Pilot Study: Concordance of Disposition for Hypothetical Medical Patients in the Emergency Department

Abstract

Emergency physicians (EPs) make dispositions for every patient in the emergency department (ED) and often require agreement from inpatient services to admit medical patients to the hospital. Sometimes disagreements arise. The authors sought to determine in their institution disposition concordance between EPs and admitting medical services of hypothetical ED patients. A second objective was to describe additional information requested by each service. Within a two-week period, physicians from emergency medicine (EM), family practice (FP), and internal medicine (IM) voluntarily completed anonymous questionnaires asking them to disposition hypothetical ED patients. Eleven case scenarios were followed by three disposition choices (admit, discharge, or cannot tell) from given information (followed by a free-text area). A total of 105 questionnaires were returned: 42 from EM (12 attending physicians/30 residents), 33 from FP (14 attending physicians/19 residents), and 30 from IM (4 attending physicians/26 residents). Admission rates were statistically different for EM (68% attending physicians/65% residents/66% total) when compared with FP (42% attending physicians/54% residents/49% total) (p < 0.01) and IM (36% attending physicians/53% residents/51% total) (p < 0.01). Discharge rates were also statistically different for EM (8% attending physicians/19% residents/16% total) versus FP (29% attending physicians/29% residents/28% total) (p < 0.01) and IM (27% attending physicians/28% residents/28% total) (p < 0.01). "Cannot tell" rates were not significantly different between groups (p > 0.05) for EM (23% attending physicians/16% residents/18% total) versus FP (25% attending physicians/14% residents/19% total) and IM (36% attending physicians/16% residents/18% total). The proportion of additional tests requested in the "cannot tell" disposition by attending EPs was about half that of the other groups. Dispositions between FP and IM were not different at any level (p > 0.05). In the authors' institution, when presented with identical hypothetical ED patients, EPs would admit more, discharge less, and ask for fewer additional tests than FP or IM physicians. The FP and IM physicians surveyed disposition patients in a similar manner.