Improving medical interviewing skills in medical students through simulation-based training with pediatric patient models

Technical Summary: Improving Medical Interviewing Skills in Medical Students through Simulation-Based Training with Pediatric Patient Models

Problem Statement
Medical history-taking and physical examination are critical skills directly linked to diagnostic accuracy. However, training medical students in these skills, particularly with pediatric patients, presents significant challenges. In the Japanese medical education system, resources for simulated patient (SP) training are limited prior to graduation. Furthermore, direct clinical training with real pediatric patients and their parents is often deemed inappropriate for students due to the sensitivity of the interactions and the risk of causing distress. While simulation-based education is recognized as a solution, existing literature predominantly relies on subjective questionnaire surveys to evaluate effectiveness. There is a notable lack of quantitative assessments measuring the actual improvement in students' diagnostic abilities, the quality of their questioning techniques, and their consideration during examinations.

Methodology
The study employed a prospective educational intervention involving 93 fourth- and fifth-year medical students, organized into 25 groups of three to four participants. The training utilized a simulation-based approach where a research physician, who is a specialist, played the role of the parent for five distinct pediatric patient scenarios (ranging from intussusception to neuroblastoma).

Key methodological features included:

• Training Structure: Each group conducted three consecutive medical interviews and physical examinations without time limits. Following each session, students presented their findings and received immediate feedback from the supervising physician.
• Standardization: The research physician underwent approximately 50 hours of role-playing training to ensure consistent responses across all groups. All interactions were audio-recorded, transcribed, and analyzed.
• Outcome Measures:
  • Primary Endpoint: The number of "essential" interview and examination items completed out of a predefined list of 15 critical items (covering basic history, disease-specific findings, and differential diagnosis points).
  • Secondary Endpoints: Total number of questions asked, the proportion of open-ended questions (OQs), the effective examination rate (ratio of required items to total questions/exams), and a "consideration points" score derived from the Liverpool Undergraduate Communication Assessment Scale (LUCAS). This score penalized unprofessional behaviors (e.g., rude language, inappropriate silence) and rewarded empathetic comments.
• Statistical Analysis: Changes from the first to the third case were analyzed using Repeated Measures ANOVA on a group basis, with statistical significance set at p < 0.05.

Key Results
The intervention demonstrated statistically significant improvements in specific areas of clinical performance:

• Essential Items: The average number of essential interview and examination items completed increased significantly from 9.2 in the first case to 12.0 in the third case (p < 0.001). The authors note that achieving approximately 12 of the 15 items allows for a fairly accurate differential diagnosis.
• Question Volume: The total number of questions asked increased significantly from an average of 21.8 to 27.7 (p = 0.0113).
• Consideration Points: There was a dramatic improvement in professional conduct, with the consideration score rising from an average of -5.6 (indicating frequent unprofessional behaviors) in the first case to -0.3 in the third case (p < 0.001).
• Non-Significant Findings: While the proportion of open-ended questions and the effective examination rate showed gradual increases, these changes did not reach statistical significance.

Key Contributions
This study distinguishes itself from prior simulation-based education research by moving beyond subjective self-assessments to provide a quantitative evaluation of clinical skill acquisition. Specifically, it:

1. Quantifies the improvement in the selection of non-exhaustive medical interview items based on chief complaints.
2. Numerically evaluates the reduction of unprofessional behaviors and the increase in considerate communication during examinations.
3. Demonstrates the efficacy of a training model combining immediate feedback with multiple consecutive simulation cases.
4. Utilizes a research physician as a standardized "parent" model, trained for 50 hours, to ensure high-fidelity role-playing and consistent data collection.

Significance and Claims
The authors claim that this study provides evidence that simulation-based training with immediate feedback and repeated practice significantly improves medical students' interviewing techniques and professional demeanor. They assert that by quantitatively evaluating these skills, the study validates a method for scoring improvements relevant to real-world clinical practice. The authors suggest that this approach is particularly useful for overcoming the difficulties of training with real pediatric patients and that the quantitative data generated could inform the development of future medical interview simulators, including their existing Virtual Reality Objective Structured Clinical Examination (VR OSCE) software.

The paper maintains a modest tone regarding its limitations, acknowledging that the study was not a two-group comparative design, did not assess individual student abilities within groups, and that the specific quantification method for interview items, while clinically practical, may not be universally applicable to all complex disease models.