Objectives: To determine the effect of simultaneous ambulance diversion at multiple emergency departments (EDs) (gridlock) on transport delays for patients with chest pain.
Methods: Retrospective data on consecutive ambulance patients with chest pain and the diversion status of EDs in Toronto were obtained from January 1998 to December 1999. Gridlock was calculated separately for the four city quadrants as the daily duration of episodes where all EDs in the quadrant were simultaneously diverting ambulances. The primary outcome was 90th percentile ambulance transport interval (scene departure to hospital arrival).
Results: Eleven thousand four hundred patients were included (mean age 67 years; female 51%; severity of illness: moderate to life-threatening 89%). Gridlock occurred an average of 1.1 hour/day, and 3,060 patients were transported on days when it occurred. Ninetieth percentile transport interval was 15.5 minutes (95% CI = 15.3 to 15.9) for patients not exposed to gridlock vs. 17.4 minutes (95% CI = 16.8 to 17.8) for patients who were exposed to gridlock. In multivariate analyses, gridlock was associated with both transport and total out-of-hospital interval delays (0.2 min/hour, 95% CI = 0.1 to 0.4 and 0.2 min/hour, 95% CI = 0.04 to 0.4, respectively). Delays were similar regardless of patient severity of illness (p = 0.5). Age (0.8 min/10 years, 95% CI = 0.5 to 1), female gender (1.9 min, 95% CI = 1.3 to 2.6), advanced care paramedics (5.3 min, 95% CI = 4.4 to 6.3), and snowfall (0.8 min/cm, 95% CI = 0.2 to 1.5) were also independently associated with delays.
Conclusions: Ambulance diversion was associated with delays in out-of-hospital ambulance transport for chest pain patients, but only when it resulted in gridlock. The magnitude of the out-of-hospital delay was the same regardless of the patient's severity of illness.