Abstract
Objective
Dismissal from pediatric residency is uncommon but high-stakes and labor-intensive. Performance deficits underlying the decision to dismiss are not well understood.Methods
We examined Milestone scores for categorical pediatric residents who started training between 2013 and 2021 stratified by resident status: graduated, dismissed, or withdrawn. We fit linear mixed effects models to mean Milestone ratings and compared estimated marginal means in each milestone domain between groups.Results
We obtained Milestones for 32,426 residents: 28,345 graduated on time, 99 withdrew, and 29 were dismissed. Dismissals were evenly distributed across PGY1 (34%), PGY2 (28%), and PGY3 (38%), while most instances of withdrawal occurred during PGY1 (44%) then decreased over time (PGY2, 37%; PGY3, 18%). In aggregate, dismissed residents had statistically significantly lower mean Milestone ratings than residents who graduated in all domains and at all time points. Residents dismissed in PGY2 or PGY3 had lower Milestone scores than those who graduated starting in PGY1 spring.Conclusions
Residents who were dismissed at any point in training had performance deficits present in most Milestone domains by PGY1 spring. Future work is necessary to better characterize the processes leading up to these dismissals and barriers to successful remediation of pediatric residents.