Menu
Menu
Menu
Menu
Introduction: Every year, approximately 260,000 children are infected with HIV in low- and middle-income countries. The timely initiation and high level of maintenance of antiretroviral therapy (ART) are crucial to reducing the suffering of HIV-positive children. We need to develop a better understanding of the background of children’s ART non-adherence because it is not well understood. The purpose of this study is to explore the background related to ART non-adherence, specifically in relation to the orphan status of children in Kigali, Rwanda. Methods: We conducted 19 focus group discussions with a total of 121 caregivers of HIV-positive children in Kigali. The primary data for analysis were verbatim transcripts and socio-demographic data. A content analysis was performed for qualitative data analysis and interpretation. Results: The study found several contextual factors that influenced non-adherence: among double orphans, there was psychological distance between the caregivers and children, whereas economic burden was the primary issue among paternal orphans. The factors promoting adherence also were unique to each orphan status, such as the positive attitude about disclosing serostatus to the child by double orphans’ caregivers, and feelings of guilt about the child’s condition among non-orphaned caregivers. Conclusions: Knowledge of orphan status is essential to elucidate the factors influencing ART adherence among HIV-positive children. In this qualitative study, we identified the orphan-related contextual factors that influenced ART adherence. Understanding the social context is important in dealing with the challenges to ART adherence among HIV-positive children.
We conducted focus group discussions (FGDs) from May to July 2011 with 121 caregivers of HIV-positive children. Kigali city was selected as the study site due to its high number of HIV-positive children and high percentage (23.8%) of Rwanda’s paediatric ART services [24]. We used FGDs because it is an optimal method for comprehending participants’ social norms, views and living conditions [17], and it is conducive to stimulating their verbalizations. The study methods and results were reported according to the checklist of consolidated criteria for reporting qualitative research (COREQ) [25] and the relevance, appropriateness and transparency (RATS) guidelines for qualitative research [26]. We recruited FGD participants from 717 respondent caregivers who participated in a survey in our previous study on their children’s ART adherence [16]. In that study, 717 of the 1301 pairs of caregivers and their children enrolled in an ART programme responded to the survey. They were recruited from 15 health facilities. Four public hospitals and one clinic were selected because they enrolled half of all HIV-positive children in Kigali; and 10 of 21 health centres providing routine paediatric ART services were randomly selected. All of the respondent caregivers were over 18 years of age and the primary caregivers of HIV-positive children. The respondent children were aged six months to 14 years and were enrolled in ART for at least 12 weeks. None of the respondent caregivers or their children had previous relationships with research personnel prior to completing the quantitative survey. Our interviewers explained the purpose and themes of the FGDs to all of the participant-caregivers in the quantitative survey. To ensure easy communication with the 717 caregivers, those without phone access were excluded, leaving 507 caregivers, of which 365 agreed to participate in the FGDs. The 365 potential FGD participants were classified as adherent or non-adherent, based on the results of our survey. Adherence was measured by pill count, which was previously announced. According to the World Health Organization (WHO) guideline for adherence, a child taking less than 85% of the monthly prescribed pills was considered non-adherent, 85% was poor adherence, and an 85% or higher adherence rate was the minimum level to achieve HIV suppression and the clinical benefits of ART [27]. Caregivers were classified into four categories, corresponding to their child’s orphan status: double orphan, paternal orphan, maternal orphan and non-orphan. We arranged two FGDs for each of the eight categories (16 FGDs) with two groups each of adherent and non-adherent caregivers composed of a mix of orphan categories (four FGDs). Two FGDs were organized for each category because two was the minimum number of groups needed to obtain different views of the same category’s participants [21]. We could form only one FGD for caregivers of adherent maternal orphans due to the lack of eligible participants (Figure 1). Flowchart summarizing participant inclusion and categories. Of the 365 caregivers, we selected those whom we reached by phone to schedule their FGD date. The caregivers were selected randomly by skipping through the list at regular intervals until the necessary number of participants was recruited. Of the 365 caregivers, 55 were excluded because they could not be reached by phone after three attempts to schedule an appointment (all appointments were made at different times), were unavailable on any of the proposed dates, had changed their minds regarding participation or had an incorrect phone number. We scheduled FGD dates with 152 randomly selected caregivers from the 310 remaining caregivers. The 158 remaining caregivers were excluded because we had recruited a sufficient number to provide the 8–12 participants needed for each FGD. However, 31 of the 152 selected caregivers did not attend the FGD due to urgent business or public transportation delays, or were absent without notice. Finally, 121 caregivers participated in the FGDs (4 to 10 participants for each FGD). Repeat interviews were not conducted. We held FGDs until we felt that data saturation was achieved. We conducted the FGDs using a guide that we developed based on a guide previously used in African settings [28]. The topics included barriers to and promoters of adherence, and difficulties perceived by non-biological caregivers. Before conducting the main FGDs, we held three practice sessions with 24 caregivers, and then modified the guide. The FGDs were conducted in the local language (Kinyarwanda) by a male facilitator (Rwandan medical doctor) with extensive experience in FGDs and a female note-taker (Rwandan nurse) with a sociology background and experience in fieldwork. The note-taker recorded the remarks and reactions of the participants. The discussions were held in a private room inside Kigali’s district hospital. Each discussion’s duration was 45 to 60 minutes. In addition to the facilitator and note-taker, the first author and her interpreter were present. We used a tape recorder after obtaining participants’ permission. The recordings and the notes were used to produce a complete transcript of each discussion in the local language, which was translated into English for analysis. The transcriptions and translations were produced by different research assistants who had not attended the FGDs. The note-taker of the FGDs reviewed the accuracy of the English translations. Before each FGD, the facilitator explained the study’s purpose, discussion themes and ethical considerations to the participants. Each caregiver provided socio-demographic information and information about the child for whom they cared. Transcripts were not returned to the participants for comments or corrections. However, feedback meetings were held with participants (63 of 121 attended) in November 2011 to obtain their validation and to search for discrepancies in evidence and negative cases. The primary data for the analyses were verbatim transcripts and socio-demographic data. The transcripts were analysed by the content analysis method according to the caregivers’ categories (based on the orphan and adherence statuses of the children). We analysed data using the “five-phase cycle,” suggested by Yin [17]: (1) compiling all verbatim transcripts and notes taken, (2) disassembling compiled data using a coding method, (3) reassembling fragmented data by sorting them into different groups, (4) interpreting data by developing data arrays and referring to socio-demographic data and (5) drawing conclusions. Phase (1) was completed by the note-taker, phases (2) and (3) were conducted by the first author and an expert in qualitative research who was not involved in the data collection process, and phases (4) and (5) were performed by the first author. NVivo software, Version 10 (QRS International Pty. Ltd., Doncaster, Australia), was used to sort and code the data. The coding classification scheme was based on the major topical headings specified in the interview guide. Two individuals reviewed and discussed the coded data segments to reach an agreement. Ethical approval for the study was obtained from the National Ethics Committee of Rwanda and the Research Ethics Committee of the University of Tokyo. Official permission was obtained from the Institute of HIV Disease Prevention and Control/Rwanda Biomedical Centre to collect data at each facility. We started the FGDs after obtaining participants’ written consent and their permission to make digital recordings of the sessions for transcriptions. The discussions were conducted anonymously in a private room, and only a number supplied on the day of the FGD identified the participants.
N/A
