TY - JOUR
T1 - Which Internal Medicine Clerkship Characteristics Are Associated With Students' Performance on the NBME Medicine Subject Exam? A Multi-Institutional Analysis
AU - Fitz, Matthew M.
AU - Adams, William
AU - Haist, Steven A.
AU - Hauer, Karen E.
AU - Ross, Linette P.
AU - Raff, Amanda
AU - Agarwal, Gauri
AU - Vu, T. Robert
AU - Appelbaum, Jonathan
AU - Lang, Valerie J.
AU - Miller, Chad
AU - Grum, Cyril
N1 - Publisher Copyright:
© by the Association of American Medical Colleges. Unauthorized reproduction of this article is prohibited.
PY - 2020/9/1
Y1 - 2020/9/1
N2 - Purpose To identify which internal medicine clerkship characteristics may relate to NBME Medicine Subject Examination scores, given the growing trend toward earlier clerkship start dates. Method The authors used linear mixed effects models (univariable and multivariable) to determine associations between medicine exam performance and clerkship characteristics (longitudinal status, clerkship length, academic start month, ambulatory clinical experience, presence of a study day, involvement in a combined clerkship, preclinical curriculum type, medicine exam timing). Additional covariates included number of NBME clinical subject exams used, number of didactic hours, use of a criterion score for passing the medicine exam, whether medicine exam performance was used to designate clerkship honors, and United States Medical Licensing Examination Step 1 performance. The sample included 24,542 examinees from 62 medical schools spanning 3 academic years (2011-2014). Results The multivariable analysis found no significant association between clerkship length and medicine exam performance (all pairwise P >.05). However, a small number of examinees beginning their academic term in January scored marginally lower than those starting in July (P <.001). Conversely, examinees scored higher on the medicine exam later in the academic year (all pairwise P <.001). Examinees from schools that used a criterion score for passing the medicine exam also scored higher than those at schools that did not (P <.05). Step 1 performance remained positively associated with medicine exam performance even after controlling for all other variables in the model (P <.001). Conclusions In this sample, the authors found no association between many clerkship variables and medicine exam performance. Instead, Step 1 performance was the most powerful predictor of medicine exam performance. These findings suggest that medicine exam performance reflects the overall medical knowledge students accrue during their education rather than any specific internal medicine clerkship characteristics.
AB - Purpose To identify which internal medicine clerkship characteristics may relate to NBME Medicine Subject Examination scores, given the growing trend toward earlier clerkship start dates. Method The authors used linear mixed effects models (univariable and multivariable) to determine associations between medicine exam performance and clerkship characteristics (longitudinal status, clerkship length, academic start month, ambulatory clinical experience, presence of a study day, involvement in a combined clerkship, preclinical curriculum type, medicine exam timing). Additional covariates included number of NBME clinical subject exams used, number of didactic hours, use of a criterion score for passing the medicine exam, whether medicine exam performance was used to designate clerkship honors, and United States Medical Licensing Examination Step 1 performance. The sample included 24,542 examinees from 62 medical schools spanning 3 academic years (2011-2014). Results The multivariable analysis found no significant association between clerkship length and medicine exam performance (all pairwise P >.05). However, a small number of examinees beginning their academic term in January scored marginally lower than those starting in July (P <.001). Conversely, examinees scored higher on the medicine exam later in the academic year (all pairwise P <.001). Examinees from schools that used a criterion score for passing the medicine exam also scored higher than those at schools that did not (P <.05). Step 1 performance remained positively associated with medicine exam performance even after controlling for all other variables in the model (P <.001). Conclusions In this sample, the authors found no association between many clerkship variables and medicine exam performance. Instead, Step 1 performance was the most powerful predictor of medicine exam performance. These findings suggest that medicine exam performance reflects the overall medical knowledge students accrue during their education rather than any specific internal medicine clerkship characteristics.
UR - http://www.scopus.com/inward/record.url?scp=85090074264&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090074264&partnerID=8YFLogxK
U2 - 10.1097/ACM.0000000000003322
DO - 10.1097/ACM.0000000000003322
M3 - Article
C2 - 32195693
AN - SCOPUS:85090074264
SN - 1040-2446
VL - 95
SP - 1404
EP - 1410
JO - Academic Medicine
JF - Academic Medicine
IS - 9
ER -