Refusal Rates to Sexual Orientation and Gender Identity Items in the Behavioral Risk Factor Surveillance System, 2014-2019

Objectives. To explore trends in sexual orientation and gender identity (SOGI) item refusal in the Behavioral Risk Factor Surveillance System (BRFSS). Methods. We used annual data from 7 US states that implemented the SOGI module of the BRFSS from 2014 to 2019 to examine prevalence of sexual orientation (n5373 332) and gender identity (n5373 336) item refusal. Analyses included the weighted Wald chi(2) test of association between refusal and year and logistic regressions predicting refusal by year. We weighted analyses to account for complex sampling design. Results. We found low SOGI item refusal rates, significant declines in these refusal rates over time, and differences in refusal rates by age, sex, race, education, and language. We also found that Hispanic group membership did not explain SOGI item refusal when accounting for interview language; interview language was strongly associated with both sexual orientation and gender identity item refusal. Conclusions. Our results indicate acceptance of SOGI measurement and empirically support continuation of SOGI in health surveillance surveys. Findings indicate a need to further investigate the association between survey translation and item refusal.