When do people increase COVID-19 protective behaviors? A longitudinal analysis of the impact of infections among close contacts and in the community

Studies examining individual-level changes in protective behaviors over time in association with community-level infection and self or close-contact infection with SARS-CoV-2 are limited. We analyzed overall and demographic specific week-to-week changes in COVID-19 protective behaviors and their association with COVID-19 infections (regional case counts and self or close contacts). Data were collected through 37 consecutive weekly surveys from 10/17/2021 - 6/26/2022. Our survey panel included 212 individuals living or working in St. Louis City and County, Missouri, U.S.A. Frequency of mask-wearing, handwashing, physical distancing, and avoiding large gatherings was self-reported (more/the same/less than the prior week). Close contact with COVID-19 was reported if the panel member, their household member, or their close contact tested positive, got sick, or was hospitalized for COVID-19 in the prior week. Regional weekly COVID-19 case counts were matched to the closest survey administration date. We used generalized linear mixed models to estimate odds ratios (ORs) and 95% confidence intervals (CIs) for associations. Evidence for effect modification was assessed using the likelihood ratio test. Increased protective behaviors were positively associated with COVID-19 case counts (OR_{highest vs. lowest case count category} = 4.39, 95% CI 3.35-5.74) and with participant-reported self or close contacts with COVID-19 (OR = 5.10, 95% CI 3.88-6.70). Stronger associations were found for White vs. Black panel members (p <.0001). Individuals modulated their protective behaviors in association with regional COVID-19 case counts and self or close contact infection. Rapid reporting and widespread public awareness of infectious disease rates may help reduce transmission during a pandemic by increasing protective behaviors.