Issues in Distinguishing Sex and Gender in Surgical Registries: A NSQIP and VASQIP Analysis.

Background: Surgical registries do not have separate sex (the biological construct) and gender identity variables. We examined procedures specific to sexually dimorphic anatomy, such as ovaries, testes, and other reproductive organs, to identify "discrepancies" between recorded sex and the anatomy of a procedure. These "discrepancies" would represent a structural limitation of surgical registries, one that may unintentionally perpetuate health inequities.

Study Design: Retrospective cohort study using 2015-2019 NSQIP and 2016-2019 VASQIP. Surgeries were limited to procedures pertaining to anatomy that is either specifically male (CPT codes 54000-55899) or female (56405-59899). The sex recorded in the surgical registries, often automatically retrieved from electronic health record data, was compared to the specified anatomy of each procedure to quantify discrepancies.

Results: 575,956 procedures were identified specific to sexually dimorphic anatomy (549,411 NSQIP; 26,545 VASQIP). Of those, 2,137 recorded a sex discordant with the anatomy specified by the surgical procedure (rates 0.4% in NSQIP; 0.2% in VASQIP). Procedures specific to female anatomy with recorded male sex were more frequent (82.6% in NSQIP; 98.4% in VASQIP) than procedures specific to male anatomy with recorded female sex.

Conclusions: Discrepancies between recorded sex and the anatomy of a surgical procedure were limited. However, because sex in surgical registries is often directly acquired from electronic health record data, these cases likely represent transgender, gender diverse, or living with a difference of sex development (intersex) patients. As these populations increase and continue to seek healthcare, precise measurement of sex, gender identity, and legal sex is necessary for adequate risk adjustment, risk prediction, and surgical outcome benchmarking for optimal care.