Treat

This part of the framework includes diagnosis, screening, triage and treatment. Our review includes five randomized controlled studies (RCTs) that all showed operationally important and statistically significant reductions in child mortality as a result of community health worker (CHW) treatment of pneumonia with antibiotics – reductions in the range of 13% to 60% [S1–5]. Throughout this article we will be referencing assessments from our database with numbers in brackets, preceded by an S prefix, to distinguish them from the references cited in the list of references at the end of this article. The number in brackets with an S prefix refers to the number of the assessment in Appendix S2 in Online Supplementary Document(Online Supplementary Document) . Many other assessments – mainly non–randomized controlled, uncontrolled and case–control studies – also observed significant operationally important decreases in pneumonia–specific mortality in children aged less than 5 years, ranging from 28% to 69% [S6–11]. Two other RCTs demonstrated that CHWs can decrease the clinical severity of pneumonia significantly by treating respiratory infections at the community level through implementing good–quality case management [S12, S13]. Over 20 other studies showed decreases in child pneumonia–specific incidence or mortality but as their pneumonia case management was part of Integrated Management of Childhood Illnesses (IMCI) or Primary Health Care (PHC), they will be discussed under those sections below. Co–trimoxazole was the antibiotic most commonly used by CHWs in these studies.

Protect

Under this component are good health practices from birth: exclusive breastfeeding during the first six months of life, adequate complementary feeding, and vitamin A supplementation. Several RCTs demonstrated the efficacy of community–based vitamin A supplementation in reducing pneumonia mortality. In one, vitamin A supplementation decreased pneumonia–specific child mortality by 26% [S14]. In another study, the incidence of pneumonia was decreased through vitamin A supplementation by 44% [S15]. Zinc supplementation and promotion of hand washing provided by CHWs were each also found to significantly decrease the incidence of both pneumonia and diarrhea [S16, S17]. In one randomized controlled trial assessment, a community–based integrated nutrition program apparently not including vitamin A or zinc supplementation demonstrated a decreased incidence of pneumonia [S18]. Studies of vitamin A and zinc supplementation will be presented in more detail under the nutrition section below. Further studies have demonstrated the strong efficacy of zinc supplementation in reducing the incidence, severity and/or duration of diarrheal episodes in children [S19–24].

Prevent

This component includes vaccinations, hand washing with soap, safe drinking water and sanitation, reducing household air pollution, HIV prevention and co–trimoxazole prophylaxis for HIV–infected and HIV–exposed children. Education of community members about diarrheal disease was a common activity carried out by trained CHWs, usually by visiting households or meeting with community groups. Randomized controlled trials found that community education focused specifically on the importance of proper disposal of animal feces from living areas produced decreases in the incidence of childhood diarrhea [S25, S26]. Randomized controlled trial assessments of education of caregivers about hand washing along with the provision of soap also decreased childhood diarrhea to an even greater degree than those mentioned in the previous sentence [S27–31]. Teaching mothers to use oral rehydration solution at home along with education about good household sanitation practices – whether by nurses working at the community level or by CHWs – was also effective [S32–34].

Purification of water within the household with sodium hypochlorite or another locally produced purifying agent was found effective in reducing childhood diarrhea in several studies [S35–38]. Solar sterilization of water was demonstrated as an effective approach to decrease the incidence of childhood diarrhea [S39–41]. Water filters such as BioSand and Lifestraw Family Filter that remove particulate matter were similarly effective in reducing the E. coli concentration in water and decreasing episodes of diarrhea [S42, S43]. The efficacy of community–based interventions concerning immunizations, HIV and nutrition are presented later in the respective sections.

Promotion and uptake through CHWs or others in routine systems

Community–based interventions involving CHWs reaching to the household level to promote participation in immunization activities and CHWs mobilizing communities have played a key role in producing high rates of population coverage for immunizations throughout the world. Peer education provided by CHWs visiting households, by community members recruited just for this purpose, by female community health education workers, and by members of mobile health teams coming from health facilities have contributed to greatly increased immunization coverage rates for children [S119–123].

