The role of integrated community case management projects in strengthening health systems: case study analysis in Ethiopia, Malawi and Mozambique

Theoretical propositions for the three iCCM case studies

Thematic area	Theoretical propositions Integrated community case management strengthens health systems to demonstrate…	References
National coordination and policy setting	(1) Signs of ownership and commitment to the policy, manifested through different levels of the health system. (2) Legal framework and health policies support the interventions. (3) Coordinating efforts of different internal stakeholders and external partners. (4) Standardized approaches across partners, aligned with MOH.	Bennett et al. (2014; 2015), Campbell et al. (2015), Dalglish et al. (2015a,b), Doherty et al. (2015); George et al. (2011; 2015), Legesse et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015) and Smith et al. (2014)
Costing and financing	(5) Progressively increased domestic funding for services with reduced financial hardship on users, without displacement of resources from other essential public goods. (6) Future funding opportunities (to local/national stakeholders) for scaling up or sustaining existing interventions. (7) Fair allocation of resources through conflict mitigation and efforts to consider power dynamics.	Bennett et al. (2014), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Dalglish et al. (2015a), Doherty et al (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Rodríguez et al. (2015b), Siekmans et al. (2017) and Sunguya et al. (2017)
Human resources	(8) More human resources for health (both skilled and lay, where one might supplement the need for the other). (9) More rational allocation of health workers’ time on expanding coverage of essential services, including task shifting/sharing. (10) Improved skills mix/role of health workers at facility and community levels, without displacement. (11) Decentralized, adaptive, evidence-based and problem solving management, balancing needs of programmes in and out of target package (no displacement).	Adamo et al. (2016), Callaghan-Koru et al. (2013), Campbell et al. (2015), Chilundo et al. (2015), Doherty et al. (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Seidman and Atun (2017), Smith et al. (2014), Sunguya et al. (2017) and Tulenko et al. (2013)
Supply chain management	(12) Adequate drugs and supplies to sustain the capacity of CHWs to correctly diagnose and treat illness with proper equipment. (13) Better coordinated supply chain management to ensure adequate drug supplies, reliant on supervision, training, reporting, financing and transportation networks.	Bagonza et al. (2015), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Daniels et al. (2015), Doherty et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Siekmans et al. (2017) and Tulenko et al. (2013)
Quality of service delivery and referral	(14) Services are responsive to community needs and adapted to context. (15) Services perceived to be available and effective among the community. (16) Cross-cutting benefits outside the scope of pneumonia, diarrhoea and malaria. (17) Rational expansion of health infrastructure to achieve effective coverage in essential services, with associated resources to manage these.	Bennett et al. (2015), Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2011; 2015), Legesse et al. (2014), Marsh et al. (2014), Munos et al. (2016), Sarriot et al. (2015), Seidman and Atun (2017) and Siekmans et al. (2017)
Supportive supervision and management capacity	(18) Availability of technical assistance to districts to provide support and identify performance gaps in the system.	Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2015), Nsona et al. (2012), Sarriot et al. (2015) and Tulenko et al. (2013)
Health information systems	(19) Coordinated health management information systems to monitor and evaluate health service data. (20) Adequate reporting and performance measures used to inform decisions and identify bottlenecks in implementation. (21) Stronger processes for programme learning, management course correction based on information.	Chilundo et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015b) and Siekmans et al. (2017)
Community organizations and societal partnerships	(22) Improved engagement with community organizations to design, implement and evaluate community health work. (23) MOH engages in effective societal partnerships to improve functionality of system for health, efficiency and resilience, for community engagement, community services and facility-based services. (24) Health systems stakeholders (MOH, non-health sectors, civil society, private sector) develop stronger accountability mechanisms from bottom to top of health pyramid. (25) Increased social capital (i.e. trust, cohesion, resources, voice, empowerment) and accountability through community engagement.	Bennett et al. (2014), George et al. (2015), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015), Seidman and Atun (2017), Smith et al. (2014) and Tulenko et al. (2013)
Inclusivity and equity of the health system	(26) Inclusion of marginalized communities in benefits of iCCM.	Adamo et al. (2016), Collins et al. (2014), Dalglish et al. (2015a), Daniels et al. (2015), George et al. (2011), Legesse et al. (2014), Nsona et al. (2012) and Smith et al. (2014)

Thematic area	Theoretical propositions Integrated community case management strengthens health systems to demonstrate…	References
National coordination and policy setting	(1) Signs of ownership and commitment to the policy, manifested through different levels of the health system. (2) Legal framework and health policies support the interventions. (3) Coordinating efforts of different internal stakeholders and external partners. (4) Standardized approaches across partners, aligned with MOH.	Bennett et al. (2014; 2015), Campbell et al. (2015), Dalglish et al. (2015a,b), Doherty et al. (2015); George et al. (2011; 2015), Legesse et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015) and Smith et al. (2014)
Costing and financing	(5) Progressively increased domestic funding for services with reduced financial hardship on users, without displacement of resources from other essential public goods. (6) Future funding opportunities (to local/national stakeholders) for scaling up or sustaining existing interventions. (7) Fair allocation of resources through conflict mitigation and efforts to consider power dynamics.	Bennett et al. (2014), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Dalglish et al. (2015a), Doherty et al (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Rodríguez et al. (2015b), Siekmans et al. (2017) and Sunguya et al. (2017)
Human resources	(8) More human resources for health (both skilled and lay, where one might supplement the need for the other). (9) More rational allocation of health workers’ time on expanding coverage of essential services, including task shifting/sharing. (10) Improved skills mix/role of health workers at facility and community levels, without displacement. (11) Decentralized, adaptive, evidence-based and problem solving management, balancing needs of programmes in and out of target package (no displacement).	Adamo et al. (2016), Callaghan-Koru et al. (2013), Campbell et al. (2015), Chilundo et al. (2015), Doherty et al. (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Seidman and Atun (2017), Smith et al. (2014), Sunguya et al. (2017) and Tulenko et al. (2013)
Supply chain management	(12) Adequate drugs and supplies to sustain the capacity of CHWs to correctly diagnose and treat illness with proper equipment. (13) Better coordinated supply chain management to ensure adequate drug supplies, reliant on supervision, training, reporting, financing and transportation networks.	Bagonza et al. (2015), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Daniels et al. (2015), Doherty et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Siekmans et al. (2017) and Tulenko et al. (2013)
Quality of service delivery and referral	(14) Services are responsive to community needs and adapted to context. (15) Services perceived to be available and effective among the community. (16) Cross-cutting benefits outside the scope of pneumonia, diarrhoea and malaria. (17) Rational expansion of health infrastructure to achieve effective coverage in essential services, with associated resources to manage these.	Bennett et al. (2015), Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2011; 2015), Legesse et al. (2014), Marsh et al. (2014), Munos et al. (2016), Sarriot et al. (2015), Seidman and Atun (2017) and Siekmans et al. (2017)
Supportive supervision and management capacity	(18) Availability of technical assistance to districts to provide support and identify performance gaps in the system.	Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2015), Nsona et al. (2012), Sarriot et al. (2015) and Tulenko et al. (2013)
Health information systems	(19) Coordinated health management information systems to monitor and evaluate health service data. (20) Adequate reporting and performance measures used to inform decisions and identify bottlenecks in implementation. (21) Stronger processes for programme learning, management course correction based on information.	Chilundo et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015b) and Siekmans et al. (2017)
Community organizations and societal partnerships	(22) Improved engagement with community organizations to design, implement and evaluate community health work. (23) MOH engages in effective societal partnerships to improve functionality of system for health, efficiency and resilience, for community engagement, community services and facility-based services. (24) Health systems stakeholders (MOH, non-health sectors, civil society, private sector) develop stronger accountability mechanisms from bottom to top of health pyramid. (25) Increased social capital (i.e. trust, cohesion, resources, voice, empowerment) and accountability through community engagement.	Bennett et al. (2014), George et al. (2015), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015), Seidman and Atun (2017), Smith et al. (2014) and Tulenko et al. (2013)
Inclusivity and equity of the health system	(26) Inclusion of marginalized communities in benefits of iCCM.	Adamo et al. (2016), Collins et al. (2014), Dalglish et al. (2015a), Daniels et al. (2015), George et al. (2011), Legesse et al. (2014), Nsona et al. (2012) and Smith et al. (2014)

