Background: During emergency medicine (EM) training, residents are exposed to a wide spectrum of patient complaints. We sought to determine how resident clinical experience changes based on training level in relation to the patient acuity levels, chief complaints, and dispositions. Methods:We performed a retrospective chart review of patients seen at a safety-net, academic hospital in Los Angeles from July 1, 2015, to June 30, 2016. Resident postgraduate year (PGY) level and specialty, patient acuity (based on the Emergency Severity Index), chief complaint (based on one of 30 categories), and disposition were abstracted. Our primary objective was to examine the progression of EM resident experience throughout the course of training. As a secondary objective, we compared the cases seen by EM and off-service PGY-1s.Results: A total of 49,535 visits were examined, and of these, 32,870 (66.4%) were in the adult ED (AED) and 16,665 (33.6%) were in the pediatric ED (PED). The median acuity level was 3, and 27.4% of AED patients and 7.3% of PED patients were admitted. Data from 126 residents were analyzed. This included 94 PGY-1 residents (16 EM and 78 off-service), 16 PGY-2 EM, and 16 PGY-3 EM residents. Residents of different training levels evaluated different types of patients. Senior EM residents were more likely to care for higher-acuity patients than junior EM residents. EM PGY-3s saw higher percentages of acuity level 1 and 2 patients (2.3 and 37.8%, respectively, of their total patients) than EM PGY-1s (0.3 and 18.7%, respectively). Conversely, EM PGY-1s saw higher percentages of acuity level 4 and 5 patients (27.9 and 1.6%, respectively) compared to EM PGY-3s (10.7 and 0.7%, respectively). There was a significant linear trend for increasing acuity with training year among EM residents (p < 0.001). EM PGY-1s saw more patients than off-service PGY-1s with slightly higher acuities and admission rates. Conclusion:The clinical experience of EM residents varies based on their level of training. EM residents show a progression throughout residency and are more likely to encounter higher volumes of patients with higher acuity as they progress in their training. When designing EM residency curriculums, this is a model of an EM residency program.

Methods: Six hundred clinical records of student visits to the ED in 6 academic years from 2009-10 to 2014-15 were randomly selected for chart review by 2 independent reviewers to identify visits with alcohol intoxication. Results were then compared with ICD-9 diagnostic codes indicating alcohol intoxication (30500, 30502, and 3030) in the hospital discharge database. Sensitivity, specificity, positive predictive and negative predictive values were calculated to evaluate the validity of diagnostic codes using the chart review as the "gold standard."Results: Over the study period, there were 9616 student visits to ED. Overall prevalence of alcohol intoxication was 10.4% based on ICD-9 diagnostic codes. Of the review sample of 600 records, the use of ICD-9 diagnostic codes in patient medical records identified 64 visits (10.6%) with alcohol intoxication, while the chart review identified 96 visits (16%) with alcohol intoxication. Sensitivity was 65%, indicating that ICD-9 diagnostic codes only captured 65% of the total ED visits with alcohol intoxication in the review sample. The specificity, positive predictive value, negative predictive value, and accuracy were 99%, 94%, 94%, and 94%, respectively (Table 1). There were 41 visits which involved both alcohol intoxication and injury or trauma, of which alcohol intoxication diagnostic codes were provided in only 18 visits (44%).Conclusions: Although code-based recording of student ED visits due to alcohol intoxication had a high level of accuracy, over one third of ED visits due to alcohol intoxication were not captured by diagnostic codes. In particular, when the visit also involved injury or trauma, only less than half of visits with alcohol intoxication were given a diagnostic code for this condition. Code-based measurement appears to severely underestimate the true burden of alcohol intoxication in the ED associated with student visits. There is a strong need to improve emergency physician coding of alcohol intoxication so that ED electronic medical records can serve as a reliable data source to evaluate the burden of alcohol intoxication in the hospital emergency setting.