Brand New Day: Newly Launched Nationwide PBF Scheme in Burundi Reflects the Hopes of a Nation

By Lindsay Morgan[1]

 

Picture Burundi in 1996. For the third time in a decade the government is turned upside down: the president has been overthrown in a military-led coup. This, even as the country is still reeling from the murder of its two previous presidents—one in 1993 and another in 1994. The coup prompts the international community to impose economic sanctions. GDP falls by about 8 percent. The Burundian franc plummets. The government institutes fuel rationing, and all over the countryside, civilians are caught up in a brutal and bloody battle. 

 

Since independence in 1962, ethnic and political conflicts have resulted in five wars, and left tens of thousands massacred and close to two million displaced or fleeing to neighboring countries. The effects on Burundi’s economic and social fabric are staggering: between 1990 and 2002, per capita income in the tiny, landlocked nation falls from $210 to $110, making Burundi the poorest country in the world. The proportion of people living below the poverty line increases from 35 to 68 percent,[2]^,[3] and the health and education systems are shattered.

 

Flash forward to 2009, and I’m sitting in the Burundian capital of Bujumbura with Olivier Basenya, a senior member of the national technical support unit for performance-based financing (PBF), within the Burundian Ministry of Health (MOH). Lake Tanganyika is only a few hundred yards away; its cool, silvery vastness conveys a tranquillity that belies the country’s troubled past. It is the last day of a week-long workshop on PBF and Basenya looks tired. “There are lots of problems in the health sector in Burundi,” he says over coffee. “Maternal and child health, not enough doctors, motivation among health workers is low.”

 

Since holding its first successful post-war democratic election in 2005, Burundi has enjoyed several years of relative calm. But health (and many other) indicators remain grim. The country’s maternal mortality ratio is stunningly high, hovering at about 615 per 100,000 live births—the second highest in sub-Saharan Africa.[4] HIV/AIDS, the second most common cause of mortality among adults, compounds health woes. Infant and under-five mortality rates are exceptionally high as well, at 108 and 180 per 1,000 live births respectively.[5] 

Of all poor countries with a legacy of war, Burundi has been close to the bottom of the pile for decades. Which makes it all the more astonishing and impressive that the country is instituting one of the most far-reaching health reform efforts in Africa.

 

Following several (mostly) successful PBF pilots, the government made an ambitious decision: to scale it up across the country, in concert with another major reform—free healthcare for pregnant women and children under the age of five. For Basenya, these efforts are about disrupting the quantifiable order of things: “There’s a lot we hope to change.” Launched in April 2010, expectations surrounding PBF in Burundi are rich—and the challenges are many.

Following elections in 2005, the Burundian government began the laborious process of tackling the many problems facing the country. Refugees began to return home. Attempts were made to address land reform, rehabilitate infrastructure, and institute financial reforms. School fees for primary education were scrapped. Mirroring what was done in education, in May 2006, in an effort to jump start improvements in maternal and child health, the government abolished health service fees for children under five and pregnant women at public health facilities. Use of health services subsequently ballooned: the number of medical consultations involving children increased by 42 percent and the health facility delivery rate increased from 22.9 percent in 2005 to 56.3 percent in 2008.[6]  

 

Facilities were ill prepared: there were shortages of equipment, drugs, and qualified staff, and government reimbursements to health facilities were often delayed. Angry and tired, health workers organized several month-long strikes, protesting poor working conditions and low wages, and many left public facilities for better-paying private sector jobs.    

 

Six months later, two Dutch NGOs—HealthNet-TPO and CORDAID—began implementing a handful of PBF pilots in three provinces in partnership with the MOH, in an effort to address those supply side constraints.[7] Jean François Busogoro, a program manager of the PBF program in Gitega province who had previously worked on the national reproductive health program, had learned about the schemes’ ability to break the back of persistent health woes in places such as Cambodia, Afghanistan and Rwanda. “The approach was very exciting for me and when CORDAID and HNI-TPO came with PBF programs, I was really interested.

