Child mortality remains a major problem in developing settings, and the world remains far from accomplishing the 2005 Millennium Development Goals for reduction of child mortality.1 HIV contributes importantly to child mortality, as infants born to mothers with HIV experience higher morbidity and mortality than those born to HIV-negative mothers.2,3 The adverse impact of HIV is compounded by traditional risk factors for child mortality, including poor maternal health, low socioeconomic status, and lack of accessibility of primary care services.4,5 Given the impact of HIV on maternal and child survival, the international community has recently stepped up calls for the elimination of mother-to-child transmission by the year 2015.6
Despite expanded efforts to scale-up services to prevent mother-to-child transmission of HIV, estimates from sub-Saharan Africa in 2010 underscore the need for prioritization of high-quality programs: only 43% of women received HIV testing, 58% of HIV-infected women received prophylactic antiretroviral (ARV) medications, and 40% of infants born to mothers with HIV received prophylactic ARVs.7 That same year, globally, only 15% of infants born to HIV-infected women were tested for HIV within 2 months of life. Effective implementation of prevention of mother-to-child transmission (PMTCT) programs has been challenged by weak and uncoordinated referral networks among HIV, maternity, and child health services; lack of adequate laboratory and diagnostic supplies; slow expansion and adaptation of ARV therapy; poor support for nutrition and infant feeding practices; failure to address socioeconomic disparities; weak monitoring and evaluation systems; and continued stigma and discrimination at the community level.7–9 PMTCT programs have been particularly ineffective in rural areas with limited diagnostic and programmatic resources.10
In 2005, in conjunction with the Rwandan Ministry of Health, Partners In Health supported and implemented a comprehensive child survival program in rural Rwanda with special emphasis on infants born to HIV-infected women. This program consisted of PMTCT services integrated with a comprehensive basic package of health services to promote child survival, including access to clean water. High HIV-free survival among a small cohort of early enrollees has been reported.11 Since that time, the program was scaled-up in 8 health clinics across eastern Rwanda, which serves a total catchment area of 480,000 people. We report programmatic and clinical outcomes among HIV-exposed infants enrolled in the program during this period of scale-up and examine sociodemographic predictors of program retention.
MATERIALS AND METHODS
Rwanda, with a HIV prevalence among pregnant women of 2.4% and overall HIV prevalence of 3%, ranks among the 25 low- and middle-income countries with the highest burden of HIV disease in women and children.7,12 More than 80% of the population lives in rural areas.12 At the beginning of the study period, the rural Eastern Province reported the highest infant and under 5 mortality rates in the country with 84 deaths/1000 live births and 174 deaths/1000 live births, respectively (compared with 80/1000 and 127/1000, respectively, across the sub-Saharan African region) and the highest total fertility rate of 5.8 per woman (compared with 5.2 across the sub-Saharan Africa region).13–15 Recent data suggest substantial and continual nationwide improvements in infant and child mortality rates over the last decade.12
Infants born to women with HIV who were enrolled in the child survival program at 8 health clinics at or before birth from March 1, 2007, until February 28, 2010, and elected to replacement feed were included for analysis. The option to replacement feed was provided to all HIV-infected women.
Elements of the Child Survival Program
Key components of the program are listed in Figure 1. Briefly, the child survival program was developed within existing public health facilities and executed by facility directors, nurses, and social workers employed by the Rwandan Ministry of Health with additional training and mentorship provided by Partners In Health. The program included routine infant vaccination, mebendazole and vitamin A administration, screening and treatment for malnutrition, tuberculosis, HIV, and a comprehensive malaria prevention and treatment program. All services were offered free of charge and included active case finding, triage, treatment, and home-based directly observed therapy by community health workers (CHWs) for mothers and infants on ARVs. CHWs were chosen by patients and were trained by Partners In Health to administer daily therapy, identify medication side effects and opportunistic infections, and offer psychosocial support, as previously described.16 CHWs received a monthly financial incentive.
