Western Journal of Emergency Medicine: Integrating Emergency Care with Population Health

Objectives: To validate and expand upon prior data suggesting that visits shared with a midlevel or resident influence EP behavior as measured by resource use and propensity to admit a patient.

Background: Variability exists in Emergency Physicians’ (EP) resource utilization as measured by ordering practices, propensity to admit patients, and whether a visit is shared with a resident or midlevel provider (nurse practitioner or physician assistant).

Methods: This is a retrospective study of routinely gathered operational data from two community, suburban hospitals within an academic emergency network. We analyzed 34 EPs with 141,433 patient visits from July 1, 2016 to June 30, 2019. We collected individual EP data on advanced imaging (CT, US, MRI), admission rates, and whether a visit was shared with a midlevel or resident for each patient encounter. To investigate whether there might be distinct groups of practice patterns relating these resources, we used a Gaussian Mixture Model (GMM), a classification method used to determine the likelihood of distinct subgroups within a larger population. The total number of groups and covariance structure were determined by Bayesian Information Criteria.

Results: Our GMM revealed three distinct groups of physicians based on their ordering practices. The largest group is characterized by a homogenous pattern of neither high or low resource utilization (n=19, 58% female, median years’ experience: 9 [IQR 2-16]; rates of Advanced Imaging: 44%, Admission: 21%, Midlevel/Resident staffing 35% with a modest group of low-resource users (n=10, 0% female, median years’ experience: 7 [IQR 5-11]; rates of Advanced Imaging: 31%, Admission: 17%, Midlevel/Resident staffing 32%), and far fewer members of a high-resource use group (n=5, 20% female, median years’ experience: 15 [IQR 5-16]; rates of Advanced Imaging: 49%, Admission: 22%, Midlevel/Resident staffing 35%) [Figure 1]. This variation suggests that use of advanced imaging and propensity to admit may be influenced by whether a patient visit is shared with a midlevel or resident provider.

Conclusion: At two community EDs, three distinct subgroups of EP ordering practices exist based on advanced imaging use, propensity to admit a patient, and whether a visit was shared with a midlevel or resident. This data validates prior work showing that resource utilization and admission rates are related, while demonstrating that more nuanced patterns of EP ordering practices exist based on whether a visit is shared with a midlevel or resident provider. Further investigation is needed to understand the impact of EP characteristics and behavior on throughput and quality of care.