OBJECTIVES: To validate the use of a script concordance test (SCT), a tool to assess clinical reasoning in contexts of uncertainty, which are common in clinical geriatrics practice, on geriatric urinary incontinence (UI) to discriminate levels of expertise in this content area. DESIGN: A reference panel (15 geriatricians) and 12 respondents (10 senior geriatrics fellows and 2 interns) completed an online 100-item SCT test covering major topics in UI. The test was then optimized by discarding items with negative item-total correlation; the remaining 70 questions covered all major topics in UI. The test was then administered to a second group of participants with different levels of experience, mostly from the University of Miami: eight geriatricians, nine junior geriatrics fellows, 53 internal medicine residents, and 26 medical students. Investigators assessed test reliability and construct validity (to discriminate between levels of expertise). SETTING: Tertiary academic medical center and affiliated medical school. PARTICIPANTS: Medical students, internal medicine residents, geriatric medicine fellows, and practicing geriatricians. MEASUREMENTS: Seventy-item SCT. RESULTS: The Cronbach alpha for the 70-item test was 0.72. Mean scores were 75.3 ± 7.9 for geriatricians (n = 23), 69.0 ± 9.3 for senior geriatrics fellows (n = 10), 66.4 ± 6.8 for junior geriatrics fellows n = (9), 66.1 ± 5.7 for residents (n = 53), and 65.6 ± 8.5 for students (n = 26). Using analysis of variance, significant differences were found between the mean scores of the geriatricians and all other participants except senior fellows. CONCLUSION: The geriatric UI SCT demonstrated moderate reliability and evidence of construct validity, discriminating between experienced and nonexperienced physicians.
OBJECTIVES: To validate the use of a script concordance test (SCT), a tool to assess clinical reasoning in contexts of uncertainty, which are common in clinical geriatrics practice, on geriatric urinary incontinence (UI) to discriminate levels of expertise in this content area. DESIGN: A reference panel (15 geriatricians) and 12 respondents (10 senior geriatrics fellows and 2 interns) completed an online 100-item SCT test covering major topics in UI. The test was then optimized by discarding items with negative item-total correlation; the remaining 70 questions covered all major topics in UI. The test was then administered to a second group of participants with different levels of experience, mostly from the University of Miami: eight geriatricians, nine junior geriatrics fellows, 53 internal medicine residents, and 26 medical students. Investigators assessed test reliability and construct validity (to discriminate between levels of expertise). SETTING: Tertiary academic medical center and affiliated medical school. PARTICIPANTS: Medical students, internal medicine residents, geriatric medicine fellows, and practicing geriatricians. MEASUREMENTS: Seventy-item SCT. RESULTS: The Cronbach alpha for the 70-item test was 0.72. Mean scores were 75.3 ± 7.9 for geriatricians (n = 23), 69.0 ± 9.3 for senior geriatrics fellows (n = 10), 66.4 ± 6.8 for junior geriatrics fellows n = (9), 66.1 ± 5.7 for residents (n = 53), and 65.6 ± 8.5 for students (n = 26). Using analysis of variance, significant differences were found between the mean scores of the geriatricians and all other participants except senior fellows. CONCLUSION: The geriatric UI SCT demonstrated moderate reliability and evidence of construct validity, discriminating between experienced and nonexperienced physicians.