Background and objectives: Rates of hospitalization are known to be high in patients with kidney disease. However, ongoing risks of subsequent hospitalization and mortality are uncertain. The primary objective was to evaluate patients with kidney disease for long-term risks of subsequent hospitalization, including admissions resulting in death.
Design, setting, participants, & measurements: Patients hospitalized in Washington State between April of 2006 and December of 2008 who survived to discharge (n=676,343) were classified by International Classification of Disease codes into CKD (n=27,870), dialysis (n=6131), kidney transplant (n=1100), and reference (n=641,242) cohorts. Cox proportional hazard models controlling for age, sex, payer, comorbidity, previous hospitalization, primary diagnosis category, and length of stay were conducted for time to event analyses.
Results: Compared with the reference cohort, risks for subsequent hospitalization were increased in the CKD (hazard ratio=1.20, 99% confidence interval=1.18-1.23, P<0.001), dialysis (hazard ratio=1.76, 99% confidence interval=1.69-1.83, P<0.001), and kidney transplant (hazard ratio=1.85, 99% confidence interval=1.68-2.03, P<0.001) cohorts, with a mean follow-up time of 29 months. Similarly, risks for fatal hospitalization were increased for patients in the CKD (hazard ratio=1.41, 99% confidence interval=1.34-1.49, P<0.001), dialysis (hazard ratio=3.04, 99% confidence interval=2.78-3.31, P<0.001), and kidney transplant (hazard ratio=2.25, 99% confidence interval=1.67-3.03, P<0.001) cohorts. Risks for hospitalization and fatal hospitalization increased in a graded manner by CKD stage.
Conclusions: Risks of subsequent hospitalization, including admission resulting in death, among patients with kidney disease were substantially increased in a large statewide population. Patients with kidney disease should be a focus of efforts to reduce hospitalizations and mortality.