Abstract
BACKGROUND: Long-term symptoms are common after the acute phase of COVID-19. We hypothesized that sick leave as an estimate of functional recovery, adjusted for confounding, differs between intensive care unit (ICU) and hospitalized COVID-19 patients and population controls. METHODS: In this cohort study, we identified all working-age individuals with COVID-19 admitted to ICUs or hospitals until July 20, 2020 from national registries. Matched population controls were randomly assigned to each ICU patient. Using logistic regression to adjust for confounding, we compared ICU patients to hospital patients and population controls on the number of sick leave-free days alive during the first year after hospital discharge and the proportion of alive individuals on sick leave after 1 year. RESULTS: We included 1020 COVID-19 ICU patients, 5306 COVID-19 hospital patients, and 4387 population controls. The ICU patients had a median of 271 (interquartile range, 33-349) sick leave-free days alive, while hospital patients had 354 (334-365) and population controls 365 (365-365). ICU patients had an odds ratio (OR) of 0.14 (0.12-0.16, 95% confidence interval) compared to hospital patients, and 0.02 (0.02-0.03, both p < 0.001) compared to population controls for at least one more sick leave-free day alive. Being on sick leave 1 year after inclusion had similar but inverse ORs. CONCLUSION: This national cohort study, in ICU and hospitalized patients with COVID-19, shows that the severity of COVID-19 disease, functional and health status before COVID-19, and demographic factors had a major impact on recovery. TRIAL REGISTRATION: NCT05054608.