Assessing Justice, Equity, Diversity, and Inclusion Best Practices for Test Item Writing at a Physician Assistant Program: A Retrospective Analysis

Abstract

Introduction Outcomes for specific racial and ethnic groups during the COVID-19 pandemic and the widespread protests demanding racial justice in 2020 renewed a focus on health equity practices and bias in medical education. Routine inclusion of sociodemographic identifiers such as sex/gender, age, race, and/or ethnicity in test items risks perpetuation of implicit bias and race-based practices in medical education. Methods A retrospective review was conducted of test items in a pathophysiology course series at an academic medical center in the Western United States. Sociodemographic identifiers in multiple choice test items were reviewed to determine clinical relevance. Results Of 795 test items, 475 referred to sociodemographic information (gender, race/ethnicity, age, sexual orientation) and in only 27.5% of questions, this inclusion was clinically relevant (131/475). Discussion Current evidence suggests that routine inclusion of sociodemographic information in test items infrequently supports the clinical diagnosis and perpetuates implicit bias and race-based practices in medical education. Thoughtful inclusion of sociodemographic information in assessment offers an opportunity to address sociostructural determinants of health.