The RSVP Project: Factors Related to Disengagement From Human Immunodeficiency Virus Care Among Persons in San Francisco

Background: In the United States, an estimated two-thirds of persons with human immunodeficiency virus (HIV) infection do not achieve viral suppression, including those who have never engaged in HIV care and others who do not stay engaged in care. Persons with an unsuppressed HIV viral load might experience poor clinical outcomes and transmit HIV. Objective: The goal of the Re-engaging Surveillance-identified Viremic Persons (RSVP) project in San Francisco, CA, was to use routine HIV surveillance databases to identify, contact, interview, and reengage in HIV care persons who appeared to be out of care because their last HIV viral load was unsuppressed. We aimed to interview participants about their HIV care and barriers to reengagement. Methods: Using routinely collected HIV surveillance data, we identified persons with HIV who were out of care (no HIV viral load and CD4 laboratory reports during the previous 9-15 months) and with their last plasma HIV RNA viral load >200 copies/mL. We interviewed the located persons, at baseline and 3 months


Introduction
In the United States, an estimated two-thirds of persons living with human immunodeficiency virus (HIV) infection do not achieve viral suppression, including both those who have never engaged in HIV care and those who linked to HIV care after diagnosis but did not stay engaged [1][2][3].Persons with an unsuppressed HIV viral load might experience poor clinical outcomes and transmit HIV [4] The National HIV/AIDS Strategy for the United States focuses on improving the HIV care continuum, including interventions that link, retain, and reengage persons in HIV care [5,6].
The goal of the Re-engaging Surveillance-identified Viremic Persons (RSVP) project in San Francisco was to use routine HIV surveillance databases to identify, contact, interview, and reengage persons living with HIV infection who appeared to be out of care because their last HIV viral load was unsuppressed.We interviewed participants about their HIV care patterns and barriers to reengagement.Understanding why people disengage from HIV care and what would help them reengage is essential for optimizing HIV clinical and public health outcomes.

Methods
RSVP began on April 20, 2012 for a 12-month period.Our project methods and implementation, including success in locating truly viremic out-of-care persons, have been previously described [19].Briefly, persons with an unsuppressed plasma HIV RNA viral load (>200 copies/mL) at their last measurement were eligible if they appeared to have disengaged from HIV care because they lacked HIV viral load and CD4 cell count laboratory reports during the 9 to 15 months prior to April 20, 2012 in the San Francisco Department of Public Health Enhanced HIV/AIDS Reporting System.We asked participants to complete a 30-minute interviewer-administered survey and offered them assistance with reengaging in HIV care from the San Francisco Department of Public Health linkage and navigation program (LINCS).Baseline interview questions included demographics, health insurance coverage, HIV testing and care experiences, treatment use, sexual activities, unmet services, and willingness to talk with LINCS.Participants were asked to complete a follow-up interview 3 months later that assessed changes since baseline and care reengagement.RSVP participants could meet with LINCS staff for health insurance assistance, appointments for HIV care reengagement, and referrals to ancillary services.The US Centers for Disease Control and Prevention and the University of California, San Francisco determined this to be a nonresearch activity; therefore, we did not require institutional review board approval.

Baseline RSVP Interview
The characteristics of the 75 (26.6%)interviewed participants were broadly representative of the 282 eligible persons [19].Most of the 75 interviewed participants (median age 45 years) were male (85%), born in the United States (77%), and current San Francisco residents (75%), and identified as gay or homosexual (69%) (Table 1).Participants were racially and ethnically diverse, one-third reported a college degree or higher, the majority were either unemployed or receiving disability benefits, and 1 in 5 reported being homeless or in unstable housing in the previous 12 months.A total of 64 (85%) were sexually active in the past year, and 19 (25%) reported having had condomless anal or vaginal sex with a person of HIV-negative or unknown status.Drug use and binge drinking were common as were symptoms of depression (Table 1).A total of 67 (89%) reported current health insurance or coverage, 59 (79%) had ever been prescribed HIV medications, and 45 (60%) reported current medication use.In the past year, 59 (79%) had seen an HIV provider for medical care, and 67 (89%) reported a CD4 cell count or viral load test.For the 11 (15%) participants who had not seen an HIV provider in the past year, the frequently reported reasons included feeling healthy (n=6), drinking alcohol or using drugs (n=5), not having enough money or health insurance (n=4), and not wanting to take HIV medicines (n=4).
Nearly half (45%) reported missing an HIV medical care appointment in the past year (Table 1).Among 59 participants who reported seeing a provider in the past year, 27 (46%) reported missing an HIV medical care appointment in the past year.These 27 participants volunteered 1 or more of these responses to "What would help you the most to get to HIV medical care appointment within the next 3 months?":transportation assistance (n=8), a stable living situation or housing (n=3), sound mental health (n=3), help getting organized to track appointments (n=3), and more appointment reminders (phone calls, text messages, email, letters) (n=3).
The 30 (40%) participants who reported not currently taking medications to treat HIV were similar in their characteristics to the overall interviewed population: 24 were men, 25 had current health insurance or coverage, 20 had seen an HIV provider for medical care in the past year, and 23 expressed interest in meeting with LINCS staff.Overall, the most frequently reported services used in the past year were HIV education or information, public benefits support, HIV case management, financial assistance, and spiritual support (Table 2).The most frequently reported needed but not accessed services were dental services, mental health services, financial assistance, transportation assistance, and HIV case management (Table 2).Of the participants, 52 (69%) accepted referral to LINCS; their linkage outcomes were previously published [19].Among the 11 participants who reported not having seen an HIV medical provider in the previous 12 months, 10 (91%) agreed to talk to LINCS staff about reengagement or ancillary services.Of these, 8 (80%) did meet with LINCS staff and, of these, 5 (63%) were already in HIV care and 3 (38%) reengaged in care within 6 months with LINCS assistance.

