Importance: The case fatality rate of SARS-CoV-2 has been high among residents of long-term care (LTC) facilities. It is important to know if the excess mortality persists beyond the acute infection.
Objective: To evaluate whether SARS-CoV-2 is associated with higher mortality after the first month from documented infection.
Design: We extended the follow-up period of a previous, retrospective cohort study based on the Swedish Senior Alert register. LTC residents infected with SARS-CoV-2 were matched to uninfected controls using time-dependent propensity scores on age, sex, body mass index, health status, comorbidities, and prescription medication use. In a sensitivity analysis, residents were also matched on geographical region and time of Senior Alert registration.
Setting: LTC facilities in Sweden.
Participants: 3731 LTC residents with SARS-CoV-2 and 3731 controls (n=3604 in each group in the sensitivity analysis).
Exposure: SARS-CoV-2 infection, documented in the SmiNet register (until September 15, 2020).
Main Outcome: All-cause mortality over 8 months (until October 24, 2020).
Results: The median age was 87 years, and 65% were women. Excess mortality was highest 5 days after documented infection (hazard ratio 19.1; 95% confidence interval [CI], 14.6-24.8), after which excess mortality decreased rapidly. After the second month, the mortality rate became lower in infected residents than in controls. Median survival of uninfected controls was 577 days (1.6 years), which is much lower than national life expectancy in Sweden at age 87 (5.05 years in men, 6.07 years in women). During days 61-210 of follow-up, the hazard ratio for death was 0.41 (95% CI, 0.34-0.50) in the main analysis and 0.76 (95% CI, 0.62-0.93) in the sensitivity analysis.
Conclusions and Relevance: No excess mortality was observed in LTC residents who survived the acute SARS-CoV-2 infection (the first month). The life expectancy of uninfected residents was much lower than that of the general population of the same age and sex. This difference should be taken into account in calculations of years of life lost among LTC residents.