Integrated Management of Childhood Illness implementation in Nepal: understanding strategies, context, and outcomes

Over the past several decades, a surge of proven evidence-based interventions (EBIs) known to reduce under-5 mortality (U5M) has led to important progress in reducing U5M especially in low- and middle-income countries (LMICs) [1‐3]. Yet countries’ goals of reducing those child deaths preventable by a quality and effective system have been slowed by the lag between scientific progress and the dissemination and implementation at scale of these EBIs with quality and equity [4‐6]. There is continued need to increase understanding of how existing and emerging EBIs in geographically diverse and complex settings were implemented, successfully or otherwise, in order to learn from and generalize this experience in other contexts [7, 8].

Implementation research (IR) studies how EBIs are adopted and integrated into community and clinical settings to improve individual and population health [9]. This work bridges the gap between what we know about the effectiveness of EBIs and their actual uptake into a health system [10]. IR facilitates understanding of the successes and challenges in implementing EBIs in LMICs and is crucial to understanding how countries are successful in taking up new innovations as new EBIs emerge [8, 11]. Yet the literature on EBIs tends to focus on effectiveness and coverage. Often, it does not capture important insights into what was actually done – implementation strategies – and what worked and why, including the barriers and facilitators – contextual factors – which may have influenced strategy choice and success [12]. As there is increasing evidence on adapting existing EBIs and integrating new EBIs, it is critical to understand how countries succeed in accelerating the uptake and integration of new innovations into their health systems, thereby narrowing the gap between research and practice [13].

To bridge this gap, between 2017 and 2020 we conducted implementation research case studies in six countries (Nepal, Rwanda, Senegal, Bangladesh, Peru, and Ethiopia) which from 2000–2015 outperformed their geographic and economic peers in reducing amenable U5M – deaths preventable through health system-delivered EBIs (the full Nepal case study is available at https://​www.​exemplars.​health/​topics/​under-five-mortality/​nepal). We developed a hybrid implementation research framework building on existing frameworks, to guide the study in identifying implementation strategies, contextual factors which served as barriers or facilitators in LMICs relevant to amenable U5M, and implementation outcomes such as acceptability, effectiveness, and fidelity (Appendix 1) [14].

Nepal has experienced challenges to reducing U5M related to geography, conflict, and deeply ingrained social inequities. The country witnessed a decade-long political conflict (1996–2006) which resulted in significant loss of lives, internal displacement, and economic loss [15]. The country’s mountainous terrain makes it difficult to deliver health services in the community [16]. Further, longstanding social inequities including unequal access to economic, social, and political representation across the country’s many ethnic groups have continued to persist despite efforts to put an end to them [17]. Yet between 2001 and 2016, Nepal saw a marked decline in U5M from 91 deaths per 1,000 live births to 39 deaths per 1,000 live births, a reduction of 57% [18]. Neonatal mortality dropped by 46% (from 39 deaths per 1,000 live births to 21 per 1,000 live births). While the decline in neonatal mortality was slower than that of U5M, both dropped faster than the average in the Southeast Asian Region (47% and 37%, respectively) for the period 2000 to 2015 [19] as well as the average decline across LMICs (44% and 36%) [20]. Important to this success was the exploration, adaptation, and integration of EBIs.

In this paper, we use IR to understand the steps that Nepal took to explore, prepare, implement, adapt, and sustain the implementation of integrated management of childhood illness (IMCI) as an example of one EBI important for U5M reduction, and the strategies adapted and adopted at each step of the process. We identify contextual factors that facilitated or hindered the adoption of IMCI and broader EBI implementation work to better understand how implementation strategies were selected or adapted and the context in which Nepal’s leaders made decisions. Understanding how Nepal was able to overcome barriers and adopt and adapt implementation strategies to implement EBIs such as IMCI can help to provide key lessons that can be replicated in other settings working to accelerate their efforts to implement or adapt new and existing EBIs as new evidence emerges through research and experience.

