Embedding a longitudinal diagnostic reasoning curriculum in a residency program using a bolus/booster approach.

Background Diagnostic reasoning skills are essential to the practice of medicine, yet longitudinal curricula to teach residents and evaluate performance in this area is lacking. We describe a longitudinal diagnostic reasoning curriculum implemented in a university-based internal medicine residency program and self-evaluation assessment of the curriculum's effectiveness. Methods A longitudinal diagnostic reasoning curriculum (bolus/booster) was developed and implemented in the fall of 2015 at the University of Iowa. R1, R2, and R3 cohorts were taught the "bolus" curriculum at the beginning of each academic year followed by a "booster" component to maintain and build upon diagnostic reasoning skills taught during the "bolus" phase. Self-administered diagnostic thinking inventory (DTI) scores were collected in the spring of pre-curriculum (baseline, 2014-2015) and post-curriculum (2016-2017). Results The overall DTI scores improved in the R1 cohort, although statistically significant differences were not seen with R2s and R3s. In the original DTI categories, R1s improved in both flexibility of thinking and structure of thinking, the R2s improved in structure of thinking and the R3s did not improve in either category. R1s showed improvement in three of the four subcategories - data acquisition, problem representation, and hypothesis generation. The R2s improved in the subcategory of problem representation. R3s showed no improvement in any of the subcategories. The R3 cohort had higher mean scores in all categories but this did not reach statistical significance. Conclusions Our program created and successfully implemented a longitudinal diagnostic reasoning curriculum. DTI scores improved after implementation of a new diagnostic reasoning curriculum, particularly in R1 cohort.