Contribution of community-based newborn health promotion to reducing inequities in healthy newborn care practices and knowledge: evidence of improvement from a three-district pilot program in Malawi

There is strong and growing evidence of inequities in maternal and newborn health worldwide. The term “inequity in health” describes unjust differences in access to care and health outcomes between advantaged and disadvantaged population groups [1]. At the global level, the populations of wealthier countries have better maternal and neonatal health outcomes than poorer countries [2, 3]. Sub-Saharan Africa, the poorest region in the world, accounts for 14 of the 18 countries with the highest neonatal mortality rates globally [2]. Socioeconomic inequality is also consistently observed within many countries. In both rich and poor countries, more neonatal deaths occur among the poor [2] and rural areas have higher mortality rates, missing out on the great declines in mortality that occurred in urban areas during the 1990s [4]. It is estimated that elimination of within-country inequities in high mortality countries could reduce annual neonatal deaths by close to 750,000 [2].

Several factors may contribute to wealth-based inequities in child and newborn health outcomes, including living conditions, knowledge and behaviors, nutritional status, and physical and financial access to health services [5–9]. Although evidence-based interventions exist to address the main causes of newborn death [10], there is inadequate coverage of these interventions among the poor (a trend sometimes referred to as the “inverse care law” [11]). For example, while virtually all deliveries in high-income countries have a skilled attendant [12], 54% is the median coverage for skilled attendance at birth among the 68 low-income countries tracked by the Countdown to 2015 [13]. Within poor countries, richer women are between 2 and 3 times more likely to utilize antenatal care and up to six times more likely to have a skilled attendant at delivery [13–15]. In some countries these equity gaps have remained consistent since 1990 [16]. There is consensus that reducing inequities is critical for making major gains in newborn and child health and reaching Millennium Development Goal 4 [2].

Although there are multiple approaches that can be used in pursuit of greater equity in health, community-based programs and outreach are considered especially important [15, 17, 18]. Community-based programs may quickly expand coverage to areas where access to facility care is limited, thereby removing key barriers for poor households such as distance and transport costs. Community-based programs are also often offered free of charge and can be used to promote healthy behaviors, to provide screening and referral for complications and illnesses, to promote utilization of facility-based services, and in some cases to provide treatment at the home or community level.

Over the past decade, Saving Newborn Lives (SNL) of Save the Children helped develop evidence that community-based programs are effective at reducing neonatal mortality through support of trials in South Asia [19–21] and ongoing trials in Africa [22–24]. SNL has also successfully advocated for policy changes to expand community-based newborn programs in many countries [25, 26]. However, little is currently known about the extent to which the community delivery strategy improved equity in newborn health.

This paper examines the contributions of community-based programming towards improving the equity of newborn health in Malawi, one of the countries included in SNL’s initiative to scale-up community-based newborn care programs in partnership with governments. Although large inequities have been reported in Malawi for utilization of maternal and child health services [27], little is known about inequities in knowledge and practices related to newborn care. The objectives of this paper are to: 1) describe baseline coverage of maternal and newborn health services, knowledge, and newborn care practices by household wealth status; 2) to assess changes in health inequities following implementation of a community-based program for promotion of newborn health.

The baseline household survey sample included 903 women, the majority of whom were between the ages of 20 and 39 (Table 2). Seventy percent of women had some primary school education, while 14% reported no education and 15% reported some secondary or post-secondary education. The majority of households had an unimproved pit latrine (75%), cooked with wood fuel (95%), lived in homes with a natural roof of grass or leaves (61%), and had walls made of bricks (88%). Few households possessed electronics or appliances other than a radio (70%), although watch and bicycle ownership was reported by around half of households. The endline household survey included a sample of 900 women. Although the sampling frame was different between the two surveys, the samples are comparable in terms of demographic indicators and most household assets. Slightly more households in the endline household survey reported inferior home construction materials, such as dirt walls (15 percentage points higher) and a grass or leaf roof (12 points higher), and fewer households reported watch or bicycle ownership. Notably, mobile phone ownership increased from 15% at baseline in 2007 to 34% at endline in 2011. The increase in mobile phone ownership may also explain some of the difference in ownership of a wristwatch between surveys, as the mobile phone may replace the wristwatch as a time keeping device.

The CBMNC package in Malawi aimed to contribute to efforts to increase coverage of facility-based care and targeted newborn care practices, and to improve the equity of these services and behaviors. To achieve this aim, SNL and the MOH trained 622 HSAs in the CBMNC package and the intervention reached 38% of women through antenatal home visits and 11% through postnatal home visits. Over the course of the study period, coverage rates significantly increased for facility-based services, newborn care practices, and knowledge. At the same time, Demographic and Health Surveys from 2004 and 2010 demonstrate an increase in use of facility-based maternal and newborn health services across the country [27, 43]. Our analysis of baseline and endline household surveys in the study area demonstrates that health equity for knowledge of danger signs and facility delivery, as well as early breastfeeding and delayed bathing, which are promoted during training of facility providers, also improved significantly over the study period. The baseline coverage differences between poorer and wealthier households were relatively small, and accordingly the decreases in inequity were small but significant. Greater improvements in inequities were observed for knowledge indicators. At endline, a greater proportion of women from the poorest quartile reported knowledge of danger signs for pregnancy, delivery, and postpartum mothers than women from the least poor quartile.

