To The Editor:
Programs for pediatric antiretroviral therapy (ART) traditionally focus on urban tertiary hospitals, but many African children have their first health care encounter at a rural primary health care center. Their access to ART depends on HIV screening and timely referral by primary health care workers (PHCWs). Rural PHCWs, however, are usually among the last to be trained for the prevention of mother-to-child transmission of HIV (PMTCT) and the integrated management of childhood illness (IMCI). Low training coverage rates result in intermittent service provision. These programs pay little attention to HIV screening among siblings of PMTCT babies and regular follow-up of HIV-infected children (older than 5 years) with minor HIV symptoms. Thus, many African children do not have access to HIV care. The urgency of this situation compels one to look for an additional strategy to improve pediatric HIV/AIDS care.
Attention to non-ART HIV care tends to be overshadowed by the attention to ART, but adequate treatment preparation is one of the most important determinants of successful ART.1 The use of a register to monitor HIV-infected patients on ART is part of the public health approach of the World Health Organization, but there are many benefits from enrolling children long before ART is initiated. A regular follow-up of HIV-exposed children from the age of 3 months on facilitates early HIV testing, timely commencement of cotrimoxazole prophylaxis, and ART preparedness. A “safety net,” through the monitoring of a limited set of indicators of vulnerability, can accompany such a care program and provide a timely alert for psychosocial problems that could interfere with the child's health care. Despite the fact that much HIV care could be provided by PHCWs, there has so far been little debate and research into their potential role. An observational study in Chimanimani district (Eastern Zimbabwe) tried to evaluate the role of PHCWs by developing a strategy to improve the identification and care of HIV-exposed and HIV-infected children at primary health care centers and analyzing the most important results of 7 months of pilot testing.
The PHCWS participated in the development of the strategy so that it was considered meaningful, feasible, and not too time-consuming for PHCWs; it included their ideas for further simplifications and gave them ownership of the program. The nurses introduced the name CFU, derived from child follow-up, but used only its abbreviated form to avoid stigmatization. Before the project, there were no data about the district's number of HIV-infected children and the pediatric HIV care provided at primary health care centers (including HIV testing and provision of cotrimoxazole prophylaxis), but the care was believed to be negligible. There were no medical doctors employed in the district. Permission for the pilot project was given by Zimbabwean Ministry of Health authorities.
PHCWs enrolled children ranging in age from birth to 16 years who met at least 1 of the following criteria: being born from a mother enrolled in the PMTCT program, being a sibling of a PMTCT baby, being referred for HIV testing after health education, exhibiting clinical symptoms suggestive of HIV, or when the parent was known to be HIV-positive or exhibited symptoms suspicious for HIV. Consent from the legal guardian for enrollment in the program was crucial and independent from consent for HIV testing. Enrolled children were followed on a monthly basis unless transport problems required longer follow-up periods.2 Children were “officially” discharged from the register only in the case of a referral to a pediatric ART program; a referral to the well-baby clinic (if they tested HIV-negative); or when investigations in the community concluded that the child had died, emigrated, or was lost. The information from each follow-up visit was recorded in the register in a concise and visual manner to limit the administrative burden (Fig. 1). Reminders helped to offer timely vitamin A, the additional measles vaccine, and an HIV test. The serostatus of the child was monthly reassessed, whereby all children were classified according to the following 4 categories: “tested,” “too young for rapid tests” (<18 months), “lack of testing facilities,” or “no consent for testing yet” (in case of a refusal or when the necessary discussions within the family were not yet finalized). Gradually, over time, the project was implemented in 15 sites: 5 operational testing sites where PMTCT was provided, 6 sites with trained staff that became operational PMTCT sites only after the pilot study, and 4 sites that were not yet included in the national PMTCT program. Each clinic completed a progress report on a monthly basis. This provided a timely alert for the need for a follow-up visit and is in line with other national systems for monitoring and evaluation. Data from the monthly progress reports were compared with the corresponding registers and entered into spreadsheets for a descriptive analysis.
Between October 5, 2005 and April 25, 2006, the strategy allowed initiating HIV care for 112 children. One third of these children were enrolled before the age of 6 months. IMCI training was needed for the identification of symptomatic HIV-infected children. The enrollment of asymptomatic PMTCT babies and their siblings, however, does not require training because it can be done by routinely inquiring about the PMTCT history of the child (eg, on the occasion of a postpartum visit or immunization). Thus, even PHCWs who were not yet trained in PMTCT and IMCI could identify HIV-affected children and apply this strategy.