Table 1

Theoretical propositions for the three iCCM case studies

Thematic area	Theoretical propositions Integrated community case management strengthens health systems to demonstrate…	References
National coordination and policy setting	(1) Signs of ownership and commitment to the policy, manifested through different levels of the health system. (2) Legal framework and health policies support the interventions. (3) Coordinating efforts of different internal stakeholders and external partners. (4) Standardized approaches across partners, aligned with MOH.	Bennett et al. (2014; 2015), Campbell et al. (2015), Dalglish et al. (2015a,b), Doherty et al. (2015); George et al. (2011; 2015), Legesse et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015) and Smith et al. (2014)
Costing and financing	(5) Progressively increased domestic funding for services with reduced financial hardship on users, without displacement of resources from other essential public goods. (6) Future funding opportunities (to local/national stakeholders) for scaling up or sustaining existing interventions. (7) Fair allocation of resources through conflict mitigation and efforts to consider power dynamics.	Bennett et al. (2014), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Dalglish et al. (2015a), Doherty et al (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Rodríguez et al. (2015b), Siekmans et al. (2017) and Sunguya et al. (2017)
Human resources	(8) More human resources for health (both skilled and lay, where one might supplement the need for the other). (9) More rational allocation of health workers’ time on expanding coverage of essential services, including task shifting/sharing. (10) Improved skills mix/role of health workers at facility and community levels, without displacement. (11) Decentralized, adaptive, evidence-based and problem solving management, balancing needs of programmes in and out of target package (no displacement).	Adamo et al. (2016), Callaghan-Koru et al. (2013), Campbell et al. (2015), Chilundo et al. (2015), Doherty et al. (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Seidman and Atun (2017), Smith et al. (2014), Sunguya et al. (2017) and Tulenko et al. (2013)
Supply chain management	(12) Adequate drugs and supplies to sustain the capacity of CHWs to correctly diagnose and treat illness with proper equipment. (13) Better coordinated supply chain management to ensure adequate drug supplies, reliant on supervision, training, reporting, financing and transportation networks.	Bagonza et al. (2015), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Daniels et al. (2015), Doherty et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Siekmans et al. (2017) and Tulenko et al. (2013)
Quality of service delivery and referral	(14) Services are responsive to community needs and adapted to context. (15) Services perceived to be available and effective among the community. (16) Cross-cutting benefits outside the scope of pneumonia, diarrhoea and malaria. (17) Rational expansion of health infrastructure to achieve effective coverage in essential services, with associated resources to manage these.	Bennett et al. (2015), Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2011; 2015), Legesse et al. (2014), Marsh et al. (2014), Munos et al. (2016), Sarriot et al. (2015), Seidman and Atun (2017) and Siekmans et al. (2017)
Supportive supervision and management capacity	(18) Availability of technical assistance to districts to provide support and identify performance gaps in the system.	Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2015), Nsona et al. (2012), Sarriot et al. (2015) and Tulenko et al. (2013)
Health information systems	(19) Coordinated health management information systems to monitor and evaluate health service data. (20) Adequate reporting and performance measures used to inform decisions and identify bottlenecks in implementation. (21) Stronger processes for programme learning, management course correction based on information.	Chilundo et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015b) and Siekmans et al. (2017)
Community organizations and societal partnerships	(22) Improved engagement with community organizations to design, implement and evaluate community health work. (23) MOH engages in effective societal partnerships to improve functionality of system for health, efficiency and resilience, for community engagement, community services and facility-based services. (24) Health systems stakeholders (MOH, non-health sectors, civil society, private sector) develop stronger accountability mechanisms from bottom to top of health pyramid. (25) Increased social capital (i.e. trust, cohesion, resources, voice, empowerment) and accountability through community engagement.	Bennett et al. (2014), George et al. (2015), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015), Seidman and Atun (2017), Smith et al. (2014) and Tulenko et al. (2013)
Inclusivity and equity of the health system	(26) Inclusion of marginalized communities in benefits of iCCM.	Adamo et al. (2016), Collins et al. (2014), Dalglish et al. (2015a), Daniels et al. (2015), George et al. (2011), Legesse et al. (2014), Nsona et al. (2012) and Smith et al. (2014)

Thematic area	Theoretical propositions Integrated community case management strengthens health systems to demonstrate…	References
National coordination and policy setting	(1) Signs of ownership and commitment to the policy, manifested through different levels of the health system. (2) Legal framework and health policies support the interventions. (3) Coordinating efforts of different internal stakeholders and external partners. (4) Standardized approaches across partners, aligned with MOH.	Bennett et al. (2014; 2015), Campbell et al. (2015), Dalglish et al. (2015a,b), Doherty et al. (2015); George et al. (2011; 2015), Legesse et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015) and Smith et al. (2014)
Costing and financing	(5) Progressively increased domestic funding for services with reduced financial hardship on users, without displacement of resources from other essential public goods. (6) Future funding opportunities (to local/national stakeholders) for scaling up or sustaining existing interventions. (7) Fair allocation of resources through conflict mitigation and efforts to consider power dynamics.	Bennett et al. (2014), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Dalglish et al. (2015a), Doherty et al (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Rodríguez et al. (2015b), Siekmans et al. (2017) and Sunguya et al. (2017)
Human resources	(8) More human resources for health (both skilled and lay, where one might supplement the need for the other). (9) More rational allocation of health workers’ time on expanding coverage of essential services, including task shifting/sharing. (10) Improved skills mix/role of health workers at facility and community levels, without displacement. (11) Decentralized, adaptive, evidence-based and problem solving management, balancing needs of programmes in and out of target package (no displacement).	Adamo et al. (2016), Callaghan-Koru et al. (2013), Campbell et al. (2015), Chilundo et al. (2015), Doherty et al. (2015), George et al. (2015), Legesse et al. (2014), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015a), Seidman and Atun (2017), Smith et al. (2014), Sunguya et al. (2017) and Tulenko et al. (2013)
Supply chain management	(12) Adequate drugs and supplies to sustain the capacity of CHWs to correctly diagnose and treat illness with proper equipment. (13) Better coordinated supply chain management to ensure adequate drug supplies, reliant on supervision, training, reporting, financing and transportation networks.	Bagonza et al. (2015), Callaghan-Koru et al. (2013), Chilundo et al. (2015), Daniels et al. (2015), Doherty et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Siekmans et al. (2017) and Tulenko et al. (2013)
Quality of service delivery and referral	(14) Services are responsive to community needs and adapted to context. (15) Services perceived to be available and effective among the community. (16) Cross-cutting benefits outside the scope of pneumonia, diarrhoea and malaria. (17) Rational expansion of health infrastructure to achieve effective coverage in essential services, with associated resources to manage these.	Bennett et al. (2015), Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2011; 2015), Legesse et al. (2014), Marsh et al. (2014), Munos et al. (2016), Sarriot et al. (2015), Seidman and Atun (2017) and Siekmans et al. (2017)
Supportive supervision and management capacity	(18) Availability of technical assistance to districts to provide support and identify performance gaps in the system.	Dalglish et al. (2015b), Daniels et al. (2015), George et al. (2015), Nsona et al. (2012), Sarriot et al. (2015) and Tulenko et al. (2013)
Health information systems	(19) Coordinated health management information systems to monitor and evaluate health service data. (20) Adequate reporting and performance measures used to inform decisions and identify bottlenecks in implementation. (21) Stronger processes for programme learning, management course correction based on information.	Chilundo et al. (2015), George et al. (2011), Legesse et al. (2014), Munos et al. (2016), Nefdt et al. (2014), Nsona et al. (2012), Rodríguez et al. (2015b) and Siekmans et al. (2017)
Community organizations and societal partnerships	(22) Improved engagement with community organizations to design, implement and evaluate community health work. (23) MOH engages in effective societal partnerships to improve functionality of system for health, efficiency and resilience, for community engagement, community services and facility-based services. (24) Health systems stakeholders (MOH, non-health sectors, civil society, private sector) develop stronger accountability mechanisms from bottom to top of health pyramid. (25) Increased social capital (i.e. trust, cohesion, resources, voice, empowerment) and accountability through community engagement.	Bennett et al. (2014), George et al. (2015), Nsona et al. (2012), Rodríguez et al. (2015a), Sarriot et al. (2015), Seidman and Atun (2017), Smith et al. (2014) and Tulenko et al. (2013)
Inclusivity and equity of the health system	(26) Inclusion of marginalized communities in benefits of iCCM.	Adamo et al. (2016), Collins et al. (2014), Dalglish et al. (2015a), Daniels et al. (2015), George et al. (2011), Legesse et al. (2014), Nsona et al. (2012) and Smith et al. (2014)

Fifth, an individual case report was written for each project using a case study outline that followed the format of the nine major themes in the codebook while highlighting places where the documentation that addressed the theoretical propositions. Sixth, following the completion of the individual case reports, we developed a cross-case country report. Each country report focused on the following three elements: country profile, population health and health services achievements and systems effects. The country profile included the country context (i.e. healthcare coverage, iCCM and CHW policies, international iCCM partners and local implementation partners) and a description of each project’s reach and implementation (i.e. including the budget, duration, beneficiaries reached and demographics of the target populations). The population health and health services achievements focused on the effects of each project on project-specific health outcomes within each country. The systems effects section provided a description of the interaction between project efforts, observed changes in health systems capacity over time, and the resulting changes in the overall health system ‘strength’ (i.e. theoretical propositions). Seventh, after developing a first draft of the country report, we obtained written comments from the technical team to fill information gaps from the documentation review and then conducted four key informant interviews with individuals who were closely involved with each project during its implementation. Each interview lasted approximately 75 min and the information gleaned from the interview was incorporated into the final country report. Finally, we conducted a cross-country analysis to determine the health systems themes that emerged from all three country reports. We created a new cross-cutting codebook that included seven themes, which were derived from our three country reports: continuum of health system strategies from gap-filling to strengthening; interdependent relationship between projects and health system; limits to influence of a project; disruptions can create positive ripples; Ministry of Health (MOH) alignment is not static and simple; evaluation gaps limit opportunities; and building stronger community systems is critical for HSS. We coded all three country reports using our new codebook using NVivo 11.0 (Supplementary File S2). This codebook was distinct from the aforementioned codebook that was used to analyse individual cases because it was designed to look at cross-cutting themes related to HSS that were common among all three country reports. An overview of the findings from the country reports and a detailed description of the cross-country findings from the final analytic step are presented below.

Results

Findings from country case studies

The three country reports were published in 2019 (Olivas and Story, 2019; Pritchard et al., 2019; Sarriot and Hejna, 2019). A summary of the findings from the individual country reports is available in Table 2. Overall, we found similar trends across all three country reports with respect to each of the nine themes and the theoretical propositions embedded within each theme.

