Truace
Subscribe

Faculty-supervised AI-supported BOPPPS-CBL and performance on a simulated neurological-emergency assessment among junior-college medical interns: a batch-stratified randomized educational study

Background County hospitals and township health centers in China provide frontline care for time-sensitive neurological emergencies, and three-year junior-college clinical medicine programs are an important pathway into this workforce. We evaluated a faculty-supervised artificial intelligence (AI)-supported BOPPPS (Bridge-in, Objectives, Pre-assessment, Participatory Learning, Post-assessment, and Summary) and case-based learning (CBL) instructional package among junior-college medical interns. Methods In this p…

By the Truvace desk2 min read1 sourcePeer-reviewedTRV-2026-1472· cite & verify
Faculty-supervised AI-supported BOPPPS-CBL and performance on a simulated neurological-emergency assessment among junior-college medical interns: a batch-stratified randomized educational study
Reading Wikipedia in the Classroom for Secondary School Students 07 by James Rhoda. CC BY-SA 4.0 · https://creativecommons.org/licenses/by-sa/4.0

In brief

In a batch-stratified randomized study, 118 junior-college clinical medicine interns in China were split between an AI-supported teaching package and conventional instruction during a one-month neurology rotation. The AI arm used standardized DeepSeek prompts for Socratic questioning, error feedback and reflection under faculty supervision, and it scored higher on a four-station simulated neurological-emergency assessment.

The result matters because county hospitals and township health centers rely on graduates of these programs for time-sensitive neurological emergencies. Still, the assessment was locally developed, the study was not prospectively registered, and the findings do not show that the gains carry over to real clinical work or patient outcomes.

Main points

  1. 118 interns from 12 monthly rotation batches were randomized: 58 to AI-supported teaching and 60 to conventional teaching.
  2. The adjusted difference in total simulated-assessment score was 5.67 points, with an adjusted Cohen's d of 0.74.
  3. Among station-specific analyses, only lesion localization met the multiplicity-adjusted threshold, at 2.58 points.
  4. Written-examination estimates were smaller and imprecise.
  5. Learners in the AI-supported group reported a greater preparation-time burden in exploratory questionnaire items.

The problem

The AI-supported group reported a greater perceived preparation-time burden, and the study design could not rule out cueing or carry-over from reused case contexts.

The rundown

The intervention combined the BOPPPS lesson structure with case-based learning. Standardized DeepSeek prompts supplied Socratic questioning, error-focused feedback and structured reflection, with faculty supervising. The control group received conventional BOPPPS-CBL.

The primary endpoint was the total score (0–100) on a four-station simulated neurological-emergency assessment. All 118 participants completed it. Regression models adjusted for rotation batch, baseline academic performance and digital literacy.

Post hoc sensitivity analyses excluding the station with reused case contexts gave standardized estimates of similar magnitude. They could not exclude cueing or carry-over. A Station 2-specific ICC was not estimated.

The authors call for multicenter studies with independent assessment cases, longer follow-up and workplace-based outcomes.

Sources

  1. Peer-reviewedFrontiers in Medicine2026-10-14

The debate