Establishment of village networks of trained traditional birth attendants and female CHWs was effective. These CHWs promoted immunizations, use of health facilities, and household diarrhea management with oral rehydration solution (ORS) and also carried out growth monitoring of children. Their activities led to a 150% increase in the coverage of 12–23 month–old children with full immunization [S124].

Village–level approaches to community mobilization

Promotion of community participation through education of village leaders, teachers, and extension workers (who in turn educated community members) was found effective, increasing full immunization completion coverage levels by 50% [S125]. Mass media using TV, radio, newspapers and leaflets, distributed and explained by community–level workers, significantly increased community awareness about immunizations with mothers. Those mothers who had increased awareness were much more likely to take their children to be vaccinated [S126]. In Lao PDR, community–based workshops promoting attendance for vaccination significantly increased all childhood vaccinations [S127].

Promotion of immunizations through microcredit programs

A case–controlled study of community health education campaigns associated with microcredit programs were found to greatly increase fully immunization coverage [S128].

Health Days

National Immunization Days, in which community mobilization and immunization at peripheral service points followed up by immunization at the household for those who did not come to the service point led to significant decreases in the incidence of acute flaccid paralysis [S129]. Annual vaccination weeks with household visits by CHWs increased vaccination completion rates from 30% to 53% [S130].

Household vaccination strategies

A case–controlled study of peer education provided by CHWs visiting households, promoting community involvement, and providing immunizations, vitamin A supplementation and growth monitoring led to not only to greatly increased immunization coverage but also to a 58% decrease in under–5 mortality compared to controls [S131]. House–to–house administration of polio vaccine significantly increased polio vaccination rates [S132].

Protein–energy undernutrition

Table 5 presents the findings from randomized controlled and non–randomized controlled studies with statistically significant and operationally large effects compared to controls with protein–energy undernutrition. Table 5 demonstrates that undernutrition can be addressed successfully at the community level through health education involving CHWs visiting households, regular monitoring of child growth in the community, and supplementation with ready–to–use therapeutic food (RUTF). Albendazole supplementation to mothers also was found to have an important effect on child growth. Even for depressed mothers with HIV, well–organized programs improved the nutrition of their children. Group learning programs associated with small loans (that may have enabled mothers to obtain more nutritious foods for their children) also improved child nutrition. Many other integrated programs were also demonstrated to contribute to good child nutrition. These will be covered below in the final section on integrated programs.

Other controlled interventions with smaller effect sizes and statistically significant results at the P < 0.05 level also were very informative. In Vietnam, among children aged less than 15 months with a weight–for–age Z score of <–2, the Hearth approach along with de–worming significantly improved growth when compared to controls who received only deworming [S143]. The Hearth approach is a process of identifying local “positive deviant” women who have well–nourished children. Mothers of malnourished children are also identified and they are guided through a process of learning how positive deviants care and feed their children and applying this knowledge in the care of their own children through hands–on cooking sessions using locally available foods [S143].

In a non–randomized controlled project that was implemented over a five–year period, the hypothesis was tested that younger siblings of older children with severe undernutrition whose undernutrition had been overcome using the Hearth approach should have better nutrition than similar children whose mothers had not been exposed to the Hearth program. Outcomes were compared for 10 different 3–month age groupings of younger siblings (6–8, 9–11, 12–14, etc.). For younger siblings whose older sibling had been severely malnourished and whose mother had been exposed to the Hearth approach (the intervention group), the younger sibling mean weight for age Z score was always higher than the older sibling (P = 0.005 or less in all age groups). For the control group (children with an older sibling who had been moderately malnourished, mildly malnourished, or of normal weight and whose mother had not been exposed to the Hearth program) the same comparison with younger siblings was carried out. The mean weight for age Z score of the younger siblings was always lower for mildly malnourished and normal weight children than their older sibling (P < 0.05 for all but one age group, 6–8 months). This study provides evidence regarding the wider family effects of nutritional education [S144]. However any conclusions need to be guarded due to the limited size of the populations studied. These results would need to be repeated in further similar studies.