Table 2

Summary of results from the county case studies in Ethiopia, Malawi and Mozambique

Thematic area	Ethiopia	Malawi	Mozambique
National coordination and policy setting	National commitment to iCCM existed, but government involvement varied between the project areas. Although Save the Children successfully helped standardize iCCM procedures in one project area, either operations or reporting periods were too short in other areas to demonstrate sustained impact.	Save the Children coordinated with the MOH to align with national policies; make decisions related to drug selection and training; and standardize reporting and supervision guidelines. However, district-level buy-in varied.	Save the Children supported the national iCCM programme through tool development and training; however, there were strong limitations in country capacity to coordinate iCCM efforts. The sustainability of these coordination efforts is uncertain.
Costing and financing	The Ethiopian government was involved in financing iCCM activities at the national level; however, there were only two notable instances of new domestic funding for iCCM. There were no future plans for financial sustainability or scale-up.	Large-scale financing improvements were beyond the scope of these projects, which allowed system weaknesses to persist. However, Save the Children was successful with some gap-filling and system strengthening activities, like purchasing drugs and supporting the transition to malaria rapid diagnostic tests (mRDTs).	There was no evidence for a clear, nationally owned plan for financing iCCM. However, Save the Children made attempts to integrate national-level financing plans into the iCCM programmes, some of which were successful.
Human resources	Development of human capacity was an area of strength for Save the Children. Amhara, Oromia, or SNNPR focused on training health workers, while the Somali region introduced performance review and clinical mentoring meetings to increase capacity for problem solving.	Save the Children most successfully built human capacity (through standardized trainings) and supported facilities (through logistical improvements). However, structural human resources issues persisted, including workload, supervision and difficulty reach isolated areas.	Save the Children contributed to improvements in human capacity through curriculum development, but there were also many complex challenges that were outside of the mandate of any individual project. These included changes in CHW roles, expanding tasks for CHWs, staff attrition and supervision capacity.
Supply chain management	Government systems remained at early stages of development of autonomous capacity, for both procurement and supply issues. iCCM projects sought to make positive, downstream contributions where and when possible, but were limited in capacity to forecast essential supplies and stock management.	National capacity for supply chain management was not shown to improve, but Save the Children sought to find opportunities to make a strategic difference through a combination of gap-filling and systems strengthening interventions.	Save the Children used two different strategies to support the supply chain management system. In the first phase, they ran a relatively successful parallel drug procurement and supply system. In the second phase (under different funding), they aligned with the MOH without mandate or resources to strengthen the system, which led to pervasive drug stockouts.
Quality of service delivery and referral	There were improvements in the perceived responsiveness of the health system to community needs, primarily through HEWs and Community Health Volunteers. However, the linkages between the health posts and PHC units were weak, which made referrals difficult.	Improvements to quality were made in terms of increasing access to close-to-home services, including mRDTs. Available quality indicators on HSA performance showed mixed results, although they compared positively with other public health services.	Quality improvements included increasing availability of care close to remote communities, access and responsiveness of the health system. APE proficiency improved and their referrals increased in number over the course of the project.
Supportive supervision and management capacity	Save the Children provided trainings for individual staff to build capacity for supportive supervision across all regions, except Somali. All projects carried a de facto role in building management capacity, but there was limited documentation.	Success with strengthening supportive supervision was limited for a variety of reasons, such as increasing workload without a corresponding investment in human resources and a lack of institutionalization of supervision within HSA job descriptions.	Save the Children provided support through mentoring district teams and the provision of supervision tools, which were adopted nationally. Some investments in supervision addressed issues of staff motivation, the value in health facility coordination meetings and the quality of data reporting.
Health information systems	Save the Children conducted trainings on data collection and reporting; however, there was no evidence on the use of data. Most data focused on activity and output reporting and were collected in a parallel system developed by United Nations International Children’s Fund.	Save the Children made efforts to transition to government-owned, digital systems, but faced difficulties with implementation. Although overall impact was hard to quantify, planning and problem-solving occurred in conjunction with the MOH, which is indicative of positive progress.	Save the Children invested in data collection tools, processes and use of information for decision-making; however, persistent gaps in the health information systems remained, including the inability integrate APE information into the HMIS.
Community organizations and societal partnerships	In the Amhara, Oromia and SNNPR regions, a lack of documentation made system-level improvements to societal partnerships and accountability mechanisms difficult to assess. In the other two project sites, accountability mechanisms and government-community linkages were successfully improved.	Community Mobilization Teams and Village Health Committees were supported and trained to mobilize community resources. However, reporting was weak overall and there were some missed opportunities with community engagement.	Community engagement through Community Health Committees (CHCs) was reported as an area where the MOH strategic plan aligned with the project designs. Project efforts with CHCs evolved towards greater systems strengthening approaches over the two project phases, providing support to APEs and potentially voice to communities.
Inclusivity and equity of the health system	In the Somali region, there was a focus on rural, poor and under-privileged communities, especially women. Unfortunately, activities were not sustained and documentation was lacking throughout all project sites.	Though the projects were designed with disadvantaged groups in mind, there was little mention of their inclusion in documentation.	The iCCM and the APE revitalization programmes aimed to reach isolated and vulnerable communities with limited access to health services, which helped improve the equity of the health system. However, persistent drug stockouts made equity in access to care difficult to attain.

Thematic area	Ethiopia	Malawi	Mozambique
National coordination and policy setting	National commitment to iCCM existed, but government involvement varied between the project areas. Although Save the Children successfully helped standardize iCCM procedures in one project area, either operations or reporting periods were too short in other areas to demonstrate sustained impact.	Save the Children coordinated with the MOH to align with national policies; make decisions related to drug selection and training; and standardize reporting and supervision guidelines. However, district-level buy-in varied.	Save the Children supported the national iCCM programme through tool development and training; however, there were strong limitations in country capacity to coordinate iCCM efforts. The sustainability of these coordination efforts is uncertain.
Costing and financing	The Ethiopian government was involved in financing iCCM activities at the national level; however, there were only two notable instances of new domestic funding for iCCM. There were no future plans for financial sustainability or scale-up.	Large-scale financing improvements were beyond the scope of these projects, which allowed system weaknesses to persist. However, Save the Children was successful with some gap-filling and system strengthening activities, like purchasing drugs and supporting the transition to malaria rapid diagnostic tests (mRDTs).	There was no evidence for a clear, nationally owned plan for financing iCCM. However, Save the Children made attempts to integrate national-level financing plans into the iCCM programmes, some of which were successful.
Human resources	Development of human capacity was an area of strength for Save the Children. Amhara, Oromia, or SNNPR focused on training health workers, while the Somali region introduced performance review and clinical mentoring meetings to increase capacity for problem solving.	Save the Children most successfully built human capacity (through standardized trainings) and supported facilities (through logistical improvements). However, structural human resources issues persisted, including workload, supervision and difficulty reach isolated areas.	Save the Children contributed to improvements in human capacity through curriculum development, but there were also many complex challenges that were outside of the mandate of any individual project. These included changes in CHW roles, expanding tasks for CHWs, staff attrition and supervision capacity.
Supply chain management	Government systems remained at early stages of development of autonomous capacity, for both procurement and supply issues. iCCM projects sought to make positive, downstream contributions where and when possible, but were limited in capacity to forecast essential supplies and stock management.	National capacity for supply chain management was not shown to improve, but Save the Children sought to find opportunities to make a strategic difference through a combination of gap-filling and systems strengthening interventions.	Save the Children used two different strategies to support the supply chain management system. In the first phase, they ran a relatively successful parallel drug procurement and supply system. In the second phase (under different funding), they aligned with the MOH without mandate or resources to strengthen the system, which led to pervasive drug stockouts.
Quality of service delivery and referral	There were improvements in the perceived responsiveness of the health system to community needs, primarily through HEWs and Community Health Volunteers. However, the linkages between the health posts and PHC units were weak, which made referrals difficult.	Improvements to quality were made in terms of increasing access to close-to-home services, including mRDTs. Available quality indicators on HSA performance showed mixed results, although they compared positively with other public health services.	Quality improvements included increasing availability of care close to remote communities, access and responsiveness of the health system. APE proficiency improved and their referrals increased in number over the course of the project.
Supportive supervision and management capacity	Save the Children provided trainings for individual staff to build capacity for supportive supervision across all regions, except Somali. All projects carried a de facto role in building management capacity, but there was limited documentation.	Success with strengthening supportive supervision was limited for a variety of reasons, such as increasing workload without a corresponding investment in human resources and a lack of institutionalization of supervision within HSA job descriptions.	Save the Children provided support through mentoring district teams and the provision of supervision tools, which were adopted nationally. Some investments in supervision addressed issues of staff motivation, the value in health facility coordination meetings and the quality of data reporting.
Health information systems	Save the Children conducted trainings on data collection and reporting; however, there was no evidence on the use of data. Most data focused on activity and output reporting and were collected in a parallel system developed by United Nations International Children’s Fund.	Save the Children made efforts to transition to government-owned, digital systems, but faced difficulties with implementation. Although overall impact was hard to quantify, planning and problem-solving occurred in conjunction with the MOH, which is indicative of positive progress.	Save the Children invested in data collection tools, processes and use of information for decision-making; however, persistent gaps in the health information systems remained, including the inability integrate APE information into the HMIS.
Community organizations and societal partnerships	In the Amhara, Oromia and SNNPR regions, a lack of documentation made system-level improvements to societal partnerships and accountability mechanisms difficult to assess. In the other two project sites, accountability mechanisms and government-community linkages were successfully improved.	Community Mobilization Teams and Village Health Committees were supported and trained to mobilize community resources. However, reporting was weak overall and there were some missed opportunities with community engagement.	Community engagement through Community Health Committees (CHCs) was reported as an area where the MOH strategic plan aligned with the project designs. Project efforts with CHCs evolved towards greater systems strengthening approaches over the two project phases, providing support to APEs and potentially voice to communities.
Inclusivity and equity of the health system	In the Somali region, there was a focus on rural, poor and under-privileged communities, especially women. Unfortunately, activities were not sustained and documentation was lacking throughout all project sites.	Though the projects were designed with disadvantaged groups in mind, there was little mention of their inclusion in documentation.	The iCCM and the APE revitalization programmes aimed to reach isolated and vulnerable communities with limited access to health services, which helped improve the equity of the health system. However, persistent drug stockouts made equity in access to care difficult to attain.

