Actionable companion · Assessment and evidence of learning

From evidence to district action: Integrating ChatGPT into knowledge-retrieval tutorials in undergraduate medical education: a prospective evaluation of higher-order learning and feasibility

AI-assisted editorial implementation guide. The district actions are proposed applications, not findings or recommendations attributed to the original authors. No additional peer review is asserted.

Evidence and context

Health-professions education context: test transfer to K–12 rather than assuming it.

Original authors’ abstract, reproduced under the linked source license:

Abstract Background Generative artificial intelligence (AI) platforms such as ChatGPT present new opportunities to strengthen competency-based medical education (CBME). While retrieval practice is a proven strategy for enhancing long-term retention, its application to higher-order domains of Bloom’s taxonomy within CBME, particularly when scaffolded by AI, remains underexplored. To our knowledge, this is the first longitudinal CBME study worldwide to evaluate supervised ChatGPT-assisted retrieval practice and its durability over time. Methods We conducted a six-month, prospective, non-randomised delayed-intervention study and all 270 third-year MBBS students (including supplementary batch) at a private medical college in South India were invited to participate in a faculty-supervised ChatGPT-assisted retrieval-practice intervention. Participants were allocated into four tutorial clusters; two received the intervention immediately, while the remaining two received it after a delay. The intervention comprised four weekly, two-hour sessions featuring higher-order multiple-choice questions, structured faculty-supervised interactions with ChatGPT, and guided metacognitive reflections. Outcomes assessed included MCQ performance at baseline, immediately post-intervention, and at one- and three-month follow-up, as well as student perceptions. Data analysis employed repeated-measures ANOVA and mixed-effects modelling. Results Of 270 students invited, 253 (93.7 %) met inclusion criteria and completed all assessments, achieving 100% follow-up at immediate, one-month, and three-month evaluations. MCQ performance demonstrated significant improvement across time points (F[3,756] = 65.14, p < 0.001). On a 20-point higher-order MCQ scale, mean scores increased from 12.8 ± 2.4 at baseline to 15.2 ± 2.1 immediately post-intervention (adjusted gain +2.37, d = 0.62, p < 0.001). Gains were sustained at one month (+3.78, d = 0.99) and three months (+3.65, d = 0.95), demonstrating durable higher-order learning retention. Both Lower-achieving and Higher-achieving students improved, though the effect was greater among Lower-achieving students (d = 0.72 vs 0.49). Student feedback revealed high levels of satisfaction (mean 4.25 ± 0.88) and cognitive engagement (4.15 ± 0.92), while clarity of AI interaction received comparatively lower ratings (3.39 ± 1.19). Conclusion Supervised ChatGPT-assisted retrieval practice produced sustained improvements in higher-order cognitive performance, with particularly strong benefits for Lower-achieving students. This scalable, standards-aligned model holds promise for advancing CBME globally and warrants further validation through multi-institutional trials incorporating performance-based and clinical outcomes.

A 90-day district pilot

1. Days 1–15 · Appraise. Have the curriculum lead and a teacher read the methods and limitations. Record the study design, sample, comparator, outcomes and evidence gaps. Define a local question that the study can help inform.
2. Days 16–30 · Design. Redesign one low-stakes assessment: require a first attempt, an AI-assisted revision, a source-check log and a short oral explanation. Use the same rubric for the comparison task. Name an accountable owner and set the baseline, success threshold and budget before implementation.
3. Days 31–60 · Test. Start with a small voluntary cohort, preserve an accessible non-AI route and require teacher review. Use synthetic or de-identified examples; approve data handling and tool terms through district processes.
4. Days 61–90 · Decide. Blind-score reasoning, factual accuracy and unaided transfer on a fresh task. Record suspected misconduct separately from demonstrated evidence. Compare with baseline, document uncertainty and stop or revise if learning worsens, access is unequal or critical errors persist.

The board decision brief

Report the learning problem, evidence quality, participating population, total cost including review time, measured benefit, errors, access differences and remaining uncertainties. Make a dated recommendation to stop, revise or expand. A positive perception survey alone does not establish learning improvement.

Attribution and reuse

Roy Arokiam Daniel; Nivethitha V; Surya BN; Antoinette Daniel; Visalakshi R. Integrating ChatGPT into knowledge-retrieval tutorials in undergraduate medical education: a prospective evaluation of higher-order learning and feasibility. Medical Education Online (2026-03-02). https://doi.org/10.1080/10872981.2026.2639203

Companion prepared 2026-10-30. Original authorship and DOI are preserved. The abstract is an attributed source extract; the implementation plan is new AI-assisted editorial material, offered under the same CC BY 4.0 license. This guide has not undergone independent peer review.