Physician bias during patient selection in the pediatric emergency department

Abstract

Study objectives: To evaluate whether pediatric or emergency medicine residents exhibit a bias when they select patients from triage based on the chief complaint, ie, medical versus surgical in the pediatric emergency department (PED). Design: A retrospective chart review of a convenience sample of consecutive patients, excluding those seen during times when both pediatric and emergency medicine residents were not simultaneously present. Setting: Urban Municipal PED with 25,000 visits annually. Type of participants : Pediatric residents, emergency medicine residents, and medical students rotating through the PED and their supervising attending physicians. Interventions: None. Measurements and main results: Five hundred and ninety-nine charts were included in the study. On the basis of the triage complaint the initial diagnosis was classified as either surgical or medical. Surgical diagnoses were assigned to those patients who required a surgical procedure, involved a surgical subspecialty or were victims of trauma and represented 151 (25.2%) of the patients seen. Medical diagnoses were assigned to the nonsurgical patients and represented 448 (74.8%) of the patients seen. There are roughly three pediatric residents to each emergency resident working in our PED. Of the 367 patients seen by the pediatric residents, 73 (19.9%) had surgical diagnoses, and 294 (80.1%) had medical diagnoses. Of the 158 patients seen by the emergency residents, 59 (37.3%) had surgical diagnoses and 99 (62.7%) had medical diagnoses, x 2 analysis was used to compare categorical variables. The P value was considered significant at <0.05. Discussion: Residents, both pediatric and emergency medicine, were instructed to see patients based upon the severity of the patient illness as judged by the triage nurse unless the patients' illnesses were of equal severity, in which case they were to be seen in the order in which they presented. The null hypothesis was that in the absence of physician bias, both pediatric and emergency medicine residents would see the same proportion of surgical and medical patients. The results showed that a bias exists. Emergency medicine residents saw a greater proportion of surgical patients, and pediatric residents saw a greater proportion of general medical patients. A limitation of this study may be the that the supervising attending physician selected residents to see certain patients to expedite PED flow. Conclusions: Recognizing that bias in the selection of patients seen exists is important in ensuring a balanced education experience.