Training and assessment in flexible sigmoidoscopy: using a novel direct observation of procedural skills (DOPS) assessment tool.

BACKGROUND AND AIMS Data supporting milestone development during flexible sigmoidoscopy (FS) training are lacking. We aimed to present validity evidence for our formative direct observation of procedural skills (DOPS) assessment in FS, and use DOPS to establish competency benchmarks and define learning curves for a national training cohort. METHODS This prospective UK-wide (211 centres) study included all FS formative DOPS assessments submitted to the national e-portfolio. Reliability was estimated from generalisability theory analysis. Item and global DOPS scores were correlated with lifetime procedure count to study learning curves, with competency benchmarks defined using contrasting groups analysis. Multivariable binary logistic regression was performed to identify independent predictors of DOPS competence. RESULTS This analysis included 3,616 DOPS submitted for 468 trainees. From generalisability analysis, sources of overall competency score variance included: trainee ability (27%), assessor stringency (15%), assessor subjectivity attributable to the trainee (18%) and case-to-case variation (40%), which enabled the modelling of reliability estimates. The competency benchmark (mean DOPS score: 3.84) was achieved after 150-174 procedures. Across the cohort, competency development occurred in the order of: pre-procedural (50-74), non-technical (75-149), technical (125-174) and post-procedural (175-199) skills. Lifetime procedural count (p<0.001), case difficulty (p