 

In the PBF pilots, health facilities received performance bonuses for both quantitative and qualitative indicators.  Quantity indicators reflected a basic health package and included such things as assisted deliveries, essential vaccinations, bed net distribution and HIV testing. The quality checklist, which contained 154 composite quality indicators, focused mostly on whether conditions for quality were met—on whether there was adequate equipment, infrastructure, hygiene, and sanitation, for example. Not counting the quality-related bonus, health centers typically received between US$700-1,500 per month while district hospitals received between US$3,500-5,000 per month. Bonuses as they related to salaries varied, from between 10-40 percent of worker salaries at hospitals to a much higher proportion in health centers, which sometimes exceeded their salaries.

 

In HNI-TPO areas, contracts were signed between health facilities and provincial steering committees, which were comprised of provincial administration and health authorities, representatives from health facilities, NGOs and social ministries, and civil society. In CORDAID pilot areas, health facilities negotiated contracts with autonomous purchasing agencies (AAPs), which also verified performance data and made the payments, and were comprised of technical, financial and medical representatives. Contracts were renewed every quarter.

 

As part of contract negotiations, each health facility created a business plan in which targets were negotiated. For the quantity indicators, there was no minimum target that must be reached for the bonuses to be paid (or under which the subsidy was not paid). Facilities were paid an amount per service they delivered. For quality indicators, facilities could earn a bonus of up to 15 percent of the amount earned for the quantity indicators, meaning that the better a facility did on quantity, the larger the potential quality payment.  Quantity-related payments were distributed on a monthly basis, while quality payments were made every quarter.  No more than 50 percent of each bonus payment could go towards individual bonuses, and the remainder was intended for service quality improvements (although there were exceptions).  

 

Evaluation of community and patient satisfaction took place in conjunction with local associations, which were also contracted by the autonomous purchasing agents. These associations validated health services, gauged patient satisfaction with services and assessed the extent of patient knowledge. Findings from these quarterly community surveys were fed back to the health care providers, and half of the quality payment depended on their results. 

 

Encouraging Results Whet Country and Donor Appetites  

Overall findings showed an average increase of 50-60 percent for each indicator compared to the period prior to the introduction of PBF. The pilots were evaluated both by an MOH unit and the Royal Tropical Institute of Amsterdam.[8]   Service utilization increased in all pilot areas.  Complete child vaccination coverage increased, as did the rate of assisted births and the number of women who used contraception.   Both found positive results.

 

Michel Bossuyt, CORDAID’s country representative in Burundi, says: “The most interesting thing about PBF is how quickly things change. The dynamic of the health system, the attitude of personnel—they had more initiative, and this was true for other countries too.”

 

In 2008, the pilots were extended to six additional provinces (for a total of 9 out of 17 provinces), and the MOH decided to make PBF an official policy and scale up throughout the country.  Planning began in mid-2009 and the national PBF scheme was launched in April 2010. With an estimated budget of just under US$100 million over three years (of which nearly US$70 million has been secured; approximately $29 million from the Burundian government, $23.5 million from the World Bank, and the rest from various donors[9]), the scheme will cover all qualified health facilities (health centers, district hospitals, tertiary hospitals) in all 17 provinces.  

 

The national scheme mirrors the pilots in several ways but changes will also be made.[10] New indicators for HIV care and services for under-fives and pregnant women will be added, and a Web-based database is being designed, in which provinces will be able to upload monthly data that will form the backbone for data analysis and produce the consolidated invoices.

 

The quality checklist has also been significantly overhauled, though most of the indicators remain focused on whether conditions for quality are met—for example, whether a facility has adequate equipment, even trash cans—but not necessarily on quality per se (staff attitudes, time spent with patients, etc). However, the revised quality checklist is based on lessons learned from Rwanda, where a rigorous impact evaluation of the national scheme showed impressive results, particularly for quality of care.

