Background : National surveys based on probability samples are crucial tools for unbiased estimates of health disparities among gender minorities. In 2014, the Behavior Risk Factor and Surveillance Survey (BRFSS) began offering an optional module to capture transgender and gender nonconforming (TGNC) identity. Although the BRFSS provides much needed data, self-identified TGNC respondents are vulnerable to misclassification of sex assigned at birth.
Methods : This study applied quantitative bias analysis to explore the magnitude and direction of the systematic bias present in four measures of reproductive health that result from sex misclassification. We use imputation models and probabilistic bias adjustments to explore bias present estimated associations between gender identity and four sex-specific outcomes: PSA testing, Pap testing, hysterectomy, and pregnancy.
Results : TGNC individuals with misclassified sex are demographically distinct from those asked sex-specific questions, suggesting that there is significant selection bias present in BRFSS measures of reproductive health. Estimates for gender nonconforming respondents are the most vulnerable to small degrees of bias, while estimates for transgender women and men are more robust to moderate degrees of bias.
Conclusions : Researchers who use BRFSS data to examine reproductive health disparities in TGNC populations should recognize the systematic bias present in the data, and avoid reporting results as population-based estimates. Our findings emphasize the importance of implementing validated sex assigned at birth and gender identity questions in national surveys.