Pre-stroke dementia and in-hospital outcomes in the Chinese Stroke Center Alliance

BACKGROUND: Little is known about the impact of prevalent dementia on in-hospital outcomes of patients with incident stroke in China. Using data from the Chinese Stroke Center Alliance (CSCA), we aim to quantify the prevalence of pre-stroke dementia and whether this group is at higher risk of adverse in-hospital outcomes compared to those without pre-stroke dementia. METHODS: We used multivariable logistic regression models to assess the associations between pre-stroke dementia and ambulation by day 2, in-hospital mortality, in-hospital complications, and being discharged home. Covariates included age, sex, comorbidities [dyslipidemia, atrial fibrillation, peripheral vascular disease (PVD), smoking, and alcohol use], medication history (antiplatelet drugs or lipid-lowering drugs), stroke severity [measured by the National Institute of Health Stroke Scale (NIHSS)], administration of intravenous tissue plasminogen activator (IV tPA) within 4.5 hours of stroke onset, and receipt of deep vein thrombosis (DVT) prophylaxis if indicated. RESULTS: In the final analytic sample of 559,070 ischemic stroke patients with no prior stroke history enrolled across 1,476 hospitals, those with pre-stroke dementia (n=1,511; 0.3%) were older and more likely to be female. Despite having received similar treatment, patients with pre-stroke dementia had lower odds of ambulating by day 2 [odds ratio (OR) =0.69; 95% confidence interval (CI): 0.62–0.78], higher odds of in-hospital mortality (OR =2.01; 95% CI: 1.35–2.99) or complications (OR =2.17; 95% CI: 1.93–2.44), and lower odds of being discharged home compared to those without pre-stroke dementia (OR =0.71; 95% CI: 0.62–0.83). CONCLUSIONS: Worse in-hospital outcomes among patients with pre-stroke dementia may be explained by pre-existing cognitive impairment that limited their ability to advocate for care needs. Further research is needed to determine whether a different care pathway or additional attention from clinicians is necessary for patients with pre-stroke dementia.