Recovery housing predictors of closure risk during COVID-19

Background COVID-19 has had widespread health and economic costs in the United States and around the world, especially for individuals with substance use disorders (SUDs). An important resource to assist those in recovery from SUDs is recovery housing, a housing model that utilizes peer support to help individuals on their path to recovery. Recovery housing has faced additional challenges due to COVID-19. Method We used cross-sectional survey data and a probit regression model to determine important predictors of recovery housing closure risk during COVID-19. Results We found that populations served, COVID-19 policies enacted, and house location are significant predictors of closure risk. We showed that, even when important differences between rural and non-rural houses are controlled for, rural houses are less likely to report being at risk of closing due to COVID-19. Conclusion Rurality remained an important predictor of closure risk, regardless of house characteristics. This may suggest inherent differences across rurality not included in our model. Recovery houses, in both rural and non-rural areas, continue to need support to offer important recovery services as the pandemic continues.