Linda Tewksbury1, Regina Richter2, Colleen Gillespie3, and Adina Kalet2
1Pediatrics, 2Medicine, NYU School of Medicine, 3NYU Robert F. Wagner Graduate School of Public Service
Background: Clinical skills exams (CSE) have been shown to be valid and reliable tools for detecting students with poor clinical skills. In order for such students to improve, they need to be able to accurately assess their ability, the basis of lifelong learning. Having a better understanding of how these students self-assess their level of competence would thus be important in developing successful remediation programs.
Objective: To explore how students with lowest performance on a CSE self-assess their level of competence.
Design/Methods: All 4th year medical students completed an 8-station comprehensive clinical skills exam during which standardized patients (SPs) rated communication skills (CS), history gathering (HG) and physical exam skills (PE). Students were requested to complete a post-exam survey including a self-assessment of competence in CS, HG and PE relative to level of training. Criteria for remediation included failure in 2 or more competencies (CS, HG or PE) or CS alone. Responses to self-assessment (lower, accurate or higher level of competence) relative to SP-rated performance (below, within or above one SD of mean for class) were compared for remediated versus passing students using Pearson chi-square tests.
Results: 145/170(85%) of students who took the exam consented to have their data analyzed anonymously. Eight students required remediation: 2 failed CS only, 1 failed CS, HG and PE, 3 failed CS and HG, and 2 failed HG and PE. All 8(100%) remediated students overestimated level of competence for CS versus 35/137(25%) passing students (p<.001). For HG, 6/8(75%) of remediated students overestimated and 2/8(25%) accurately estimated competence versus 18/137(13%) and 105/137(77%) of passing students (p<.001). For PE: 4/8(50%) of remediated students overestimated, 3/8(37.5%) accurately and 1/8(12.5%) underestimated competence versus 21/137(15%), 78/137(57%) and 38/137(28%) of passing students (p=0.04).
Conclusions: Medical students with the lowest performance on a clinical skills exam were significantly more likely to overestimate their level of competence compared to their peers, particularly in communication skills. Such findings should be taken into consideration when developing remediation strategies to enable students to more accurately self-assess their abilities.