Integrated management of childhood illness was chosen as it was designed to treat children for diseases responsible for the leading amenable causes of death: in 2000, acute respiratory infections (ARI), diarrhea, and malnutrition combined were accountable for more than one-third of under-5 deaths in Nepal [32]. The country began piloting a combined facility- and community-based IMCI in districts between 1999–2000 and it was rolled out nationally to all 75 districts by 2009. Care-seeking for children under 5 – a marker of the EBI’s acceptability – between 2001 and 2016 increased: for example, for diarrhea, rising from 21 to 64%; for fever, rising from 24 to 80% [18, 27] (Appendix 3). There was variability in equity of improvements based on geography. While care-seeking for diarrhea increased in both urban and rural settings from 2001 to 2016, there was a larger increase in rural areas (from 21% to 70%, up 49 points) than urban (from 23% to 60%, up 37 points) [18, 27]. There were also increases in care-seeking by wealth category. For example, care-seeking for fever rose from 19% and 33% among the lowest and highest quintiles in 2001 to 59% and 84% in 2016 – but the absolute gap between the quintiles also grew during this period [18, 27].

We used a hybrid implementation research framework to understand how Nepal implemented IMCI as an example of health system-delivered EBIs designed to contribute to the reduction of amenable U5M between 2000–2015. We found that the country chose implementation strategies which often either addressed barriers or leveraged facilitating contextual factors. These strategies included engagement of community and local stakeholders, engagement of partners in preparation and for implementation, integration of programs into existing structures, and pilot testing. Following steps from exploring the need, preparing, and pilot testing, IMNCI programming was implemented nationwide. This work was also facilitated by contextual factors including culture and capacity of data use, prioritization of local research, and existing community health system and structure.

Implementation research has long recognized that adaptation is necessary to achieve impact of EBIs across settings – but often research does not specify whether adaptations were made to planned EBIs [8]. Nepal deployed adaptation as a critical strategy and step in the implementation pathway. It made adaptations of strategies based on local research and available health systems and structure; adaptations to local context to leverage its existing FCHVs, culture of data use and local research capacity, as well as to face the barrier of geographical challenges; and adaptations targeting implementation outcomes, to improve reach and effectiveness. The government expanded FCHVs’ roles as community providers for CB-IMCI, adapted and expanded WHO training materials and guidelines for FB-IMCI and CB-IMCI, and eventually adapted the program again to include the neonatal component. These were data-driven adaptations including of existing training, guidelines, and programs to fit the local context. They were not always successful however, and there has been debate about whether the adaptation to IMNCI was the right call given lack of training of FCHVs, low demand from families, and existence of other programs that may have been a better fit [47]. A study of IMNCI implementation in India found that increased supervision and standardization of training were important for CHWs given the added newborn services [48], further adaptations that may be necessary in Nepal to increase the quality and sustainability of its IMNCI programming.

Nepal’s investment in community- and facility-based IMCI implementation was successful in part because of its leveraging of an important facilitating contextual factor: the existing strong cadre of community-based healthcare champions known as female community health volunteers. This finding is supported by others who found that community health interventions implemented with FCHVs had high coverage during the study period[49].

Yet the FCHV system has also began to experience strains, related to expansions to its scope of activities and an aging workforce. Increases in program demands, service quality issues, and growing access of the population to alternative sources of health care as the country addressed geographic and financial barriers have led to concerns around the sustainability of this system that will need to be attended to[45, 50]. Duplications across what FCHVs provide at the community level by different divisions contributes to overburdening of FCHVs who might be required to perform tasks supporting multiple program activities [49], in a context of insufficient supervision and training, found to have historically been challenges in Nepal [51]. This experience of growing pressure on Nepal’s core of FCHVs is reflected in literature on the experiences of CHWs in health systems elsewhere [52‐54]. WHO recommendations have been updated to further institutionalize the role of CHWs through salary and supervision recommendations and additional steps to ensure CHWs operate in an official capacity including certification, community engagement, and mobilization of community resources to more firmly anchor CHWs as a community resource [55]. In recognition of these challenges, the MOH has partnered with local researchers to trial a program that will inform the development of a new CHW system, including through professionalization of the workforce and improving the quality of service delivery [56].