 

The Burundi PBF scheme also works to ensure equity among vastly unequal provinces and health facilities. The most disadvantaged (i.e., the poorest) provinces will receive a per capita allocation that is 40 percent greater than the per capita allocation of the richest provinces. Similarly, there is a 40 percent variation in the fees each facility can receive: the most disadvantaged health facilities (i.e., those located in poor and/or remote locations) receive unit fees that are 40 percent higher than the most advantaged facilities. 

 

 

One key way in which the national scheme will diverge from the pilots is in the body responsible for negotiating contracts, verifying performance and making payments.

In the new scheme, a mixed group comprised of NGO representatives, medical and technical experts, and crucially, political appointees, will perform these functions. Called the Comités Provinciaux de Vérification et de Validation (CPVV), these bodies will contract health facilities and local associations, verify and validate results and approve the payments, acting as a sort of decentralized governing board for PBF.

 

Having several key roles reside in one body that includes political appointees has provoked controversy. Says CORDAID’s Piet Vroeg: “The biggest challenge is to avoid that the institutional set up will allow the Ministry of Health to verify its own results. It is our opinion that the main functions in a health care system (finance, delivery and verification) should be neatly separated. No institute should be allowed to verify its own results. This should be done by an independent party. In Burundi there is a reflex of the MOH to keep the function of result verification within the Ministry, often for very noble reasons.”

 

Because of concerns, in the provinces where CORDAID and HNI-TPO are already working, the AAPs will continue to work alongside these CPVVs, and will provide assistance to the new provinces in the form of training and technical assistance. After a year, the partners will assess the CPVV and associated structures, and may propose changes.

 

Others are less concerned with the makeup of the CPVV, pointing out that a similar institutional set up worked in neighboring Rwanda, where PBF was scaled up in 2006. “Creating a decentralized decision making platform in which government, together with civil society, look at health system performance is a new and powerful way of working together,” says World Bank senior health specialist Gyuri Fritsche.

 

The CPVVs will also be subject to a performance framework as well as the health facilities they monitor—they will be assessed on contract management, auditing, data verification and validation, regularity in data submission and invoice preparation, among other things. Incentives will be paid each quarter, conditional on good performance. This will create a new kind of civil servant, says Fritsche: “the incentivized bureaucrat.”

 

PBF uses a systematic approach to verification, which the free healthcare policy did not. In addition to the CPVV’s performance framework, an independent third party will be contracted to do counter-verification at all levels of the system: from the community/client level all the way up to the MOH. No formal impact evaluation is planned at this time.

 

“What they [the government of Burundi] are doing is very ambitious,” says Bruno Meessen, a health economist at the Institute of Tropical Medicine. “Free healthcare and PBF.”

 

There are a cascade of issues that arise from trying to merge the two reforms, which amount to a major shift for the country’s entire health system. For example, facilities were being paid more by the government for certain services provided to under-fives and pregnant women than they will be paid for PBF indicators, therefore, the former are being subsumed into PBF and their prices adjusted. 

 

Most health center and hospital staff seem supportive of the new system despite the major changes it implies. But some have raised concerns that their facilities’ revenues may be reduced, perhaps even to the point of endangering the survival of the facilities. One point of contention, for example, is the reduced tariff for Caesareans, which have become a significant source of income for some hospitals. The government was paying health facilities 200,000 Burundian francs ($163) per caesarian section, but under the new PBF system, they will pay 60,000 francs ($49).

 

“One of the reasons to decrease the payment for caesarian sections, for example, is to avoid sending the wrong financial signals to providers,” says Fritsche. “PBF attempts to ‘balance’ the payments to providers, focusing their attention on those essential public health services that are frequently neglected by both providers and their clients.” In other countries, high payments for Caesarians have led to an excessively high number of Caesarians being reported, due either to their being performed when not needed, or to falsification, or both.

 

World Bank Task Team Leader Sunil Rajkumar says concerns about revenue are understandable but exaggerated: “If they [facilities] reform and perform they will get more money than before.” Indeed, simulations have shown that the reduced income for Caesareans would be more than compensated for by increased income from other interventions, assuming adequate performance.

