Methodological recommendations for future clinical reasoning education studies: report effect sizes with confidence intervals, control confounders, use ITT analysis, and determine MCID¶

Strategy · All strategies
Evidence · 6 claims (6 for) · 1 study (1 review), q3 · 1 of 1 report an effect size · 6 claims rest on one study

Description¶

The review recommends that future researchers "include confidence intervals for effect sizes to clarify the precision of their findings alongside p-values", rigorously identify and control confounders (e.g., via multivariable regression or stratification), report adherence and compliance, employ per-protocol and intention-to-treat analyses, use CONSORT standards, perform a priori sample size calculations, and determine the minimal clinically/practically important difference (MCID) as a basis for significance levels and sample size calculation.

Design Implications¶

Context¶

Requirements¶

• Future studies should prospectively estimate sample size and power targeting the smallest educationally important difference

Constraints¶

• The review notes RCTs and meta-analysis have notable limitations in medical education, and a washout period is impossible for learning interventions in crossover designs

Target Learners¶

• undergraduate medical students

Target Learning Goals¶

• clinical reasoning skill
• diagnostic accuracy

Claims¶

• Blended teaching (TBL plus feedback and reflection) outperformed traditional bedside teaching on clinical reasoning performance [+W]
• Immediate feedback combined with reflection improved diagnostic accuracy on training cases versus outpatient clinic attendance or lecture [+W]
• Live case-based discussion outperformed paper cases on clinical reasoning performance at posttest and follow-up [+W]
• Evidence on clinical reasoning teaching strategies is inconclusive, so no specific teaching strategy can currently be recommended [+W]
• Resident self-explanation with prompts produced a large improvement in diagnostic performance compared with solving word puzzles [+W]
• SNAPPS outperformed the one-minute preceptor and both outperformed traditional clinical teaching for clinical reasoning skill [+W]

Related Strategies¶

• Future active-learning research should use randomised controlled designs with larger samples and standardised outcome measures
• Use large samples, true experimental designs, and moderate treatment durations when evaluating SRL scaffolds
• Future psychiatry education technology studies should use active comparators, standardized assessment tools, and long-term follow-up
• Use cluster randomized clinical trials and blinded simulation-based behavior assessment to strengthen NTS training evidence

Examples¶

-

Key Sources¶

• Delavari S, Barzkar F, Rikers MJP, Pourahmadi M, Soltani Arabshahi SK, Keshtkar A, Dargahi H, Yaghmaei M, Monajemi A. (2024). Teaching and learning clinical reasoning skill in undergraduate medical students: A scoping review. PLOS ONE. https://doi.org/10.1371/journal.pone.0309606