The benefits of promotion of agriculture and voucher programs on childhood nutrition have also been demonstrated. In a population including 130 000 children younger than 5 years of age in Nepal, promotion of increased household production of food through training Village Model Farmers, and subsequently village women, led over a 2–year period to a decrease of 10% in the prevalence of underweight in children aged 0 to 4 years [S145]. A community development and livestock promotion project in Nepal for 307 children produced similar results. Although the results in the latter study were not statistically significant after 2 years, the intervention group was more likely to have indoor access to water, treat their water and have a latrine. Longer participation in the program was strongly associated with a better mean height–for–age score (P < 0.00001) [S146].

Giving vouchers to mothers along with health education and a community household health package was found not to result in statistically significant improved child nutrition in the short term but if the program for those children was extended for 2 years more until the children were aged 8 to 10 years, then the mean height for weight Z scores of these children increased by 23% (P = 0.029) compared to controls of the same age [S147]. Other studies demonstrated a statistically significant association of mothers receiving vouchers with greater use of nutrition monitoring at the community level and improved nutrition of their children [S148–151].

Breastfeeding and complementary feeding

Exclusive breastfeeding (BF) during the first 6 months of age with continued BF through the first two years of life is an important contributor to good childhood nutrition, reduced morbidity, and improved mortality in resource–constrained settings. Promotion of exclusive breastfeeding for the first 6 months of life has been estimated to be one of the most effective preventive strategy for saving the lives of young children in low–income settings [9]. Complementary feeding (CF) to supplement breastfeeding is needed from 6 months of age onwards for children to sustain normal growth. Findings from randomized and non–randomized controlled community–based assessments included in our review are presented in Table 6 .

The data from Table 6 indicate that exclusive breastfeeding can be effectively promoted at the community level by CHWs, by trained home peer counsellors, by community outreach health professionals from the nearest health facility, and by mothers’ community health clubs. Of note is that the strongest effects were found when the CHWs and home peer counselors rather than more highly trained health professionals reaching out from local health facilities were doing the education. Education about complementary feeding was found to produce statistically significant improvements in mean height and weight. The Hearth approach mentioned in the section on protein energy undernutrition was also found to be effective in undernourished children younger than 15 months of age, in the study cited in Table 5 [S141], and in other studies with similar results [S143, S159, S160].

Micronutrient supplementation

Types of micronutrient supplementation that were included in projects whose assessments qualified for our review included vitamin A, zinc, iron and multivitamins. Table 7 contains details about randomized and non–randomized controlled studies that have been included in this review and that have operationally important effects. Table 7 shows that vitamin A supplementation provided at the household level to mothers, to newborns, and especially to children 6–59 months of age leads to decreased child mortality. Even fortifying market monosodium glutamate with vitamin A leads to a decrease in the rate of xerophthalmia (a condition of eye dryness and eventual scarring produced by vitamin A deficiency) and all–cause child mortality. It also decreases child mortality from pneumonia and measles.

Daily zinc supplementation decreased all–cause mortality in children 12–48 months of age, but not to the same extent as vitamin A. A decrease in the incidence of diarrhea in children receiving zinc has also been demonstrated in other controlled studies [S174, S175]. Of particular note is that in one study of children 1 to <6 months of age in a malaria–prone area, the risk of death or severe morbidity increased significantly in those who received iron supplementation [S176]. While other studies in non–malaria–endemic areas confirmed the value of iron supplementation for treating anemia, this finding provides reason for caution in providing iron supplementation to children aged 1 to <6 months of age in malaria–endemic areas.

Integrated Management of Childhood Illness (IMCI) and Integrated Community Case Management (iCCM)

Integrated Management of Childhood Illness (IMCI) integrates the prevention and treatment of all childhood illness at health facilities. Its community component, called Community IMCI (or C–IMCI), usually consists of preventive activities and early recognition of potentially serious acute illness that can be performed in the community by trained CHWs going door–to–door and meeting with groups, usually without treatment of illnesses other than ORS for diarrhea. CHWs are taught to recognize children with danger signs and refer or even escort patients to the nearest health facility for treatment. CHWs also facilitate outreach activities from the local health center such as immunizations.

Integrated Community Case Management (iCCM) enables CHWs to diagnose and treat serious acute illnesses of childhood (acute respiratory infection, diarrhea, malaria and in some cases acute malnutrition).

