Evaluating the Quality of Federal SARS-CoV-2 Diagnostic Testing Data

Abstract

Purpose: In April 2020, the US Department of Health and Human Services (HHS) and the US Centers for Disease Control and Prevention established the COVID-19 Electronic Laboratory Reporting program (CELR) to collect data on SARS-CoV-2 laboratory tests. Over the course of the following year, the federal government, partnering with the Association for Public Health Laboratories, onboarded every state to submit laboratory results to this system—the first of its kind in the US. We set out to evaluate the quality of data collected by CELR. Methods & Materials: We compared jurisdiction-level data collected through CELR and published by HHS to the testing data published by jurisdictions on their health department webpages. Because jurisdictions define their testing data differently, we anticipated some differences from federal testing data. However, jurisdictions also tend to prioritize their dashboard reporting—since it is what is used for policy decisions like reopening—so we hypothesized that differences from federal data absent a definitional explanation could point to problems with federal data. Where we found differences between jurisdictional and federal data, we conducted interviews with public health officials to understand their cause. Results: Of the 56 states and territories, as of April 2021 (the first month when all states were onboarded to CELR), 38 had federal total data that diverges from state data by more than 5%. Of those states, the differences of 27 could not be explained by definitional factors. Based on our interviews, we identified three problems: non-electronic reporting streams, out-of-date surveillance systems, and deduplication of laboratory data. Conclusion: The federal testing dataset displays major unresolved quality problems, and because states present testing data so differently, state-published data forms a poor alternative to federal datasets. The federal government, which is uniquely positioned to provide testing data on infectious diseases, must work to improve the quality of laboratory data submissions by states. To support better national laboratory data, the United States should invest in updating state and laboratory data surveillance infrastructure—including updates to state surveillance systems and laboratory system updates to eliminate outdated reporting methods like faxes—and in creating more national laboratory data infrastructure.