There is variation in emergency physician (EP) resource consumption, as indicated by ordering habits, consultation rate, and proclivity to admit patients. For a study, researchers sought to confirm and build on earlier data demonstrating that resource use, as assessed by EP ordering patterns, is positively associated with admission rates. It was a retrospective analysis of routinely collected operational data from an urban academic tertiary care hospital’s emergency department. Individual EP data on advanced imaging, consultation, and admission rates were recorded for each patient visit. They employed a Gaussian mixture model, a classification approach used to evaluate the likelihood of unique subgroups within a larger population, to examine whether there may be separate groups of practice patterns related to these three resources. 

Based on their ordering habits, the Gaussian mixture model identified three unique groups of EPs. The largest group was distinguished by a consistent pattern of neither high nor low resource utilization (n = 37, 27% female, median years’ experience: 6 [interquartile ratio IQR 3–18]; rates of advanced imaging, 38.9%; consultation, 45.1%; and admission 39.3%), with a smaller group of low-resource users (n = 15, 60% female, median years’ experience: 6 [IQR 5–14]; rates of advanced imaging, 37%; and far fewer members of a high-resource use group (n = 6, 0% female, median years’ experience: 6 [IQR 4–16]; rates of advanced imaging, 42.2%; consultation, 45.8%; and admission 40.6%). The variance implied that not “all testers are admitters,” but that EPs’ practices vary widely. 

There were three unique subgroups of EP ordering patterns at the academic tertiary institution, based on consultation rates, advanced imaging utilization, and tendency to admit a patient. The findings supported earlier research that found a link between resource use and admission rates, while also suggesting that more sophisticated patterns of EP ordering procedures exist. More research was needed to understand the influence of EP features and behavior on throughput and care quality.