Task-Level Variability in a Stepwise Surgical Training Framework: A Retrospective Observational Study Using the Step Ladder System

Background Stepwise competency frameworks aim to clarify clerkship expectations by organizing tasks across domains and hierarchical steps. In surgical clerkships, however, actual attainment is shaped by local workflow and case mix, making alignment between intended task placement and observed attainment uncertain. Routinely collected clerkship logs provide a practical means of describing task-level attainment patterns, including tasks not observed during a defined period. Methods We retrospectively analyzed de-identified Step Ladder System (SLS) logs from 24 sixth-year medical students who completed a four-week surgical clerkship in 2025. The SLS comprises 159 tasks across five domains (General Surgery [GES], Hepato-Biliary-Pancreatic Surgery [HBP], Upper Gastrointestinal Surgery [UGI], Lower Gastrointestinal Surgery [LGI], and Pediatric Surgery [PES]) and three hierarchical Steps. Tasks were classified as non-zero attainment (achieved by at least one student; n = 124) or zero attainment (n = 35). For non-zero tasks, attainment rates were summarized by domain, Step, and domain by Step categories. Because attainment was recorded as a binary outcome, the coefficient of variation (CV = SD/mean) was used solely as a visualization-friendly transformation of attainment rather than as an independent measure of variability. Exploratory comparisons were performed using Kruskal–Wallis tests with Dunn’s post hoc tests (Bonferroni adjustment). Results Thirty-five tasks (22.0%) demonstrated zero attainment, occurring predominantly in higher Steps, with only one such task in Step 1. Among tasks with non-zero attainment, GES exhibited lower CV-derived values, whereas HBP, UGI, LGI, and PES showed higher values. Stepwise summaries indicated lower CV-derived values in Step 1 compared with Steps 2 and 3, while separation between Steps 2 and 3 was less distinct. In the domain by Step matrix, PES Step 3 comprised exclusively zero-attainment tasks and was labeled “CV not calculable.” Conclusions Task attainment patterns varied across domains and hierarchical Steps, with clearer separation between Step 1 and higher Steps than between Steps 2 and 3. Routinely collected clerkship logs revealed uneven task exposure and a substantial subset of tasks not observed during the clerkship. These findings are descriptive and should not be interpreted as evidence of latent task difficulty, differential access, or structural deficiencies of the SLS. Trial registration: Not applicable.