Background: Race and racism are social determinants of health resulting in marginalized groups often experiencing worse health outcomes. These racial differences are often due to physician bias as well as biased processes. To mitigate this, it is recommended that clinicians avoid reifying racial differences as biological differences. This study describes the use of racial identifiers and potentially stigmatizing language in a medical journal.

Methods: We reviewed adult case reports in the Journal of Hospital Medicine between 2006-2018 using PubMed searches yielding 234 results. 82 were excluded due to lack of case report formatting or international location and 152 were reviewed using an advanced search function for: born, birth, White, Caucasian, Black, African, Asian, Latino, Hispanic, American, Chinese, Japanese, Mexican, Korean, alcohol, tobacco, cigarette, drug, cocaine, heroin, amphetamine, marijuana, cannabis, MDMA, ecstasy, opiate, opioid, abuse, adherence, compliance, obese, overweight. Search results were reviewed by two authors (DG, ZP), and the decision was made a posteriori to consider mention of the above search terms to connote substance misuse, substance use disorders, and stigmatized medical diagnoses, excepting alcohol use not meeting criteria for an alcohol use disorder. Frequencies of racial specification were calculated, as were the presence or absence of additional potentially stigmatizing language.

Results: Of 152 case reports, 34 (22.4%) included race. Of these, 12 (35.3%) were Black/African American, 11 (32.4%) White/Caucasian, 7 (20.6%) Hispanic, and 4 (11.8%) Asian. There was over-representation of racial specification for Black/African Americans (35.3% vs. 13.4%, p < .001), but not for other races. Race was more often specified in cases where potentially stigmatizing language was used (12/35 versus 22/118 , 35.3% vs. 18.6%, p = 0.040).

Conclusions: These data suggest case reports disproportionately identify Black/African American patients when mentioning race, and that race is more often mentioned concurrently with stigmatized diagnoses and behaviors, possibly worsening clinician biases and health inequities. Strategies for reducing this include robust peer review and providing prospective authors guidance on avoiding the use of racial identifiers and stigmatizing language in medical scholarship. Limitations of these findings include the a posterori classification of potentially stigmatizing terms. Future research can identify how often racial specification occurs in other medical scholarship, with a priori identification of language that is stigmatizing with incorporation of concepts from current literature on the subject.