Abstract Body

Although pre-exposure prophylaxis (PrEP) across San Francisco is expanding, significant age, gender, and racial/ethnic disparities in uptake persist. Via a data-driven approach, we created a PrEP registry and a high-risk non-PrEP registry to identify patients in primary care who may benefit from targeted outreach.

Starting in January 2016, patients receiving PrEP within the San Francisco Department of Public Health Primary Care (SFPC) clinics were included in a registry if they had received a PrEP prescription from a SFPC medical provider and were confirmed HIV-negative via laboratory and medical chart review. In the absence of structured medical record data on HIV-risk, we used available laboratory data to identify high-risk non-PrEP patients if they were HIV-negative; were not prescribed PrEP; and either were screened for a rectal sexually transmitted infection, were diagnosed with syphilis in the past 12 months, or received ≥3 HIV tests in a 24 month period. This analysis compares PrEP patients to non-PrEP patients as of 5/31/2017. Chi-square tests measure the bivariate association between PrEP initiation, demographics, and active panel status (assigned to a SFPC clinic and primary care provider, with at least one primary care visit in the last 24 months). A multivariate logistic regression model with an outcome of not initiating PrEP was created.

Overall, 451 patients were confirmed to have started PrEP and 2,109 patients were identified as high-risk non-PrEP patients. Non-PrEP patients were more likely to be female (45% vs 16%, p<0.01), Black (32% vs 14%, p <0.01), and ≥50 years (24% vs 18%, p=0.02); and less likely to be active on a care panel (52% vs 86%, p<0.01). In a multivariate analysis controlling for active panel status, non-PrEP patients were more likely to be women [adjusted OR (aOR) 5.64; 95%CI:4.22-7.54)], Black (aOR 2.85; 95%CI:2.03-4.00), Latino (aOR 1.38; 95%CI:1.02-1.87), and ≥50 years (aOR 2.18; 95%CI:1.60-2.95) than PrEP patients. Active panel status was a strong negative predictor of being a non-PrEP patient (aOR 0.13; 95%CI:0.09-0.17).

Age, gender, and racial/ethnic disparities remain in PrEP uptake across DPH-funded primary care clinics in San Francisco, suggesting that access to care is not sufficient to address these disparities. Additional interventions, informed by this data-driven approach, are needed to help both primary care providers and patients identify risk and, respectively, prescribe and initiate PrEP.