Free and Public

A closer look at the role of Community Based Health Insurance in Rwanda’s success

A woman talks to a doctor about her twin boys at a health centre in Rwanda. Credit: Stringer/OxfamRwanda’s mutuelle health insurance scheme has been consistently held up as an example of how community health insurance can be scaled up to achieve large scale improvements in access and health outcomes. However, the role of the mutuelle scheme in achieving recent health improvements in Rwanda has often been exaggerated without consideration of other important factors (including a five-fold increase in health spending).

This post takes an in-depth look at the mutuelle scheme, raising critical questions about its impact on access to health care and the extent to which mutuelles are really financially sustainable. It shows that while insurance premiums and co-payments are less prohibitive than traditional point-of-service charges, they continue to pose a financial barrier for many Rwandans. International commentators should be careful not to disproportionately underscore the mutuelle’s role in improving health outcomes and oversell community health insurance as a health financing panacea.

User fees for healthcare in poor countries have been an issue of intense controversy since their adoption in the 1980s. In recent years, fees have drawn criticism for the inequitable barrier they create for the poor and the lack of improvement in service delivery that they were expected to produce.

As an alternative, community health insurance programs that shift fees from the point-of-service to prepayments have received increasing attention in Sub-Saharan Africa. In addition to alleviating financial barriers at the time of illness, community insurance programs have also been touted as a ‘sustainable’ approach to health financing.

In practice, however, few countries employing community insurance schemes have achieved meaningful coverage with most having fewer than 10% of the population enrolled.[1]  Government-led programs in Ghana and Rwanda are frequently cited exceptions with Rwanda’s mutuelle reaching 91% coverage in 2010[2]  (although a recent report  has challenged the claimed impact of Ghana’s health insurance scheme). In light of Rwanda’s dramatic improvement in health indicators, the mutuelle is often offered as an impetus for introducing insurance to other countries in the region.

However, the mutuelle deserves closer scrutiny before being promoted as a model for health financing. While Rwanda’s insurance program has surely improved access compared to traditional point-of-service charges,[3]  it is not clear how much of Rwanda’s success can actually be attributed to the mutuelle. Insurance coverage improved from 7% in 2003 to 91% in 2010.[4]  Over the same period, utilization of health services jumped from 0.31 outpatient visits per capita to 0.95.[5] Health indicators made an unprecedented improvement during this time with under-five mortality dropping by half from 15.2% in 2005 to 7.6% in 2010.[6,7]  This is a staggering reduction in just five years and puts Rwanda almost on par with far wealthier countries like India and South Africa.[8]

The correlation between mutuelle coverage and improvements in utilization and indicators, however, is confounded by massive concurrent increases in health spending. In 2002, Rwanda spent just 10 USD per capita on health; by 2010, this increased almost five times to 48 USD per capita.[9,10] Much of this growing investment in health comes from donors who contributed 53% of Rwanda’s health budget in 2006, up from 33% in 2002.[11,12] And mutuelle has had very little to do with this expanding resource envelope; it accounted for only 5% of all health spending in 2006 (and now probably closer to 3%).[13,14] Moreover, at least 20% of mutuelle funds, in fact, come from donor and government subsidies.[15]   

In a recently published study based on a small pilot in 2007, several colleagues and I examined utilization patterns at a health facility in rural Rwanda.  Service delivery at the facility was aggressively upgraded with increased staffing, reinforced drug supplies, new equipment, and improved management. Despite these enhancements, utilization rates remained unchanged compared to neighboring sites that received no upgrades. After mutuelle enrollment was subsidized to achieve full coverage and associated point-of-service co-payments removed, utilization skyrocketed, literally tripling overnight and eventually leveling to a sustained rate of 1.2 to 1.6 visits per capita up from a baseline of 0.65. Statistical analysis showed that the removal of financial barriers was responsible for an increase of 0.6 visits per capita, or a doubling of utilization.

So what does this tell us?

First, premiums and co-payments, while less harmful than traditional point-of-service fees, remain a financial barrier without whose removal true universal access to healthcare cannot be achieved. Even with rising incomes, Rwanda’s mutuelle has not achieved universal coverage and, even at high coverage, access remains constricted by co-payments charged at the point-of-care.

