High Retention and Adherence With Rapid Long-acting Injectable Preexposure Prophylaxis Implementation in an Urban Safety Net Clinic Population

Abstract

In a low-barrier, long-acting preexposure prophylaxis (PrEP) program in a safety net setting, permitting same-day or next-day initiation, 85% of injections were on time, and 6-month retention was 83%, surpassing outcomes from most previously reported oral PrEP studies. With drop-in, wraparound services, similar retention among housing-insecure populations was seen. Long-acting PrEP expansion is urgently needed.

Despite the availability of effective oral preexposure prophylaxis (PrEP) regimens, less than one-third of Americans with PrEP indications receive it [1]. Furthermore, 37%–62% discontinue oral PrEP within 6 months [2–4]. Stigma, difficulty with taking daily pills, substance use, mental health challenges, and housing insecurity all contribute to PrEP discontinuation [2–5].

Injectable cabotegravir-based PrEP has the potential to address some of these barriers [6]. Despite initial promising findings, with cabotegravir showing superiority to oral PrEP [7], cabotegravir accounts for only 1% of U.S. PrEP use [1]. Impediments to implementation include burdensome prior authorizations, need for dedicated staff to provide injections, and complicated procedures to procure cabotegravir [1]. Little is known about the feasibility of cabotegravir PrEP rollout, particularly in safety net clinics that serve populations that could benefit the most. In this cabotegravir implementation program performed within a municipal primary care health system in San Francisco, which permitted rapid PrEP initiation following negative human immunodeficiency virus (HIV) rapid testing at 1 referral site, we examined retention and on-time injections among a population with a high prevalence of substance use, mental health concerns, and housing instability.

METHODS

Patients who received injectable cabotegravir PrEP within the San Francisco Department of Public Health Primary Care Clinics (SFPCC) verified via electronic health record (EHR) were included. Injectable cabotegravir was offered onsite within 8 of the 14 San Francisco Department of Public Health Primary Care Clinics clinics, and patients could be referred to or drop-in for wraparound PrEP services at Ward 86's PrEP clinic. The injectable PrEP program at Ward 86 includes: support by nurses, pharmacists, and a navigator who helps coordinate follow-up; availability of same or next-day PrEP following a negative HIV fourth-generation antigen/antibody rapid test with viral load testing pending, and an e-consult program staffed by nurses and a physician. The second largest site, Castro Mission Health Center, has access to a PrEP navigator and a focus on lesbian, gay, bisexual, transgender, and queer health, whereas the third largest site, Maria X. Martinez Resource Center offers drop-in services and low-barrier addiction treatment, with a focus on people experiencing homelessness.

Patients were included who initiated cabotegravir PrEP from December 2021 through May 2024, with follow-up ascertained through June 2024. The study was reviewed and approved by the University of California San Francisco institutional review board. We obtained dates of PrEP initiation and discontinuation, age, race/ethnicity, housing status, substance use, and mental health disorders via electronic health record abstraction. Individuals were considered to have on-time injections if they occurred ±7 days from the target injection date (4 weeks for the loading dose and 8 weeks for subsequent doses), and to have discontinued cabotegravir if it was not restarted by the end of observation.

Mental health diagnoses were recorded via International Classification of Diseases, 10th revision, codes in the active problem list and contemporaneous chart review including anxiety; depression; posttraumatic stress disorder; schizophrenia, schizoaffective disorder, or psychotic disorder; and bipolar disorder. Data were captured on stimulant use (any amphetamine or cocaine/crack use documented by a medical provider), nonprescribed opioid use, and hazardous alcohol use or alcohol use disorder (documented by a medical provider) if reported in the medical chart at the time of cabotegravir use. Patients with documentation from a provider of homelessness or unstable housing were classified as having housing instability.

We examined predictors of late injections via adjusted generalized mixed-effects models and used Cox regression to identify predictors of discontinuation. Predictors included in both models were housing status (unstable/stable), comorbid mental health diagnosis (yes/no), a substance use composite consisting of any reported hazardous alcohol (Alcohol Use Disorders Identification Test-Concise > 4 or documented by physician), opioid, or stimulant use, prior PrEP history (no prior PrEP use, previous oral PrEP but no current PrEP, and current oral PrEP), clinic, age, gender, and race/ethnicity.

RESULTS

Overall, 111 patients (mean age = 37 years ±11.5) taking cabotegravir as PrEP from December 2021 to May 2024 were included; 65% were cisgender male, 19% cisgender female, 10% nonbinary gender, 5% transgender female; 35% were Hispanic; 37% White, 15% Black, 6% Asian; and 36% had unstable housing. One-third (33%) had used PrEP previously but had stopped, 28% were initiating PrEP for the first time, and 39% switched from oral PrEP (Supplementary Table 1); 41% were actively using stimulants, 15% using opioids, and 21% had hazardous alcohol use. Of the 63 patients seen at the site offering rapid PrEP start (Ward 86), 20 (32%) received cabotegravir within 24 hours of an initial appointment.

Across the sample, there were a total of 507 injections, and most injections (85%) were on time. When examining predictors of late injections, only a mental health diagnosis significantly predicted late injections (adjusted odds ratio = 1.93; 95% confidence interval [CI], 1.03–3.60; P = .040; Supplementary Table 2).

