HIV-exposed infant follow-up in Mozambique: Formative research findings for the design of a cluster randomized controlled trial to improve testing and ART initiation

Background: Early infant diagnosis (EID) of HIV-exposed and initiation of HIV-positive infants on anti-retroviral therapy (ART) requires a well-coordinated cascade of care. Loss-to-follow-up (LTFU) can occur at multiple steps and effective EID is impeded by human resource constraints, difficulty with patient tracking, and long waiting periods. The objective of this research was to conduct formative research to guide the development of an intervention to improve the pediatric HIV care cascade in central Mozambique. The study was conducted in Manica and Sofala Provinces where the adult HIV burden is higher than the national average. The research focused on 3 large clinics in each province, along the highly populated Beira corridor. Methods: The research was conducted in 2014 over 3 months at six facilities and consisted of 1) patient flow mapping and collection of health systems data from postpartum, child-at-risk, and ART service registries, 2) measurement of clinic waiting times, and 3) patient and health worker focus groups. Results: HIV testing and ART initiation coverage for mothers tends to be high, but EID and pediatric ART initiation are hampered by lack of patient tracking, long waiting times, and inadequate counseling to navigate the care cascade. About 76% of HIV-positive infants were LTFU and did not initiate ART. Conclusions: Effective interventions to reduce LTFU in EID and improve pediatric ART initiation should focus on patient tracking, active follow-up of defaulting patients, reduction in EID turn-around times for PCR results, and initiation of ART by nurses in child-at-risk services. Trial registration: Retrospectively registered, ISRCTN67747315, July 24, 2019. Quantitative and qualitative data were gathered to identify barriers and facilitators in the EID and treatment cascade, and solicit experiences, perceptions, and recommendations of HIV-positive mothers and health care workers to reduce LTFU to help in the design of a pilot intervention. The formative research study included six health facilities, selected in collaboration with the Provincial Health Directorates (DPSs) of Manica and Sofala using criteria of high patient volume and a mix of urban and more rural facilities along the highly populated Beira corridor. The selected facilities included three in Sofala province (Macurungo, Munhava, and Dondo) and three in Manica (Nhamaonha, 1° de Maio, and Gondola). All are public facilities in the National Health Service that have provided the full range of PMTCT services, including HIV testing (rapid test and access to PCR test for EID, with 5 of 6 facilities shipping the samples to a referral lab), access to CD4 testing (2 of 6 sites shipped the samples to a referral lab) and ART. The post-partum clinic visits (known as CPP, consulta pós-parto) and child-at-risk (CCR, consulta de criança em risco) clinic visits were staffed by mid-level maternal-child health (MCH) nurses, and ART for infants and children was provided by physician assistants, known as tecnicos de medicina (referred to as tecnicos), in Mozambique, and medical doctors. Data were collected to identify inefficiencies and bottlenecks in the follow up of mothers and their infants in maternity, post-partum, CCR and retention in ART services, to guide the identification of key workflow modifications, and develop an enhanced adherence and retention package to test in an intervention. Data were collected from September to November 2014 at the six sites. The research consisted of 1) patient flow mapping, 2) time-motion studies, 3) collection of health systems data per clinic, and 4) focus group discussions with mothers and health staff to identify facilitators and barriers to patient flow and access. Individual interviews with health workers were used to map patient flow patterns from the maternity and CCR, to ART services at each of the target sites to produce flow diagrams [28]. To measure specific clinic waiting and consult times, researchers used a time-motion study method in which they were stationed at clinics and measured the waiting time and clinic visit duration from the moment mothers arrived at the CPP, CCR, and ART services through provider consultations [29]. Individual mothers were continually observed through their clinic visits to measure time spent waiting and in consultation. Two trained researchers followed 20 mothers to measure waiting time in CCR and 10 in pediatric ART in each of the six clinics (for a total of 120 for CCR and 60 for pediatric ART) over a full week per clinic. Two researchers also collected data over 3 months from health systems resources, such as maternity, CPP and CCR clinic registries, and ART patient charts and clinic registries. In maternities, the team collected data for number of births, HIV positives, ART status and prophylaxis. In CPP, data were collected on number of clinic visits, HIV positives, and children referred to CCR. In CCR, data were collected on the number of exposed HIV infant’s clinic visits, month of first PCR sample, PCR received in health facility, PCR results given to a mother, positive PCR tests, and ART initiation and retention over 3 months. Focus groups discussions (FGDs) were conducted at each site as described below. FGDs were conducted with mothers and health workers by research team members with one note taker [30]. One FGD with mothers (5–8 participants) was completed at each facility (total of six FGDs). With approval and support from health facility staff and leadership, mothers were purposively sampled and recruited through pre-existing mother-to-mother peer support groups organized for HIV-positive women already receiving HIV care and treatment at each facility. A research team member attended a regular group meeting and asked for volunteers to participate in a one-hour FGD. The goal of the FGDs was to capture consensus among the patients about what they experienced as the most important barriers and facilitators to accessing and continuing in care. The pre-existing peer groups provided a sample of mothers who had regularly visited the facility and had become more comfortable speaking about their experience with others they already knew in their groups. The FGDs used a semi-structured interview guide with open-ended questions developed to (1) assess patient experience with ANC, post-partum care, EID, and pediatric ART, (2) identify barriers and facilitators in accessing and navigating care, and (3) solicit suggestions to improve patient follow-up in care [30]. All volunteers were consented, and FGDs took place at the health facilities in private settings. Since the FGDs sought to identify broad consensus on key facilitators and barriers to care among current users, extensive individual demographic data were not requested from these groups. The focus groups were conducted by trained interview teams in Portuguese, with the occasional use of local terms as needed from local languages including Sena, Ndau, and Tewe. Additional FGDs were conducted at each facility (six FGDs in total with 5–8 participants) with MCH nurses, counselors for those services, pharmacists, laboratory technicians, tecnicos, and medical doctors (where available) engaged in EID and pediatric HIV related activities. These FGDs used semi-structured interview guides with open-ended questions on three categories of information to: (1) understand the perceptions of the importance to follow up for the HIV-positive mothers and infants, (2) identify facilitators and barriers to retention, and (3) identify potential strategies to improve retention [30]. Since many participants preferred not to be recorded in both sets of groups, extensive detailed notes were taken by trained note-takers, and then entered into an Excel spreadsheet to analyze and identify key themes focusing on barriers and facilitators to care and follow-up, and recommendations for improvement [31]. For both sets of FGDs, the lead researchers individually coded for themes based on the three categories and then systematically compared codes to identify, discuss, and resolve code discrepancies in the final code lists [30, 31]. The research team also conducted 36 in-depth individual interviews (IDIs) with health workers (6 at 6 sites) to help explain and diagram work and patient flow, and help identify possible bottlenecks [30]. Respondents were drawn from the same health workers who participated in the FGDs including MCH nurses, tecnicos, doctors, and facility directors. In each of the IDIs, one interviewer with a note taker used a semi-structured interview guide to ask respondents to (1) describe their role in the postpartum, EID, and/or ART services process, (2) describe work and patient flow in their specific segment of those services, and (3) describe what bottlenecks and barriers they identify in patient flow and retention in their area of work. IDI’s lasted from 30 to 60 min and were conducted in Portuguese. Extensive notes were taken and used to help develop patient flow diagrams for each facility. Notes were also analyzed to identify key themes concerning bottlenecks and barriers to service delivery as described above for FGDs. All data from qualitative interviews were stored on the lead researchers’ computers in password protected folders following the IRB-approved protocol. The overall process was reviewed for data reporting using the COREQ checklist for qualitative research [32].