 

And despite lower payments for some services, the overall budget has been substantially increased under PBF and will be spread among multiple services that cover the entire population, instead of only the slice of the population covered by the free healthcare policy. PBF also uses standardized prices for each service, whereas the basis of payment for subsidized health services was simply the cost of inputs based on receipts, which placed a burden on health facilities’ accounting and was vulnerable to fraud.

 

There are also concerns that the government’s payments, if it not delivered quickly enough, could de-motivate health facility staff. Under the NGO-administered pilots, facilities were typically reimbursed in about two weeks; the government hopes it will be able to do so in 51 working days, but their track record in the previous system, of reimbursing facilities for providing free health care, is not encouraging: reimbursements to health facilities were often delayed by as much as six months. (According to Basenya, 30 percent of invoices have not been reimbursed at all.)^[11]

 

Further, PBF requires more work to receive payment than facilities are used to. Under the subsidy system, all facilities must do to receive payment is submit a bill (although the cost for invoicing can be high—facilities have to photocopy invoices and document all drugs and other materials exhaustively). In the new system, data must be collected and verified.

 

“The biggest concern with the PBF program in Burundi is this high ambition of PBF scaling up in the country,” says Busogoro. “[Some facilities] may say: your money is not easy to get.”

 

Others are worried about what they see as a dearth of technical capacity at the national level. Though the MOH has developed a procedures manual that describes merging PBF with free health care; a training guide and user manual whose content reflects the manual procedures; and trained national and provincial trainers (four per province), the national technical PBF support unit within the Ministry of Health needs strengthening. “Certainly, the implementation will require more expertise with new PBF structures at national and provincial level,” one observer cautioned.

 

But the team working on PBF are determined and persistent—Basenya, Rajkumar and many others have worked tirelessly for months, hammering out details and addressing obstacles as they emerge. The World Bank task team’s presence and leadership in Bujumbura have been particularly vital. Says Benjamin Loevinsohn, Health, Nutrition, & Population Cluster Leader for the Africa Region at the World Bank: “the roll-out simply would not have happened without Bank support.”

 

Just days before the PBF conference began in Bujumbura, there were whispers of an attempted coup: some agitators had been arrested on the shores of the lake and there was a bristle on the streets. Despite the peace deal and several years of relative calm, there is a ripple in Burundi, just below the surface, of uncertainty, a sort of brace for what might happen. You can feel it in the torn up streets, sudden road blocks, the industrial wasteland feel of parts of the city.

 

The International Crisis Group, a Brussels-based think tank, recently warned that tensions could escalate in the run-up to elections—communal, presidential, then legislative—which begin in May and continue through September[12] “ruining the electoral process’s credibility and endangering a fragile democracy and, ultimately, many gains of the peace process.”[13] Opposition parties face harassment and intimidation from police and the ruling party’s youth wing appears dangerously volatile, the report goes on to say.

 

Concerns about stability have insinuated themselves into negotiations on the scheme’s design and implementation. Take the CPVV. Some donor representatives expressed concern about the fragility of a body with such huge responsibilities being led by government political appointees. “The government plan is risky,” says one observer. “If [committee members] have a political function, things could change quickly.”

 

Claude Sekabaraga, formerly of the Rwandan Ministry of Health and now a World Bank official based in Nairobi, says that political leadership was critical to the success of the Rwandan PBF program): “To run a national PBF program you need strong political commitments. There are risks [in Burundi] to have new leaders come and stop the system, because she/he doesn’t believe on linking payment and results.”

 

Michel Bossuyt of CORDAID thinks PBF will survive a change in administrations, “but only if results are positive.”

PBF holds great promise for Burundi. At a reception during the Bujumbura workshop, the Burundian Minster of Health said that PBF is a core priority for the Ministry and requires significant long-term support. But potential pitfalls remain and details of design and implementation must be carefully addressed and revised as lessons are learned along the way. Says Meessen: “PBF has experienced a long maturation process…these programs have to be very carefully designed, piloted, tested, and consensus must be carefully built.”

 

 

---