Rapid Covid-19 tests are being deployed by the millions across the nation. The federal government is sending these tests, which can provide results in minutes, to states for educators, students, nursing home patients, first responders, and other sites.
That’s a good thing. But in a rush to get individual test results, we’re making a dangerous public health mistake: We’re losing critical data about Covid-19.
For months, the U.S. has struggled to get accurate information about Covid-19 cases and testing about different demographic groups. As rapid tests surge — and are performed at sites that don’t follow specific Covid-19 data reporting processes — even more information will be lost.
Accurate data are essential for fighting public health crises like Covid-19. Policymakers need complete, timely information about who is getting the disease to track its spread, identify hot spots, and form effective, equitable solutions. Without data, public health strategies will be haphazard and ineffective.
The Covid-19 pandemic has been a data horror story from the beginning. For months, Americans struggled to get Covid-19 tests, leaving cases of the infection undetected. The data that did exist were woefully incomplete, often lacking important information like a patient’s race and ethnicity.
Data collection has improved since then, but is still lacking. The Centers for Disease Control and Prevention website has race and ethnicity data only for half of Covid-19 cases. And the Trump administration is under investigation for tampering with the CDC’s Covid-19 reports.
Without significant changes, rapid tests will make this data deficit worse.
Laboratories are required to provide extensive data to accompany Covid-19 tests. According to federal guidance published in June, they must collect and report for each test demographic information such as the patient’s age, race, ethnicity, sex, and ZIP code.
Those are sound reporting requirements. But thousands upon thousands of testing sites nationwide are beginning to operate outside of laboratory facilities and are not equipped to comply with state and federal reporting regulations. That includes the vast majority of nursing homes — major hot spots for Covid-19 — as well as many urgent care and physician offices. It also includes unconventional sites like community testing pop-ups.
So a site could test hundreds of people in a day and not report comprehensive data to local and federal authorities. And if reporting does occur, it could take a while to be entered into a digitized system. That means missing crucial data, particularly information on racial inequities.
Gaps in data reporting will become more common in the months to come. Rapid at-home Covid-19 tests that give on-the-spot results will soon hit the market. These tests provide immediate results, but it’s unknown if they will require people to report their results or demographic information to health departments. That means we could lose millions of data points about Covid-19.
To be sure, quick test results are critically important to combat Covid-19. But those tests should come with the requirement to report comprehensive, accurate data back to local or state public health agencies.
We should be moving forward on data collection efforts, not backward. Fortunately, several legislators and organizations are leading the way. Sen. Elizabeth Warren (D-Mass.) and Rep. Ayanna Pressley (D-Mass.), for example, with others have proposed several bills aimed at accurate, timely data collection. The Data: Elemental to Health Campaign aims to secure $100 million in federal funding over 10 years for the CDC and state, local, tribal, and territorial health departments to modernize health data and create a more effective data system for public health.
And the We Must Count coalition — a collection of racial and health equity organizations for which I serve as a senior adviser — is pushing for uniform data collection and reporting on Covid-19 testing, cases, and health outcomes disaggregated by race, ethnicity, socioeconomic status, and other factors.
We must stop at nothing to get accurate data about the results of Covid-19 tests and other aspects of the pandemic. Otherwise, we will lose our battle against it.
Joia Crear-Perry is an OB-GYN and founder and president of the National Birth Equity Collaborative.
Data collection is easy. The tech giants (Google, Facebook) do it all the time, and they know how to keep that data anonymous. For once, the big guys can be relied upon to do something really socially positive, if government provides oversight and accountability. The old model of taking 10 years (!) to do something that could be done in something closer to 10 days with the right technology and proper oversight is what’s really wrong with the data collection horror story. New Zealand seems to be doing the job just fine with their remarkably diverse population. It’s time for the USA to stop thinking of itself as top dog and start learning some new tricks from other nations with better governmental relationships to public health and big data.
A broader strategy will use rapid tests asymptomatically and potentially outside of hot spots. In these low incidence settings a confirmatory test will be needed once the result of the test is positive. This is due to the low positive predictive (PPV) value in these kinds of settings. Basically once you have PPVs below 50% a positive rapid test is meaningless, and the person taking it will need to have the result confirmed by e.g. by PCR. This result will then go into the statistics, soothing the concerns you mention in the article.
What most people do not understand is, that RTs still make sense, since most of the results will be negative (no confirmatory test needed). Even if we get 50% false positives, a positivity rate of 50% in the PCR Test is way higher the the current results. In other terms, if you send someone with a positive rapid test to take a PCR Test, their “chance” is 50% to get a positive result (bad for them, good for the public health strategy of test-isolate). If you send a random person (without prior taking of a rapid test) to take a PCR test their “chance” might only be 1-5% (incidence dependent). With these differences it becomes clear, that using rapid tests is an important screening strategy that will even help to save PCR-test resources.
Well said, Alex B! The benefits will be greater still if more data can be captured prior to PCR testing, which is the point of this article. Setting aside for the moment active data tampering by government itself, even a wholly committed public health system that relies on conventional top-down approaches will operate far too slowly to catch up in a pandemic. We have the raw data collection capacity, but it lies outside of government in places like Silicon Valley. The time has come to call on the tech giants nurtured in the USA to serve the public that gave them crucial funding and tax breaks to get them up and running. If Google even remembers that they promised in writing “not to be evil”, now is the time for them to step up and prove they meant what they wrote.