Do emergency physicians admit more or fewer patients on busy days?

Background: Although the issues concerning the impact of emergency department (ED) overcrowding have been the subject of much recent concern, there are few data regarding the effect of ED census on emergency physician behavior with respect to the decision to admit patients. Admission rates might either increase or decrease on busy days, when the system and the physician are under stress.

Study objective: The purpose of this study was to determine if ED physicians change their admitting behavior depending on ED census.

Methods: This was a retrospective review of 3 months' data (92 consecutive days, July 9-October 9, 2006) in a community ED with an annual census of approximately 70,000 patients and an emergency medicine residency program. We defined each of the 92 days to be either "busy" (> 180 patients seen), "slow" (< 147 patients seen) or "medium" (147-180 patients seen). We then compared the rates of admission to the hospital on the "busy," "medium," and "slow" days. We also compared each attending physician's personal rates of admission on slow days to his or her rate of admission on medium or busy days. ED staffing was constant throughout the study period. All comparisons were with chi-squared.

Results: There were 14,969 patients seen in the ED during the 92 study days. On "busy" days, 20.1% of the 3400 patients were admitted to the hospital; on "medium" days, 20.6% of the 9057 patients were admitted; on "slow" days, 19.7% of the 2512 patients were admitted. There was no significant association between the level of patient volume in the ED and rate of admission (p = 0.55). When comparing each of 14 attending physicians to him- or herself, there was no significant association found between rate of admission and ED census (all p values > 0.3). All three categories of days, "busy," "medium," and "slow" did not differ in terms of acuity as judged by triage level distribution.

Conclusion: The likelihood of a patient's admission vs. discharge is not affected by ED patient volume. Furthermore, we found no evidence that an individual physician's admitting behavior was associated with ED patient volume.