Trends, patterns and outcomes in hospitalisations for Alzheimer's disease: a 14-year retrospective analysis in Southwest China (2011-2024).

OBJECTIVES: To analyse temporal trends, clinical outcomes and healthcare costs of Alzheimer's disease (AD)-related hospitalisations in Southwest China over 14 years and identify modifiable risk factors for adverse outcomes. DESIGN: Retrospective time-series analysis of electronic health records. SETTING: A tertiary referral centre in Southwest China (covering Sichuan, Guizhou, Yunnan and Xinjiang provinces); single-centre study. PARTICIPANTS: 4970 hospitalisations of patients with a confirmed diagnosis of AD (ICD-10: G30) discharged between January 2011 and December 2024, with a mean age of 77.8±0.14 years, and 51.7% were female. INTERVENTIONS: Not applicable (observational study). PRIMARY AND SECONDARY OUTCOME MEASURES: Primary outcomes were poor hospital outcomes, defined as prolonged length of stay (>14 days), frequent readmissions (≥4 admissions over the study period) and in-hospital mortality. Secondary outcomes included annual trends in hospitalisation rates, total costs and prevalence of comorbidities. RESULTS: Comorbidities (64.75% of admissions, mainly infections and fractures) and admission to non-neurology departments independently increased risks of prolonged stay (OR=5.248, 95% CI 4.287 to 6.425; p<0.001), frequent readmissions (OR=1.735, 95% CI 1.236 to 2.435; p=0.001) and in-hospital mortality (OR=2.176, 95% CI 1.506 to 3.146; p<0.001). Hospitalisations rose annually by 7.65% (95% CI 3.00% to 12.31%; p=0.001) and total costs by 8.32% (p<0.001), yet the composite poor-outcome rate declined from 70% to 50% after 2018 (annual percent change, -6.32%, p=0.004), coinciding with fewer comorbidity-driven admissions and increased neurology department care. CONCLUSIONS: Comorbidities and non-neurology admissions were associated with increased risks of adverse outcomes in this cohort. The post-2018 temporal improvements coincided with a decline in comorbidity-driven admissions and increased neurology department use. Further prospective studies are needed to confirm these associations. TRIAL REGISTRATION: Not applicable (observational study; not registered).