Comprehensive and integrated district health systems strengthening: the Rwanda Population Health Implementation and Training (PHIT) Partnership

Utilization and health outcomes in Rwanda, 2005-2009

By 2009, the Rwandan government had improved access to primary health services mainly through infrastructure development and expansion of a community-based health insurance program. Geographic access to care was relatively well distributed, with 75% of the nation’s population living within 5 kilometers of a health facility [6]. Although one of the world’s poorest countries, Rwanda created the Mutuelle de Santé health insurance program, which provided access to basic primary health care services for 85% of its citizens by 2009 [7]. Nevertheless, there was a marked shortage of trained health care workers with 0.04 generalist physicians and 0.62 nursing professionals per 1,000 people [6]. Due, in part, to increased health care access, health outcomes in Rwanda had improved since 2000 [8]. Under-5 mortality, a useful gauge of health system impact on overall child health [9], fell from 152 deaths per 1,000 live births in 2005 to 103 in 2008 (five-year mortality estimates) [10]. Consistent with patterns in other low- and middle-income countries, child and infant mortality rates declined more steeply than neonatal mortality; from 2001-2005, neonatal mortality was 37 per 1,000 live births annually [1], compared with 28 per 1,000 live births from 2004-2008 [10]. Concurrent with increased uptake of family planning services, the total fertility rate declined from 6.1 in 2005 to 5.5 in 2008. Ninety-six percent of pregnant women received some antenatal care and 45% delivered at a health facility in 2008, an increase from 26% who delivered at a health facility in 2000. Utilization by children was lower, with 35% of children with a fever and 33% of children with diarrhea treated at a health facility [10].

The Rwandan health system

Rwanda is divided into 30 districts, and 85% of the country’s 10.8 million people live in rural areas [11]. The architecture of the district health system is typical for a developing country setting (Figure 1). Each district is served by a network of community health workers (CHWs) — three per village — offering health education, basic preventive and curative services, and family planning. CHWs are supported by local health centers, which serve approximately 20,000 people and are staffed by nurses, most of whom have a secondary school education level. Health centers provide vaccinations, reproductive and child health services, acute care, and diagnosis and treatment of HIV, tuberculosis, and malaria. District hospitals, staffed in part by 10-15 generalist physicians, provide more advanced care, including basic surgical services, such as caesarean sections. District pharmacies procure essential medicines and consumables from a central agency and distribute them to all health facilities within the district. District health units are responsible for administrative management of the district health system, while district hospitals are responsible for clinical supervision and monitoring and evaluation of all district health facilities.

The Rwanda Population Health Implementation and Training (PHIT) Partnership

In 2009, the Rwanda Population Health Implementation and Training (PHIT) Partnership was established with the support of the Doris Duke Charitable Foundation’s African Health Initiative. The PHIT Partnership designed a multidimensional and integrated district-level HSS intervention that aligned with Rwanda’s 2009 National Health Sector Strategic Plan and the WHO Health Systems Framework. The partnership endeavors to improve the capacity and performance of the health system in southern Kayonza and Kirehe districts with targeted financial and technical investments focused on the health centers, coupled with quality improvement (QI) initiatives designed to improve both service delivery and strengthen monitoring and evaluation (M&E) systems. We hypothesize that the intervention will result in demonstrable improvements in population health, exceeding the already considerable health gains being made nationally.

The PHIT Partnership aspires to create a replicable, evidence-based road map for district-level primary health care delivery in rural Africa and to disseminate lessons learned within and beyond Rwanda’s borders. Accordingly, the partnership has developed an integrated model of implementation, operational research, and impact evaluation to facilitate refinement and rigorous evaluation of the intervention.

The PHIT Partnership is a public-private-academic collaboration comprised of six institutions. The Rwandan MOH and PIH lead the implementation and jointly fund the health system in the intervention area. The Ministry of Health sets national health policies and oversees the health management information system (HMIS) and planned national data warehouse, which provide data for both routine M&E and the impact evaluation. Harvard Medical School (HMS) and staff from Brigham and Women’s Hospital (BWH) faculty support evaluation activities, as well as capacity building for M&E and research components of the partnership. The National University of Rwanda School of Public Health (NUR-SPH) takes part in the evaluation team and leads the research capacity building program. Finally, the National Institute of Statistics of Rwanda (NISR) coordinates and implements the demographic and health surveys (DHS) and other key data collection activities, including the national census and the economic living conditions survey used for the impact evaluation.

Guiding principles of the Rwanda PHIT Partnership

The Rwanda PHIT Partnership is guided by five key principles that we believe are essential elements of successful HSS interventions. First, health systems interventions should be comprehensive, as enshrined in the 1978 conference on International Primary Health Care in Almaty, Kazakhstan, through the Alma Ata Declaration [12]. Second, integrated interventions that apply a “systems thinking” approach deliver more value than isolated vertical interventions [13]. Third, interventions should be locally responsive, reflecting the engagement of local government leaders, civil society, and the community. Fourth, sustainability of a public health approach can only be achieved with efforts to strengthen institutions. Fifth, health systems should place a premium on equity by addressing issues of access, quality, and cost.

Geographic scope of the intervention

The Rwanda PHIT intervention area comprises two PIH-supported rural districts — southern Kayonza and Kirehe, which are contiguous over an area of roughly 3000 km² in the Eastern Province of Rwanda (Figure 2). The intervention catchment area comprises 480,000 people, with nearly 2,800 CHWs, 23 health centers, and two district hospitals.

Health and social indicators in the Eastern Province, which includes the PHIT Partnership catchment area, were markedly worse than the rest of the country at the beginning of the project period. In the 2008 interim DHS, southern Kayonza and Kirehe had a total fertility rate of 6.2, compared to 5.6 in other rural areas. The intervention area also had a lower percentage of women completing four standard antenatal care visits (15.6% versus 24.0%). Under-5 mortality remained higher in the intervention area than other rural areas (125.5 vs. 107.9) [10].

