-19 are now likely overestimated by at least quadruple. A recently published investigation in Denmark documented that, following the emergence of Omicron a year ago, an astonishing 65-75% of deaths officially attributed to COVID-19 were simply incidental to the coronavirus, according to the exercise hypothetical above. Yet even if only half of the deaths currently reported in the United States are not actually caused by the virus, that would mean a real daily COVID-19 toll of around 200, roughly the number of deaths during a bad flu season.
In addition to the overestimated number of COVID-related hospitalizations and deaths, another reason to maintain a public health emergency is the allegedly massive wave of ongoing long COVIDs. Yet almost all of the long COVID reports are based on tables of the number of people who self-report persistent symptoms after infection, rather than controlled studies that carefully compare the prevalence of persistent symptoms in people who have been infected to those that were not. An advertisement on San Francisco Bay Area rapid transit trains warns that any of a number of common illnesses, including headaches, anxiety, diarrhea, muscle aches and concentration problems, can be caused by a long COVID. But case-control studies have so far found, at most, only modest differences in the prevalence of symptoms between previously infected and uninfected people (and new research suggests that most symptoms dissipate in a year). While the long COVID is undeniably a significant issue, as are the deaths still caused by the coronavirus, careful analysis is needed to more accurately estimate prevalence.
The unintended exaggeration of COVID-19 deaths and long COVID is leading not only to misplaced policy decisions, such as new mask mandates and recall recommendations for 6-month-old babies, but also to a climate of needlessly lingering fear, especially in bluer regions (like my hometown of San Francisco, where mask-wearing remains commonplace, even outdoors). After three long years, it is high time to base public health statements and policies on solid scientific evidence rather than on well-intentioned but often misleading assumptions.
More must-reads from TIME