The culture of data use and leveraging existing local research capacity were important contextual factors to facilitate pilot programs such as the Morang Initiative Neonatal Intervention program. This pilot ultimately led to Nepal’s adaptation to include neonatal care provision in its IMCI program. Similar to our findings, evidence from IR studies in several countries found that policymakers valued local implementation research and that international research evidence was found useful for identifying policy options that were relevant in local settings [7]. Other studies of Nepal’s work to reduce U5M found the government to have been progressive in the adoption of EBIs, and has been important in fostering the culture of data use to inform program design [57]. However, the culture of data use was also associated with some challenges. While prioritization of local research has been a facilitator to the country’s successful implementation of many EBIs, it has also caused some delays between decisions to explore a new EBI and the actual nationwide scaling of an intervention. Further, some EBIs found to be successful in local pilots were never scaled in the course of the study period [40]. Finding a balance of commitment to local research and decisions to adopt national rollout of interventions based on existing evidence could have an impact on U5M outcomes in the future.

The Government of Nepal has been working towards universal health coverage (UHC) through expanded social insurance and free primary health care programs in recent years. This leadership and vision of health care as a fundamental right has persisted across national political transitions, and in 2016 Nepal introduced the National Health Insurance Program [58]. A driving goal of Nepal’s UHC expansion has been to increase access in health policy and delivery – supporting its equity strategy [59]. While ensuring equity was a core strategy, and national coverage increased for many of the IMCI markers and metrics between 2001 and 2016, there was not always uniform benefit across urban and rural areas and across wealth quintiles. These differences demonstrate that there are opportunities to improve equity between groups while continuing to make substantial progress for the population as a whole. Thapa and colleagues looked at equity and coverage of many of the leading U5M EBIs in Nepal between 2001 and 2016 and noted that as coverage increased inequitable coverage decreased steadily – if slowly – across many EBIs [60]. There are examples of more effective strategies to reduce inequities across income and geography, including Nepal’s immunization program, which has made continuous ongoing outreach efforts to ensure an uninterrupted supply of vaccines for children under 5 [60].

We found other important factors which facilitated implementation of EBIs including IMCI and contributed to Nepal’s dramatic drop in U5M included those less explicitly focused on health care-targeted preventative and curative efforts. These include improvements to infrastructure and communications – such as to roads, schools, electricity, and communication technology; addressing social determinants and other contributors to U5M – such as water, sanitation, and hygiene (WASH); and access to family planning, as well as women’s education and empowerment through initiatives such as expanding economic opportunities and increasing asset ownership. A study of success factors in countries on track to meet MDG goals for U5M found non-health system investments responsible for as much as half the U5M reductions [61]. Infrastructure and other non-health system investments can contribute to the momentum needed in countries working to accelerate progress in health outcomes.

Despite its successes, challenges remain for Nepal. Overcoming service gaps and equity barriers that result from the country’s difficult geography is an important persistent challenge as most areas in the country are hills and mountains. Strengthening local service provision has happened but ongoing gaps in access and coverage remain. Adapting existing programs that have been successful in the past, such as the system of FCHVs, to be sustainable going forward will be a critical undertaking.

This paper has a number of limitations. Key informant interviews took place after the period of interest, and so results depended on recall about strategies and contextual factors as well as the pathway to implementation dating back to 2000–2015. Resources and timing of the study also limited the range of key informants, so we were not able to interview frontline implementers or communities about the time period of interest. Due to limitations of the research scope we did not include all potentially relevant contextual factors including those adjacent to or beyond the health system. Data availability and quality limited findings on implementation outcomes as key factors such as cost, quality, and reasons for variability in outcomes were often absent in the published reports and quantitative data. The limitations on data availability meant that we were unable to include subnational variability on implementation outcomes and context, an area for further work. The design of this work was specifically focused on the implementation of known EBIs in a real-world setting. As such the attribution of specific EBIs to overall amenable U5M reduction was not directly measurable.

Using implementation research we broadened understanding of how and why Nepal used a number of chosen implementation strategies to implement IMCI as an EBI known to reduce amenable U5M in LMICs. This approach was important in overcoming the know-do gap which has hindered progress in achieving child health goals, through adaptation, sustainment, and outcomes beyond effectiveness. The importance of understanding context and then choosing strategies that leverage facilitating factors such as community health programs, culture of data use, and leadership, and ones which address barriers such as geography, are useful lessons for countries working to accelerate expansion of quality EBIs to continue work to achieve child health targets.