Growing use of at-home tests could lead to an undercount of omicron cases

The omicron surge has triggered a rapid increase in the use of at-home coronavirus tests. Supplies of those tests are now running low, even as access to PCR tests (which are sent to a lab for results) is strained amid rising demand.

The Biden administration pledges to resolve this serious problem through a program to increase the number of testing sites and the production of at-home tests. Further, the administration plans to provide at least 500 million at-home tests free starting in January.

This plan is welcome, although it’s unclear whether the ramp-up can meet the demand in time. That depends on the trajectory of the omicron outbreak. If the outbreak extends well into February or March even a billion free tests will not be enough in a nation of 334 million people.

But there’s another challenge associated with the at-home tests: The results they yield are not widely reported to health authorities. I asked the Centers for Disease Control and Prevention (CDC) about this and got an honest but incomplete answer.

Until now the dominant form of testing involved going to a pharmacy, provider or pop-up vaccination site to get a PCR test. In all states, the results of those tests, which are processed in a lab, are reported to authorities. That allows for counting test-confirmed cases, detecting where outbreaks or super-spreader event are occurring, as well as assessing the general state of the pandemic by geographic area.

Those test results are also transmitted to public health departments and their contractors to perform contact tracing.

With home test kits this protocol does not exist. People who test positive on an at-home test do not have their result counted in local, state or federal databases unless they choose to get a confirmatory PCR test.

Until recently, the vast majority of people with a positive at-home test result did get a confirmatory PCR test — since the PCR tests were widely accessible and use of the at-home tests was limited. They only became widely available this past summer. 

That is all now changing. Demand for the home tests is soaring. Pharmacies and PCR test sites are overwhelmed. And more people are understandably reluctant to go to a pharmacy or test site and stand in line — possibly for hours and multiple times. Plus, experience with and trust in the home tests — and their convenience — is leading many more people to skip the confirmatory test. If you get a “positive” result with two or more at-home tests over a day or two, it’s pretty certain you’re infected whether you have symptoms or not.

Some eight different at-home test kits are now on the market. A few require interaction with online apps. With these, the result is documented and reported. But the most popular home tests — such as Binax, FlowFlex and iHealth — don’t have an easy reporting mechanism.

The reality is that the majority of people using such tests now are unlikely to report a positive test result to anyone. In fact, in November, CDC removed a long-standing recommendation that people report self-test results to a health provider. A CDC spokesperson told me via email, “This was done to intentionally remove barriers to self-test usage,” noting CDC now “encourages anyone with a positive self-test to contact their healthcare provider.”

But there’s no federal requirement for doctors or health systems to report the at-home test results either. Most doctors have not been doing so because most people were getting confirmatory PCR tests.

Acknowledging that the situation is now fluid, CDC spokesperson Jasmine Reed said, “CDC is working through the challenges of reporting self-test data to consider creative ways of collecting this data in the future.” 

That’s good news but when is “in the future?” The agency declined to specify. Reed also added, “CDC does not need to count every case to be able to make judgments about the spread or path or progress of the pandemic.” 

That’s not entirely accurate. Gaps in COVID-19 testing and data gathering have plagued the pandemic response since the beginning. Poor access to PCR testing, with frequent delays in results, were widespread from February to August 2020 and then again during last winter’s three-month surge in cases. 

This, along with other factors, has contributed to a substantial undercount of infected people, and thus the actual spread of the virus in the United States. Simply put, lots more people than you might think have been infected with coronavirus, and the majority of infected people over the past two years who had no symptoms or very mild symptoms never got tested.

Knowing this, CDC has used statistical methods to estimate the real number of people ever infected. As of the agency’s latest update on Nov. 16, for every infection confirmed through PCR testing three infections went unconfirmed or unreported between February 2020 and September 2021. Said another way, only one in four infections were confirmed through PCR testing during that period.

Extrapolated to the current U.S. PCR-confirmed case count of around 50 million people, the estimated cumulative undercount (through September 2021) of ever-infected people is between 135 million and 185 million. Thus, 185 million to 235 million have ever been infected. In a population of 334 million, that’s 55 to 70 percent. 

The estimated undercount of deaths due to covid-19 is smaller, according to CDC data — on the order of 20 to 30 percent. Still, that would put the number of people who have died due to COVID as of this month at over 1 million, not the official figure of 815,000.

I’ve long believed that more emphasis on these numbers could and should have led to more aggressive efforts to mitigate exposure and contact trace. I know from other reporting and personal experience (as a contact tracer in Maryland) that at most only about half of the close contacts of infected people were ever reached nationwide.

Would people in, say, a city of 50,000 have behaved differently if they had been informed a year ago that instead of 10 percent of their community being infected the real number was likely closer to 25 or 30 percent? I think so.

The issue now, with soaring use of at-home tests, is that CDC could be partially blind to the magnitude of the omicron surge. By how much is not yet knowable. Too many factors are in play. The Biden administration and CDC should, however, heed the lessons of past mistakes and immediately resolve how to best track positive at-home test results.

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.