We should stop obsessing over COVID numbers — but count infections better

Federal and state officials are debating whether to stop putting so much emphasis on the number of people infected with coronavirus. That’s warranted. But an updated approach to tracking the virus and infections is what’s really needed.

The media have widely reported — you might even say obsessively reported — the rate of infection, the changes in that rate over time, and the cumulative case count since the beginning of the pandemic. During surges, those numbers are especially closely watched — affecting public health policy, behavior, anxiety levels and the national mood.

The argument now is that the infection rate and case count have outlived their usefulness as meaningful barometers — and may even be doing harm. Less focus on these numbers could help shift us toward “living with the virus” and reducing social restrictions and mandates.

Four main reasons are cited:

1) The numbers are increasingly inaccurate anyway

As we shift from lab-processed PCR tests, which are routinely reported to authorities, to rapid home tests, which are not, we’re losing track of how many people are infected at any one time and over time.

Home tests are fast eclipsing PCR testing. Some 60 million (about half) of American households in the last month have requested their initial allotment of four free tests from the new federal website, covidtests.gov. States, employers, schools and local libraries are also distributing free home tests and millions have been purchased online and at stores.

2) Better ways to measure infection prevalence exist

These include routine wastewater testing, surveys of the public and targeted PCR testing. The Centers for Disease Control and Prevention (CDC) this month signaled a nationwide expansion of wastewater monitoring. The agency gets data from around 400 sites already.

3) It’s close to endemic

Once an infectious agent becomes commonplace over years (endemic), estimates based on routine surveillance and statistical modeling suffice to track its spread and impact. That’s what we’ve done with flu and others infectious diseases for decades.

4) More meaningful measures

Hospitalizations and deaths are now more meaningful measures on which to base public health policy. Preventing infection is a lesser priority given that more than half the population has already been infected and most people who contract the virus have no symptoms or a minor course of illness. 

These are compelling reasons to change course. I would add another: The cumulative infection count has been misleading all along. The CDC estimated last fall, for example, that for every infection confirmed through PCR testing between February 2020 and September 2021 three other infections went undetected and unreported. The reasons: the slow ramp-up of testing in 2020; spotty access to testing, including during the recent surge; a low rate of testing among asymptomatic people; and resistance to testing.

If the CDC estimates are right, and most experts believe the undercount is very high, between 180 million and 250 million Americans have been infected to date — not the official PCR-confirmed number of 77.6 million.

In addition to the measures above, an improved approach to tracking infections and COVID-19, the illness the virus causes, should involve the following steps.

First, the more expensive PCR tests should now be used mostly among vulnerable populations, to confirm rapid test results or on their own. That includes nursing and retirement home residents; hospitalized patients; workers in crowded workplaces; and selected groups of people over age 65. Rates of infection among these populations are more important to know with the precision the PCR tests allow. Also, unburdened from having to process millions of tests a day, labs should be able to deliver results in 24 hours or less. 

Second, genetic sequencing of PCR samples to identify variants and their mutations should be greatly expanded. The U.S. lags other countries in sequencing. The Biden administration has committed $1.7 billion to expand it. More money could be needed.

Third, a much larger percentage of positive home test results need to be reported and counted. As it stands now, if another highly infectious variant comes along, public health officials will be at a significant disadvantage in assessing the rate of spread. 

The government, manufacturers and medical groups currently recommend that people call their doctor to report a positive result from a rapid test. That’s not remotely happening at the level needed. The CDC, White House and American Medical Association acknowledged in emails to me that this was a problem, but all declined to suggest ways to enhance home test reporting.

One idea is to offer incentives. Those could include additional free home tests or even cash. Another idea would be to create a website dedicated to securely reporting such results, anonymously if needs be. Parents nationwide have been reporting their kids’ home test results to school websites for months with little problem.  

At the very least, public messaging on reporting home test results should be ramped up. It’s been nil so far. To be sure, many people still won’t report. But then we don’t need everyone to do it to yield more accurate estimates.

Fourth, breakthrough infections (among the vaccinated) and reinfections need to be tracked far more closely. Only a handful of states do so now, and the CDC is not counting such cases nationally. That agency last year initially pooh-poohed counting breakthrough infections then got caught short when omicron led to a spike in both breakthrough infections and reinfections. Knowing the real rate sooner could have led to earlier warnings to the public. 

Matching up state PCR-result databases with vaccination databases, along with surveys, could get this job done — providing important measures of virus prevalence and strength of immunity in the population.

Lastly, the U.S. needs a national disease registry for people diagnosed with “long covid.” Current estimates are that from 10 to 30 percent of ever-infected people develop some form of long covid. Over the next year we simply must narrow that estimate and better understand what we’re dealing with.

Disease registries allow doctors and researchers to monitor the clinical care and outcomes of a patient population. The National Institutes of Health has about $1 billion to study long covid. If more money is needed for a registry, Congress should fund it.

The U.S. has a mixed record on tracking the pandemic’s course. SARS-CoV-2 is too wily and unpredictable a foe to permit complacency as infections decline. Vigilant and improved tracking will better position us to respond quickly should a new, and possibly more dangerous, variant arise.

Steven Findlay, MPH, is an independent health policy analyst and journalist. He previously worked as a senior health policy analyst at Consumers Union, as well as director of research and policy at the National Institute for Health Care Management.