Predictors for readmission of acute ischemic stroke in Taiwan

Background/PurposeHospital readmission rates are usually higher for stroke than for other chronic conditions. To prevent readmission, effective clinical services and accurate estimates of the absolute readmission rates are required. This study examined the patterns of care received by patients with ischemic stroke, estimated stroke readmission rates, and identified predictors related to readmission in Taiwan.MethodsA retrospective analysis of the claims database of the Bureau of National Health Insurance in 2004–2007 was performed. This study included ischemic stroke patients who survived hospitalization and whose initial admission occurred in 2006. Time-dependent Cox regression models were developed separately to identify predictors of readmission within 1 month, 6 months, and 1 year after index discharge.ResultsWe identified 1194 patients from the data set. At the initial stroke, the care provided was almost fully concordant with evidence-based practice guidelines, and the prevalence of antiplatelet therapy was 87.8%. The percentage of patients regularly taking antiplatelet agents within 1 month, 6 months, and 1 year after the index discharge was 65.0%, 18.8%, and 8.0%, respectively. The stroke readmission rates for survivors at 1 month, 6 months, and 1 year after the index discharge were 9.9%, 23.0%, and 30.7%, respectively. Older age, diabetes, longer length of stay for the index admission, and continuous use of antiplatelet agents less than 9 months after the index discharge were all predictors of readmission for acute ischemic stroke.ConclusionStroke readmissions are not related to receipt of less than optimum or below standard health care during index admission in Taiwan. Additional stroke readmissions in Taiwan might be avoided if more patients used antiplatelet agents for a longer period.