Table 2

Summary of results from the county case studies in Ethiopia, Malawi and Mozambique

Thematic area	Ethiopia	Malawi	Mozambique
National coordination and policy setting	National commitment to iCCM existed, but government involvement varied between the project areas. Although Save the Children successfully helped standardize iCCM procedures in one project area, either operations or reporting periods were too short in other areas to demonstrate sustained impact.	Save the Children coordinated with the MOH to align with national policies; make decisions related to drug selection and training; and standardize reporting and supervision guidelines. However, district-level buy-in varied.	Save the Children supported the national iCCM programme through tool development and training; however, there were strong limitations in country capacity to coordinate iCCM efforts. The sustainability of these coordination efforts is uncertain.
Costing and financing	The Ethiopian government was involved in financing iCCM activities at the national level; however, there were only two notable instances of new domestic funding for iCCM. There were no future plans for financial sustainability or scale-up.	Large-scale financing improvements were beyond the scope of these projects, which allowed system weaknesses to persist. However, Save the Children was successful with some gap-filling and system strengthening activities, like purchasing drugs and supporting the transition to malaria rapid diagnostic tests (mRDTs).	There was no evidence for a clear, nationally owned plan for financing iCCM. However, Save the Children made attempts to integrate national-level financing plans into the iCCM programmes, some of which were successful.
Human resources	Development of human capacity was an area of strength for Save the Children. Amhara, Oromia, or SNNPR focused on training health workers, while the Somali region introduced performance review and clinical mentoring meetings to increase capacity for problem solving.	Save the Children most successfully built human capacity (through standardized trainings) and supported facilities (through logistical improvements). However, structural human resources issues persisted, including workload, supervision and difficulty reach isolated areas.	Save the Children contributed to improvements in human capacity through curriculum development, but there were also many complex challenges that were outside of the mandate of any individual project. These included changes in CHW roles, expanding tasks for CHWs, staff attrition and supervision capacity.
Supply chain management	Government systems remained at early stages of development of autonomous capacity, for both procurement and supply issues. iCCM projects sought to make positive, downstream contributions where and when possible, but were limited in capacity to forecast essential supplies and stock management.	National capacity for supply chain management was not shown to improve, but Save the Children sought to find opportunities to make a strategic difference through a combination of gap-filling and systems strengthening interventions.	Save the Children used two different strategies to support the supply chain management system. In the first phase, they ran a relatively successful parallel drug procurement and supply system. In the second phase (under different funding), they aligned with the MOH without mandate or resources to strengthen the system, which led to pervasive drug stockouts.
Quality of service delivery and referral	There were improvements in the perceived responsiveness of the health system to community needs, primarily through HEWs and Community Health Volunteers. However, the linkages between the health posts and PHC units were weak, which made referrals difficult.	Improvements to quality were made in terms of increasing access to close-to-home services, including mRDTs. Available quality indicators on HSA performance showed mixed results, although they compared positively with other public health services.	Quality improvements included increasing availability of care close to remote communities, access and responsiveness of the health system. APE proficiency improved and their referrals increased in number over the course of the project.
Supportive supervision and management capacity	Save the Children provided trainings for individual staff to build capacity for supportive supervision across all regions, except Somali. All projects carried a de facto role in building management capacity, but there was limited documentation.	Success with strengthening supportive supervision was limited for a variety of reasons, such as increasing workload without a corresponding investment in human resources and a lack of institutionalization of supervision within HSA job descriptions.	Save the Children provided support through mentoring district teams and the provision of supervision tools, which were adopted nationally. Some investments in supervision addressed issues of staff motivation, the value in health facility coordination meetings and the quality of data reporting.
Health information systems	Save the Children conducted trainings on data collection and reporting; however, there was no evidence on the use of data. Most data focused on activity and output reporting and were collected in a parallel system developed by United Nations International Children’s Fund.	Save the Children made efforts to transition to government-owned, digital systems, but faced difficulties with implementation. Although overall impact was hard to quantify, planning and problem-solving occurred in conjunction with the MOH, which is indicative of positive progress.	Save the Children invested in data collection tools, processes and use of information for decision-making; however, persistent gaps in the health information systems remained, including the inability integrate APE information into the HMIS.
Community organizations and societal partnerships	In the Amhara, Oromia and SNNPR regions, a lack of documentation made system-level improvements to societal partnerships and accountability mechanisms difficult to assess. In the other two project sites, accountability mechanisms and government-community linkages were successfully improved.	Community Mobilization Teams and Village Health Committees were supported and trained to mobilize community resources. However, reporting was weak overall and there were some missed opportunities with community engagement.	Community engagement through Community Health Committees (CHCs) was reported as an area where the MOH strategic plan aligned with the project designs. Project efforts with CHCs evolved towards greater systems strengthening approaches over the two project phases, providing support to APEs and potentially voice to communities.
Inclusivity and equity of the health system	In the Somali region, there was a focus on rural, poor and under-privileged communities, especially women. Unfortunately, activities were not sustained and documentation was lacking throughout all project sites.	Though the projects were designed with disadvantaged groups in mind, there was little mention of their inclusion in documentation.	The iCCM and the APE revitalization programmes aimed to reach isolated and vulnerable communities with limited access to health services, which helped improve the equity of the health system. However, persistent drug stockouts made equity in access to care difficult to attain.

Thematic area	Ethiopia	Malawi	Mozambique
National coordination and policy setting	National commitment to iCCM existed, but government involvement varied between the project areas. Although Save the Children successfully helped standardize iCCM procedures in one project area, either operations or reporting periods were too short in other areas to demonstrate sustained impact.	Save the Children coordinated with the MOH to align with national policies; make decisions related to drug selection and training; and standardize reporting and supervision guidelines. However, district-level buy-in varied.	Save the Children supported the national iCCM programme through tool development and training; however, there were strong limitations in country capacity to coordinate iCCM efforts. The sustainability of these coordination efforts is uncertain.
Costing and financing	The Ethiopian government was involved in financing iCCM activities at the national level; however, there were only two notable instances of new domestic funding for iCCM. There were no future plans for financial sustainability or scale-up.	Large-scale financing improvements were beyond the scope of these projects, which allowed system weaknesses to persist. However, Save the Children was successful with some gap-filling and system strengthening activities, like purchasing drugs and supporting the transition to malaria rapid diagnostic tests (mRDTs).	There was no evidence for a clear, nationally owned plan for financing iCCM. However, Save the Children made attempts to integrate national-level financing plans into the iCCM programmes, some of which were successful.
Human resources	Development of human capacity was an area of strength for Save the Children. Amhara, Oromia, or SNNPR focused on training health workers, while the Somali region introduced performance review and clinical mentoring meetings to increase capacity for problem solving.	Save the Children most successfully built human capacity (through standardized trainings) and supported facilities (through logistical improvements). However, structural human resources issues persisted, including workload, supervision and difficulty reach isolated areas.	Save the Children contributed to improvements in human capacity through curriculum development, but there were also many complex challenges that were outside of the mandate of any individual project. These included changes in CHW roles, expanding tasks for CHWs, staff attrition and supervision capacity.
Supply chain management	Government systems remained at early stages of development of autonomous capacity, for both procurement and supply issues. iCCM projects sought to make positive, downstream contributions where and when possible, but were limited in capacity to forecast essential supplies and stock management.	National capacity for supply chain management was not shown to improve, but Save the Children sought to find opportunities to make a strategic difference through a combination of gap-filling and systems strengthening interventions.	Save the Children used two different strategies to support the supply chain management system. In the first phase, they ran a relatively successful parallel drug procurement and supply system. In the second phase (under different funding), they aligned with the MOH without mandate or resources to strengthen the system, which led to pervasive drug stockouts.
Quality of service delivery and referral	There were improvements in the perceived responsiveness of the health system to community needs, primarily through HEWs and Community Health Volunteers. However, the linkages between the health posts and PHC units were weak, which made referrals difficult.	Improvements to quality were made in terms of increasing access to close-to-home services, including mRDTs. Available quality indicators on HSA performance showed mixed results, although they compared positively with other public health services.	Quality improvements included increasing availability of care close to remote communities, access and responsiveness of the health system. APE proficiency improved and their referrals increased in number over the course of the project.
Supportive supervision and management capacity	Save the Children provided trainings for individual staff to build capacity for supportive supervision across all regions, except Somali. All projects carried a de facto role in building management capacity, but there was limited documentation.	Success with strengthening supportive supervision was limited for a variety of reasons, such as increasing workload without a corresponding investment in human resources and a lack of institutionalization of supervision within HSA job descriptions.	Save the Children provided support through mentoring district teams and the provision of supervision tools, which were adopted nationally. Some investments in supervision addressed issues of staff motivation, the value in health facility coordination meetings and the quality of data reporting.
Health information systems	Save the Children conducted trainings on data collection and reporting; however, there was no evidence on the use of data. Most data focused on activity and output reporting and were collected in a parallel system developed by United Nations International Children’s Fund.	Save the Children made efforts to transition to government-owned, digital systems, but faced difficulties with implementation. Although overall impact was hard to quantify, planning and problem-solving occurred in conjunction with the MOH, which is indicative of positive progress.	Save the Children invested in data collection tools, processes and use of information for decision-making; however, persistent gaps in the health information systems remained, including the inability integrate APE information into the HMIS.
Community organizations and societal partnerships	In the Amhara, Oromia and SNNPR regions, a lack of documentation made system-level improvements to societal partnerships and accountability mechanisms difficult to assess. In the other two project sites, accountability mechanisms and government-community linkages were successfully improved.	Community Mobilization Teams and Village Health Committees were supported and trained to mobilize community resources. However, reporting was weak overall and there were some missed opportunities with community engagement.	Community engagement through Community Health Committees (CHCs) was reported as an area where the MOH strategic plan aligned with the project designs. Project efforts with CHCs evolved towards greater systems strengthening approaches over the two project phases, providing support to APEs and potentially voice to communities.
Inclusivity and equity of the health system	In the Somali region, there was a focus on rural, poor and under-privileged communities, especially women. Unfortunately, activities were not sustained and documentation was lacking throughout all project sites.	Though the projects were designed with disadvantaged groups in mind, there was little mention of their inclusion in documentation.	The iCCM and the APE revitalization programmes aimed to reach isolated and vulnerable communities with limited access to health services, which helped improve the equity of the health system. However, persistent drug stockouts made equity in access to care difficult to attain.