For iCCM to be effective, CHWs need to be well–trained, to have the confidence and support of their community, to be well–linked to their local health facility staff for referral of patients, to receive regular supervision to maintain their skills, and to be well–supplied with the drugs and equipment necessary to perform their tasks [12]. These CHWs often also have community health education roles, perform household visiting, and may also be responsible for such activities as promotion and distribution of ITNs. Studies of IMCI and iCCM are often concerned with maintaining the quality of all the above tasks. Table 8 summarizes the findings of assessments of C–IMCI and iCCM interventions.The studies described in Table 8 show that iCCM can be implemented successfully at the community level and indeed may lead to a decrease in under–5 mortality. A large assessment of children younger than 5 years of age in 15 districts in Rwanda with complete mortality data further supports this. This assessment found that the number of children receiving community–based treatment for diarrhea and pneumonia increased significantly in the 1–year period after iCCM implementation, from 0.83 cases/1000 child–months to 3.80 cases/1000 child–months (P < 0.001) and from 0.25 cases/1000 child–months to 5.28 cases/1000 child–months (P < 0.001), respectively. On average, total under–5 mortality rates declined significantly by 38% (P < 0.001), and health facility use declined significantly by 15%. These decreases were significantly greater than expected based on baseline trends [S192].

In many parts of rural Uganda with limited access to trained health staff, up to 50% of cases of childhood illnesses are managed by drug sellers. One study in which private drug sellers were trained to treat patients using iCCM protocols revealed a strong adherence to the iCCM protocol in terms of testing, examining and treating children. On follow up evaluation after training, 88% of children diagnosed with diarrhea received ORS. 88% of children presenting with a fever received a RDT for malaria and 94% of children who were diagnosed as RDT–positive received artemisinin combination therapy. Of those who were diagnosed with pneumonia, 91% of them received amoxicillin treatment. Overall performance (defined as correct treatment) showed a 27% (P = 0.001) increase compared with baseline levels [S188]. The other studies cited in Table 8 demonstrate that monthly community–level supervision by trained supervisors from the local health facility can lead to maintenance of CHW skills in iCCM diagnosis and treatment and that iCCM leads to more children receiving treatment for these common illnesses [S186, S189–191].

Care Groups

Care Groups were included in the review through the publication of the results of the evaluation of several projects. A Care Group is a group of 10–15 community volunteers who act as community–based health educators. The Care Group meets every two weeks with a project facilitator for two hours or so to learn some new education messages. Each volunteer is responsible for regularly visiting 10–15 of her neighbors, sharing the new messages they just learned. With this structure and basic approach, scaling up is readily possible [13,14].

In a 5–year Care Group project in Sofala Province in Mozambique, the project area was divided into two sub–areas (A and B) since project activities began several years later in Area B after activities in Area A had begun. Major improvements were achieved across most indicators of child health comparing baseline with endline findings. Key outcomes were that the overall proportion of children with undernutrition (WAZ<–2.0 SD) decreased by 6% in Area A and by 10% in Area B; insecticide–treated bed net (ITN) use increased by 45% in Area A and by 71% in Area B; rates of exclusive breastfeeding increased by 60% in Area A and 25% in Area B; the percentage of children 9–23m of age who ate three or more meals per day increased from by 42% in Area A and by 20% in Area B. Based on findings obtained with the Lives Saved Tool (LiST), the project saved an estimated 6848 lives and the cost per life saved, the cost per disability–adjusted life year (DALY) averted, and the annual cost per beneficiary were US$ 441, US$ 14.72 and US $2.78, respectively [S193].

Another Care Group project in the rural part of the Chokwe District in Mozambique also incorporated a community–based vital events registry system as part of the activities of the Care Groups. The assessment of this project demonstrated not only the efficacy of Care Groups but also the quality of a community–based vital events registration system. This assessment demonstrated that the Care Group approach resulted in a 49% decrease in the infant mortality rate and a 42% decrease in the under–5 mortality rate over the five year period of project implementation, confirmed by an independent retrospective morality assessment based on maternal birth histories [S194]. Similar results were found in another Care Group project in our database in Burundi [S195].

