How did Rwanda implement?

The community health worker (CHW) program

Rwanda inaugurated its community health worker (CHW) program in 1995 to compensate for a severe shortage of doctors, nurses, and midwives in the immediate aftermath of the genocide.

At the start of the program, there were 12,000 CHWs. Within a decade, there were 45,000 - three for each of Rwanda’s approximately 15,000 villages. A CHW trio typically includes one male and one female generalist, plus an animatrice de santé maternelle (ASM) dedicated to maternal health.

Villages - which consist of roughly 100 to 200 households each - elect their own CHWs, who must meet minimum requirements for age and literacy. Once elected, they undergo a standardized six-week training program, with regular refresher sessions thereafter.

CHWs are responsible for general health interventions for the entire community, including for children under five. CHWs are trained in a cascading, train-the-trainer model on their assigned community-health activities, which have varied over time. By 2007, CHWs nationwide had undergone training at their local health centers on home-based management of fever (HBMF; see here for more detail).

For this program, the CHWs were taught signs and symptoms of malaria, treatment of disease, warning signs requiring referral to a health facility, drug management, and reporting systems. They used prepackaged medications with instructions in Kinyarwanda, illustrated with drawings to assist both the CHWs and the children’s care givers in fidelity and adherence to recommended treatment. Their training has expanded over time to address other significant causes of U5M.

CHWs report each month to the local health center on a series of indicators, the data from which is then fed to district and national reporting systems. When dropping off these monthly reports, CHWs also pick up medications and supplies for their villages .¹²

Although the program is embedded within the national health system, and the system is deeply reliant on their work, the CHWs remain unpaid volunteers. Because of this status, there is concern about the program’s sustainability, and worker motivation has been a concern from the start.

To improve retention and morale, Rwanda created a network of CHW-managed cooperatives in 2005, financed through a CHW performance-based financing (PBF) scheme. However, few CHWs have training in finance or management, and the cooperatives have generated little benefit .¹³ To address this problem, the MOH has asked the cooperatives to recruit accountants. In a perhaps more substantive move, the ministry is also overseeing an external evaluation of the CHW model and looking into ways to strengthen it as the nation works to attain the health-related Sustainable Development Goals.

An early and consistent emphasis on primary care and U5M

A consistent theme of the research and interviews was the clarity of the Rwandan national government’s public health priorities and its insistence on disciplined coordination among ministries, regional officials, donors, and other system participants to achieve the twin goals of a strengthened primary care system and reduced rates of U5M.

An emphasis on primary care generally - and U5M specifically - was evident in the government’s Vision 2020 document, which was released in 2000. This ambitious plan was a blueprint for transforming Rwanda into a middle-income nation within two decades, setting goals across multiple sectors. The health portion of Vision 2020 included a specific mandate to reduce U5M and had clear targets against which Rwanda assessed progress annually.¹

According to a former Minister of Health, the intent was to “create a system to build health infrastructure with the objective of equitable geographic distribution, grounded on a strong primary-care platform.”

Some elements would require novel collaborations across ministries. The government created innovative governance structures - such as the Social Cluster for senior officials from the Ministries of Health, Education, and Gender - to meet regularly to develop cross-sectoral policies and plans. This Social Cluster collaborated to address key priorities, including gender-based violence and teenage pregnancies, which required coordination and input across these ministries.

From interviews with key informants and a study of available literature, one feature of Rwanda’s primary-care strategy stands out as critically important. The government held fast to a system-wide, “horizontal” vision of primary-care improvement, placing a priority on such fundamentals as clinic construction; personnel training and retention; high-quality data systems; and vaccination delivery infrastructure.

When well-meaning donors earmarked money for disease-specific “vertical” interventions, Rwanda sought wherever possible to steer those funds toward uses that were consistent with broader systemic improvements.

In time, the government’s focus on primary care would yield immense dividends, including the reduction of U5M. For example, Rwanda’s implementation of the pneumococcal conjugate vaccine was also designed to strengthen future vaccination campaigns, including through cold-chain improvements, monitoring, CHW training, and community engagement. This contributed to the rapid rollout of rotavirus and other vaccines soon afterward. When the government ramped up its efforts against HIV, it designed laboratory and delivery systems so that there was immediate integration with primary-care and other health services.

Beyond the strengthening of its primary-care systems, the payoff of Rwanda’s system-minded approach could be quite literal: By meeting basic primary care benchmarks, Rwanda could better meet donors’ funding parameters and thereby garner still more funding for general health improvements. For example, by meeting Gavi standards for child vaccinations, Rwanda earned additional funds to strengthen its health system at the government’s discretion.² 

A governmental commitment to steering donors and partners toward a unified vision

The relationship between the Rwandan national government and its outside donors and partners is central to understanding the country’s U5M achievements.

One of the most distinctive features of that relationship is Rwanda’s insistence that foreign donors and partners follow - and contribute to - a nationally-led agenda.

This has enabled Rwanda to retain national control of its own health policy while benefitting from foreign insight and technical assistance. It can be difficult for recipient nations to maintain true “country ownership” while also receiving outside funding and expertise, but Rwanda has developed a variety of means for sustaining an effective balance.

