Asymptomatic health-care worker screening during the COVID-19 pandemic.

We applaud the establishment of the COVIDsortium by Thomas Treibel and colleagues as a bioresource focusing on asymptomatic health-care workers (HCWs). However, we disagree with the authors' conclusion that “the rate of asymptomatic infection among HCWs more likely reflects general community transmission than in-hospital exposure”. This report was an ecological study subject to the ecological fallacy. Moreover, the figure compares symptomatic inpatients who were tested in hospital with asymptomatic HCWs. It is possible that HCWs were exposed to asymptomatic patients who were infectious in the hospital or to patients with false-negative tests. The number of infections among HCWs was also most likely underestimated, as HCWs with symptoms or who were self-isolating were excluded.

The role of hospital exposures versus community exposures has been problematic for assessments of the occupational risk of other infectious diseases (eg, multidrug-resistant tuberculosis). Valid risks for disease were not appreciated until HCWs were compared with controls with similar educational and economic status, such as medical students versus chemical-engineering students.

Finally, evidence exists that the risk to HCWs from severe acute respiratory syndrome coronavirus 2 is not only from community exposures but also from other types of exposures. Among 1423 HCWs in the USA with COVID-19, 780 (55%) HCWs reported contact with a patient with confirmed COVID-19 in the 14 days before the onset of their symptoms, whereas 384 (27%) reported contact only with a household member, 187 (13%) in a community setting, and 72 (5%) in more than one of these settings. Some HCW exposures confer a higher risk than do others, and personal protective equipment and infection control training are associated with a decreased risk of infection.