A theoretically informed instrument to assess clinical communication in objective structured clinical examinations.

Huntley, CD and Salmon, P and Fisher, P and Fletcher, Ian and Young, B (2012) A theoretically informed instrument to assess clinical communication in objective structured clinical examinations. Medical Education, 46 (3). pp. 267-276. ISSN 0308-0110

Abstract

Objectives  We reviewed papers describing the development of instruments for assessing clinical communication in undergraduate medical students. The instruments had important limitations: most lacked a theoretical basis, and their psychometric properties were often poor or inadequately investigated and reported. We therefore describe the development of a new instrument, the Liverpool Undergraduate Communication Assessment Scale (LUCAS), which is intended to overcome some of these limitations. We designed LUCAS to reflect the theory that communication is contextually dependent, inherently creative and cannot be fully described within a conceptual framework of discrete skills. Methods  We investigated the preliminary psychometric properties of LUCAS in two studies. To assess construct and external validity, we examined correlations between examiners’ LUCAS ratings and simulated patients’ ratings of their relationships with students in Year 1 formative (n = 384) and summative (n = 347) objective structured clinical examination (OSCE) samples. Item–total correlations and item difficulty analyses were also performed. The dimensionality of LUCAS was examined by confirmatory factor analysis. We also assessed inter-rater reliability; four raters used LUCAS to rate 40 video-recorded encounters between Year 1 students and simulated patients. Results  Simulated patient ratings correlated with examiner ratings across two OSCE datasets. All items correlated with the total score. Item difficulty showed LUCAS was able to discriminate between student performances. LUCAS had a two-dimensional factor structure: we labelled Factor 1 creative communication and Factor 2 procedural communication. The intraclass correlation coefficient was 0.73 (95% confidence interval 0.54–0.85), indicating acceptable reliability. Conclusions  We designed LUCAS to move the primary focus of examiners away from an assessment of students’ enactment of behavioural skills to a judgement of how well students’ communication met patients’ needs. LUCAS demonstrated adequate reliability and validity. The instrument can be administered easily and efficiently and is therefore suitable for use in medical school examinations.