One is a quarterly meeting between donors and a group of government representatives led by the Ministry of Finance and Economic Planning (MINECOFIN). These meetings assess the quantity and quality of donors’ development assistance.

Donors who demonstrate particularly high levels of direct financial support for Rwanda’s health sector are permitted to attend the entire budget review process. This has had the dual benefit of keeping important partners apprised of Rwanda’s priorities, while also giving them a forum for sharing useful knowledge.

A related donor-coordination mechanism was the Joint Health Sector Reviews (JHSR), which the Minister of Health co-chairs twice a year along with a leading health donor. The forum brings together ministry officials, academics, hospital directors and other partners to assess progress toward health objectives.

Yet even as the national government held its donors close, it took unambiguous steps to make sure everyone understood who wielded ultimate authority. Even during the fragile first decade following the genocide, when much donor funding for health care was earmarked for disease-specific projects, the nascent government strove to ensure that resources were used to build and strengthen primary health care systems.

One interviewee said that if a donor objected to Rwanda’s terms, the government “would say no and not take the money.” Another described the underlying ethos as “We are in charge and we will tell you what to do. This is a Rwandan program.”

One partner who had worked in several other sub-Saharan nations said such an uncompromising emphasis on donor and partner coordination was far from the regional norm; several observers have identified this approach as an important lesson for other countries.

Development and sustainment of a strong culture of data usage for decision-making

Rwanda’s campaign to reduce U5M has relied heavily on collecting, disseminating, and evaluating data. From an early stage, the post-genocide government has emphasized data as an indispensable tool for identifying and addressing a wide range of public health challenges, including U5M.

One interviewee who had worked both within the MOH and with a partner organization said the shift toward a more data-minded health system took some time. Even though health centers and district hospitals had routinely gathered data and sent it to the MOH, “It [was] not really in the culture to analyze that data, to see what happened, and so on.”

The ministry set out to build a more sophisticated data culture, beginning at the top. It pushed all department heads at the Ministry of Health and the Rwanda Biomedical Center (RBC) to attain master’s degrees, a measure that created a strong internal cadre of technical experts. The ministry has also worked to improve district-level data capabilities, and to empower localities to exploit data for their own decision-making.

These efforts eventually paid off, building data awareness throughout the health policy chain and - as one interviewee put it - enabling the ministry to “produce evidence and science when it may not be available” by generating sound, data-grounded inquiries. In addition, the government was able to draw upon the data expertise of donors and partners through technical working groups.

With these strengthened human and technical capacities in place, the MOH began using data more aggressively to identify gaps and set priorities. In 2010, for example, it reviewed district data to understand where maternal and neonatal deaths were still occurring despite an increase in antenatal care.

This analysis led to an identification of lower-performing districts, whose leaders were then invited to explain both the reasons for their faltering numbers and their plans for improvement.

But the MOH has not merely employed data as a means of exposing underachieving jurisdictions. Consistent with the values of decentralization, the ministry has also worked to improve district-level data capabilities, and to empower localities to exploit data for their own decision-making.

Rwanda’s digital strategy took another important step forward in 2012, when the MOH migrated its Health Management Information System onto a new web-based platform using District Health Information System open-source software (DHIS-2). This made current health data visible to officials and practitioners at all levels of the health system. More than 700 managers and staffers have been trained in DHIS-2, which may be entered from almost any site with an internet connection.³ 

The development of a data usage culture in turn informed Rwanda's development of a strong institutional culture of accountability. For example, the national government invoked the concept of imihigo (meaning "to vow to deliver") through creating performance contracts with districts and ensuring districts met the health-related goals specified in these contracts. As such, the culture of data collection and usage enabled the national government to hold districts accountable to key metrics in line with national priorities. (For more information on Imihigo, see the section in Context here.)

A further example of developing a culture of accountability was through Rwanda's application of health data initiatives into performance-based financing (PBF) mechanisms. The MOH piloted a PBF plan in two districts from 2003 to 2005. The success of this trial led to an expanded study of incentive programs in 166 rural health facilities.⁴

That study found significant improvements in the intervention facilities compared with controls, including deliveries at health facilities (a 23 percent improvement), the number of preventative care visits for children under 23 months old (a 56 percent improvement), and the number of such visits for children 24 to 59 months old (a 132 percent improvement).⁴

However, the study also showed that PBF had no significant impact in certain other areas, such as the number of women completing their recommended cycle of four antenatal care appointments. The bottom line was that PBF seemed to work best for services that had the highest payment rates and were the easiest to deliver.⁴

Based on this and other evidence, Rwanda adopted PBF for district hospitals, public health centers, and government-supported nonprofit health centers as part of the 2005-2009 Health System Strategic Plan. Government payments were based on 13 qualitative and 13 quantitative measures, including some directly related to U5M.

