Cost-recovery in the health sector: continuing the debate
by Richard Hands, ECHO; Egbert Sondorp, Tim Poletti and Olga Bornemisza, London School of Hygiene and Tropical Medicine; and Linda Doull, Merlin November 2004

In the March 2004 edition of Humanitarian Exchange (issue number 26), Timothy Poletti of the London School of Hygiene and Tropical Medicine (LSHTM) took issue with cost-sharing in healthcare programmes in complex emergencies, arguing that it raised little revenue and threatened access for the poorest. Below, we print a response to his article from Richard Hands of ECHO, along with a rejoinder from Poletti and colleagues at the LSHTM.

Cost-recovery in humanitarian heathcare programmes: a donor’s prespective
Richard Hands, ECHO

No humanitarian would take issue with the substantive thesis of Timothy Poletti’s recent HPN article – that cost-recovery systems should not be used in humanitarian healthcare programmes. Nor would many NGOs disagree that donors like the European Commission Humanitarian Aid Office (ECHO) ‘have significant leverage over health policy’ in such situations. However, the author is incorrect in his assertion that ECHO requires its health partners in the Democratic Republic of Congo (DRC) to include cost-sharing mechanisms in their programmes.

In attempting to rectify this misapprehension, it might be helpful to start by reviewing some of the strategic and operational constraints facing humanitarians working in complex emergencies.

  1. ‘Complex emergency’ is a euphemism for ‘nasty civil war which goes on for a very long time’. One of the defining features of humanitarian action since the end of the Cold War is precisely its chronic nature: Sudan, Angola, Burundi, Somalia, Coastal West Africa and the DRC have been in a state of complex emergency for years. In such cases, with normal development cooperation (implemented by governments) suspended or much reduced in scope, humanitarian assistance (implemented by international agencies) faces challenges with which the traditional, top-down relief model is not designed to cope. If we are to avoid crushing what remains of the public health system in countries like the DRC, we must adopt a longer-term approach, working with, and through, local capacity.
  2. is important to bear in mind how the health economies of very poor countries actually work in practice. With many African states in particular finding it difficult if not impossible to fund public services, the laudable objective of la santé pour et par tous has progressively degenerated into a perverse system which, in the absence of outside assistance of some kind, is financed solely by the ever-dwindling resources of the paying customer. In most cases this twisted mutation of ‘cost recovery’ is the established practice in African countries in which humanitarian/health agencies are now intervening. It is therefore disingenuous to refer (as Poletti does) to the ‘introduction’ of such policies in complex emergencies; cost recovery is a given, already entrenched when humanitarians engage.
  3. It is also worth recalling that international law recognises no humanitarian ‘right to intervene’; NGO staff by and large work on tourist visas, and agencies have to deal with national policies – and recalcitrant local authorities – as best they can. This has significant practical consequences, which in some cases can undermine not only the implementation of healthcare programmes, but also the safety of humanitarian personnel. For example, try explaining to a district medical officer in some remote bush location, who for years has survived on his wits without receiving a penny from the Ministry of Health, that he must stop charging his patients because it offends our humanitarian principles. However regularly we supply the drugs and pay his ‘incentives’, he knows that one day we will leave, and he does not believe there will be a government salary to replace us. He may well be right.
  4. By no means all healthcare NGOs are well-resourced emergency specialists; some are religious bodies, others are small, single-issue associations, and many have been where they are for so long that they have taken root. Such partners understandably react with hostility to the attempts of johnny-come-lately humanitarians to put the sacred cow of cost recovery out to pasture. But a country like DRC is the size of Western Europe, and perhaps half its population of 60 million lives in areas directly and continuously affected by conflict since 1997. As the capacity of the humanitarian machine is finite, the relief donor’s choice is straightforward: engage with these reluctant partners and attempt to cajole them into reflecting on the public health consequences of maintaining a cost recovery system in a conflict zone – or abandon the population of that zone to its fate.
  5. Despite its bad press, cost recovery in fact works pretty well, even in emergencies – as long as the aim is to ensure the survival of a system that provides healthcare of doubtful quality to the most affluent 10% of the population. This was the system that ECHO inherited in the DRC in the late 1990s – and Congolese healthcare staff and their partners were on the whole quite happy with it. The system is of course perfectly indefensible even in a stable, ‘development’ environment; in a ‘humanitarian’ (conflict) situation, with mortality rates at unprecedented heights, it is downright immoral.
  6. The practice of ‘cost recovery’ as it has evolved requires that the patient be charged the real cost of a given service, plus whatever margin is required to keep the structure going. Given that the average all-in cost of a disease episode at peripheral level in the DRC is around €1.5, and that this sum represents the monthly family revenue of the poorest half of the four million people of North Kivu (itself a relatively well-to-do province), ECHO has never advocated or operated a ‘cost recovery’ system in the DRC.
  7. On the other hand, not everything about the Bamako Initiative is obsolete or inappropriate in crisis situations; after a great deal of reflection, ECHO and its partners have concluded that there is merit in retaining and indeed encouraging the concept of ‘community participation’. In doing so, our main aim is to preserve the concept of ownership – the idea of a public health system as a thing of value, which belongs to the community. One key component of this approach is the reinvigoration of the Comité de Santé (COSA), a committee of local notables responsible for running the local health centre. The other is the policy of charging patients a fee which is used to contribute to the health delivery system’s running costs.

