The Truly Heroic “COVID Tracking Project” at The Atlantic Magazine has Ended

A year ago, one of the biggest challenges with the COVID-19 pandemic was simply trying to figure out how many people In the United States were infected with the disease. It was startling to find out that no one in government, at either the state or federal level, seemed to know exactly how many people were being tested, infected, or hospitalized. The Centers for Disease Control (CDC), which should have been the authoritative source for this information, was having a difficult time even assembling data from different states. 

• 

• 

One of the few sources of useful information was a volunteer effort organized by two staff writers at the Atlantic magazine. Robinson Meyer and Alexis Madrigal stepped into the breach in March of 2020 and created the “COVID Tracking Project,” a volunteer effort to assemble daily information on infections and testing. In the first few months of the pandemic, it was one of the only legitimate sources of data and became a trusted source not just for scientists and physicians like me interested in the pandemic, but also for the federal government as it continued to have difficulty assembling the numbers. The Project stopped reporting data on March, 7, 2021.

This effort was all the more important because most academic and public health resources spent their energy on making projections about the pandemic rather than acquiring data. Unfortunately, many of the academic public health resources were making faulty projections because the data they were using to base their computational models on was itself flawed. Only the Atlantic magazine’s effort was focused on accurately acquiring data and understanding the problems with data reporting that were causing difficulties in understanding the pandemic. 

In an article entitled “Why the Pandemic Experts Failed”, written as somewhat of a goodbye note in The Atlantic, Meyer and Madrigal reflect on “the painstaking labor of its more than 550 contributors” which allowed the project to be “among the first to identify virus surges in the Sun Belt and the Midwest,” to determine “the outsize importance of nursing homes in driving COVID-19 deaths,” and find “widespread evidence of overwhelmed hospitals during the harsh winter surge.”

Maybe more importantly, they grew to understand the limitations of public health data reporting in federal government and in states that make the numbers so unreliable. They also saw the “allure” of using the government’s flawed statistics to direct every issue in the pandemic response from lockdowns and school closures, to blocking interstate travel and quarantines. Often, flawed decisions based on inaccurate data undermined the public’s confidence in the government and led to conflicts that have needlessly divided the country,  

Meyer and Madrigal point out that these issues remain and continue to bedevil public officials. They call out CDC director Rochelle Walensky’s March 1^st comments that “cautioned the public about new coronavirus variants,” while the “data at the COVID Tracking Project showed this narrative of a variant-driven surge didn’t hold.” Instead, what was being observed were huge backlogs in death certificates in Texas and several other states that had been immobilized by a winter storm. When reports caught up, they acutely increased death numbers; however, these numbers have subsequently fallen by almost 25 percent. Bad data make well intention public officials appear disingenuous and feed the conspiracy theories that have been a backdrop to the entire pandemic response.

Other significant problems with the country’s public health data persist. Meyer and Madrigal report that the “COVID Tracking Project has shown that at least five states have disturbingly incomplete testing data. In some states, 80 percent of tests are missing from the equivalent federal data set. Yet the CDC is referring leaders of those states to its own test-positivity-rate data—which are calculated from these inaccurate data—when they consider reopening their schools.”

Most importantly, these two writers argue that, “To avoid another data calamity, our public-health system must expend as much energy on understanding the present as it does on modeling the future.” They point out that, “Governing through a pandemic—or any emergency—is about making the least-bad decisions with the best information available.” And they advise, “if you do look at the data, then you must understand how each point, each cell, was made; otherwise, you’re likely to be misled.”

It is hard to reward writers for the difficult job of acquiring data without commenting or trying to interpret it. Yet, the generation of hard factual information is the basis of every decision we make as a country and as individuals. The fact that a band of private citizens came together to provide this service for our country at its time of need is truly heroic. While it may not garner a Pulitzer Prize, it deserves a Presidential Medal of Freedom!