Follow-Up Interview
Of the 75 participants, 64 (85%) completed the 3-month follow-up interview.Among the 64, 97% (n=62) had health insurance or coverage, 73% (n=47) reported having "seen a doctor, nurse or other health care provider for HIV medical care" since baseline interview, 81% (n=52) reported having a CD4 cell count test, 77% (n=49) had an HIV viral load test, and 47% (n=30) were prescribed HIV treatment.HIV risk behaviors since the baseline visit included using "any drugs that were not prescribed by a doctor" (n=30, 47%,) and injecting drugs that were not prescribed (n=9, 14%).A total of 33 (52%) reported vaginal or anal sex, including 7 (11%) who reported condomless vaginal or anal sex with an HIV-negative or unknown-status partner.

Discussion
The RSVP project sought to identify a high-priority population of persons living with HIV who appeared to be out of care and viremic based on HIV surveillance data.However, the majority of interviewed participants reported both having seen an HIV provider in the past 12 months and having health insurance, thus having means and opportunity to access the HIV care needed to achieve viral suppression.The frequent self-reported care engagement was corroborated by HIV surveillance data: XSL • FO RenderX over 80% had at least one viral load or CD4 cell count test during the 12 months after they met RSVP project eligibility [19].
Nevertheless, RSVP participants identified personal and structural barriers to HIV care and many unmet needs.Almost half reported missing a scheduled medical appointment in the previous year, and 40% reported no current HIV treatment.Deficits in comprehensive health care and social support included unmet needs for legal services, dental care, transportation assistance, mental health services, financial assistance, and practical support.Notably, about one-fourth of participants reported engaging in condomless sex with negative or unknown HIV-status partners.Given that all RSVP participants had an unsuppressed HIV viral load at their last measurement, the risk of further HIV transmission was possible, indicating that we reached persons who would benefit from assistance remaining in HIV care.Similar to our findings, a myriad of social, behavioral, and structural factors have previously been found to be correlated with poor retention in care [20,21], including substance and alcohol use, poor mental health, homelessness, and low socioeconomic status [8,10,22]; low care satisfaction, medical establishment distrust, stigma, and lack of social or ancillary support services [15,23]; and early HIV disease or feeling well [8,17].Patient perceptions of HIV care engagement, including when accessing care only sporadically, may also differ from standard care engagement metrics [24].The multifaceted barriers to staying engaged in HIV care point to the need for comprehensive, ongoing, innovative, client-centered approaches to support retention (eg, case management, wraparound services, and contingency management).We were unable to analyze specific associations between barriers participants faced and the services they used, and their likelihood of reengaging in care, due to the heterogeneity of barriers and care patterns in our relatively small interviewed population.
In summary, among persons presumed to be out of HIV care, there were self-reported indicators of at least intermittent HIV care, as well as frequently reported barriers to care and unmet needs.As previously described, locating persons who are truly out of HIV care is difficult [19].Nevertheless, surveillance-based public health efforts support the HIV care continuum, as most RSVP project participants returned for a follow-up interview, seemed willing to engage with health department staff, and accepted assistance with ancillary services and HIV care reengagement, a prerequisite for ongoing HIV treatment and viral suppression.
High school, General Equivalency Diploma, or less (40) Some technical or college training (31) College degree or more Current housing situation, n (%) (57) Person's own house or apartment (21) Someone else's house or apartment (15) Single room, rented room, motel, single-room occupancy (7) All other (shelter, transitional housing, homeless) Current work situation, n (%) (33) Working full-time or part-time XSL • FO RenderX Median (IQR a ) or n (%) Characteristics (33) Unemployed, laid off, looking for work (21) Disabled (receiving disability income) In jail, detention, or prison in the past 12 months, n (%) Sexual risk, drug use, and other behaviors (past 12 months) Had any sex (anal, vaginal, or oral), n (%) Had anal or vaginal sex without a condom with a person of HIV b -negative or unknown status, n (%) (25) Yes (60) No (15) Not applicable (not sexually active) Injected any nonprescription drugs, n (%) Prescription drugs or painkillers without a prescription Binge drinking (≥5, if male, and ≥4, if female, alcoholic drinks in one sitting), n (%) in particular where usually goes for medical care not related to HIV infection, n (%) Years since first HIV-positive test, n (%) other HIV-related services a Ordered by highest count; percentage calculated among those who did not use a service in past 12 months.bThe top 5 most frequently used services, and the top 5 most frequently needed services (unmet needs).
c HIV: human immunodeficiency virus.

Table 2 .
Services used, and services that were needed but not used, among the Re-engaging Surveillance-identified Viremic Persons (RSVP) project participants at baseline, San Francisco, 2012-2013 (n=75).