BackgroundMedical billing and coding are critical components of residency programs since they determine the revenues and vitality of residencies. It has been suggested that residents are less likely to bill higher evaluation and management (E/M) codes compared with attending physicians. The purpose of this study is to assess the variation in billing patterns between residents and attending physicians, considering provider, patient, and visit characteristics.MethodA retrospective cohort study of all established outpatient visits at a family medicine residency clinic over a 5-year period was performed. We employed the logistic regression methodology to identify residents’ and attending physicians’ variations in coding E/M service levels. We also employed Poisson regression to test the sensitivity of our result.ResultsBetween January 5, 2009 and September 25, 2015, 98,601 visits to 116 residents and 18 attending physicians were reviewed. After adjusting for provider, patient, and visit characteristics, residents billed higher E/M codes less often compared with attending physicians for comparable visits. In comparison with attending physicians, the odds ratios for billing higher E/M codes were 0.58 (p = 0.01), 0.56 (p = 0.01), and 0.63 (p = 0.01) for the third, second, and first years of postgraduate training, respectively. In addition to the main factors of patient age, medical conditions, and number of addressed problems, the gender of the provider was also implicated in the billing variations.ConclusionResidents are less likely to bill higher E/M codes than attending physicians are for similar visits. While these variations are known to contribute to lost revenues, further studies are required to explore their effect on patient care in relation to attendings’ direct involvement in higher E/M-coded versus their indirect involvement in lower E/M-coded visits.