Comparison of COVID-19 mitigation and decompression strategies among homeless shelters: a prospective cohort study

•Guests in the shelter that utilized temporary outdoor tents for decompression were 6.21 times more likely to be tested positive for SARS-CoV-2 than guests in the shelter that decompressed their guests to other indoor facilities.•Daily temperature and symptom checks performed poorly alone and should be combined with routine SARS-CoV-2 RT-PCR surveillance.•The study calls on local governments to assist homeless shelters in decompressing their population with suitable indoor facilities and in improving testing capacities to protect this vulnerable population during the pandemic. To compare the effectiveness of COVID-19 mitigation strategies in two homeless shelters in Massachusetts during the pandemic. We conducted a prospective cohort study that followed guests in two Massachusetts homeless shelters between March 30 and May 13, 2020, which adopted different depopulation strategies. One set up temporary tents in its parking lot, while the other decompressed its guests to a gym and a hotel. The outcome was assessed by comparing the odds ratios of positive SARS-CoV-2 RT-PCR assays. Guests residing at the shelter that used temporary tents had 6.21 times (95% CI = 1.86, 20.77) higher odds of testing positive for SARS-CoV-2 at follow-up after adjusting for loss to follow up, age, gender, and race. The daily COVID-19 symptoms checklist performed poorly in detecting positive infection. The study highlights the importance of depopulating shelter guests with stable and adequate indoor space to prevent SARS-CoV-2 transmission. Daily temperature and symptom checks should be combined with routine testing. With the rising homelessness due to mass unemployment and eviction crisis, our study supports further governmental assistance in decompressing homeless shelters during this pandemic.