Factors associated with 30-day unplanned readmission after first-ever ischemic stroke: a retrospective cohort study in China

Unplanned readmission is an important post-stroke outcome, but clinically detailed data from Chinese cohorts remain limited. We evaluated the frequency of 30-day unplanned readmission after first-ever ischemic stroke and examined clinical factors associated with this outcome. We conducted a single-center retrospective cohort study of consecutive adults with first-ever acute ischemic stroke admitted from January 2022 to December 2024. After exclusions, 962 patients were included. The primary outcome was 30-day unplanned readmission, and cumulative 90-day readmission was analyzed as a secondary outcome. Clinical variables were extracted from electronic medical records. Two complementary multivariable logistic-regression models separately evaluated diabetes mellitus and HbA1c while adjusting for age, sex, admission NIHSS, discharge mRS ≥ 3, atrial fibrillation, and eGFR; sensitivity analyses excluded NIHSS ≤ 3 and accounted for death as a competing event. Among 962 patients (median age, 67 years; 58.6% male), 98 (10.2%) experienced a first unplanned readmission within 30 days, and an additional 55 were first readmitted during days 31–90, yielding 153 cumulative readmissions (15.9%) by day 90. In the diabetes model, older age (per 10 years: aOR 1.18, 95% CI 1.03–1.36), higher NIHSS (per point: aOR 1.07, 95% CI 1.02–1.14), discharge mRS ≥ 3 (aOR 1.61, 95% CI 1.09–2.51), diabetes mellitus (aOR 1.49, 95% CI 1.01–2.21), and atrial fibrillation (aOR 1.85, 95% CI 1.10–3.11) were associated with 30-day readmission; in the complementary glycemic model, HbA1c was also associated with the outcome (per 1%: aOR 1.22, 95% CI 1.02–1.46). Associations were generally similar in direction for the cumulative 90-day endpoint and in prespecified sensitivity analyses. Approximately one in ten patients experienced 30-day unplanned readmission after first-ever ischemic stroke. Greater stroke severity, discharge functional dependence, diabetes mellitus, atrial fibrillation, and higher HbA1c were associated with early readmission and may identify clinical features warranting closer post-discharge assessment; external replication in multicenter cohorts is needed before these findings are used for formal prediction or intervention decisions.