Payments were made quarterly, with 80 percent going to the providers themselves, and the remaining 20 percent covering health centers’ non-labor costs such as medical equipment, ambulances, and facility maintenance.⁵

Between 2003 and 2007, PBF funding increased from $800,000 to $8.9 million, and wages increased by 60 percent to 100 percent depending on the facility.⁵  From 2005 to 2008, personnel in publicly funded facilities almost doubled, with much of the increase occurring at rural facilities that could now offer better salaries.

A focus on health care equity and access

From the start, equity was one of the top health policy priorities for the post-genocide government. For a nation contending with both deep poverty and a legacy of civil strife, this was an understandable and necessary area of emphasis.

This concern manifested itself in relatively high levels of national health expenditure, including the ambitious rollout of a national health insurance plan. Through these and other measures, the Rwandan government sought to ensure that public health progress touched all economic classes and extended to every corner of the nation’s geography.

To a degree that is unusual among low-income nations, Rwanda has significantly reduced the inequality of U5M outcomes among differing income groups.

As of 2000, the U5M rate for Rwanda’s highest economic quintile was 156 per 1,000 live births, compared with 229 per 1,000 for the lowest economic quintile - an equity gap of 73 deaths per 1,000 live births.¹⁶ By 2014, the U5M rate had sharply decreased for both quintiles: 38 per 1,000 for the wealthiest; 83 per 1,000 for the poorest - a differential of 45 per 1,000.¹⁶

Similar reductions in inequality appeared in the coverage rates for specific interventions, according to the Countdown to 2030 composite coverage index (CCI), a weighted average of the coverage of eight interventions along four stages of the continuum of care: reproductive health; maternal health; immunizations; and childhood-illness management.

Rwanda’s vaccination campaigns, in particular, achieved high levels of equity across economic strata from an early stage. This is because these campaigns scaled up quickly nationwide, drew upon needs assessments from every village; and integrated closely with community-health programs. As a result of these factors, there tends to be little geographic or socioeconomic variation in vaccination coverage.

For example, measles vaccination rates vary little between urban and rural populations (96.4 percent and 94.9 percent respectively in 2014) or wealth quintile (91.3 percent for the lowest and 97 percent for the highest in 2014).¹⁶

Data from national demographic health surveys for 2000 to 2014 show impressive geographic parity across Rwanda. IHME maps show that no single region of the country is long left behind the others; while U5M continued to decline, the locations of districts with highest rates of U5M in Rwanda have also shifted over time as gains have progressed across Rwanda. The only area that has consistently shown relatively high rates of U5M is the south-central tier along the border with Burundi.

Thanks in part to its emphasis on decentralization and rural health outreach, Rwanda has narrowed one disparity that bedevils many nations, rich and poor alike - the gap between urban and rural residents.

In 2000, 66 percent of women in Rwanda’s urban areas delivered their babies in a health facility, but only 20 percent of rural women had done so.⁸ By 2014, this margin narrowed as care improved for all groups: 97 percent of women in urban areas were now giving birth in hospitals or clinics - and so were 90 percent of their rural counterparts. ⁶

One factor that has supported geographic equity in Rwanda is the government’s consistent emphasis on health infrastructure development. The 2005 National Health Sector Policy identified such development as a top priority for addressing disparities in access. This led to the construction and expansion of hospitals and health centers nationwide.

By 2012, there were five national referral hospitals, 40 district hospitals, 450 health centers, and 157 private health facilities - a respectable level of coverage for a nation only about the size of Haiti, Macedonia, or the U.S. state of Maryland.⁷

On average, each district now has at least one hospital and one health center per 20,000 residents. The average time to reach a health center was 56.5 minutes in 2014, down from 95.1 minutes in 2006.⁸  All told, 60 percent of Rwanda’s population now lives within five kilometers of a health facility, and 85 percent lives within 10 kilometers.⁹

Mutuelle de Santé health insurance

One of the most important ways that Rwanda has sought to ensure health care equity and access is through Mutuelle de Santé, the national community-based health insurance (CBHI) plan. After conceiving of the idea in the late 1990s, the government tested it for feasibility and impact before rolling it out nationwide in the middle of the following decade.¹⁰ 

In the years since, Rwanda has adapted the system to emphasize coverage of the poorest. Through a Ministry of Local Government-led system known as ubudehe, the families in a village are classified into socioeconomic categories. The families in the lowest categories have their insurance premiums and co-payments funded at all points of care by the government or its development partners. The costs for other enrollees are determined on a sliding scale.¹¹ 

Rwanda’s CBHI is not a universal-coverage program. Some residents (less than 8 percent of the population) remain enrolled in private health-insurance plans, while civil servants, the military, and law-enforcement personnel have their own coverage under a separate scheme. 

Nonetheless, the scope of CBHI is impressive. At peak enrollment, more than 90 percent of Rwanda’s population was in the program; current estimates are around 75 percent. By any standard, Rwanda’s plan has one of the highest enrollment rates of any CBHI among low-income and lower-middle-income nations worldwide.¹⁴

The success of the program has had several positive implications for U5M, both direct and indirect. These include significant reductions in the cost of maternal, newborn, and child health, including for family-planning services.¹⁵