Rehabilitation and In-Hospital Mortality in COVID-19 patients.

BACKGROUND: COVID-19 guidelines endorse early rehabilitation to improve outcomes in hospitalized patients, but the evidence-base to support this recommendation is lacking. We examined the association between early rehabilitation and in-hospital deaths in COVID-19 patients. METHODS: Single center retrospective study involving 990 COVID-19 patients (42·4% women, mean age 67.8 years) admitted between March 1, 2020 and May 31, 2020 to a community hospital. Association of rehabilitation during hospitalization with in-hospital mortality was examined using logistic regression analysis adjusted for demographics, length of stay, body mass index, comorbid illnesses, functional status as well as for COVID-19 presentations, treatments, and complications. RESULTS: Over the 3-month study period, 475 (48·0%) in-patients were referred for rehabilitation. Patients who received rehabilitation were older (73·7 ± 14·0 vs. 62·3 ± 17·2). There were 61 hospital deaths (12·8%) in the rehabilitation group and 165 (32·0%) in the non-rehabilitation group. Receiving rehabilitation was associated with an 89% lower in-hospital mortality (OR 0·11, 95% CI 0·06-0·19) after adjusting for multiple confounders and COVID-19 disease markers. In sensitivity analyses, the results were significant in sub-populations defined by age group, sex, race, length of hospitalization, or pulmonary presentations. Each additional rehabilitation session was associated with a 29% lower risk of in-hospital mortality (OR per session 0·71, 95% CI 0·64-0·79) in the fully adjusted model. CONCLUSION: Among hospitalized COVID-19 patients, receiving early rehabilitation was associated with lower in-hospital mortality. Our findings support implementation of rehabilitation services for COVID-19 patients in acute care settings, but further research from randomized clinical trials is needed.