The Effects of Postacute Rehabilitation on Mortality, Chronic Care Dependency, Health Care Use, and Costs in Sepsis Survivors.

Abstract

Rationale: Sepsis often leads to long-term functional deficits and increased mortality in survivors. Postacute rehabilitation can decrease long-term sepsis mortality, but its impact on nursing care dependency, health care use, and costs is insufficiently understood. Objectives: To assess the short-term (7-12 months postdischarge) and long-term (13-36 months postdischarge) effect of inpatient rehabilitation within 6 months after hospitalization on mortality, nursing care dependency, health care use, and costs. Methods: An observational cohort study used health claims data from the health insurer AOK (Allgemeine Ortskrankenkasse). Among 23.0 million AOK beneficiaries, adult beneficiaries hospitalized with sepsis in 2013-2014 were identified by explicit codes from the International Classification of Diseases, Tenth Revision. The study included patients who were nonemployed presepsis, for whom rehabilitation is reimbursed by the AOK and thus included in the dataset, and who survived at least 6 months postdischarge. The effect of rehabilitation was estimated by statistical comparisons of patients with rehabilitation (treatment group) and those without (reference group). Possible differential effects were investigated for the subgroup of ICU-treated sepsis survivors. The study used inverse probability of treatment weighting based on propensity scores to adjust for differences in relevant covariates. Costs for rehabilitation in the 6 months postsepsis were not included in the cost analysis. Results: Among 41,918 6-month sepsis survivors, 17.2% (n = 7,224) received rehabilitation. There was no significant difference in short-term survival between survivors with and without rehabilitation. Long-term survival rates were significantly higher in the rehabilitation group (90.4% vs. 88.7%; odds ratio [OR] = 1.2; 95% confidence interval [95% CI] = 1.1-1.3; P = 0.003). Survivors with rehabilitation had a higher mean number of hospital readmissions (7-12 months after sepsis: 0.82 vs. 0.76; P = 0.014) and were more frequently dependent on nursing care (7-12 months after sepsis: 47.8% vs. 42.3%; OR = 1.2; 95% CI = 1.2-1.3; P < 0.001; 13-36 months after sepsis: 52.5% vs. 47.5%; OR = 1.2; 95% CI = 1.1-1.3; P < 0.001) compared with those without rehabilitation, whereas total health care costs at 7-36 months after sepsis did not differ between groups. ICU-treated sepsis patients with rehabilitation had higher short- and long-term survival rates (short-term: 93.5% vs. 90.9%; OR = 1.5; 95% CI = 1.2-1.7; P < 0.001; long-term: 89.1% vs. 86.3%; OR = 1.3; 95% CI = 1.1-1.5; P < 0.001) than ICU-treated sepsis patients without rehabilitation. Conclusions: Rehabilitation within the first 6 months after ICU- and non-ICU-treated sepsis is associated with increased long-term survival within 3 years after sepsis without added total health care costs. Future work should aim to confirm and explain these exploratory findings.