Save the Children maintained consistent coordination with the MOH in Ethiopia and Malawi by aligning with current policies and by standardizing procedures for the implementation of iCCM. In Mozambique, iCCM revealed limitations to coordination at different levels and with various stakeholders. Costing and financing efforts by Save the Children were largely missing at the national level across all three countries. One of the strongest contributions by Save the Children was building human resource capacity by training existing cadres of CHWs in all three countries on iCCM procedures [i.e. Agentes Polivalentes Elementares (APEs) in Mozambique, Health Extension Workers (HEWs) in Ethiopia and Health Surveillance Assistants (HSAs) in Malawi]; however, CHW workload and retention were common struggles in each project. Supply chain management was challenging for most countries due to the complex nature of drug forecasting, procurement and delivery. In Ethiopia, United Nations International Children’s Fund (UNICEF) was responsible for drug procurement; however, Mozambique and Malawi tried various strategies (both parallel systems and systems that aligned with the MOH) to increase the capacity to manage drug supplies at the national level, with limited success. With respect to quality of service delivery and referral, the main contribution across all three countries was improving access to service in hard-to-reach areas by building the capacity of CHWs. Mozambique and Ethiopia reported improvements in quality of care by APEs and HEWs (respectively), whereas Malawi reported mixed HSA performance results. Save the Children also built supervision and management capacity through training and the provision of supervision tools in all three countries; however, it was difficult to improve supervision rates in Malawi. Health information systems primarily focused on using data for decision-making; however, gaps persisted in terms of data collection and processing in Mozambique and Ethiopia. In Malawi, Save the Children used successive projects to strengthen the health information system for iCCM. Engagement of community organizations was evident in Mozambique to advance iCCM efforts; however, strong engagement with community structures was not well documented in Ethiopia and Malawi. With respect to inclusivity and equity, Mozambique and one project in Ethiopia focused on reaching isolated, vulnerable communities; however, inclusion and equity were not mentioned in Malawi or in the other Ethiopia reports.

Findings from Cross-Country analysis

We focus on findings that have the potential to help rethink or optimize project designs and advance our theoretical understanding of HSS. These fall under three main themes.

Projects had circumscribed influence on health systems and operated on a continuum of health system strategies, ranging from gap-filling to strengthening

Very few functions of the health system, if any, were fully addressable by these iCCM projects, especially when there was no mandate to strengthen the health system. Projects attempted to maximize their contribution to HSS, but only within the strategic space defined by the donor or enabled by the MOH. The same strategic space created incentives to address different levels of immediate vs long-term outcomes. Projects were also limited by idiosyncratic factors in their performance, such as the non-governmental organization’s (NGO) position in the community of development partners, key staff changes and relationships with the MOH. In Malawi, there was a recruitment freeze on HSAs in March 2017, which limited the overall staffing of the project sites. One of the key informants reported that Save the Children was unable to influence the recruitment system nor criteria for recruitment. Also, despite extensive trainings for HSAs and supervisors, the HSAs ultimately stopped providing services, mostly due to transfers to new areas that were not part of Save the Children’s mandate to focus on hard‐to‐reach areas.

The iCCM projects belonged to a loosely linked system of influencing agencies interacting with the health system, including donors, implementers and the government. Positioning of the project and insufficient visibility within the eco-system of DAH partners and government leaders was sometimes seen as a constraint for projects to address challenges. There were often signs of tension between: (i) strengthening efforts not linked to official objectives of the project, but tacitly guiding project decisions, vs (ii) trying to meet the donor-driven objectives of the project as efficiently as possible. In Mozambique, during the first phase of the project [funded by the Canadian International Development Agency (CIDA)], Save the Children established a parallel drug procurement and supply system that led to timely delivery of essential supplies. During the next phase, alignment to MOH systems—required by the government and donor, without mandate or resources to strengthen the system—led to pervasive drug stockouts, especially in more remote regions, and ultimately failure of the project in reaching its objectives.

Successive iCCM projects used a patchwork of strategic and opportunistic (trial-and-error) approaches to engage health systems. Their stance was pragmatic, sometimes gap-filling (which can also be considered a duplicative role, using parallel systems), to genuine strengthening efforts. Key informants described an underlying, but sometimes implicit, intention of programmes to strengthen and sustain health system improvements, even when sub-optimal choices were made. This resulted in a corresponding mix of achievements in terms of strengthening some elements of the health systems at times and filling emergent gaps in the health system at different times. This relationship was not unidirectional along the HSS continuum but went both forward and backward. In Malawi, some project investments were clearly gap-filling or substitutive (e.g. purchasing antimalarial drugs after withdrawal of the Global Fund), while others could be seen as helping the health system become more efficient in its performance (e.g. supporting the transition to malaria rapid diagnostic tests used by HSAs).

While there were limits to the influence of projects on health systems, some contributions were consistent, such as developing human capacity (e.g. CHWs, supervisors), building understanding and skills in data management and use of information, and supporting community systems (e.g. community groups and CHWs) in relation to subnational health systems. Projects, however, had to operate without always being able to influence the coherence of the overall trajectory of the health system.

There was an interdependent relationship between projects and health systems

Projects operated with frequent adaptations of focus, based on systems conditions and events. The health systems provided both opportunities for strengthening the iCCM delivery infrastructure, as well as significant constraints—some of which were unresolvable, while others guided adaptations to the health system. The interactions between projects and health systems elements might in the end matter far more for long-term impact, than initial evaluations of objectives. The Mozambique country report illustrates this interdependence. We observed both system effects of the project, and effects of the system’s behaviours on the project. Another way to frame this is the identification of elements revealing adaptation of the system to project interventions, and adaptation of the project to systems conditions and behaviours. Adaptation of the project to systems conditions included the decision to take drug procurement and supply in hand during the first phase of the project (funded by CIDA), and the Rapid Access Expansion project modification of its training programme to address the high turnover rate of APEs. Adaptation of the system to project interventions followed when facilities were found to lack the drugs now available at community level (through the iCCM program), ultimately resulting in the revisions of the essential drug list.

Given this interdependency, alignment with the MOH was neither static nor simple. Projects generally sought to align with country policies and strategic directions for the health sector. However, national governments and health ministries were themselves pluralistic and alignment was complicated and frequently changed. Projects supported strategic changes in implementation to align with a policy at one level, while bringing a disruption at another level. The revitalization programme in Mozambique required APEs to be mobile, rather than attached to a health post. With the introduction of iCCM, this mobility became an issue: APEs were now unable to properly store and transport iCCM medications and the lack of a fixed post made it difficult for district health facility staff to find them for supervision visits. Communities vocalized the need for fixed health posts so they, too, could find the APEs and bring sick children. A key informant suggested that, in fact, most APEs had kept their posts despite the MOH revitalization policy to close them.

Imperfect alignment was not always entirely negative. Projects created disruptions in the health system, such as introducing a new product or approach, which did not necessarily align with all MOH policies or stakeholders' operations (as these perspectives may not have been internally coherent at a given time in a long process of change). This may be considered sub-optimal from a grand-strategy perspective, but these disruptions also influenced positive changes in parts of the health system, which the project could not directly affect. In Mozambique, iCCM‐trained APEs were provided dispersible amoxicillin and zinc to treat iCCM illnesses (as warranted by policy), although these medications had not yet been introduced into the National Health System and were not even available in most health facilities (a clear operational gap upstream from community services). Ultimately, the MOH was prompted to include these drugs in the public system to avoid conflicts where an APE would have better quality products than the health facility. A local disruption caused a positive central response in this case.

Evaluation gaps limited opportunities to assess health systems effects

In all three countries (and for nearly all projects reviewed), limitations in the evaluation limited opportunities to assess contributions to HSS. This came with notable exceptions, when complementary research efforts took place during implementation of early projects, focusing on how the intervention (i.e. iCCM) was being integrated in the health system, as a contribution to the global community seeking to scale up iCCM.