Additionally, HIV-infected mothers and their infants were given nutritional supplementation, infant formula, and supplies to facilitate clean water preparation.11,17 During biweekly visits to receive supplies, nurses regularly trained mothers on clean water and formula preparation, hygiene, family planning, recognition of diarrheal disease and dehydration, and other health topics and infants had regular physical examinations. CHWs conducted routine home visits to address problems and support formula preparation. Social workers visited mother-infant pairs who missed more than 1 distribution and closely monitored losses-to-follow-up and mortality. Adherence to breastfeeding was routinely assessed by self-report, physical examination for evidence of lactation, or infant affinity for breastfeeding.
In accordance with national protocols, women received ARV medications to prevent vertical HIV transmission (Fig. 1). Infants received cotrimoxazole and were tested for HIV after 6 weeks of age using polymerase chain reaction testing of dried blood spot samples at the Rwanda National Reference Laboratory.
We extracted routinely collected clinical and program data from the program database. Program intake forms were modified in January 2008 to include additional sociodemographic variables and information on infant ARV prophylaxis. As part of routine program evaluation, the prevalence and management of diarrheal disease and family planning uptake were assessed over a 6-month period in 2008.
Infant deaths were confirmed by a program social worker. An infant was considered to be lost-to-follow-up (LTFU) from the program if the mother–infant pair missed 2 consecutive distribution visits and could not be located by the clinician, social worker, or CHW by home visit or phone contact. An infant was considered retained unless s/he died, became LTFU, or was suspended from the program due to evidence of mixed feeding. Two-week prevalence of diarrhea was recalled by mothers, with diarrhea defined as 3 or more watery stools in a 24-hour period.
We calculated descriptive statistics for programmatic indicators and Kaplan–Meier 18-month infant survival and retention probabilities. Infants were censored at program graduation at 18 months or on their last visit date if they exited the program for a reason other than death, LTFU, or suspension. Among the 634 for whom sociodemographic data were collected, we used the Cox regression models to examine whether maternal CD4 cell count, infant gender, or any of the sociodemographic characteristics reported in Table 1 predicted retention at 18 months. Variables associated with retention at a P value <0.20 in univariable analysis were considered candidates for the final multivariable model and retained at a P value <0.05. This project was approved by Partners Human Research Committee, Boston, Mass, and Rwanda National Ethics Committee, Kigali, Rwanda.
Overall, 1038 infants were newly enrolled during the study period. Of these, 634 (61.1%) were enrolled during 2008 or later and therefore had extended demographic data collected as part of routine care. These characteristics are reported in Table 1. Gender was available for nearly all infants (N = 1008; 97.1%) and 49.2% were female (n = 496). Median maternal CD4 cell count was 546.5 (interquartile range: 261–768; N = 918). During the expansion and scale-up phase of the program from 2007 to mid-2008, there was an approximately 4-fold increase in the number of infants participating in the program at a given time (Fig. 2).
With respect to HIV testing, 92.5% of infants were HIV tested within 120 days of birth and 94.9% were tested by the end of the program. Of the 53 infants who were not tested, 23 (43.4%) died, were LTFU, or moved from the catchment area before initial HIV tests were conducted at 6 weeks. Of the remaining 30 infants who were not tested, most exited the program prematurely: 14 died, 6 were LTFU, 4 moved, and 5 were suspended. Only 1 infant graduated the program without being tested for HIV. Information on place of delivery and perinatal ARV prophylaxis was available for approximately 70% of the 634 women for whom these data were collected. Of these, 93.9% (426/458) of the mothers delivered in a hospital setting, 98.6% (442/451) of infants received single-dose nevirapine at birth, and 98.0% (430/436) received zidovudine prophylaxis. Of the 370 mothers enrolled in 2008 when the assessment on diarrhea treatment and contraception uptake was conducted, 74 (20.7%) reported that their infant had experienced diarrhea in the past 14 days. Of infants with diarrhea, 74.3% were managed at a health facility and 37.8% received treatment with oral rehydration solution. Of the mothers interviewed, 50.8% reported using a modern method of family planning (Table 2).