Second, even with high enrollment, mutuelle generates minimal financing. In order to increase the funds collected, Rwanda is now introducing higher premiums. While robust economic growth in recent years may enable some to pay more, others may be priced out of the system. In addition, administering the insurance program requires considerable overhead costs and diverts donor and government funds that could perhaps be more efficiently injected directly into the health system. Further, the need for heavy subsidies to garner sufficient coverage begs the question of how much more ‘sustainable’ – one of the reasons made for the adoption of insurance in the first place – these programs actually are.

Third, Rwanda has made unparalleled progress in health by doing what its leadership has felt best for the country and its people. Despite the limitations of the mutuelle strategy, the country’s collective policies have helped it achieve historic gains. However, some international commentators disproportionately underscore the mutuelle’s role in these achievements and oversell community insurance as a financing panacea for others to adopt solely on its basis. It is important for all aspects of Rwanda’s success to be acknowledged and studied for broader adaptation and, in particular, its increasing and strategic investments in health, strong economic performance, uniquely effective public administration, and popular buy-in to government initiatives. Indeed, these other factors are part of the reason why the mutuelle as a program has been as successful as it has.

Rwanda’s leadership should be lauded for their impressive accomplishments. Policymakers in countries looking to follow in their footsteps need to take the Rwanda model as a whole and look at the mutuelle program more critically to understand its relative merits and many limitations rather than simply buying the hype.

Ranu S. Dhillon M.D. is a Health Advisor with The Earth Institute, Columbia University and a Clinical Fellow in the Division of Global Health Equity at Brigham and Women’s Hospital/Harvard Medical School

[1] Walsh  C and Jones N. 2009. Maternal and Child Health: The Social Protection Dividend. UNICEF.
[2] Ministry of Health (MOH), Republic of Rwanda. 2010. Annual Report July 2009 – June 2010. Kigali: MOH.
[3] Diop, F., Leighton, C., and Butera, D. 2007. Health financing task force discussion paper: policy crossroads for mutuelles and health financing in Rwanda. Washington, DC: Health Financing Task Force.
[4] MOH 2010 op.cit.
[5] MOH 2010 op.cit.
[6] National Institute of Statistics of Rwanda (NISR). 2006. Rwanda demographic and health survey 2005. Calverton, Maryland, USA: NISR and ORC Macro.
[7] National Institute of Statistics of Rwanda (NISR). 2011. Rwanda demographic and health survey 2010: Preliminary Report. Kigali, Rwanda: NISR and ICF Macro.
[8] World Bank, 2011. Data: Mortality rate, under-5 (per 1,000) [online]. Accessed 9 September 2011.
[9] Ministry of Health (MOH), Republic of Rwanda, 2008. National health accounts Rwanda 2006 with HIV/AIDS, malaria, and reproductive health subaccounts. Kigali: MOH.
[10] World Health Organization (WHO), 2011. National health accounts: country health information [online]. Accessed 9 September 2011.
[11] MOH 2008 op.cit.
[12] The most recent National Health Accounts is for 2006. A more recent figure is not available. However, using budget information in the MOH Annual Report and WHO estimates for recent health spending, this percentage is likely consistent with current trends.
[13] MOH 2008 op.cit.
[14] The 3% figure was calculated based on World Bank estimates for population in 2009,  mutuelle coverage of 91%, and WHO estimates for health spending in 2010.
[15] MOH 2008 op.cit.


5 Responses to “A closer look at the role of Community Based Health Insurance in Rwanda’s success”

  1. Health services in low income countries cannot be financed with health insurance premiums. You don’t need to be an economist to understand that. So far the analysis is correct. Beyond this, Dhillon somewhat misses the point.

    The question is not how much money can be generated through mutual fund contributions, but rather how best to channel public funds into a health care system to provide equitable services.

    User fees were established to increase the power of patients, to make services more responsive. Nobody ever expected them to finance national health care. In many countries they contributed to the expansion of a network of first level services, especially in rural areas. The income from consultation fees, laboratory charges and drug margins paid for the operating expenses, auxiliary staff salaries, infrastructure expansion of first and second level health facilities. But they were clearly a barrier to access to care, especially for anything more complex and expensive than a case of intestinal worms.

    So governments and international donor agencies started a wave of user fee exemptions motivated by considerations of public health and of social equity. Exemptions were granted for children under five, for seniors over 65, for pregnant women, for people living with HIV, for tuberculosis, for malaria, for sexually transmitted infections, for Caesarean sections, for emergency ambulance services etc.

