Social Network Intervention to Engage Out-of-Care PLH Into Treatment

Formative Research Phase (Months 1-6) We will undertake qualitative formative studies to: (1) identify barriers to HAART and strategies currently used to engage PLH in care; (2) identify access points and ways to reach a diversity of PLH social networks; (3) gain an understanding of PLH views, motivations, barriers, and facilitators of care entry, maintenance, and adherence; (4) examine the structure and segments of the PLH community in St. Petersburg; and (5) elicit input from members of the PLH community and its stakeholders concerning the planned network recruitment, assessment, and intervention procedures and content. We will refine protocols used in our intervention pilot study based on findings of the formative research phase.

Key informant interviews. In-depth individual interviews will be conducted with approximately 30 persons with a high level of understanding about the PLH community. They will include persons occupying formal and informal roles with PLH such as care providers in clinics, testing programs, and AIDS centers; and leaders of non-governmental organizations (NGOs) that serve PLH. Interviews will be recorded, conducted by staff experienced in open-ended interviewing, last about 2 hours, and follow a topic guide with latitude to pursue new themes that emerge. The guide will elicit information on: (1) current strategies to link, maintain, and promote adherence of PLH to care, and the success and shortcomings of these strategies; (2) structural, psychosocial, service availability, and other barriers and facilitators of care and adherence; (3) reasons PLH avoid, drop out, or interrupt care and adherence; (4) characteristics of PLH who are most and least likely to enter, remain in, and adhere to care; (5) access points for reaching PLH not in care; and (6) recommendations concerning planned study methods.

In-depth interviews with PLH. We will also carry out in-depth interviews with at least 50 PLH, purposively selected to maximize diversity with respect to gender, age, mode of exposure, longstanding versus recent infection, involvement in care systems, and reported HAART adherence. To achieve this heterogeneity, PLH will be recruited in both clinical and community settings and use PLH who were successfully recruited to recruit PLH they know who are not in care. In addition to the topics above, PLH interviews will assess characteristics of interactions with other PLH in their social networks including where they interact with PLH not in care, including virtual and physical locations; ways they talk about issues related to HIV status, how these may differ between males and females, experiences with care systems, adherence, perceived stigma, and coping. We will also elicit feedback related to recruitment, assessment, and intervention methods planned for the main trial. 2-hour interviews, conducted by experienced interviewers, will be audio recorded.

Qualitative phase data analysis and integration of formative findings into intervention planning. All interview tapes will be transcribed in Russian and translated into English by bilingual study staff experienced in AIDS-related vernacular, coded, and analyzed for key themes and patterns of responses. Data analysis will look for reasons why PLH are or are not in care for their HIV disease; reasons why PLH adhere or do not adhere to antiretroviral regimens; experiences with HIV care systems; beliefs, attitudes, and intentions regarding treatment as held by individuals and among friends who are also PLH; how and when PLH talk with HIV-positive friends about the disease and treatment; existing peer norms regarding HAART and adherence; access points to reach PLH not in care including community venues and virtual sites; and other themes relevant to intervention planning. Analysis will take place inductively and deductively in relation to the study's major analytic domains and constructs. Coding of data will occur in three stages. The first stage will code the data by participant type (key informant or PLH) and, for PLH, by key participant background characteristics (gender, mode of exposure, in treatment, or not in treatment). The second stage will apply codes to reflecting the main analytical domains (engagement in care, adherence to HAART, network norms regarding HAART, barriers to care systems, and others). Sub-codes will be identified through careful reading of interview transcripts to identify themes, event sequences, and relationships among factors within the major study domains. Creation of sub-codes is the third stage of the coding process, allowing discovery of unexpected patterns and relationships.

Qualitative data analysis will first summarize the variability in each domain. We will then look for patterned differences and relationships among codes such as differences between PLH who did not link or remain in care and those who did, and those who are or are not adherent to HAART. For example, we will look at differences in mode of exposure, gender, experiences with the health care system, and other life circumstances among those who linked into care compared to those who did not. We will then form initial hypotheses regarding reasons why some PLH engage in care while others do not, which we will test by comparing participants with the hypothesized explanatory factor (e.g. gender, negative experiences with health care systems) to see if they differ in engagement in care. This iterative and comparative process will continue until the research team is comfortable with the explanatory model developed. We will then present findings to the Community Advisory Committee in order to translate findings into development of intervention content including messages, attitudes, motivations, and skills needed to promote entry and maintenance in care; strategies and skills for maintaining HAART adherence; benefits of care to the health of oneself and others; handling barriers to care involvement; areas in which the intervention needs to be tailored by gender; ways to handle issues that may arise in cross-gender communication; and development of a positive self-identity as a PLH. We will also identify access points to recruit initial PLH seeds. We will specifically elicit Advisors' recommendations concerning recruitment, assessment, and intervention content and delivery.

Community Advisory Committee. We have extensively discussed the project with PLH community advocates and stakeholders in St. Petersburg. A 7-member Advisory Committee has already been formed, including leaders of NGOs and agencies that serve PLH: Doctors to Children, Humanitarian Action, The Society of People Living with HIV/AIDS, The Unity of People Living with HIV, Astra, and the Leningrad Oblast AIDS Center. We will meet at least annually to solicit committee guidance about project activities, findings, community perspectives and concerns.

Main Trial Phase (Months 7-60)

Overview of the main intervention outcome trial's experimental design. The main trial is a two-arm randomized outcome study. A total of 32 sociocentric social networks of PLH will be recruited by first identifying initial seeds-always PLH who are either out-of-care or treatment nonadherent-in multiple access points that were identified in the formative phase. We will then enroll three rings of HIV+ friends outward beginning with each seed. Each sociocentric network is expected to consist of approximately 16 to 18 individuals (expected n=32x18=576 participants). This estimate is based on the size and density of participants' personal networks observed in our pilot studies. Each network member will be assessed at baseline using measures to be described shortly and will receive individual counseling about the benefits and availability of HIV care and of antiretroviral therapy adherence. This session will "prime" participants to an understanding about the availability, accessibility, and benefits of care. The counseling will be given to all study participants, including those in the comparison group. Members of the 16 PLH networks randomized to the experimental condition will then additionally receive the network intervention. Cadres of empirically identified influence leaders within each experimental condition network will be identified, trained, and engaged to reinforce other network members for engagement and adherence. At 6- and 12-month followup points, assessment data will again be collected to determine intervention impact on the primary outcomes.