Analyzing SARS-CoV-2 case numbers and clustering to predict a nursing home outbreak

Abstract

The COVID-19 pandemic devastated nursing homes, highlighting the urgent need for effective outbreak control measures. This study analyzed twice-weekly PCR surveillance data from 134 Veteran Affairs Community Living Centers (December 2021-June 2022) to identify early predictors of SARS-CoV-2 outbreaks. Among 16,353 residents (mean age 74, 96% male, 68% white), we identified 1,868 infections and evaluated neighborhood ward-level case counts and their association with subsequent infections over two-week periods. Epidemic unit-days with no initial cases had an 87.49% likelihood of remaining case-free, while those with ≥ 4 initial cases demonstrated a 38.5% probability of developing ≥ 4 additional cases. These findings indicate that early case clusters strongly predict larger outbreaks, underscoring the importance of rapid detection and intervention. Study limitations include demographic homogeneity and reliance on frequent PCR testing, potentially limiting generalizability. This research provides a valuable framework for refining outbreak definitions and improving infection control strategies for respiratory virus outbreaks in nursing homes.