At 18 months, 867 (83.5%) children had graduated from the program, 30 (2.9%) were LTFU, 74 (7.1%) had died, 28 (2.7%) had moved from the catchment area, 22 (2.1%) graduated early due to the birth of a sibling, 8 (0.8%) were suspended due to evidence of mixed feeding, and 9 (0.9%) were suspended for other reasons, such as selling formula. The 18-month survival probability and retention probability were 0.93 (95% confidence interval: 0.91 to 0.94; Fig. 3A) and 0.88 (95% confidence interval: 0.86 to 0.90; Fig 3B), respectively. The proportion of infants retained was similar among infants for whom sociodemographic data were and were not collected (558/634; 88% and 359/404; 89%, respectively). Twenty-seven children (2.6%) tested positive for HIV, 26 of whom were detected before 120 days. Of these, 1 infant died and none were LTFU from the program by 18 months of age. Five-hundred children who tested negative for HIV before 120 days had a subsequent HIV test after 9 months of age, and all results were negative.
Predictors of Program Retention
Among the 634 women with sociodemographic data, we did not identify any statistically significant predictors of retention at a P value <0.20 in univariable analyses (see Table S1, Supplemental Digital Content,http://links.lww.com/QAI/A374). We therefore did not conduct multivariable analyses.
Relative to HIV-unexposed infants, HIV-exposed infants have been shown to be 2 to 4 times more likely to die, even in the absence of HIV infection.18,19 The HIV-exposed infants in the present cohort survived at a rate that was comparable to the most recently reported rate for infants in eastern Rwanda (infant mortality rate: 63/1000 live births).12 The survival probability reported here is similar to that which we reported among a small cohort of early enrollees in this program, suggesting that the effectiveness of the program was sustained throughout the scale-up phase.11 The survival probability is also comparable to that previously reported in another closely monitored cohort in Rwanda and by a similar programmatic intervention in rural Haiti.20,21 We attribute the high child survival and retention in this program to the multifaceted package of services, which included improved access to health care, successful PMTCT, clean water sources, home visits by CHWs, prevention and treatment of major childhood illness, nutritional support, and social and economic support for the extremely vulnerable. These program elements have been previously identified as key components of HIV-free child survival programs and were provided without interruption from the initiation of maternal prenatal services though the infant development period.5,22,23
Utilization and quality of services provided by this child survival program were high: nearly all infants received HIV testing and ARV prophylaxis and approximately half of mothers used family planning services. By comparison, the Rwandan Interim Demographic Health Survey published in 2009 reported that only 26.3% of married women used some form of family planning in the Eastern Province.13 The incidence of diarrheal disease, a leading cause of childhood mortality, has previously been reported as higher in nonbreastfed infants.24,25 In this cohort of nonbreastfed infants, the 14-day incidence of diarrheal disease (20.6%) seemed comparable to national survey data for children in this age group (<6 months: 5.6%, 6–11 months: 23.1%, and 12–23 months: 22%).13 Furthermore, the percentage of mothers in this program that reported seeking care from a formal health provider for infant diarrhea was higher than the percentage reported for Eastern Rwanda (74.3% versus 29.4%, respectively).13 We attribute these findings to improved access to clean water sources, home visits by CHWs, and anticipatory guidance for parents regarding signs, symptoms, and management of diarrheal illness.