Gaps in evaluation were evident and constrained important potential achievements of the projects, notably because it is difficult for health systems to know what to sustain if there is no evaluation. These weaknesses were clear, but also reflected donor and national-level priorities. In Ethiopia, e.g. there was no robust learning and evaluation plan developed for the iCCM projects; they mostly focused on reporting about activities and outputs. In Malawi, gaps in evaluation put limits on the diffusion of lessons from practice, outside of the direct implementers’ institutional memory, and raised questions about future institutionalization by national stakeholders.

Discussion

Based on the findings from the country case studies and our cross-country analysis, our study made substantial contributions to the theoretical understanding of the role of projects in HSS by not only documenting evidence of systems strengthening in multi-year iCCM projects but also emphasizing important deficiencies in systems strengthening efforts. We found consistent coordination with the MOH at national and district levels, the capacity of CHWs and supervisors was strengthened and the quality of services was improved, especially in hard-to-reach areas. However, there were limitations to these projects individually and as a whole. Namely, costing and financing information was largely missing, supply chain management was complex and challenging, health information systems did not establish robust evaluation and learning plans, and not all projects intentionally engaged communities or explicitly address equity and inclusion. Our cross-country analysis provided another level of analysis and theory-building related to HSS, which revealed the limits of projects’ influence on health systems as well as the interdependence between projects and health systems, which led to a broad range of health system approaches—from gap-filling to strengthening. Our discussion largely spans beyond the specifics of iCCM programmes and focuses on opportunities for improving the systems effects of programmes in general, which may be relevant to other PHC interventions.

Everybody’s business

Our findings support the WHO’s declaration that HSS is ‘everybody’s business’ [World Health Organization (WHO) 2007]. The dynamic interdependency between countries, DAH partners and projects themselves defines the space for systems strengthening. In fact, while DAH and projects seek to influence national systems, there is a hierarchy of influences between these different actors, which starts with the country’s leadership. Country systems are the fundamental basis on which everything is built. National cohesion, clarity of vision and strong governance will create conditions reducing the natural risk of fragmentation, or perhaps entropy of DAH (OECD, 2008). Consequently, DAH projects will have greater potential for systems strengthening when country systems are already strong enough to provide leadership—a ‘rich-get-richer’ principle illustration. The in-country DAH architecture’s coherence will in turn condition the project’s fit and potential for systems’ strengthening. This creates the space where projects’ strategic choices are expressed. The project implementing agency can also influence how the project makes its mark, notably by its access to the State and the DAH influencers, and by providing the capacity for constancy of performance of the project (e.g. staffing, problem solving and management systems). While this hierarchy of influences does not entirely pre-empt projects from having positive systems effects, this creates a lot of uncertainty and sub-optimal options for implementers. At a minimum, projects and DAH partners must make every effort to be explicit about the boundaries of their respective spheres of influence. HSS efforts are bounded within each sphere by determinants addressed in another space or level. As national systems gain strength and coherence, project opportunities for strengthening the system will evolve, while the allowance for disruptive innovations will become smaller, and fragmentation can be mitigated through stronger national leadership. This ‘chicken-and-egg’ issue might be mitigated, however, if evaluation efforts are carried out, and processes set in place to allow DAH and national partners to develop and share actionable lessons from these efforts. This is tantamount to calling for even more collaborative learning and shared accountability, something not easily achieved beyond statements of intention. It may however be a requirement to improve the systems strengthening orientation of projects.

Finding the fit of mid-size projects in solving the systems strengthening equation

First, we need to develop a dynamic view of projects and their systems effects. Before developing solutions, we need to understand the context and the stakes. Health systems are not simply the objects of interventions by projects but evolve at the interplay between prior conditions, decisions of national actors, DAH partners, including donors, and the choices of projects along the way. In addition, projects cannot simply ‘align with the country’; they must negotiate and resolve tensions in this alignment. This also entails an element of unavoidable risk, e.g. in introducing disruptions. Systems strengthening is fundamentally disrupting to a status quo. This does not mean that innovations and disruptions should be pursued indiscriminately. However, since they can shift institutional norms with positive ripples on the health system, they should be approached through a learning and adaptation perspective between stakeholders (Story et al., 2017).

Given the complexity of health systems, it would be tempting to propose giving the ‘systems strengthening’ mandate to large scale, heavily financed projects (i.e. ranging from $5 to $10 million USD per year) and leave mid-size projects (i.e. ranging from $500 000 to $3 million USD per year) to focus on donor-driven deliverables. We question the ethical and strategic value of this possible division, given the cumulative mass of effects and influences that mid-size projects can have on evolving health systems. Countries and DAH partners cannot just bet on heavily financed projects to solve the systems strengthening equation. Research into DAH mechanisms may provide a perspective on the cumulative weight of these different types of projects. However, having no collective effort between countries, donors and projects to make sure that these projects find their fit (Chambers et al., 2013), and at least contribute to systems strengthening, would be a tremendous wasted opportunity for lower- and lower-middle-income countries.

Our overarching conclusion leads to four principles for action (Box 2), which emphasize that even mid-size projects—apart from the most limited, pinpoint interventions and acute humanitarian emergencies—should optimize their contribution to the edification of viable health systems. This may sound like an excessively noble aspiration, and we offer suggestions for moving forward, beyond acknowledging the enduring centrality of the principles of policy alignment, respect for country ownership, and doing no harm.

Box 2

Principles for action

All mid-size projects should and can contribute to HSS.
Projects cannot, however, promise everything; they may need to limit, but should clearly define, their contribution.
Strengthening community platforms is integral to HSS.
Both design and evaluation must address the ‘messiness’, or complexity, of the implementing environment.

Projects will need guidance to improve the fit of their interventions with the overall process of the development of health systems. Design approaches should help to map what is their central point of entry into the health system, what is in and what is out of their sphere of influence, what responsibilities are shared with which stakeholders, what are direct and secondary systems relations they commit to monitoring and what are the essential upstream (e.g. central MOH departments for a district) and downstream (e.g. health post, CHWs and citizens’ groups for a PHC facility) systems relations to consider. In a way, this direction goes beyond iCCM projects and can build on the existing literature about using systems thinking to approach system strengthening (Savigny et al., 2009; Peters, 2014; Reynolds et al., 2018).

Technical and service innovations should come with ample attention to operational requirements (rationalization of processes, task sharing and workload distribution, management at the subnational level), long-term cost (both efficiency and resourcing) and national governance. Our study showed examples of these efforts, including an implicit orientation of projects towards opportunities for improving decision processes and streamlining data collection tools, supervision and coordination mechanisms. The question is whether these efforts were intensive enough, with sufficient attention to the ability of the health system to evaluate and learn from them. At a time when global health language has moved from ‘strength’ to ‘resilience’ of systems (Kruk et al., 2017), we may need more research and better guidance on the space that projects need to allow for these operational issues—to be specific, the improvement of operational capabilities within national health systems. Projects, implementing agencies, or coordination mechanisms within which projects operate should proactively engage MOH and partners in considering efficiencies and scenarios for long-term resourcing. This implies at least consideration of national governance issues, either directly by the project, through advocacy by its implementation agency, or through DAH-government coordination bodies. Donors and implementing agencies may want to harmonize and coordinate the distribution of responsibilities, but decentralized management capabilities such as participatory planning, budgeting and human resource management should receive greater and more explicit attention in the era of ‘transition’ and USAID’s ‘journey to self-reliance’ (USAID, 2020).

Realism and transparency

The counterpart to these recommendations is the need to combine realism with transparency. Projects should be intentional and transparent about their positioning choices, between systems support and systems strengthening (and occasionally gap-filling operations). Projects should intentionally plan and implement—possibly localized and partial, but clearly defined—system strengthening contributions, based on context and the implementing agency’s sphere of influence. This will require defining explicitly what a project can and cannot do and goes against the marketing and proposal tendencies to ‘oversell’ projects.

All stakeholders may benefit from a better acknowledgement (in design, ambitions and evaluation) of the spectrum of intervention needs, ranging from gap-filling interventions to support, to actual strengthening (Chee et al., 2013; Save the Children, 2017). This acknowledgement could feed into more shared learning, accountability and benchmarking of stages of transition away from ‘aid dependency’, to avoid the type of performance failure due to stockouts observed in the second phase of the Mozambique project, e.g. project contributions should be mapped against stages of maturity of the local and national health system in their different components. Projects could be better integrated within health sector plans, not only at national, but also subnational, levels. Creating the collective learning space where implementing agencies are incentivized to share plans with other stakeholders and report on challenges, needed adaptations and even failures represents obvious challenges but is at the heart of many global calls for partnership, adaptation and learning (Naimoli and Saxena, 2018).

Without evaluation, there is no link between ‘strengthening’ and ‘strengthened’

Whether projects had or did not have the mandate to be more ambitious with their evaluation, we must observe that weaknesses in evaluation undermine the potential for project achievements to provide learning by the health system (Sarriot et al., 2021). This raises complex questions for projects, health systems duty bearers, implementing agencies and DAH partners about who should generate that information.

Claims of ‘systems strengthening’ are bound within the quality of evaluation and learning about changes in the health system provided by projects. The long-term value of the best designed and implemented project is limited by its contribution to the health system’s own learning and evolution, and this depends heavily on the quality of evaluation. Project evaluation should itself become clearly nested within health systems and HSS evaluation questions (Adam et al., 2012; Carden, 2013; Bennett et al., 2015). There is an important role for governments, donors and DAH coordination mechanisms to invest in evaluation and collective learning processes. Implementing agencies have a stake in advancing best practices, influencing donors and showing their added value to the country. NGOs and other implementers should continue to partner with the growing community of researchers and evaluators developing approaches to defining their system-in-focus, carrying out political economy analysis, mapping of actors, participatory learning processes, use of theory of change, group modelling, scenario-based prospective assessments and realist evaluations to guide adaptive programmatic choices (Sarriot et al., 2015; Paina et al., 2017; Peters, 2018; Reynolds et al., 2018; Reich, 2019). If health systems depend on becoming learning systems (Naimoli and Saxena, 2018), then the field of HSS certainly faces the same requirement for collective learning and sense-making (Adam et al., 2012; Cleary et al., 2018; Schneider et al., 2020).