    These “free services” are supported with training packages, equipment, infrastructure rehabilitation, drugs, laboratory reagents and sometimes even staff salaries. But they have disastrous effects on the balance sheets of health facilities. Their discretional income is starting to break in. Staff members are doing more and more “volunteer service” to provide care for all those who were legislated to receive “free health care”.

    Health insurance, as introduced in Rwanda, is one way out of this crisis. Hospitals and health centres can now bill for their services; for all their services. They bill the Mutual Fund.

    Obviously, the Mutual Funds do not collect enough money from contribution fees to pay for these services. They require massive subsidies from national and international sources. But this public money will now be used to pay for services rather than to build white elephant hospitals in political constituencies.

    A year ago, when I visited a District Hospital in Rwanda, the director told me that the Mutual Fund owed his hospital four months of fees. But he also stated that even with this large unpaid debt, hospital revenues have tripled since the times when user charges provided the only discretional income. This was clearly visible in the many construction and improvement projects, from new roofs to a waste incineration plant.

    Although Mutual Fund contributions seem insignificant in the balance sheets of national health accounts, they can start playing an increasingly more important role. Linking the contributions to income is an unpopular move, but it will assure that Mutual Funds will grow in step with economic growth. It will also contribute to greater social equity. The urban elites have, for a long time, had a free ride on policies such as free anti-retroviral treatment paid for by international agencies. There is nothing wrong in asking them now to contribute to their publicly provided health care through indexed health insurance premiums. Meanwhile, for the poor, contributions can be kept to a minimum or even be waived. The poor were never able to pay for their health care, but now they have a membership card in their pocket that asserts their right to receive it.

    Establishing a national third party payment system for health services is not an easy task. Some countries will never achieve it. The Rwanda model is still work in progress and it may be unique to Rwanda. But it certainly deserves a good run to prove itself and to generate lessons that may be applicable elsewhere.

  2. Ranu S. Dhillon says:

    The commentator makes some important points and I thank him for stimulating further discussion.

    First, I think he may misunderstand how the financial mechanics of community insurance schemes and the mutuelle, in particular, work. If I understand him correctly, he seems to think that the mutuelle draws subsidies from a larger pool of health funds above and beyond what is available in the mutuelle pool which is approximately 1.50 USD per capita (or 5% of total health spending). The only subsidies the mutuelle includes (and the ones that I refer to) are a sub-component of this 1.50 USD per capita. Therefore, when a health facility bills the mutuelle for services, the bill is paid only from this very limited pool of funds. As a result, the mutuelle is not channeling significantly greater resources and the lion’s share of Rwanda’s 48 USD per capita health spending is, in fact, still being allocated centrally independent of the mutuelle. The minimal financing capacity of community insurance schemes, therefore, does have implications for how meaningfully it channels health funds.

    Second, the commentator agrees that, in its current conception, community insurance can only generate minimal funds and that this is not a point that needs emphasis. However, in the current policy debate around health financing in poor countries, this is very much an issue of contention with many proponents of insurance-based mechanisms touting these schemes as capable of ‘sustainably’ producing significant funds. So, while the limitations of this argument are obvious to some, it is still very much the premise upon which health insurance is being advocated for in other countries.

    Third, as the commentator points out, user fees were proposed to play several roles including creating community ownership, compelling better facility management where central oversight is difficult, and containing utilization so it does not overrun limited resources. Alas, user fees variably achieve any of these goals other than the latter and at the cost of constricting access for the poor. Health insurance is a potential mechanism to do this without limiting access to the same extent and affording greater social protection. However, the point to consider is that insurance does not altogether sidestep this issue and does still curtail access for many. Further, given that health insurance requires considerable administrative and financial effort to work, the question of whether this can be done more efficiently through other mechanisms (i.e., centrally financed free packages of care or centrally financed free care for priority populations) is important to ask. Independent of the financial dynamics, this consideration is especially critical in countries where the administrative capacity needed to operationalize an insurance program with many transactions is not present. Insurance in such settings will be particularly problematic. Other mechanisms (such as central financing) have their shortcomings (i.e., politicized allocations); however, no financing option is without its disadvantages and all come with trade-offs.

