Uptake of care and treatment amongst a national cohort of HIV positive infants diagnosed at primary care level, South Africa

BACKGROUND: Loss to follow-up after a positive infant HIV diagnosis negates the potential benefits of robust policies recommending immediate triple antiretroviral therapy initiation in HIV positive infants. Whilst the diagnosis and follow-up of HIV positive infants in urban, specialized settings is easier to institutionalize, there is little information about access to care amongst HIV positive children diagnosed at primary health care clinic level. We sought to understand the characteristics of HIV positive children diagnosed with HIV infection at primary health care level, across all provinces of South Africa, their attendance at study-specific exit interviews and their reported uptake of HIV-related care. The latter could serve as a marker of knowledge, access or disclosure. METHODS: Secondary analysis of data gathered about HIV positive children, participating in an HIV-exposed infant national observational cohort study between October 2012 and September 2014, was undertaken. HIV infected children were identified by total nucleic acid polymerase chain reaction using standardized procedures in a nationally accredited central laboratory. Descriptive analyses were conducted on the HIV positive infant population, who were treated as a case series in this analysis. Data from interviews conducted at baseline (six-weeks post-delivery) and on study exit (the first visit following infant HIV positive diagnosis) were analysed. RESULTS: Of the 2878 HIV exposed infants identified at 6 weeks, 1803 (62.2%), 1709, 1673, 1660, 1680 and 1794 were see at 3, 6, 9, 12, 15 and 18 months respectively. In total, 101 tested HIV positive (67 at 6 weeks, and 34 postnatally). Most (76%) HIV positive infants were born to single mothers with a mean age of 26 years and an education level above grade 7 (76%). Although only 33.7% of pregnancies were planned, 83% of mothers reported receiving antiretroviral drugs to prevent MTCT. Of the 44 mothers with a documented recent CD4 cell count, the median was 346.8 cell/mm3. Four mothers (4.0%) self-reported having had TB. Only 59 (58.4%) HIV positive infants returned for an exit interview after their HIV diagnosis; there were no statistically significant differences in baseline characteristics between HIV positive infants who returned for an exit interview and those who did not. Amongst HIV positive infants who returned for an exit interview, only two HIV positive infants (3.4%) were reportedly receiving triple antiretroviral therapy (ART). If we assume that all HIV positive children who did not return for their exit interview received ART, then ART uptake amongst these HIV positive children < 18 months would be 43.6%. CONCLUSIONS: Early ART uptake amongst children aged 15 months and below was low. This raises questions about timely, early paediatric ART uptake amongst HIV positive children diagnosed in primary health care settings. Qualitative work is needed to understand low and delayed paediatric ART uptake in young children, and more work is needed to measure progress with infant ART initiation at primary care level since 2014. Secondary analysis of data from HIV positive infants was undertaken. The data were drawn from a nationally representative prospective observational cohort study of HIV exposed infants (HEI), followed up between October 2012 and September 2014 to establish 18-month mother to child transmission of HIV (MTCT) and HIV-free survival. This cohort was established from a nationally representative survey conducted at 580 primary health care clinics / community health centres across all nine provinces of South Africa to measure MTCT at 4–8 weeks postpartum. During the survey, multistage probability proportional to size sampling methodology was used to establish a nationally and provincially representative sample. Hospitals were excluded from the sampling frame to reduce the bias posed by sick populations, and because hospitals do not routinely offer immunization; our sample was drawn from 6 week immunization service points, which have > 95% immunisation coverage. All infants receiving their six-week immunization in selected facilities and their mothers/caregivers were recruited, either consecutively (in facilities without long immunization queues) or systematically (in facilities with long immunization queues based on desired sample size for the day and number of eligible mother/infant pairs in the queue). Infant dried blood spot samples were collected from all consented infants (and not only from infants brought by self-reported HIV positive mothers or whose mothers received interventions to prevent MTCT) and tested at the National Institute for Communicable Diseases (NICD), National Health Laboratory Services for HIV antibody to ascertain HIV exposure. Antibody positive infants or those born to self-reported