Do medical students' scores using different assessment instruments predict their scores in clinical reasoning using a...

Area of Science:

• Medical Education
• Health Professions Education
• Clinical Reasoning Assessment

Background:

• Clinical problem-solving skills develop with structured training.
• Computer-based case simulations (CCS) are used for teaching and assessing clinical reasoning.
• Previous studies on CCS psychometric properties have yielded controversial results, and their integration into curricula is limited.

Purpose of the Study:

• To examine the psychometric properties of a specific CCS software (DxR Clinician) for assessing medical students.
• To evaluate the reliability and validity of CCS scores in a problem-based medical curriculum.

Main Methods:

• Assessed 130 medical students using DxR Clinician software.
• Calculated internal consistency reliability using Cronbach's alpha.
• Analyzed relationships between CCS scores and other assessment tools (MCQs, short-answer, OSCE, real patient encounters) via regression analysis.

Main Results:

• High internal consistency reliability for CCS scores (α=0.862).
• Objective structured clinical examination (OSCE) scores significantly predicted clinical reasoning (32.7%) and patient management (35.1%) in CCS.
• Multiple-choice question scores predicted diagnostic performance (15.4%), but real patient encounters did not predict CCS scores.

Conclusions:

• OSCE scores are strong predictors of clinical reasoning and patient management performance in CCS.
• CCS appears to assess different constructs than real patient encounters.
• Findings support the use of CCS as a valid assessment tool when correlated with specific other assessments.