Limitations

Our series of country case studies and cross-country analysis was not meant to be an evaluation of a project or group of projects, but an exploratory examination of health systems effects of successive projects. This study is a retrospective assessment of the manifestation of health system effects in iCCM projects and the interplay between the influence of iCCM projects on health systems and vice versa. As such, it offers limited opportunity to promote generalizable solutions to future iCCM projects as well as PHC projects as a whole.

Each case study was based on existing project documentation, which was difficult to track down at times and had its own limitations. We found that documentation was focused on the projects’ mandates and donors’ reporting guidelines, and therefore, we could not always find conclusive evidence for the systems strengthening theoretical propositions. Even processes and procedures that may have strengthened the health system, such as end-of-project transitions and sustainability plans, were not always described in the documentation that we reviewed and, therefore, corresponding measures of success were not available. To overcome this limitation, we employed key informant interviews to fill information gaps, make corrections about the sequence of events and help us understand the overall health system impacts of a group of iCCM projects. These proved to be helpful; however, finding informants with a clear memory of these projects proved to be difficult given high project staff turnover and the remote implementation of the study. In the future, more field-based research and additional key informants, notably from national institutions, would be helpful.

Regarding the case study methodology, the field of HSS is struggling to advance a clear evaluation model (Adam et al., 2012). We used theoretical propositions, expanding on evaluation questions from previous authors (Chee et al., 2013), to construct a multidimensional narrative on the health system effects of multiple iCCM projects; however, we did not have a set of reliable, quantifiable measures of change. We feel the case study methodology is useful but would be strengthened by the addition of a set of standardized HSS metrics.

Conclusion

Ultimately, national system actors have the leading role in articulating the overall vision for the health system and its operationalization. Governments and health ministries need to be aware of their own complexity and the process of change across bureaucratic layers, to ensure that development contributions are coherent and positive. They need to embrace a learning agenda, which includes a fair dose of uncertainty. They need to accept that DAH partners are themselves trying to find their way towards systems strengthening and that countries and governments can optimize the value of projects for their own benefit.

Projects, such as Save the Children’s iCCM interventions in our study, have potential for increasing their contribution to the health systems they serve. This potential is linked to their fit with national and DAH systems and they face a number of challenges. They also depend on our ability to move from the rhetoric of systems strengthening and resilience to innovations for collective learning and shared accountability. New design and embedded evaluation and learning approaches are critically needed to tilt the scale towards systems strengthening and to avoid missed opportunities.

Supplementary data

Supplementary data are available at Health Policy and Planning online.

Conflict of interest statement. SP and ES were employees of Save the Children (respectively UK and US), which funded this project. We declare no other potential conflict of interest.

Ethical approval. Since the study used secondary data without identifiers, ethical review was not necessary.

Funding

This study was funded by Save the Children USA’s Department for Global Health (DGH). The authors’ views expressed in this publication do not necessarily reflect the views of Save the Children.

Acknowledgement

Our thanks to our key informants for their valuable insights about each of the iCCM projects. The documentation we used in the case studies would not have existed, if not for the support of donors and implementing partners, mentioned in the individual country reports. Thank you to Robert Clay for the support of our ambitious efforts to unpack the strengthening of health systems throughout Save the Children’s iCCM projects. We also extend our appreciation to iCCM technical experts who reviewed many versions of this study, including Tanya Guenther, Jeanne Koepsell, and Eric Swedberg.

References

Adam

T

,

Hsu

J

,

De Savigny

D

et al.

2012

.

Evaluating health systems strengthening interventions in low-income and middle-income countries: are we asking the right questions?

Health Policy and Planning

27

:

iv9

–

iv19

.

Adamo

M

,

Rivera

D

,

Shah

R

et al.

2016

.

Time volunteered on community health activities by brigadistas in Nicaragua

.

Revista Panamericana de Salud Publica

40

:

388

–

95

.

Bagonza

J

,

Rutebemberwa

E

,

Eckmanns

T

,

Ekirapa-Kiracho

E.

2015

.

What influences availability of medicines for the community management of childhood illnesses in central Uganda? Implications for scaling up the integrated community case management programme Health policies, systems and management in low and middle-income countries

.

BMC Public Health

15

.

Bekker

LG

,

Alleyne

G

,

Baral

S

et al.

2018

.

Advancing global health and strengthening the HIV response in the era of the Sustainable Development Goals: the International AIDS Society—Lancet Commission

.

The Lancet

392

:

312

–

58

.

Bennett

S

,

George

A

,

Rodriguez

D

et al.

2014

.

Policy challenges facing integrated community case management in Sub-Saharan Africa

.

Tropical Medicine & International Health

19

:

872

–

82

.

Bennett

S

,

Dalglish

SL

,

Juma

PA

,

Rodríguez

DC.

2015

.

Altogether now ⋯ understanding the role of international organizations in iCCM policy transfer

.

Health Policy and Planning

30

:

ii26

–

ii35

.

Black

RE

,

Taylor

CE

,

Arole

S

et al.

2017

.

Comprehensive review of the evidence regarding the effectiveness of community-based primary health care in improving maternal, neonatal and child health: 8. summary and recommendations of the expert panel

.

Journal of Global Health

7

.

Callaghan-Koru

JA

,

Gilroy

K

,

Hyder

AA

et al.

2013

.

Health systems supports for community case management of childhood illness: lessons from an assessment of early implementation in Malawi

.

BMC Health Services Research

13

.

Campbell

J

,

Admasu

K

,

Soucat

A

,

Tlou

S.

2015

.

Maximizing the impact of community-based practitioners in the quest for universal health coverage

.

Bulletin of the World Health Organization

93

:

590

–

590A

.

Carden

F.

2013

.

Evaluation, not development evaluation

.

American Journal of Evaluation

34

:

576

–

9

.

Chambers

DA

,

Glasgow

RE

,

Stange

KC.

2013

.

The dynamic sustainability framework: addressing the paradox of sustainment amid ongoing change

.

Implementation Science

8

.

Chee

G

,

Pielemeier

N

,

Lion

A

,

Connor

C.

2013

.

Why differentiating between health system support and health system strengthening is needed

.

The International Journal of Health Planning and Management

28

:

85

–

94

.

Chilundo

BGM

,

Cliff

JL

,

Mariano

ARE

,

Rodríguez

DC

,

George

A.

2015

.

Relaunch of the official community health worker programme in Mozambique: is there a sustainable basis for iCCM policy?

Health Policy and Planning

30

:

ii54

–

ii64

.

Cleary

S

,

Erasmus

E

,

Gilson

L

et al.

2018

.

The everyday practice of supporting health system development: learning from how an externally-led intervention was implemented in Mozambique

.

Health Policy and Planning

33

:

801

–

10

.

Collins

DH

,

Jarrah

Z

,

Gilmartin

C

,

Saya

U.

2014

.

The costs of integrated community case management (iCCM) programs: a multi-country analysis

.

Journal of Global Health

4

.

Dalglish

SL

,

George

A

,

Shearer

JC

,

Bennett

S.

2015a

.

Epistemic communities in global health and the development of child survival policy: a case study of iCCM

.

Health Policy and Planning

30

:

ii12

–

ii25

.

Dalglish

SL

,

Surkan

PJ

,

Diarra

A

,

Harouna

A

,

Bennett

S.

2015b

.

Power and pro-poor policies: the case of iCCM in Niger

.

Health Policy and Planning

30

:

ii84

–

ii94

.

Daniels

K

,

Sanders

D

,

Daviaud

E

,

Doherty

T.

2015

.

Valuing and sustaining (or not) the ability of volunteer community health workers to deliver integrated community case management in northern Ghana: a qualitative study

.

PLoS One

10

:

e0126322

.

Doherty

T

,

Zembe

W

,

Ngandu

N

et al.

2015

.

Assessment of Malawi's success in child mortality reduction through the lens of the Catalytic Initiative integrated health systems strengthening programme: retrospective evaluation

.

Journal of Global Health

5

.

Fernandes

QF

,

Wagenaar

BH

,

Anselmi

L

et al.

2014

.

Effects of health system strengthening on under-5, infant, and neonatal mortality: 11-year provincial-level time-series analyses in Mozambique

.

The Lancet Global Health

2

:

e468

–

e477

.

George

A

,

Menotti

EP

,

Rivera

D

,

Marsh

DR.

2011

.

Community case management in Nicaragua: lessons in fostering adoption and expanding implementation

.

Health Policy and Planning

26

:

327

–

37

.

George

A

,

Rodríguez

DC

,

Rasanathan

K

,

Brandes

N

,

Bennett

S.

2015

.

ICCM policy analysis: strategic contributions to understanding its character, design and scale up in sub-Saharan Africa

.

Health Policy and Planning

30

:

ii3

–

ii11

.

Kruk

ME

,

Gage

AD

,

Arsenault

C

et al.

2018

.

High-quality health systems in the Sustainable Development Goals era: time for a revolution

.

The Lancet. Global Health

6

:

e1196

–

e1252

.

Kruk

ME

,

Ling

EJ

,

Bitton

A

et al.

2017

.

