Estimating the Undetected Burden of Respiratory Syncytial Virus Hospitalizations in Adults Through Capture-Recapture Methods.

Abstract

Traditional surveillance systems may underestimate the burden caused by respiratory syncytial virus (RSV). Capture-recapture methods provide alternatives for estimating the number of RSV-related hospitalizations in a population. Capture-recapture methods were used to estimate the number of RSV-related hospitalizations in adults in Middle Tennessee from two independent hospitalization surveillance systems during consecutive respiratory seasons from 2016-2017 to 2019-2020. Data from the Hospitalized Adult Influenza Vaccine Effectiveness Network (HAIVEN) and the Emerging Infections Program (EIP) were used. Annual RSV hospitalization rates were calculated using the capture-recapture estimates weighted by hospitals' market share divided by the corresponding census population. Using capture-recapture methods, the estimated overall adult hospitalization rates varied from 8.3 (95% CI: 5.9-15.4) RSV-related hospitalizations per 10,000 persons during the 2016-2017 season to 28.4 (95% CI: 18.2-59.0) hospitalizations per 10,000 persons in the 2019-2020 season. The proportion of hospitalizations that HAIVEN determined ranged from 8.7% to 36.7% of the total capture-recapture estimated hospitalization, whereas EIP detected 23.5% to 52.7% of the total capture-recapture estimated hospitalizations. Capture-recapture estimates showed that individual traditional surveillance systems underestimated the hospitalization burden in adults. Using capture-recapture allows for a more comprehensive estimate of RSV hospitalizations.