The volume of emergency department (ED) visits has been steadily increasing over the past decade. Also, in parallel, the number of board certified emergency medicine (EM) physicians has been growing. One of the most important decisions made in a patient’s course in the ED is the development of a disposition plan by the physician provider. Many factors influence the development of an appropriate plan of care, such as, patient medical and social needs. Literature from other specialties such as internal medicine and general surgery have shows us that practice patterns and outcomes differ at times by non-patient factors, more specifically, by physician sex bias. Studies in EM have shown non-patient variables that potentially have an influence such as ED crowding and years of physician experience. No paper to date has looked at sex of EM providers in the United States as a function of adult admission rates. The current study seeks to investigate admission rates of adult patients who are treated by male versus female emergency physicians, in order to identify whether or not a practice pattern bias exists. This was a multi-center retrospective cross-sectional study of one health care system in the Midwest. All patient encounters by EM-trained physicians in the emergency department of 4 hospitals over the academic year, July 1, 2016 to June 30, 2017 were analyzed. The primary outcome was admission rate of each physician with admission rate defined as the percentage of encounters that resulted in admission to the hospital as a portion of all encounters. Admission defined by patients who were admitted to observation, inpatient or emergent surgery from the ED. Any physician with less than 100 patient encounters was excluded. Secondary analysis assessed physician metrics and benchmarks with potential confounding effects: patient acuity, average length of stay, return visits, patient age and years of practice by physician, using independent samples t-tests. 110,230 patient encounters were analyzed, treated by 80 emergency physicians; 49 male, 31 female. The average admission rates were male 28%, female 29%, t = .504 with equal variances assumed. Acuity of patients seen by providers: male .741, female .763, t = -2.862. Average length of stay in minutes: male 264, female 292, t = -2.158. Average returns visits per 1,000: male 8.24, female 8.37, t = -.222. Patient age: male 49.74, female 51.12, t = -1.439. Years of practice: male 20.24, female 15.77, t = 2.097. In this analytic retrospective observational study, we found that the difference of admission rates of male and female EM providers was not statistically significant. We did find significance between the average length of stay, patient acuity seen and years of practice. The etiology of these significances can be postulated to be interrelated. Further studies are needed with a greater sample size of physicians to potentially find sex bias practice patterns and differences.Table 1Descriptive statistics with independent samples t-testsAdmission RatePatient Acuity (0-1)Average LOS (minutes)Returns per 1,000Patient AgeYears of PracticeMale28%0.7412648.2449.7420.24Female29%0.7632928.3751.1215.77T value0.504-2.862-2.158-0.222-1.4392.097 Open table in a new tab
Read full abstract