Background <p>Door-to-needle (DTN) performance in acute ischemic stroke is commonly evaluated using aggregate treatment times or isolated predictors. These approaches may not fully capture how delays occur across sequential emergency department processes. Evidence describing the interaction between upstream activation and downstream workflow intervals in secondary-level settings remains limited.</p> Methods <p>We conducted a retrospective cohort study of adults receiving intravenous thrombolysis at a secondary-level hospital. Multivariable logistic regression was used to identify upstream factors associated with delayed DTN (&gt; 60&#xa0;min). Interval-level decomposition and stratified quantile regression analyses were performed to examine where time differences occurred within the in-hospital pathway and to assess results according to prehospital stroke alert status.</p> Results <p>Among 245 patients, 30.60% had DTN times exceeding 60&#xa0;min, although the overall median DTN was 50&#xa0;min (IQR 40–63). Absence of prehospital stroke alert, triage misclassification, and fluctuating neurological presentation were independently associated with delayed treatment. The largest median time differences were observed during coagulation processing, final test-to-needle intervals, and patient or family decision-making. In stratified analyses, final test-to-needle and decision intervals remained prolonged in delayed cases across both alert strata, whereas coagulation delay was significant only in the alert-positive group.</p> Conclusions <p>Delayed DTN was associated with both upstream activation factors and downstream workflow intervals. Combining determinant modeling with interval-level analysis provides a structured approach to identify operational intervals for evaluation in secondary-level stroke systems.</p> Trial registration <p>Not applicable.</p> Graphical Abstract <p></p>