Using Sexual Orientation and Gender Identity Data to Identify Disparities in HIV Preexposure Prophylaxis

By
Dhruv Khurana
Maryah Garner
Sonam Khurana
Bijan Sasaninia
Anthony Firek
October 14, 2026

Jama Network Open

Overview

This cross-sectional analysis examined how restricting or missing sexual orientation and gender identity (SOGI) information affects surveillance of documented HIV preexposure prophylaxis (PrEP) orders. The supplied material describes the study design and population but does not provide its numerical results; it reports that SOGI-specific differences were not estimable when analyses were limited to sex assigned at birth and that simulated SOGI missingness reduced precision of disparity estimates.

Background

SOGI data can support assessment of whether preventive services reach populations with differing HIV risk and prevention needs. In June 2025, HRSA finalized changes removing SOGI measures from Uniform Data System Table 3B for calendar-year reporting, citing alignment with administration priorities. Surveillance systems that use sex assigned at birth alone may combine groups with substantially different HIV incidence, obscuring within-group differences relevant to equity monitoring. Incomplete SOGI documentation in electronic health records can further constrain or bias disparity analyses.

Data Highlights

Study element	Details reported
Design and setting	Cross-sectional analysis using structured electronic health record data from a Southern California safety-net integrated delivery system.
Source population	180,662 patients registered on October 28, 2025.
Final sample	165,198 adults with classifiable sex assigned at birth, sexual orientation, and gender identity.
PrEP outcome period	Medication orders dated January 2, 2019, through October 27, 2025.
Surveillance scenarios	Complete SOGI information; monitoring by sex assigned at birth only; and incomplete SOGI ascertainment with joint missingness varied from 0% to 90%.
Reported result in supplied context	Among patients assigned male at birth, adjusted probabilities of documented PrEP orders were higher for sexual-minority patients and transgender and gender-diverse patients; these SOGI-specific differences were not estimable under sex-assigned-at-birth-only analyses. Simulated SOGI missingness reduced precision of disparity estimates.

Key Findings

The analysis held the population and outcome constant while varying the identity information available to surveillance.
It compared complete SOGI ascertainment, sex-assigned-at-birth-only monitoring, and simulated joint SOGI missingness from 0% to 90%.
Structured SOGI collection began in 2021, and completeness for both SOGI fields exceeded 95% by 2023 in the study system.
Among patients assigned male at birth, adjusted probabilities of documented PrEP orders were higher for lesbian, gay, bisexual, or other sexual-minority patients and for transgender and gender-diverse patients, according to the supplied web research context.
When analyses were restricted to sex assigned at birth, the reported SOGI-specific differences were not estimable; simulated missingness reduced the precision of disparity estimates.
The supplied study context notes that registration and first-encounter dates were unavailable, so observation duration could vary and a uniform follow-up period could not be established.

Clinical Implications

The supplied material indicates that surveillance based only on sex assigned at birth cannot estimate the reported SOGI-specific differences in documented PrEP orders, while missing SOGI data reduces estimate precision. These findings concern measurement and disparity monitoring; the provided text does not establish effects on individual treatment decisions or clinical outcomes.

Conclusion

In this safety-net health system analysis, restricting or missing SOGI information limited the ability to observe PrEP-order disparities by identity group. The supplied evidence supports interpreting SOGI data availability as a determinant of what disparity surveillance can measure.