For present purposes, the salient features of this policy are:

  • ECHO is supporting healthcare in some 55 of the worst-affected of the DRC’s 306 health districts in 2004. The funds allocated to these programmes cover the full cost of providing a clearly defined package consisting of all curative and preventive healthcare at primary level, and selected activities at secondary (referral) level. Costs covered include all drugs and consumables, logistical and technical support, essential equipment and refurbishment, and staff ‘incentives’ (the word used to describe civil service salaries when they are paid by humanitarian agencies).
  • With the health delivery system thus in effect fully funded, the fees charged to patients for the purposes of promoting ownership are established not in relation to the actual cost of the service, but on the basis of the purchasing power of the community. This is continuously assessed by tracking the impact of tariff changes on utilisation rates, and by means of the socio-economic and health-seeking behaviour surveys routinely conducted by ECHO’s health partners. Encouragingly, these surveys tend to confirm that even the poorest communities are willing to contribute financially to a healthcare system that they consider to be accessible and effective.
  • The fee charged is generally little more than symbolic (15 euro-cents or less in many ECHO-supported programmes), and typically takes the form of a flat-rate, per-episode payment agreed in MoUs concluded with local health authorities and displayed in all health centres. Such is the simplicity of this system, little or no cost is attached to collecting the fees, which are managed centrally by the COSA and redistributed in line with agreed priorities – mainly to supplement the staff ‘incentives’ paid by ECHO. In some more stable areas, it may indeed be the case that revenues generated by patients’ contributions are sufficient to cover part or even all of the structure’s running costs. However, this has never been a condition of, or an objective for, ECHO support; the overarching priority is to guarantee access (both physical and economic) to healthcare. Where circumstances require, tariffs are waived altogether.
  • The MoUs also stipulate that preventive healthcare services (including antenatal and EPI clinics) must be provided free of charge to all comers. Similarly, the poorest sections of the community are recognised and registered as ‘indigents’ by their COSAs, and are entitled to free healthcare (all services). Non-paying ‘indigents’ must account for a minimum of 10% of patients treated in ECHO-funded programmes. This level is frequently exceeded.
  • Finally, a further safety net is provided by the Pool d’Urgence Congo (PUC). The PUC is a standing, nationwide surveillance and rapid-reaction capacity ECHO has been funding in the DRC for several years through Médecins Sans Frontières. It is a classic relief instrument, which intervenes in unashamedly top-down fashion when local healthcare capacity is overwhelmed by large-scale public health emergencies.

The pertinence of this strategy can be gauged by the fact that, in the health districts supported by ECHO, average utilisation rates have risen from fewer than 0.1 new contacts per person per year in 2000 to around 0.6 now, and exceed 1.0 in some areas. These figures have been achieved in a conflict situation, by working with and through existing Congolese healthcare staff and facilities, and compare extremely favourably with utilisation rates in African countries unaffected by crisis.

ECHO recognises (and indeed would emphasise) that many aspects of this approach push at the outer limits of the humanitarian mandate. Indeed, given the institutional constraints within which ECHO is obliged to operate, it has taken some ingenuity and a certain amount of determination to develop and implement this model. The work – and the debate – goes on, but ECHO and its partners believe that the DRC healthcare strategy is appropriate, fair and properly ‘humanitarian’ in intent and outcome.

We are, however, under no illusions as to the ‘sustainability’ of this system. On the one hand, ECHO has, in effect, obliged disgruntled Congolese health staff to work ten times as hard for the same income; on the other, the Congolese population plainly lacks the means to support the cost of even a pared-down public health system without substantial help, and the DRC government will not be in a position to provide such help for some years to come.

By way of illustration, for planning purposes ECHO uses an all-in unit cost bracket of €3–4 per direct beneficiary (disease episode) per year. On the basis of a target utilisation rate of one new contact per person per year, using the ECHO model it would cost approaching €200 million per annum to provide a minimal healthcare package to the Congolese population as a whole – considerably more if the full range of referral-level activities and central (ministry) costs is included. In the absence of a ‘Marshall Plan’ for the DRC, this level of funding is simply not available, nor will it be in the foreseeable future.

As the DRC enters a transitional phase which will, hopefully, put a permanent end to the conflict and create a stable political environment, ECHO thus finds itself actively engaged in a debate with development donors with a view to agreeing a viable ‘transitional model’ for funding healthcare in the DRC. It is broadly acknowledged that cost-recovery in countries with weak or absent state structures is unjust and impractical, and that if there is to be anything resembling a public health system in the DRC, it will need to be strongly subsidised for years to come. That ECHO has achieved this informal consensus around the ‘humanitarian language’ of access is in itself quite a success.