    Fourth, the commentator makes the point that exemptions may be a way to circumvent the equity effects of fees or insurance programs. However, there is a well-tread literature documenting the shortcoming of exemption strategies thus far. This literature articulates the challenge for people to pay and the difficulty in determining people’s ability to pay in communities where there is ubiquitous extreme poverty, intermittent food insecurity, and subsistence survival.

    Fifth, the commentator raises the role insurance can potentially play in channeling funds more effectively. As I have already discussed above, with the relatively small amount of financing in insurance pools, the funds – including subsidies – that get injected into insurance mechanisms is insignificant. In all cases, the brunt of funding is still centrally allocated.

    Further, in most poor communities, there is not a choice of health service providers such that the ‘democratization’ of health funds that insurance/fee mechanisms may introduce can matter so much. People usually have geographical access to only one primary care facility, if that. If anything, this means that all facilities in these settings need to be built up to provide quality care. That can be done in other ways without the same efficiency and access concerns as insurance such as legislating facility funding to be determined by population catchments or, without premiums and co-pays, giving people vouchers which they can give to their facility of choice for funding from centrally held sources. In any of these scenarios, including community insurance, the concern that health funds will be allocated in politicized ways will still persist until there is effective public administration and leadership.

    As the commentator alludes to, one additional consideration for community insurance is if it is built as a precursor to a future tax or social insurance system in which people can pay more over time and the administrative and financial costs – and inefficiencies and access implications – of insurance are taken as growing pains of this larger process. Many poor countries where insurance is being proposed are still very far from this stage (i.e., GDP per capita is less than 700 USD) with other pressing priorities that should hold greater influence over health financing policy (i.e., poor health indicators, limited access). Rwanda may be in a better position to take a shot at this and wants this to be their priority. However, we need to be very honest about the implications of this kind of approach before it is paraded as the way to go for other poor countries without as much health spending, strength of public administration, economic growth, and general social development.

    The practical challenge is what to do when there is just not enough money available for health. User fees are problematic. Centrally financed free care is likely not feasible within current resource envelopes. Insurance is a potential option with some advantages compared to user fees. However, it is not a clear-cut solution. And the administrative and financial effort required to make it work can just as easily be used to pursue other potentially effective strategies with less operational challenges and perhaps greater access. At the end of the day, there is no easy answer to this predicament other than getting more health funds. My point is not to say insurance is bad per se, but that it is not the clear answer that many are presenting it to be.

  3. Ravi Duggal says:

    I agree with Josef that health insurance in any of its avatar cannot be a model for universal access to healthcare. Insurance is based on risk pooling and contributions and in a poor country it is invariably at the cost of basic consumption. While one would laud the political will within Rwanda which has provided reasonable success it is still fledgling and dependent on too many unreliable factors, the biggest being external assistance which constitutes a huge proportion of the health budget, including for subsidizing the Mutual Fund. At best Rwanda is a model in transition which has demonstrated limited results because of the strong political commitment for healthcare but this cannot sustain in the long run. Tax based financing and strengthening of public health services across the board is the most effective option and hence to continue its success story Rwanda would have to soon shift gears.

  4. […] scheme; rather, the five-fold increase in health spending, most if it from foreign donors, seems to more clearly correspond to mortality improvements, rather than  mutuelle funds that amounted to covering only 5% of […]

    • Salma says:

      Health insurance shluod be considered a human right as it is in every other western nation. When people are well they can properly participate in a society as contributors, workers, etc. When they are ill they cannot and they will be a drain on our resources one way or the other. Should we let people die because they have no money to afford health care? On the other hand, why shluod any corporation make big money on another person’s misfortune (ill health)? The health of no person shluod be dependent on his/her monetary resources. We shluod, however, be required to make payments for health insurance according to our income or ability to pay. It can only be completely free to those who have no resources.References :

  5. Baldev Gara says:

    Hello! This is my first comment here so I just wanted to give
    a quick shout out and say I truly enjoy reading through your articles.
    Can you recommend any other blogs/websites/forums that cover the same
    topics? Many thanks!

  6. […] other regions. Among the most interesting papers already exploring the issue include the review of Rwanda’s program by Ranu Dhillon (which questions the role of community-based health insurance) and an […]

Leave a Reply

Global Health Check was created by Anna Marriott and is currently edited by Mohga Kamal-Yanni