Published (epub 2026-08-10) in SAGE Open Medicine.
Median door-to-needle time got 13.5 minutes slower. That may have been good news, and the reason is worth five minutes of any stroke-systems leader's time.
- In February 2022, a merger in the Central Denmark Region moved a Primary Stroke Center and the region's largest Emergency Department into one hospital, making their catchment areas nearly overlap. No new drug, device, or software.
- Researchers compared 1,950 suspected-stroke patients across matched six-month periods (1,059 before, 891 after), using the Danish Stroke Registry linked to prehospital records.
- Among 914 patients with ischemic stroke, the crude IV thrombolysis rate rose 4.6 percentage points from a 12.2% baseline, though after adjustment, the confidence interval for the overall effect included zero.
- The durable signal was in one subgroup: patients arriving by EMS after an emergency call. There, IVT rose 10.2 points from a 19.8% baseline, and the effect held after full adjustment (+11.1 points, 95% CI 0.1–22.2).
- Median door-to-needle time in that subgroup lengthened from 26.5 to 40.0 minutes (p=0.003), but the decomposition matters more than the median. Patients treated within 45 minutes held roughly steady (29 → 26), while those treated after 45 minutes rose from 3 to 18, and total treated went from 32 to 44. The slower median reflects additional patients being treated who previously weren't, not slower care for those already eligible. NIHSS at treatment and at ~24 hours did not differ between periods.
Two operational lessons. First, catchment geometry, where EMS delivers versus where thrombolysis lives, is a first-order lever on reperfusion access, upstream of any imaging or triage layer. Second, and more uncomfortable: a single workflow KPI can invert the story. A dashboard tracking only median DTN would have scored this change as a failure.
This is a single-region, retrospective before-and-after study. Three-month functional outcomes, mortality, and safety data weren't available, and the merger bundled several concurrent organizational changes the authors couldn't disentangle. Stroke-mimic admissions fell from 26.3% to 17.9%, but the authors explicitly decline to attribute that to better prehospital triage accuracy.
The broader trend: as stroke pathways modernize with telestroke, prehospital triage, and AI decision support, structural redesign is the control condition we rarely measure against, and registry-linked national data keeps making that comparison cheap.
A question for stroke-network leaders: if access improves but your speed metric degrades, does your current scorecard tell you that you succeeded, or that you failed?
#Telestroke #PrimaryStrokeCenter #ComprehensiveStrokeCenter #AcuteStrokeReadyHospitals #EmergencyMedicine #Neurology #StrokeAI
[1] Gude MF, Basse T, Rønde Jensen AS, Væggemose U, Al-Jazi MA, von Weitzel-Mudersbach P, Blauenfeldt RA. Aligning emergency admission and stroke center catchment areas: Association with intravenous thrombolysis rates. SAGE Open Med. 2026;14:20503121261477313. DOI:10.1177/20503121261477313. PMID:42582578. PMCID:PMC13458115. (Epub 2026-08-10; full text verified via PubMed Central, accessed 2026-08-17.)