Previously reported predictors for infant mortality or loss-to-follow-up such as indicators of poverty and poor maternal health status did not seem to be risk factors in our study. We believe that ensuring free or reduced cost health care and the utilization of CHWs greatly improved access to care for mothers and their infants and mitigated the impact of traditional obstacles to care for children in marginalized rural communities.26,27 Given that in 2005, 83% of women in rural Rwanda reported difficulty accessing health care services and 73% reported lack of money for treatment as a major obstacle to care, the reduction or elimination of fees is an instrumental component of clinical programming for child survival in rural resource poor settings and has been recently demonstrated in Rwanda.28,29
In contrast to outcome reports of HIV-exposed children in a research context, this study represents an evaluation of the effectiveness of a large-scale, long-term programmatic intervention conducted within the public sector of a rural under-resourced health care system. By relying on routinely collected clinical data, which changed over time, there were incomplete data for some variables. For example, collection of sociodemographic data did not commence until 2008; therefore, early enrollees lacked these data. If mother–infant pairs that enrolled before that time were different than those who enrolled later, the data would not be missing completely at random, and the descriptive statistics we report here may not be representative of the entire cohort. Similarly, maternal contraception and infant diarrhea data were only available for pairs enrolled at the time of the assessments in 2008. We believe that the uptake of contraception and diarrheal treatment reported among this subset of pairs is representative of that which would have been observed had assessed the entire study population; however, it is not possible for us to verify this assumption. Data on ARV prophylaxis were missing for about 30% of infants. We believe that this is because clinical charts may not always have been accessible when program staff completed the intake form. If infants missing data on ARV prophylaxis were less likely to have received it, our reported uptake of prophylaxis would be an overestimate. However, the extremely low incidence of HIV infection among the children in this cohort suggests that this is not the case. Finally, although initial HIV testing within the first few months of life was nearly complete, follow-up testing was less so, precluding our ability to calculate HIV-free survival throughout the 18-month follow-up period. Nonetheless, not a single transmission was observed among the 500 infants who did have a follow-up test, an expected result given that infants were replacement fed with intensive adherence support and monitoring.
In conclusion, our findings demonstrate that a PMTCT program with integrated comprehensive interventions to maximize HIV-free survival can produce high rates of service utilization, program retention, and patient survival in a vulnerable population of HIV-exposed infants in a rural resource-limited setting, even during a period of rapid program growth. These findings are likely generalizable to similar programs using effective PMTCT interventions, including ARV prophylaxis throughout the breastfeeding period for PMTCT. It is critical that PMTCT programs incorporate key elements of integrated and comprehensive care to improve child survival in rural resource poor settings and achieve virtual elimination of mother-to-child transmission within Rwanda and beyond.30,31
The authors would like to acknowledge Sidney Atwood for statistical and graphic assistance with this manuscript.
2. Kurewa EN, Gumbo FZ, Munjoma MW, et al.. Effect of maternal HIV status on infant mortality
: evidence from a 9-month follow-up of mothers and their infants in Zimbabwe. J Perinatol. 2010;30:88–92.
3. Hong R. Association of maternal HIV infection with increase of infant mortality
in Malawi. J Paediatr Child Health. 2008;44:291–296.
4. Villamor E, Misegades L, Fataki MR, et al.. Child mortality in relation to HIV infection, nutritional status, and socio-economic background. Int J Epidemiol. 2005;34:61–68.
5. Rollins NC, Becquet R, Bland RM, et al.. Infant feeding, HIV transmission and mortality at 18 months: the need for appropriate choices by mothers and prioritization within programmes. AIDS. 2008;22:2349–2357.
8. De Cock KM, Fowler MG, Mercier E, et al.. Prevention of mother-to-child HIV transmission in resource-poor countries. JAMA. 2000;283:1175–1182.
9. Stringer JS, Sinkala M, Maclean CC, et al.. Effectiveness of a city-wide program to prevent mother-to-child HIV transmission in Lusaka, Zambia. AIDS. 2005;19:1309–1315.
10. Perez F, Orne-Gliemann J, Mukotekwa T, et al.. Prevention of mother to child transmission of HIV: evaluation of a pilot programme in a district hospital in rural Zimbabwe. BMJ. 2004;329:1147–1150.
11. Franke MF, Stulac SN, Rugira IH, et al.. High human immunodeficiency virus-free survival of infants born to human immunodeficiency virus-positive mothers in an integrated program to decrease child mortality in rural Rwanda
. Pediatr Infect Dis J. 2011;30:614–616.
12. Ministry of Health of Rwanda
, National Institute of Statistics of Rwanda
and IFC Macro. Rwanda
Demographic and Health Survey 2010. Calverton, Maryland: MOH, NISR, and IFC Macro; 2011.