These are serious and difficult matters, and if ECHO’s continuing efforts to grapple with them are to succeed, we will need the constructive, open-minded and informed input of all parties.

Richard Hands is Head of Sector, ECHO Central Africa/Great Lakes.

Health may be a human right, but someone has to pay
Egbert Sondorp, Tim Poletti and Olga Bornemisza, London School of Hygiene and Tropical Medicine, and Linda Doull, Merlin

A recent World Health Organisation (WHO) report states that ‘health is an intrinsic human right as well as a central input to poverty reduction and socioeconomic development’. If we agree with this statement, it follows that everyone in the world should have access to a ‘Basic Package of Health Services’ addressing the major causes of illness and premature death. The World Bank has estimated that the annual cost of implementing a basic package of priority health interventions in the least-developed countries would be $11 per capita annually. However, it has found that the median annual per capita government expenditure on health in Africa is only $6, with an average $3 per capita in the lowest-income countries. Someone will have to pay this cost if such a package is to be made available to everyone.

For any country, there are only two possible sources of funds to pay for a basic package of healthcare: funds that are generated within the country (taxation, insurance revenue and user-fees) or funds that come from outside the country (funds from donors). Taxation and insurance systems have the advantage of risk-pooling and the promotion of equity. However, they require good governance, sophisticated management systems and a reasonable tax base. Unfortunately, these conditions are rare in many developing countries, and non-existent in complex emergencies.

Out-of-pocket payments or user fees are the only other option for generating funds within a country to pay for healthcare. They have been advocated since the mid-1980s to supplement scarce public resources, with the aim of generating 15–20% of total expenditure and increasing efficiency. However, these hoped-for benefits have not been achieved; revenue is usually less than 5% of total recurrent expenditure, efficiency does not improve and, despite attempts to design and implement effective exemption mechanisms, access to healthcare for the poorest and most disadvantaged decreases. It can be concluded that only outside sources will be able to ensure some form of basic, equitable healthcare for the majority of African countries, and in particular for those that are in conflict, or just emerging from it.

This health financing gap applies to the Democratic Republic of Congo. Large parts of the country are highly unstable, with no effective governance, poverty is widespread and the health status of the population is extremely poor. Tax revenues, insurance and user fees will not generate sufficient resources to pay for healthcare, and so external funding is required if a basic package of essential services is to be made accessible to the majority of the people.

Fortunately, in parts of the country, there is significant external funding for healthcare. ECHO now provides the full cost, excluding salary costs, of a clearly defined package consisting of ‘all curative and preventive health care at primary level and selected activities at secondary level’ in 55 health districts. However, ECHO defends cost-sharing of a flat fee of about 15 euro-cents per disease episode. It argues that such user fees are only symbolic, and justifiable on the basis that they promote community ownership. It also argues that abolishing them would deny health staff income, and that their use ensures that some form of healthcare will continue should ECHO withdraw or hand over to development donors.

While enhancing community ownership is clearly a worthwhile objective, there are ways to promote it, such as reinvigorating the Comités de Santé, without introducing user fees. People are also unlikely to feel ‘ownership’ of a service that they cannot afford to access. Even at 15 euro-cents, user fees may well be unaffordable for a large proportion of the population; this represents 10% of a monthly family income for the poorest half of the population.

Arguing that such a fee is affordable based on increasing utilisation rates is suspect for two reasons. First, there is significant evidence that people will pay to access healthcare even when they cannot afford to, as many sell valuable household assets resulting in catastrophic health expenditure. Second, a utilisation rate is an average. Community surveys by Merlin have revealed that many people could not afford to access care, and that the ‘well off’ were the main users of their health services.

Instead of being dependent on user fees to supplement their income, health staff are likely to welcome direct payment to cover their basic cost of living. Staff incentives are probably the most important factor in the provision of good-quality health services, which are accessible to those most in need. Providing such services should be the prime objective for humanitarian agencies in conflict and transitional areas.

Any health service in this kind of environment will collapse if it is not supported by external funds. ECHO is providing funds in parts of the DRC. However, we feel it could maximise its resources to reach those most in need if it discouraged user fees. In the meantime, let us hope that development donors fill the health financing gap in years to come. Recent experience in ‘transitional countries’ like Burundi, Sierra Leone and Afghanistan, however, does not offer much hope that this will happen.

Egbert Sondorp is Senior Lecturer at the London School of Hygiene and Tropical Medicine (LSHTM), researching issues around conflict and health with Research Fellow Olga Bornemisza. Timothy Poletti is also a Research Fellow at LSHTM, involved in the evaluation of community insurance schemes. Linda Doull is Health Director of Merlin. Any correspondence about this article should be sent to