Building resilient health systems: a proposal for a resilience index

.

BMJ (Online)

357

.

Kutzin

J

,

Sparkes

SP.

2016

.

Health systems strengthening, universal health coverage, health security and resilience

.

Bulletin of the World Health Organization

94

:

2

.

Legesse

H

,

Degefie

T

,

Hiluf

M

et al.

2014

.

National scale-up of integrated community case management in rural Ethiopia: implementation and early lessons learned

.

Ethiopian Medical Journal

52

:

15

–

26

.

Marsh

DR

,

Tesfaye

H

,

Degefie

T

et al.

2014

.

Performance of Ethiopia's health system in delivering integrated community-based case management

.

Ethiopian Medical Journal

52

:

27

–

35

.

McGorman

L

,

Marsh

DR

,

Guenther

T

et al.

2012

.

A health systems approach to integrated community case management of childhood illness: methods and tools

.

The American Journal of Tropical Medicine and Hygiene

87

:

69

–

76

.

Munos

M

,

Guiella

G

,

Roberton

T

et al.

2016

.

Independent evaluation of the rapid scale-up program to reduce under-five mortality in Burkina Faso

.

The American Journal of Tropical Medicine and Hygiene

94

:

584

–

95

.

Naimoli

JF

,

Saxena

S.

2018

.

Realizing their potential to become learning organizations to foster health system resilience: opportunities and challenges for health ministries in low- and middle-income countries

.

Health Policy and Planning

33

:

1083

–

95

.

Nanyonjo

A

,

Counihan

H

,

Siduda

SG

et al.

2019

.

Institutionalization of integrated community case management into national health systems in low- and middle-income countries: a scoping review of the literature

.

Global Health Action

12

:

1678283

.

National Statistical Office (NSO) [Malawi] and ICF

.

2017

.

Malawi Demographic and Health Survey 2015-16

.

Zomba, Malawi, and Rockville, Maryland, USA

.

NSO and ICF

.

Nefdt

R

,

Ribaira

E

,

Diallo

K.

2014

.

Costing commodity and human resource needs for integrated community case management in the differing community health strategies of Ethiopia, Kenya and Zambia

.

Ethiopian Medical Journal

52

:

137

–

49

.

Nsona

H

,

Mtimuni

A

,

Daelmans

B

et al.

2012

.

Scaling up integrated community case management of childhood illness: update from Malawi

.

The American Journal of Tropical Medicine and Hygiene

87

:

54

–

60

.

OECD

.

2008

. The Paris Declaration on Aid Effectiveness and the Accra Agenda for Action 2005/2008. https://www.oecd.org/dac/effectiveness/34428351.pdf, accessed 21 December 2020.

Olivas

E

,

Story

W.

2019

. Systems effects of integrated community case management projects. Country report 2 of 3: Ethiopia. Save the Children. https://resourcecentre.savethechildren.net/node/15091/pdf/hss_iccm-ethiopia_final.pdf, accessed 21 December 2020.

Paina

L

,

Wilkinson

A

,

Tetui

M

et al.

2017

.

Using Theories of Change to inform implementation of health systems research and innovation: experiences of Future Health Systems consortium partners in Bangladesh

.

Health Research Policy and Systems

15

.

Peters

DH.

2014

.

The application of systems thinking in health: why use systems thinking?

Health Research Policy and Systems

12

.

Peters

DH.

2018

.

Health policy and systems research: the future of the field

.

Health Research Policy and Systems

16

.

Pritchard

S

,

Ume

A

,

Hinds

G

,

Sarriot

E.

2019

. Systems effects of integrated community case management projects. Country report 3 of 3: Malawi. Save the Children. https://resourcecentre.savethechildren.net/node/15092/pdf/hss_iccm-malawi_final.pdf, accessed 21 December 2020.

Reich

MR.

2019

.

Political economy analysis for health

.

Bulletin of the World Health Organization

97

:

514

.

Reynolds

M

,

Sarriot

E

,

Swanson

RC

,

Rusoja

E.

2018

.

Navigating systems ideas for health practice: towards a common learning device

.

Journal of Evaluation in Clinical Practice

24

:

619

–

628

.

Rodríguez

DC

,

Banda

H

,

Namakhoma

I.

2015b

.

Integrated community case management in Malawi: an analysis of innovation and institutional characteristics for policy adoption

.

Health Policy and Planning

30

:

ii74

–

ii83

.

Rodríguez

DC

,

Shearer

J

,

Mariano

ARE

et al.

2015a

.

Evidence-informed policymaking in practice: country-level examples of use of evidence for iCCM policy

.

Health Policy and Planning

30

:

ii36

–

ii45

.

Sacks

E

,

Morrow

M

,

Story

WT

et al.

2019

.

Beyond the building blocks: integrating community roles into health systems frameworks to achieve health for all

.

BMJ Global Health

3

:

e001384

.

Sadruddin

S

,

Pagnoni

F

,

Baugh

G.

2019

.

Lessons from the integrated community case management (iCCM) Rapid Access Expansion Program

.

Journal of Global Health

9

:

020101

.

Sarriot

E

,

Hejna

E.

2019

. Systems effects of integrated community case management projects. Country report 1 of 3: Mozambique. Save the Children. https://resourcecentre.savethechildren.net/node/15090/pdf/hss_iccm-mozambique_final_0.pdf, accessed 21 December 2020.

Sarriot

E

,

Olivas

E

,

Khalsa

S

,

Ahraf

S

et al.

2021

.

Health systems effects of successive emergency health and nutrition projects: an embedded retrospective case study analysis in Sudan and Pakistan

.

Health Policy and Planning

. doi: 10.1093/heapol/czaa096.

Sarriot

E

,

Morrow

M

,

Langston

A

et al.

2015

.

A causal loop analysis of the sustainability of integrated community case management in Rwanda

.

Social Science & Medicine (1982)

131

:

147

–

155

.

Save the Children

.

2017

. Building Viable and Resilient Systems for Health: Save the Children Department of Global Health Strategic Orientations 2017–2020. https://resourcecentre.savethechildren.net/library/building-viable-and-resilient-systems-health-save-children-department-global-health, accessed 21 December 2020.

Save the Children US

.

2020

. Our Reach in 2019. Internal Reporting.

Savigny

D

,

Adam

T.

;

Alliance for Health Policy and Systems Research, WHO

.

2009

.

Systems Thinking for Health Systems Strengthening

.

Geneva

:

WHO

.

Schneider

H

,

George

A

,

Mukinda

F

,

Tabana

H.

2020

.

District Governance and Improved Maternal, Neonatal and Child Health in South Africa: pathways of Change

.

Health Systems and Reform

6

.

Seidman

G

,

Atun

R.

2017

.

Does task shifting yield cost savings and improve efficiency for health systems? A systematic review of evidence from low-income and middle-income countries

.

Human Resources for Health

15

.

Siekmans

K

,

Sohani

S

,

Boima

T

et al.

2017

.

Community-based health care is an essential component of a resilient health system: evidence from Ebola outbreak in Liberia

.

BMC Public Health

17

.

Smith

S

,

Deveridge

A

,

Berman

J

et al.

2014

.

Task-shifting and prioritization: a situational analysis examining the role and experiences of community health workers in Malawi

.

Human Resources for Health

12

.

Story

WT

,

LeBan

K

,

Altobelli

LC

et al.

2017

.

Institutionalizing community-focused maternal, newborn, and child health strategies to strengthen health systems: a new framework for the Sustainable Development Goal era

.

Globalization and Health

13

.

Sunguya

BF

,

Mlunde

LB

,

Ayer

R

,

Jimba

M.

2017

.

Towards eliminating malaria in high endemic countries: the roles of community health workers and related cadres and their challenges in integrated community case management for malaria: a systematic review

.

Malaria Journal

16

:

14

.

Tulenko

K

,

Møgedal

S

,

Afzal

MM

et al.

2013

.

Community health workers for universal health-care coverage: from fragmentation to synergy

.

Bulletin of the World Health Organization

91

:

847

–

852

.

UN Inter-agency Group for Child Mortality Estimation

.

2018

. Under-five, infant and neonatal mortality rate [Data set]. http://www.un.org/en/development/desa/population/publications/mortality/child-mortality-report-2018.shtml, accessed 21 December 2020.

USAID

.

2020

. The journey to self-reliance. Helping countries to solve their own development challenges. https://www.usaid.gov/selfreliance, accessed 21 December 2020.

van Olmen

J

,

Marchal

B

,

Van Damme

W

et al.

2012

.

Health systems frameworks in their political context: framing divergent agendas

.

BMC Public Health

12

:

774

.

WHO

.

2017

.

Healthy Systems for Universal Health Coverage A Joint Vision for Healthy Lives

.

Geneva

:

WHO and International Bank for Reconstruction and Development/The World Bank

.

WHO

.

2018

.

WHO Guideline on Health Policy and System Support to Optimize Community Health Worker Programmes

.

Geneva

:

WHO

.

Witter

S

,

Palmer

N

,

Balabanova

D

et al.

2019

.

Health system strengthening—reflections on its meaning, assessment, and our state of knowledge

.

The International Journal of Health Planning and Management

34

:

e1980

–

e1989

.

World Health Organization (WHO)

.

2007

.

Everybody’s Business: Strengthening Health Systems to Improve Health Outcomes: WHO’s Framework for Action

.

Geneva

:

WHO

. https://www.who.int/healthsystems/strategy/everybodys_business.pdf, accessed 21 December 2020.

Yin

RK.

2014

.

Case Study Research: Design and Methods

.

Thousand Oaks, CA

:

Sage

.