ABSTRACT
OBJECTIVES: Burnout afflicts emergency physicians (EPs) to a significant degree. The impact of burnout spans from decreased clinical efficiency to increased medical errors to heightened risk of physician suicide. This large-scale study captures responses from emergency medicine (EM) residents regarding two burnout items and examines the correlation between in-training examination (ITE) scores and burnout risk as well as that between residency year and burnout risk. METHODS: This was a prospective, mixed-methods, cross-sectional cohort study. All residents in U.S. categorical EM residents who took the 2019 ITE were included. At the end of the ITE, residents were invited to complete a voluntary survey that included two items from the Maslach Burnout Inventory (MBI) that have been found to be strongly indicative of burnout: one about self-perception of being burned out and one about feelings of callousness. Responses were on a 7-level Likert scale (1-7), ranging from very low frequency (1) to very high frequency (7). Measurements included the number of residents in each year-level of training (EM1-EM4), the MBI item ratings, and the ABEM ITE score. Performance, as measured by the scaled, equated score, was compared to the MBI item responses. A corrected Spearman's correlation coefficient (ρ) was used to compare continuous data (score) against a discrete ordinal variable (MBI Likert response). RESULTS: There were 2,501 EM1 residents, 2,389 EM2 residents, 2,206 EM3 residents, and 616 EM4 residents in the study group. There were 7,206 (93.4%) physicians who completed the first MBI question about burnout; 7,172 (93%) completed the second MBI question about callousness. There was no statistically significant association between the burnout item response and ITE performance (ρ = -0.03; p = 0.015). There was a positive, statistically significant association between the callousness item response and higher ITE performance (ρ = 0.07; p < 0.001). There was a statistically significant association between the response to the burnout item and training level (ρ = 0.07; p <0.001). There was also a statistically significant association between the response to the callousness item and training level (ρ = 0.15; p < 0.001). The overall prevalence of burnout risk in various training levels were EM1, 28.2%; EM2, 39%; EM3, 41.1%; and EM4, 43.3%. CONCLUSIONS: Our study found no significant correlation between ITE score and burnout risk. There was a weakly positive correlation between ITE scores and callousness. Based on our study results, ITE scores may not be useful in prognosticating burnout risk for EM residents and, interestingly, higher ITE scores correlated to stronger feelings of callousness. Our study indicates that EM residents at higher levels of training reported stronger self-perceptions of burnout and callousness. Further investigation into why residents at higher levels of training may experience greater burnout risk is warranted.
ABSTRACT
BACKGROUND: Health care systems often implement changes within the electronic health record (EHR) to improve patient safety and reduce medical errors. OBJECTIVE: To compare the proportion of emergency department (ED) encounters with laboratory tests resulting subsequent to patient discharge before and after a clinical decision support was implemented. METHODS: In 2015, our institution added an EHR dialogue when placing ED discharge orders, requiring providers to declare whether all laboratory results had been reviewed. To determine the effectiveness of this initiative, we searched the EHR to identify the proportion of ED encounters with laboratory tests resulting after discharge in pre- (January to June 2015) and post-intervention (January to June 2016) periods. RESULTS: There were 67,287 discharged patients during the study periods. In the pre- and post-intervention periods, respectively, 6.9% (95% confidence interval [CI] 6.7-7.2%) and 7.9% (95% CI 7.6-8.2%) of encounters had laboratory tests resulting after discharge, with an absolute difference of 0.9% (95% CI 0.5-1.3%). Of these patients with laboratory tests resulting after ED discharge, in 92% the provider inaccurately marked "yes" or "not applicable" to the EHR dialogue prompt. CONCLUSIONS: This workflow intervention was associated with an increase in the proportion of laboratory tests resulting after ED discharge; inaccurate answers to the EHR dialogue were pervasive. EHR workflow interventions do not always accomplish their intended goals, and their implementation should be considered thoughtfully.