Cabotegravir PrEP retention was 83% at 6 months (95% CI, 75–91), and median time on cabotegravir was 283 days at the end of observation (95% CI, 534–665; Figure 1). Most common reasons for discontinuation were low self-perceived HIV risk (6), injection site reactions (4), loss to follow-up (4), and substance use relapse (2). Of the 19 who discontinued, 4 transitioned to oral PrEP. In Cox regression, substance use (adjusted hazard ratio [AHR] = 2.23; 95% CI, 1.15–4.33; P = .017), a mental health diagnosis (AHR = 5.36; 95% CI, 1.31–21.8; P = .019), and prior oral PrEP but no current PrEP use (AHR = 4.66; 95% CI, 1.11–19.6; P = .036) predicted discontinuation (Supplementary Table 3). Ward 86 patients had lower discontinuation (AHR = 0.18; 95% CI, 0.03–0.95; P = .043), whereas housing-insecure individuals had similar discontinuation (AHR = 1.06; 95% CI, .33–3.36; P = .92). No seroconversions occurred in this cohort.

Figure 1.

Retention on injectable cabotegravir for preexposure prophylaxis (PrEP) within the San Francisco Primary Care Clinics. The Kaplan-Meier curve shows individuals who were initiated on cabotegravir injectable PrEP and remained on injectable PrEP until the end of observation, with the yellow highlighting representing 95% confidence intervals.

DISCUSSION

In a U.S.-based cabotegravir injectable PrEP program that serves a population with high prevalence of substance use, mental health conditions, and housing stability, 85% of injections were on time and retention at 6 months was 83%, surpassing outcomes previously reported in most settings for oral PrEP [4]. Potential factors supporting high retention included rapid access to the medication at initiation at Ward 86, which has previously been shown to be associated with higher long-term retention [8], and coordination of follow-up by nurses, pharmacists, and a navigator to ensure patients returned for follow-up. Most patients were not currently using oral PrEP, indicating that the program predominantly recruited a population that could benefit from this novel delivery strategy. Previous modeling analyses have indicated that expansion of PrEP uptake rather than converting oral PrEP users to injectables will have the most impact on the HIV epidemic [9].

The association of substance use with higher PrEP discontinuation has been documented in prior studies [3]. Administration of integrated substance use care, including long-acting buprenorphine and naltrexone, was offered at the 3 largest sites, but not used by all patients with substance use. Contingency management has been shown to have high efficacy in reducing methamphetamine use and improving adherence to HIV antiretroviral therapy [10], and was available intermittently in 1 location, which may have limited impact. Innovative programs to improve PrEP and substance use outcomes are urgently needed for this population.

The association of prior, but not current, PrEP use with discontinuation likely reflects the inclusion of a population at-risk of HIV exposure, but with challenges attending clinic to receive PrEP care. Innovative programs, including mobile administration of injectable PrEP by clinicians could be considered for this group [11, 12]. Ward 86's drop-in Health Access Point and the Maria X. Martinez Resource Center provide low-barrier PrEP integrated with substance use and mental health resources, including coadministration of long-acting substance use or antipsychotic treatment with injectable PrEP. Remarkably, despite more than one-third of our sample experiencing unstable housing, high retention, and on-time injections was achieved in this group; and care at the Ward 86 site was associated with lower discontinuation. Health Access Point is colocated with and based on the Positive-health Onsite Program for Unstably-housed Populations (POP-UP) program [12], which has been successful in engaging individuals with unsuppressed viral load using low-barrier HIV care. Expansion of these programs in a status-neutral fashion is needed urgently to address the HIV epidemic among people with unstable housing.

We found that the only predictor of late injections was a mental health diagnosis, which also predicted discontinuation. Depression was diagnosed in one-quarter of patients, which has been associated with suboptimal PrEP adherence in studies of oral PrEP [5]. Integrated mental health care has been suggested as a solution to improve adherence among PrEP users with depression [5]. Although all sites offered comprehensive primary care, some mental health services, including long-term counseling, were only available off-site due to carve-outs for mental health services separate from primary care in the context of insurance funding constraints. Integrated mental health/PrEP case management could also potentially improve adherence in this population [8].

Limitations of this analysis include the inability to exclude patient transfer, although most cases would be ascertained through the use of Epic “Care Everywhere” medical record exchange. In addition, we cannot exclude presence of covariates not documented within the medical record such as for housing status, substance use, or mental health diagnoses. Finally, Ward 86 is better resourced than some primary care clinics, which may limit external generalizability.

In conclusion, a low-barrier injectable PrEP program employing rapid starts demonstrated high retention and adherence in a safety-net setting, despite the high prevalence of housing instability and substance use. Long-acting PrEP has the potential to make dramatic impacts on the HIV epidemic, although uptake has been insufficient to date [1]. Innovative programs to offer injectable, rapid start, PrEP to populations at high risk of HIV exposure but not currently using oral PrEP should be expanded.

Supplementary Data

Supplementary materials are available at Clinical Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Note

Financial support. This work was funded by National Institutes of Health (NIH)/National Institute of Mental Health K23MH122286 (M. A. S.), NIH/National Institute of Allergy and Infectious Diseases P30AI027763 (M. G.), and NIH/NIAID 2R01AI143340 (M. G. and M. A. S.).