HIV positive mothers were tested for HIV infection using total nucleic acid polymerase chain reaction (TNA PCR). A closed cohort of HIV antibody positive infants (HEI) was established, and all HEI were subsquently seen at 3-monthly intervals from three until 18 months. At each study visit, mothers / primary carrgivers were interviewed and HEI were tested for HIV (TNA PCR). As per national policy, the laboratory returned all infant HIV test results to their clinic of origin within 1 month. Routine clinic (not research) nurses were required to return results to mothers/parents, confirm the diagnosis and initiate paediatric ART. As per national policy, all children with confirmed HIV infection were eligible for ART, regardless of their CD4 cell count. Results at the 6 week time point have been reported elsewhere[15–17]. and details about how the HIV infected population differs from the rest of the cohort are addressed in the overall 18-month MTCT and HIV-free survival paper, which is currently under review. More information about loss to follow-up in the HIV exposed cohort, from which the HIV infected children came, are provided in the paper by Ngandu et.al. in this series. For this paper, given that confirmation of the infant’s HIV status would have been complete within 1 month of the blood-taking visit, we expected that HIV positive infants should have received HIV-related care (including ART) before they were seen at the next visit, which was scheduled 3 months later. The study visit 3 months after an infant HIV positive diagnosis, or as soon as possible to the 3 month post HIV diagnosis time point if the mother could not be contacted at this time point, was the study exit interview for HIV positive infants. This visit aimed to check that the infant was in HIV-related care, as infant HIV diagnosis was the study endpoint. HIV exposed infants enrolled in the follow-up study at 4–8 weeks postpartum, and diagnosed with HIV infection at six-weeks, or at 3, 6, 9, 12 or 15 months were eligible for this analysis. All infant HIV diagnoses were made at primary health care level, as hospitals were excluded from the sampling frame. Maternal and infant baseline information was obtained from the 4–8 week interview. Data on infant birth weight and gestational age was obtained from the infant’s Road to Health booklet (RTHB), a patient-held booklet issued at birth to all infants born in South Africa. Questions on infant medication were asked at every interview, specifically about infant antiretroviral drug use. A food and medication diary was issued to all HEI at the enrolment visit to assist with recall about medicine and feeding. This diary allowed the mother/caregiver to keep a daily log of medicine and food ingested. The data collector reviewed and collected the diary during each interview. All study exit interviews were conducted at primary health care clinics / community health centers until January 2014; thereafter a non-health facility follow-up site, based on mother’s choice, was allowed. If an HIV positive infant missed the scheduled study exit interview they were invited for interviews at all subsequent time points until they attended the study exit interview, or until the infant reached 18 months, whichever came first. Trained research nurses conducted the study exit interview. All study exit interviews were scheduled to coincide with routine child follow-up visits. For mothers who did not return for study exit interviews, data collectors documented the reason for non-return, where possible. Three attempts (telephone calls on three different occasions) were made to contact each mother to return for a study exit interview. Data collectors searched ART clinic records or registers to ascertain whether children with missed study exit interviews were on ART. Data were analysed using STATA version 14. As numbers were small, unweighted descriptive analysis was conducted to explore and understand baseline characteristics of HIV infected infants (at 4–8 weeks postpartum) and uptake of study exit interviews and ART. Chi-square tests were used for categorical variables (Fisher’s exact test, if expected cell count< 5; Cochran-Armitage, for ordinal data with one binary variable) and t-tests or Wilcoxon two-sample test for continuous parametric or non-parametric data, respectively. All p-values are 2-sided. Ethics approval for this secondary analysis was obtained from the University of Pretoria Ethics Committee. The main study was approved by the South African Medical Research Council Ethics Committee and the Centers for Disease Control and Prevention. All HIV positive infants who were not on ART during the study exit interview were referred to routine services for ART initiation. Although visits were synchronized with routine clinic visits, caregiver-infant pairs received R100 (approximately 8USD) per visit, as an inconvenience allowance to cover their travel and opportunity costs.