Beyond the golden hour: prehospital delay and in-hospital stroke mortality in an elderly cohort reaching a national emergency-medicine center in Kazakhstan

Stroke remains a leading cause of death and disability worldwide, and the “time-is-brain” paradigm frames prehospital speed as the principal modifiable determinant of outcome. Data from Kazakhstan, a country of about 2.72 million km 2 with a population density of roughly 7 inhabitants per km 2 , are scarce, and it is unclear whether time metrics retain their prognostic role where most patients present outside reperfusion windows. We retrospectively analyzed 149 elderly patients (aged 57–75 years) with neuroimaging-confirmed acute stroke admitted to the National Coordination Center for Emergency Medicine (Astana, Kazakhstan) in 2025–2026; transient ischaemic attacks were excluded. Prehospital time, transport mode, reperfusion, neurological severity (NIHSS), comorbidity and in-hospital mortality were assessed using Spearman correlation, ANOVA and logistic regression. Ischemic stroke predominated (122/149; 81.9%). The median onset-to-door time was 9.5 h (IQR 3.0–48.3 h; range 25 min to 264 h), and only 36.5% arrived within the 4.5-h thrombolysis window; delays were longest in hemorrhagic stroke (median 24.8 h). Reperfusion therapy reached 24.6% of ischemic-stroke patients (thrombolysis 9.0%, thrombectomy 15.6%), whereas 67.2% were managed conservatively. Despite these delays, in-hospital mortality was low (6.8%). Onset-to-door time predicted neither in-hospital death (adjusted OR 1.00 per hour; p = 0.99) nor discharge functional status (modified Rankin Scale r = −0.013, p = 0.87); in a pre-specified Firth penalized model the predictors of death were baseline NIHSS (penalized OR 1.22 per point; 95% profile-likelihood CI 1.12–1.37; p < 0.001) and cardiac arrhythmia (penalized OR 5.51; 95% CI 1.07–29.71; p = 0.041). The crude mortality contrast (29.4% vs. 3.8%) is descriptive and hypothesis-generating. Baseline NIHSS was inversely correlated with onset-to-door time ( r = −0.193; p = 0.019), indicating that later-presenting patients were less severely affected. These findings apply specifically to elderly patients who survived to reach the center and are not causal: the severity–time and time–mortality associations reflect who reaches the center. Within this hospital-admitted cohort, in-hospital mortality appears governed by neurological severity and cardiac comorbidity rather than transport time, supporting risk-based triage, arrhythmia detection, regionalized routing and air-ambulance capacity alongside efforts to compress door times.