An experimental study of implicit racial bias in recognition of child abuse

Objectives: We evaluated whether implicit racial bias influences pediatricians' suspicion of child abuse. Methods: Child abuse experts developed 9 injury vignettes. Pediatricians (N = 342) were randomly assigned one of 2 versions to rate for suspicion of abuse, with the child's race in each vignette varying between white and black. Data were collected online and anonymously. Results: There were no statistically significant differences in suspicion for an abuse-related injury based on the race of the child. We adjusted for pediatrician race/ethnicity, years since graduation, location, and gender and did not find race effects. Conclusions: We demonstrated an experimental approach to study the influence of implicit racial bias on recognition of child abuse. Though we failed to find an effect, it is too early to conclude that none exists. The relationship among human cognition, behavior, and healthcare disparities is complex. Studies are needed that incorporate diverse approaches, clinical contexts and scenarios, patient and physician characteristics, and validated measures if we are to understand how it might be used to reduce healthcare disparities.