The effect of emergency department expansion on emergency department overcrowding

Objectives: To examine the effects of emergency department (ED) expansion on ambulance diversion at an urban, academic Level I trauma center. Methods: This was a pre-post study performed using administrative data from the ED and hospital electronic information systems. On April 19, 2005, the adult ED expanded from 28 to 53 licensed beds. Data from a five-month pre-expansion period (November 1, 2004, to March 1, 2005) and a five-month postexpansion period (June 1, 2005, to October 31, 2005) were included for this analysis. ED and waiting room statistics as well as diversion status were obtained. Total ED length of stay (LOS) was defined as the time from patient registration to the time leaving the ED. Admission hold LOS was defined as the time from the inpatient bed request to the time leaving the ED for admitted patients. Mean differences (95% confidence interval [CI]) in total time spent on ambulance diversion per month, diversion episodes per month, and duration per diversion episode were calculated. An accelerated failure time model was performed to test if ED expansion was associated with a reduction in ambulance diversion while adjusting for potential confounders. Results: From pre-expansion to postexpansion, daily patient volume increased but ED occupancy decreased. There was no significant change in the time spent on ambulance diversion per month (mean difference, 10.9 hours; 95% CI = -74.0 to 95.8), ambulance diversion episodes per month (two episodes per month; 95% CI = -4.2 to 8.2), and duration of ambulance diversion per episode (0.3 hours; 95% Cl = -4.0 to 3.5). Mean (+/- SD) total LOS increased from 4.6 (+/- 1.9) to 5.6 (+/- 2.3) hours, and mean (+/- SD) admission hold LOS also increased from 3.0 (+/- 0.2) to 4.1 (+/- 0.2) hours. The proportion of patients who left without being seen was 3.5% and 2.7% (p = 0.06) in the pre-expansion and postexpansion periods, respectively. In the accelerated failure time model, ED expansion did not affect the time to the next ambulance diversion episode. Conclusions: An increase in ED bed capacity did not affect ambulance diversion. Instead, total and admission hold LOS increased. As a result, ED expansion appears to be an insufficient solution to improve diversion without addressing other bottlenecks in the hospital.