Timing and Predictors of Loss of Infectivity Among Healthcare Workers With Mild Primary and Recurrent Coronavirus Disease 2019 (COVID-19): A Prospective Observational Cohort Study

Abstract

Abstract Background There is a need to understand the duration of infectivity of primary and recurrent coronavirus disease 2019 (COVID-19) and identify predictors of loss of infectivity. Methods Prospective observational cohort study with serial viral culture, rapid antigen detection test (RADT) and reverse transcription polymerase chain reaction (RT-PCR) on nasopharyngeal specimens of healthcare workers with COVID-19. The primary outcome was viral culture positivity as indicative of infectivity. Predictors of loss of infectivity were determined using multivariate regression model. The performance of the US Centers for Disease Control and Prevention (CDC) criteria (fever resolution, symptom improvement, and negative RADT) to predict loss of infectivity was also investigated. Results In total, 121 participants (91 female [79.3%]; average age, 40 years) were enrolled. Most (n = 107, 88.4%) had received ≥3 severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) vaccine doses, and 20 (16.5%) had COVID-19 previously. Viral culture positivity decreased from 71.9% (87/121) on day 5 of infection to 18.2% (22/121) on day 10. Participants with recurrent COVID-19 had a lower likelihood of infectivity than those with primary COVID-19 at each follow-up (day 5 odds ratio [OR], 0.14; P < .001]; day 7 OR, 0.04; P = .003]) and were all non-infective by day 10 (P = .02). Independent predictors of infectivity included prior COVID-19 (adjusted OR [aOR] on day 5, 0.005; P = .003), an RT-PCR cycle threshold [Ct] value <23 (aOR on day 5, 22.75; P < .001) but not symptom improvement or RADT result. The CDC criteria would identify 36% (24/67) of all non-infectious individuals on day 7. However, 17% (5/29) of those meeting all the criteria had a positive viral culture. Conclusions Infectivity of recurrent COVID-19 is shorter than primary infections. Loss of infectivity algorithms could be optimized.