After 7 months of pilot testing, 4 children had migrated to another catchment area, 7 (6%) had died, and 2 were discharged to the well-baby clinic. Although there was no official loss to follow-up, 8 children (7%) had not been seen for 2 or more consecutive monthly visits. The selected indicators of vulnerability were: being a maternal orphan (16 children), being brought to the clinic by 3 or more different caregivers on separate occasions (1 child), and gaps in compliance with clinic appointments. There was anecdotal evidence of successful investigations in the community resulting in nonattendees resuming follow-up, suggesting that some kind of safety net had been established. Thus, 91 (81%) of the registered children were still on monthly follow-up by April 2006.
By the end of the study, there was documented evidence that 92% of the followed children had received their cotrimoxazole supply on a monthly basis. PHCWs reported that the register facilitated the procurement of adequate cotrimoxazole supplies. The study results cannot determine how many children actually ingested the cotrimoxazole daily but suggest that the implementation of cotrimoxazole guidelines improved. Although several of the suggestions expressed by Zachariah et al3 seem to be confirmed, the CFU strategy is more integrated, and may thus allow better optimal use of resources allocated to scaling up cotrimoxazole prophylaxis.
Because the project aimed at sustainable improvement of the implementation of national HIV guidelines within the resources of the Ministry of Health, infant diagnostic tests were not introduced at this stage, but the use of rapid tests in older children was promoted. Within 7 months, 34% of all registered children older than 18 months had been tested for HIV, 21% still lacked consent for testing, and 45% were not tested because of the lack of testing facilities. Among the children enrolled at the 5 operational testing sites, 61% were tested despite gaps in the provision of rapid tests through the national PMTCT program (Fig. 2). The findings emphasize how pediatric HIV screening can also substantially be increased through logistic support and further decentralization of PMTCT programs.
The data allowed assessing the impact of the national PMTCT program on this group of children: a complete PMTCT regimen (nevirapine given to the mother and the infant) was given to 25 (22%) of the enrolled children. This low figure is explained by the delay in the national PMTCT roll-out to this district and by the fact that not only PMTCT babies but also older children were enrolled (eg, through IMCI). For 77 children, no PMTCT was done or the guardian was unable to provide this information. Ten children received an incomplete regimen. In 7 of these 10 cases, the nevirapine was given to the baby but not to the mother. Further investigations should clarify whether this phenomenon is caused by mothers' simply forgetting to take nevirapine, late enrollment in the PMTCT program as a result of home deliveries (BCG immunizations possibly functioning as a safety net for PMTCT), or resistance against counseling mothers during labor. HIV transmission rates among babies with incomplete PMTCT regimens have not been equally studied but may plead for an alternative PMTCT regimen. Also, lost opportunities for PMTCT in mothers who seroconverted after a negative routine HIV test result early in pregnancy become apparent only through an analysis of HIV test results in infants.4 Thus, more attention to the evaluation of PMTCT interventions from the side of the infant is urgently needed.
The number of enrolled children in need of ART cannot be obtained from the registers and could only be estimated by the percent of malnourished children (29%). PHCWs can diagnose malnutrition but cannot necessarily differentiate HIV-related malnutrition from other types of malnutrition. More research is necessary to investigate how PHCWs with limited training in pediatric HIV clinical staging and little or no access to CD4 cell counts could most effectively identify children who need referral to ART clinics, where the demand for pediatric ART usually exceeds its availability. Low weight for age could possibly be such a simple indicator.5
Despite the many limitations and short duration of this pilot project, the results demonstrate that it is possible to achieve greater involvement of PHCWs in pediatric HIV/AIDS care than is currently achieved and emphasize the critical need for more debate about indicators of primary pediatric HIV care in rural Africa and how the follow-up at primary health care centers can be better used for improved non-ART HIV care and ART preparedness in children.
The authors thank the Ministry of Health and Child Welfare, especially Dr. A. Mahomva and Dr. R. Mudyaradima, for facilitating this study and the staff of the French Ministry of Foreign Affairs, the French Embassy at Harare, the French Agency for Development Aid (AFD), and the University of Bordeaux. They especially thank Dr. H. Mujuru, Prof. F. Dabis, Dr. F. Perez, D. Marchand, M. Baherle, B. Shumba, R. Jana, the district health teams and nurses of Chimanimani and rural Mutare, the traditional chiefs of Chimanimani, the staff from Chimanimani hotel, Dr. W. Schrooten, and M. Van Kesteren.
Anniek J. De Baets, MD, MPH, Dip. HIV Man.*†
Sibongile Sifovo, RN‡
Isidore E. Pazvakavambwa, MD§
*FSP/ISPED Zimbabwe Université Victor Segalen Bordeaux, France
†Department of Pediatrics University of Antwerp Antwerp, Belgium
‡District Health Team Chimanimani District Zimbabwe
§Department of Pediatrics and Child Health University of Zimbabwe College of Health Sciences Parirenyatwa Hospital Harare, Zimbabwe