Evaluating Resident Screening for Social Determinants of Health: The Development and Assessment of a Structured Direct Observation Tool

Background: Physicians must gain competency engaging and screening families from all socioeconomic back- grounds with the growing number of children living in poverty. Traditional medical training lacks sufficient education in screen- ing for social determinants of health. We developed a direct observation tool to evaluate resident screening for social determi- nants of health, assessed its reliability, and determined if the tool could detect practice change after an educational intervention. Methods: A quasi-experimental interventional study of pediatric residents in a primary care setting using a direct observation tool during patient encounters was undertaken. Descriptive test statistics of individual question frequency and discrimination as well as overall test reliability were completed. Changes in the total time discussing social history and number of questions asked after an educational intervention were analyzed. Results: Increase in mean number of questions from 6.6 pre intervention to 8.5 post- intervention (p=.04). Residents spent a statistically significant greater amount of time (97 to 184 seconds, p=.01) discussing the social history with patients after education. Point biserial calculations for all items in the tool were positive. Conclusions: This direct observation tool was reliable and detected practice changes in this innovative pilot study. This structured tool may improve compliance with direct observation requirements and help provide residents with objective feedback.