Integrated community–based primary health care (CBPHC)

Primary health care (PHC) includes the provision of a comprehensive range of essential preventive and treatment actions aimed at meeting all the common health needs of community members (especially those of women of childbearing age and children but also of men and older women) using practical and affordable approaches. For integrated CBPHC to be effective at the community level outside of health facilities, CHWs need to have good linkages to the local health facility to which patients with severe illness, injuries and uncommon or more severe illnesses can be referred and where mothers can give birth. Services such as immunizations that require outreach from health facilities also need to be provided at the community level in order to make essential services readily available. Our review includes a number of community–based PHC programs that are presented in Table 9 .

Table 9 demonstrates that primary health care with strong community–based components can decrease under–5 mortality. Promotion of community involvement and training/deployment of CHWs is also shown to be a recurring element of these successful programs. Assessments S196–198 are three studies from the Navrongo experiment in Ghana. In the Navrongo experiment in Ghana, there were four groups compared: (1) community health nurses alone–called Community Health Officers, (2) community volunteers and community mobilization without community health nurses; (3) both community health nurses and community volunteers with community mobilization, and (4) a control group. The group that only had community volunteers did not reduce child mortality but did significantly improve child nutrition [S196]. The community–based nurses provided curative care and were effective in decreasing child mortality but did not improve child nutrition or contraceptive coverage [S197]. The best results were achieved when nurses worked with community volunteers and mobilized community members improving child mortality, child nutrition and contraceptive use, together with a 15% improvement in contraceptive coverage [S198].

The census–based, impact–oriented (CBIO) methodology includes mapping and community registering to ensure that all beneficiaries are documented and included in the project information system so that they are included in all community–based PHC programs [S200, S201]. The CBIO approach was pioneered in Haiti in the 1970s. Assessment by retrospective maternal birth histories and household anthropometric surveys demonstrated a 68% reduction in under–5 mortality and reduced prevalence of stunting compared to national rural indicators [S202, S203].

The last assessment in Table 9 is the earliest one in our database and was reported in 1951 [S204]. It was carried out at a time when there had not yet been many experiences with CHWs and when CHWs were used only for health promotion and referral for provision of health services at a health center.

One important study in our database that does not lend itself to incorporation into Table 9 is the Narangwal Project, which pioneered many elements of CBPHC [S205]. It operated from 1967 to 1973 in the rural Punjab of Northern India. The nutrition and health–care aspects of this study are of direct relevance to CBPHC and child health. There were four cells in the nutrition aspect of this non–randomized controlled study: (A) a nutrition–only cell, (B) a health–care–only cell, (C) a combined nutrition–and–health–care cell, and (D) a control cell (in which routine government services without outreach were provided). Promotion of community participation was a key aspect of the design of this study. Each cell contained approximately 200–300 children. Child nutrition services included growth monitoring and promotion as well as food supplementation twice daily. The child health care services included infectious disease surveillance and early treatment, immunizations, and education concerning disease prevention. In the nutrition study, mortality rates were significantly reduced during the perinatal, neonatal, post–neonatal, and 12–23 month age groups in both the nutrition cell as well as in the nutrition + health care cell compared to the control cell. In addition, the weight–for–age and height–for–age of children beyond 17 months of age were significantly greater in the nutrition cell and in the nutrition + health care cell compared to control cell [S205].

Key CBPHC aspects of this project were that Family Health Workers provided treatment in the home for dehydration from diarrhea and for childhood pneumonia. The children 0–3 years of age with pneumonia who were treated with penicillin had a 42% reduced risk of overall mortality [S206]. Other key findings based on a qualitative review of data were that: one–on–one education of mothers was essential for improving practices related to breastfeeding, infant feeding, rehydration and feeding of sick infants and also for overcoming traditional beliefs about not feeding a child with diarrhea; weekly home visits were necessary in order to achieve a reduction in infant mortality; delegation of services as far to the periphery as possible improved coverage and effectiveness; rehabilitation of malnourished children through special feeding programs was best accomplished at home or near the home; having a curative health care service was an essential element of building trust, and developing a quality health care program required active community participation and building trust with the community [S207].

Several other assessments included in our database are of particular note since they document the evidence of the long–term benefits of CBPHC projects on child health. These projects are:

These projects are discussed in detail elsewhere in this supplement [15].