13. Ministry of Health of Rwanda
, National Institute of Statistics of Rwanda
and ORC Macro. Rwanda
Interim Demographic and Health Survey 2007-08. Calverton, Maryland: MOH, NISR, and ORC Macro; 2009.
15. Population Reference Bureau. 2010 World Population Data Sheet. Washington, DC: Population Reference Bureau; 2010. Available at: http://www.prb.org/pdf10/10wpds_eng.pdf
. Accessed October 14, 2012.
16. Rich ML, Miller AC, Niyigena P, et al.. Excellent clinical outcomes and high retention in care among adults in a community-based HIV treatment program in rural Rwanda
. J Acquir Immune Defic Syndr. 2012;59:e35–e42.
17. Lim Y, Kim JY, Rich M, et al.. Improving prevention of mother-to-child transmission
of HIV care and related services in eastern Rwanda
. PLoS Med. 2010;7:e1000302.
18. Shapiro RL, Lockman S, Kim S, et al.. Infant morbidity, mortality, and breast milk immunologic profiles among breast-feeding HIV-infected and HIV-uninfected women in Botswana. J Infect Dis. 2007;196:562–569.
19. Marinda E, Humphrey JH, Iliff PJ, et al.. Child mortality according to maternal and infant HIV status in Zimbabwe. Pediatr Infect Dis J. 2007;26:519–526.
20. Peltier CA, Ndayisaba GF, Lepage P, et al.. Breastfeeding with maternal antiretroviral therapy or formula feeding to prevent HIV postnatal mother-to-child transmission in Rwanda
. AIDS. 2009;23:2415–2423.
21. Ivers LC, Appleton SC, Wang B, et al.. HIV-free survival and morbidity among formula-fed infants in a prevention of mother-to-child transmission
of HIV program in rural Haiti. AIDS Res Ther. 2011;8:37.
22. Becquet R, Bequet L, Ekouevi DK, et al.. Two-year morbidity-mortality and alternatives to prolonged breast-feeding among children born to HIV-infected mothers in Cote d'Ivoire. PLoS Med. 2007;4:e17.
23. Kerber KJ, de Graft-Johnson JE, Bhutta ZA, et al.. Continuum of care for maternal, newborn, and child health: from slogan to service delivery. Lancet. 2007;370:1358–1369.
24. Creek TL, Kim A, Lu L, et al.. Hospitalization and mortality among primarily nonbreastfed children during a large outbreak of diarrhea and malnutrition in Botswana, 2006. J Acquir Immune Defic Syndr. 2010;53:14–19.
25. WHO Collaborative Study Team on the Role of Breastfeeding on the Prevention of Infant Mortality
. Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: a pooled analysis. Lancet. 2000;355:451–455.
26. Mukherjee JS, Eustache FE. Community health workers as a cornerstone for integrating HIV and primary healthcare. AIDS Care. 2007;19:S73–S82.
27. Mukerhee JS, Farmer PE, Ivers LC, et al.. The Intersection of Access and Adherence-Improving Retention in HIV Programs in Resource Poor Settings. Presented at: XVII International AIDS Conference; August 3–8, 2008; Mexico City, Mexico. Abstract MOPE0146.
28. Ministry of Health of Rwanda
, National Institute of Statistics of Rwanda
, and ORC Macro. Rwanda
Demographic and Health Survey 2005. Calverton, Maryland: MOH, NISR, and ORC Macro; 2006.
29. Dhillon RS, Bonds MH, Fraden M, et al.. The impact of reducing financial barriers on utilisation of a primary health care facility in Rwanda
. Glob Public Health. 2012;7:71–86.
30. Becquet R, Ekouevi DK, Arrive E, et al.. Universal antiretroviral therapy for pregnant and breast-feeding HIV-1–infected women: towards the elimination of mother-to-child transmission of HIV-1 in resource-limited settings. Clin Infect Dis. 2009;49:1936–1945.
31. Ruton H, Mugwaneza P, Shema N, et al.. HIV-free survival among nine- to 24-month-old children born to HIV-positive mothers in the Rwandan national PMTCT programme: a community-based household survey. J Int AIDS Soc. 2012;15:4.