Background: Current best-practice guidelines and national quality improvement efforts emphasize reduction in door-to-needle (DTN) time. Stroke center certifying bodies depend on hospital-reported DTN to appraise compliance and award designations. However, the validation of hospital reported arrival times against ambulance arrival times has not been previously reported.
Objective: To determine differences in Emergency Medical Services (EMS) reported arrival time of stroke patients versus hospital documented arrival times at 10 primary and comprehensive stroke centers in Chicago.
Methods: We performed a retrospective analysis of stroke admissions using the Get With The Guidelines (GWTG) Stroke registries of 10 hospitals in Chicago between July 1 and December 31, 2013. We matched ambulance run numbers from EMS records to those in the GWTG-Stroke EMS special initiative tab. GWTG-Stroke data included arrival time, age, sex, race, initial NIHSS score, and whether tPA was administered. EMS data include hospital arrival time in addition to age, sex, hospital name, and clinical impression. We compared GWTG-Stroke and EMS reported arrival times overall and whether any clinical factors were associated with time discrepancy.
Results: There were 308 matched EMS to GWTG-Stroke records over the study period and of those, 51 (16.6%) received tPA. Overall, EMS documented arrival time was a median 5 minutes earlier than hospital documented arrival time (range -14 to +35m). There was no significant difference in time discrepancy by tPA administration, stroke type, NIHSS score, age, sex, race, insurance status, or hospital. Median DTN based on EMS data was 88m versus 82m for hospital-reported data (p>0.05). The proportion with DTN times <60 minutes was 22% based on EMS versus 29% based on hospital arrival times.
Conclusions: EMS reported arrival time is on average 5 minutes earlier than hospital documented arrival time at 10 Chicago stroke centers. These data suggest that current hospital reported DTN times may be lower than DTN times had EMS arrival been recorded as door time. Further quality improvement should ensure using EMS source data for more accurate documentation of arrival times.