PURPOSE: The objective structured clinical examination (OSCE) is considered to be one of the most robust methods of clinical assessment. One of its strengths lies in its ability to minimise the effects of examiner bias due to the standardisation of items and tasks for each candidate. However, OSCE examiners' assessment scores are influenced by several factors that may jeopardise the assumed objectivity of OSCEs. To better understand this phenomenon, the current review aims to determine and describe important sources of examiner bias and the factors affecting examiners' assessments. METHODS: We performed a narrative review of the medical literature using Medline. All articles meeting the selection criteria were reviewed, with salient points extracted and synthesised into a clear and comprehensive summary of the knowledge in this area. RESULTS: OSCE examiners' assessment scores are influenced by factors belonging to 4 different domains: examination context, examinee characteristics, examinee-examiner interactions, and examiner characteristics. These domains are composed of several factors including halo, hawk/dove and OSCE contrast effects; the examiner's gender and ethnicity; training; lifetime experience in assessing; leadership and familiarity with students; station type; and site effects. CONCLUSION: Several factors may influence the presumed objectivity of examiners' assessments, and these factors need to be addressed to ensure the objectivity of OSCEs. We offer insights into directions for future research to better understand and address the phenomenon of examiner bias.
Bias (Epidemiology) ; Humans ; Leadership ; Patient Selection ; Problem Solving ; Recognition (Psychology)

PURPOSE: The biases that may influence objective structured clinical examination (OSCE) scoring are well understood, and recent research has attempted to establish the magnitude of their impact. However, the influence of examiner experience, clinical seniority, and occupation on communication and physical examination scores in OSCEs has not yet been clearly established. METHODS: We compared the mean scores awarded for generic and clinical communication and physical examination skills in 2 undergraduate medicine OSCEs in relation to examiner characteristics (gender, examining experience, occupation, seniority, and speciality). The statistical significance of the differences was calculated using the 2-tailed independent t-test and analysis of variance. RESULTS: Five hundred and seventeen students were examined by 237 examiners at the University of New South Wales in 2014 and 2016. Examiner gender, occupation (academic, clinician, or clinical tutor), and job type (specialist or generalist) did not significantly impact scores. Junior doctors gave consistently higher scores than senior doctors in all domains, and this difference was statistically significant for generic and clinical communication scores. Examiner experience was significantly inversely correlated with generic communication scores. CONCLUSION: We suggest that the assessment of examination skills may be less susceptible to bias because this process is fairly prescriptive, affording greater scoring objectivity. We recommend training to define the marking criteria, teaching curriculum, and expected level of performance in communication skills to reduce bias in OSCE assessment.
Australia ; Awards and Prizes ; Bias (Epidemiology) ; Curriculum ; Humans ; New South Wales ; Occupations ; Physical Examination

PURPOSE: Optimal methods for communication skills training (CST) are an active research area, but the effects of CST on communication performance in objective structured clinical examinations (OSCEs) has not been closely studied. Student roleplay (RP) for CST is common, although volunteer simulated patient (SP) CST is cost-effective and provides authentic interactions. We assessed whether our volunteer SP CST program improved OSCE performance compared to our previous RP strategy. METHODS: We performed a retrospective, quasi-experimental study of 2 second-year medical student cohorts' OSCE data in Australia. The 2014 cohort received RP-only CST (N=182) while the 2016 cohort received SP-only CST (N=148). The t-test and analysis of variance were used to compare the total scores in 3 assessment domains: generic communication, clinical communication, and physical examination/procedural skills. RESULTS: The baseline characteristics of groups (scores on the Australian Tertiary Admission Rank, Undergraduate Medicine and Health Sciences Admission Test, and medicine program interviews) showed no significant differences between groups. For each domain, the SP-only CST group demonstrated superior OSCE outcomes, and the difference between cohorts was significant (P<0.01). The superiority of volunteer SP CST over student RP CST in terms of OSCE performance outcomes was found for generic communication, clinical communication, and physical examination/procedural skills. CONCLUSION: The better performance of the SP cohort in physical examination/procedural skills might be explained by the requirement for patient compliance and cooperation, facilitated by good generic communication skills. We recommend a volunteer SP program as an effective and efficient way to improve CST among junior medical students.
Australia ; Clinical Competence ; Cohort Studies ; Humans ; Non-Randomized Controlled Trials as Topic ; Patient Compliance ; Patient Simulation ; Retrospective Studies ; Students, Medical ; Volunteers

PURPOSE: The biases that may influence objective structured clinical examination (OSCE) scoring are well understood, and recent research has attempted to establish the magnitude of their impact. However, the influence of examiner experience, clinical seniority, and occupation on communication and physical examination scores in OSCEs has not yet been clearly established. METHODS: We compared the mean scores awarded for generic and clinical communication and physical examination skills in 2 undergraduate medicine OSCEs in relation to examiner characteristics (gender, examining experience, occupation, seniority, and speciality). The statistical significance of the differences was calculated using the 2-tailed independent t-test and analysis of variance. RESULTS: Five hundred and seventeen students were examined by 237 examiners at the University of New South Wales in 2014 and 2016. Examiner gender, occupation (academic, clinician, or clinical tutor), and job type (specialist or generalist) did not significantly impact scores. Junior doctors gave consistently higher scores than senior doctors in all domains, and this difference was statistically significant for generic and clinical communication scores. Examiner experience was significantly inversely correlated with generic communication scores. CONCLUSION We suggest that the assessment of examination skills may be less susceptible to bias because this process is fairly prescriptive, affording greater scoring objectivity. We recommend training to define the marking criteria, teaching curriculum, and expected level of performance in communication skills to reduce bias in OSCE assessment.

PURPOSE: Optimal methods for communication skills training (CST) are an active research area, but the effects of CST on communication performance in objective structured clinical examinations (OSCEs) has not been closely studied. Student roleplay (RP) for CST is common, although volunteer simulated patient (SP) CST is cost-effective and provides authentic interactions. We assessed whether our volunteer SP CST program improved OSCE performance compared to our previous RP strategy. METHODS: We performed a retrospective, quasi-experimental study of 2 second-year medical student cohorts' OSCE data in Australia. The 2014 cohort received RP-only CST (N=182) while the 2016 cohort received SP-only CST (N=148). The t-test and analysis of variance were used to compare the total scores in 3 assessment domains: generic communication, clinical communication, and physical examination/procedural skills. RESULTS: The baseline characteristics of groups (scores on the Australian Tertiary Admission Rank, Undergraduate Medicine and Health Sciences Admission Test, and medicine program interviews) showed no significant differences between groups. For each domain, the SP-only CST group demonstrated superior OSCE outcomes, and the difference between cohorts was significant (P<0.01). The superiority of volunteer SP CST over student RP CST in terms of OSCE performance outcomes was found for generic communication, clinical communication, and physical examination/procedural skills. CONCLUSION The better performance of the SP cohort in physical examination/procedural skills might be explained by the requirement for patient compliance and cooperation, facilitated by good generic communication skills. We recommend a volunteer SP program as an effective and efficient way to improve CST among junior medical students.