Although these results indicate promising improvements for newborn health in Malawi, the extent to which CBMNC contributed to these improvements in coverage and equity are not known. The design of this study would be considered an adequacy design [44], with the lack of a comparison area affecting our ability to attribute coverage and equity changes to the CBMNC program. Program implementation challenges presented additional study limitations. First, the training of HSAs in the pilot area was interrupted for six months due to funding constraints, and the program achieved only moderate coverage of pregnancy home visits and low coverage of postnatal home visits within 3 days of delivery. Additionally, HSAs in Malawi are recruited centrally rather than from the communities in which they serve, and many HSAs were not living at their posts. The HSAs are also expected to spend one or more days per week helping out at health facilities. These factors would likely have reduced their effectiveness at identifying and visiting pregnant women and new mothers. The response of the program to these challenges was to advocate with district and national officials to reinforce residency requirements for HSAs and to encourage communities to provide housing for HSAs in areas where this was a barrier to their residency and to help notify HSAs of pregnant women and newborns in their communities.

At endline, less than 40% of women received a home visit from an HSA during pregnancy. It is not known whether other health communications programs, such as those delivered by clinicians at health facilities or over the radio or text message, influenced the measured outcomes. The decision by the Government of Malawi to ban the use of traditional birth attendants in 2007—a ban which was relaxed in 2010 [45]—likely also affected the utilization of facility-based services.

An additional important component of the program was efforts to improve facility-based care. It should be noted that these quality improvement efforts were limited to government health facilities, and some households, particularly the less poor, may choose other providers, such as the Christian Health Association of Malawi. Nonetheless, there is strong evidence that clients’ perceptions about the quality of care provided at health facilities influence utilization of maternal health services [46–49]. Facility assessments conducted by SNL demonstrated that, although SNL trained and regularly supervised providers, some supply side barriers to quality remained. For example, at one assessment during the study period, less than half of intervention facilities had all four of the following essential newborn commodities available and functioning: injectable gentamycin, thermometer, infant scale, and resuscitation bag and mask [50]. Still, the rates of antenatal care attendance and facility delivery in the endline sample for this study were comparable or higher than those reported by the 2010 Malawi DHS [27].

The results of this study suggest that community-based promotion of key maternal and newborn health behaviors, accompanied by strengthening of facility-based care and counseling, may help reduce inequities in maternal and newborn health. Similar studies in other settings have shown mixed results regarding the ability of community-based programs to improve equity. While evaluations of some community-based programs for maternal and child health report reductions in inequities [51–53], others show that the richest households receive greater benefits from these programs [54, 55]. Richer households in the endline survey for this study were more likely to receive a home visit during pregnancy, but were no more likely to have a postnatal home visit or to know of a core group in their community. The observation that the poorest households are not the primary beneficiaries of HSA services is consistent with other assessments of HSA activities in Malawi [56]. The MOH has noted that HSAs posted to the most remote areas are less likely to reside in their catchment areas, and in response to this evaluation and other evidence, is identifying strategies to improve coverage by increasing residency within the hard-to-reach-areas. These results also suggest that the approaches through which community-based programs are implemented likely play an important role in their equity and deserve further attention. For example, the selection criteria for CHWs, their service location, and the presence or absence of pro-poor targeting strategies are all likely to affect equity [6]. The conditions under which community-based programs are successful at reaching the poorest households remains an important research area [17], and evidence suggests that community-based interventions are not automatically guaranteed to improve equity to the extent that policy makers and program implementers may expect.

The implementation context of community-based programs is important when considering program equity results, including the degree of overall wealth inequality in the program setting. Malawi is ranked 70^th among nations in terms of wealth inequality [57], and 90% of the population is considered poor by global standards [29]. The differences between households in the poorest categories will be minor compared to other African countries with high wealth inequality, such as Namibia, South Africa, and Botswana [57]. The patterns of wealth inequality also vary by setting, as does the health care system. For example, in Latin America, a greater proportion of the poor population lives in urban areas compared with Africa and Asia [58]. Therefore, local context should guide pro-poor targeting efforts.

As governments in low-income countries increasingly turn towards community health worker programs to improve health outcomes, the appropriate expectations for community-based health workers’ (CBHWs) scope of work and coverage targets should become a focus of policy makers and researchers [59–61]. In Malawi, HSAs, who are traditionally responsible for health education and sanitation activities, are being utilized in task shifting efforts to provide community case management [62], voluntary counseling and testing for HIV[63], and other services. CBHWs in other countries are receiving a similar increase in the scope of their work, which has implications for the impact of individual interventions as well as equity—the workload on CBHWs will likely affect their ability to reach out to the poorest households in the communities they serve.

Between 2007 and 2011 the study areas is rural Malawi demonstrated a significant improvement in the coverage of knowledge of danger signs and some health behaviors. Some of these general increases reflected small improvements in equitable coverage. The CBMNC package aimed to improve equity by building demand for facility-based services in communities and reducing barriers to counseling for newborn care and maternal and newborn complications. The coverage of home visits by HSAs during pregnancy and the postpartum period was lower than expected, and generally did not significantly favor the poorest households. Strategies for further increasing pro-poor coverage are needed to further reduce inequities in maternal and child health.