Although the two intervention districts are poor, with only 4.4% of households having electricity and the majority of households relying on wood as the main cooking fuel, 78.8% of women and 90.7% of men were employed, predominantly in the agricultural sector [14]. Within southern Kayonza and Kirehe districts, 59.8% of women aged 15-49 had completed some primary education, compared to 62.6% of men [15]. However, due to a strong government commitment to education access, the net primary school attendance ration had reached 87% of children nationally by 2010 [15].

Component 1. Targeted instrumental support for health facilities

Without adequate infrastructure — equipment, staffing, financial resources, procurement and information systems, ambulance network, and management capacity — process interventions designed to improve service delivery at health facilities are likely to fail. Therefore, we designed this intervention component to promote a flexible, evidence-based approach to improve health system readiness.

Component 2. Quality improvement initiatives

The PHIT intervention emphasizes the improvement of service quality through two main pathways: the Mentorship and Enhanced Supervision of Health Centers (MESH) program and support for M&E systems, focusing on ensuring effective routine monitoring and improved data quality and utilization.

Component 3. Strengthened network of community health workers

A cornerstone of the Rwanda PHIT Partnership is the strengthening of the community health system based on a network of multidisciplinary CHWs. The emphasis on a strong network of CHWs reflects the Rwandan government’s recognition of the importance of decentralizing health care delivery to the community level as a means of encouraging communities to take responsibility for their own health, as well as promoting prevention activities, early detection of illness, and increased utilization of health services. The PHIT Partnership targets three major areas of the community health system: a strong supervision system to ensure CHW quality of care and reporting; implementation of a household register to improve quality through systematic documentation of household visit activities; and the expansion of home-based care provided for chronic disease patients, including those with HIV.

Under the national system, CHWs conduct monthly home visits to households in their village, providing health education, screening for specific diseases, encouraging health center visits, and monitoring child growth and development. During household visits and through routine community sensitization events, CHWs educate communities on key health messages, including hygiene and sanitation, accessing care at facilities, malaria prevention, and general health.

For logistical reasons, additional support to the community health worker network is being conducted in only one of the two intervention districts: southern Kayonza. CHWs in the intervention district receive ongoing training and supervision focused on primary health care, reproductive health, family planning, and malnutrition. During routine household visits, the CHWs receive regular supervision visits in which they are given feedback on quality of care delivered and on the quality of the data recorded during the visit. In each health center catchment area, CHWs belong to a cooperative that receives modest performance-based financial incentives.

Project timeline

The timeline of implementation and evaluation activities is presented in Figure 4.

Implementation progress to date

By the end of the project’s second year (June 2011), all components of the PHIT intervention were successfully implemented. All 21 health centers and two district hospitals in the intervention area were adequately resourced, and the three core quality initiatives — the MESH program, joint M&E and data quality initiatives, and enhanced community health delivery — were operational. Based on early operational evidence of success, several intervention components are being evaluated by the Ministry of Health for potential integration into the national framework.

Overall impact evaluation

The impact evaluation includes four major components: an analysis of the difference in differences in key population health indicators between 2010 and 2015; assessment of trends in routinely collected indicators from the national HMIS; assessment of changes in the causes of death in children measured through verbal autopsies; and analysis of contextual factors that may impact district-level differences.

Economic and costing analysis

The main purpose of the economic analysis is to understand how the PHIT program contributes to local health system financing. The evaluation measures 1) the costs per capita of PHIT investment in each of the six WHO health system building blocks; 2) total cost per capita of the health system in the two intervention districts; 3) health facility spending by WHO building block; and 4) financial contributions to the local health system made by government, the PHIT Partnership, other partners, and patients. We have designed survey instruments for each facility type to measure funding sources, facility expenditure, and existing capital. Cost data have been collected for the baseline year (2010) and the surveys will be repeated annually through 2014.

We will conduct trend analyses in key variables over time, including the level and trends of total health system and PHIT spending per capita in the two districts; percentage of total health expenditures contributed by PHIT, government investment, and other sources; and the level and trends of expenditure on each of the six WHO building blocks by funding source. We will use regression analysis to evaluate the impact of PHIT spending on health care provision using the provision of child health services as the outcome variable.

Component evaluations of the PHIT intervention

Operational evaluations will focus on key components of the PHIT intervention — targeted instrumental support to health facility building blocks, quality improvement through the MESH program and strengthened M&E systems, and enhancements to the community health system. These evaluations will serve two purposes: use of the data for modification and improvement of the interventions throughout the implementation period and exploration of the relative contributions of each component intervention on health outcomes to inform replication and scale-up of the Rwanda PHIT intervention.

Research capacity building through operational research

In order to address a critical skills gap in Rwanda, an operational research capacity building effort has been embedded within the PHIT evaluation plan. Led by National University of Rwanda School of Public Health and Harvard Medical School, we launched an MPhil and PhD research training program, known as Rwanda PHIT Scholars, for 10 Rwandan health professionals involved in the PHIT Partnership. Candidates underwent a competitive selection process that involved the development of operational research proposals to complement the PHIT impact evaluation and intervention component evaluations. Each student has been assigned an advisory team, including a NUR-SPH advisor, an HMS faculty advisor, and a PHIT Rwanda site supervisor.

Among the student-led operational projects are assessments of family planning uptake, performance of the electronic medical record (EMR) currently under development by the PHIT Partnership, the impact of Rwanda’s e-health architecture, the impact of HIV care delivery on maternal and child health services, the impact of the PHIT intervention on childhood nutritional status and health facility staff retention, and ethnographic assessment of family planning utilization.