Intended for healthcare professionals

Analysis And Comment Health policy

Contracting out health services in fragile states

BMJ 2006; 332 doi: https://doi.org/10.1136/bmj.332.7543.718 (Published 23 March 2006) Cite this as: BMJ 2006;332:718
  1. Natasha Palmer, lecturer in health economics and policy (Natasha.palmer{at}lshtm.ac.uk)1,
  2. Lesley Strong, research fellow1,
  3. Abdul Wali, research associate1,
  4. Egbert Sondorp, senior lecturer in public health and humanitarian aid1
  1. 1 Health Policy Unit, London School of Hygiene and Tropical Medicine, London WC1E 7HT
  1. Correspondence to: N Palmer
  • Accepted 23 January 2006

Non-governmental organisations are contracted to provide most of Afghanistan's health services. What can we learn from their approach and is it sustainable in the longer term?

Many Western health systems contract out healthcare services, including the NHS. Contracts are less common in low income countries, but contracts with non-governmental organisations (NGOs) to deliver health services are increasingly being promoted in so called fragile states—countries affected by conflict, emerging from conflict, or otherwise lacking the will or capacity to implement pro-poor policies.1 Contracts with NGOs are seen as an effective way to expand services quickly. This is important to reach many of the poorest people living in these countries and thus to make progress towards the millennium development goals for health, but many questions about contracting remain unanswered.


Embedded Image

The challenge of outsourcing

Use of contracts

In a pilot project in Cambodia, NGOs were contracted to provide district health services on behalf of the government. An extensive evaluation showed that districts with health services that were contracted out to NGOs delivered care more efficiently and equitably than those that remained under government control.2 These findings have encouraged promotion of the contracting out approach in weaker health systems.3

Many low income countries that are implementing or discussing contracting of health care belong to the group of around 40 countries currently referred to as fragile states—for example, Cambodia, Afghanistan, Pakistan, Southern Sudan, and Democratic Republic of Congo.1 The contracts are usually funded by a donor in response to the need to expand services rapidly and the lack of functioning government infrastructure and workforce to deliver these services. As a result, perhaps paradoxically, the weaker the country's government capacity, the more likely it is that contracting is adopted. A notable example of this phenomenon is Afghanistan.

The Afghan case

After over 20 years of conflict, Afghanistan's health indicators make unhappy reading. It has one of the highest maternal death rates in the world, and one child in four dies before it is 5 years old.4 5 Even without ongoing security and political difficulties, Afghanistan's geography, climate, and infrastructure offer an immense challenge for delivering even basic health services.

The removal of the Taliban regime in 2001-2, establishment of a new government, and promises of liberal external aid have created new chances to tackle these problems. Before 2002, at least 70% of the country's limited healthcare services were provided by about 20 NGOs, many of whom had been in the country for years.6 As reconstruction began, it became a political imperative to provide basic health services as fast as possible. A joint mission of donors, largely influenced by the World Bank,7 proposed the use of non-state organisations as the main providers for a basic package of health services. The result was an ambitious programme of contracting out basic health care to both international and local NGOs.

Although some established NGOs expressed some reservations,8 most eventually bid for contracts under the new scheme. Both the Ministry of Health and the NGOs recognised that it offered the best potential for rapidly scaling-up services.

Afghan contracts

In collaboration with the Ministry of Public Health, the World Bank, the US Agency for International Development (USAID), the European Union, and the Asian Development Bank are now funding contracts with NGOs worth over $140m (£80m; €118m). These contracts nominally cover an expanding proportion of the population (currently estimated at 77%). Contracts exist in all 34 provinces, covering either the full province or clusters of districts. The remaining 23% of the country has been included in recent calls for proposals. The contracts are all based on a standardised package of care (box 1). Contracts with NGOs last from 12 months to 36 months, with an average of 26 months. NGOs are paid according to individual budgets, which they draw up as part of the bidding process.

Three provinces are run under contract to the Ministry of Public Health itself—in a scheme known as the strengthening mechanism, which is funded by the World Bank. The same services are delivered but using existing government mechanisms.

With the exception of Médicins Sans Frontières all the major NGOs that were active in the health sector in Afghanistan in 2001 bid for contracts. These include international and Afghan organisations such as Save the Children, the Swedish Committee for Afghanistan, HealthNet International, and Ibn Sina, a large Afghan NGO. There are now 27 NGOs with contracts, 17 international and 10 Afghan. In addition, some contracts were awarded to consortia of national and international NGOs. National NGOs currently have 38% of the volume of grants awarded. Since 2002 many new Afghan NGOs have been established, and other large international NGOs such as the Bangladesh Rural Advancement Committee and the Aga Khan Development Foundation have established operations in Afghanistan and won contracts.

Box 1: Afghanistan's basic package of health services at primary care level

Maternal and newborn health—Antenatal care, delivery care, postpartum care, family planning, care of the newborn

Child health and immunisation—Expanded programme on immunisation (EPI), integrated management of childhood illness

Public nutrition—Micronutrient supplementation, treatment of clinical malnutrition

Communicable diseases—Control of tuberculosis, control of malaria

Supply of essential drugs—such as antibiotics, analgesics, contraceptives, and antituberculosis drugs

Advantages of contracting

The approach has obvious advantages practically and politically. NGOs were already running most facilities, and are experienced in the difficulties of delivering services in Afghanistan. Possibly the bulk of public health expertise in Afghanistan currently resides in the NGO community. NGOs are often more flexible than government in their ability to recruit new staff and set up services rapidly. In contrast, the Ministry of Public Health has been struggling with bureaucratic procedures for hiring new staff. (One official described how recruitment of female staff for remote areas was delayed by government procedures that can take up to two months. In some cases this has resulted in potential staff being recruited by NGOs.)

In addition, some NGOs have the financial and logistical backing of large international organisations; they may supplement contract funds with their own resources. Lastly, the motivation of NGOs is generally expected to be closer to that of public providers than that of the for-profit private sector, and contracts with NGOs are argued to take advantage of the voluntary sector's greater flexibility, innovation, and morale.3

Box 2 highlights some of the theoretical arguments for and against contracting out health services.3 9 Using contracts to achieve a rapid expansion of capacity in fragile states seems an effective short term strategy. However, it also raises questions for health planners over the possible directions for restructuring health systems in the longer term.

Questions about competition

Markets that are contestable, or offer the threat of competition, are argued to encourage providers to maintain efficiency and quality. In Afghanistan, a review of bidding for the initial contracts suggested varying levels of competition. More accessible and secure areas had relatively plentiful bids, but the competition for more remote areas was low.7 In one province, Badghis, the contract was awarded without competition.

Box 2: Arguments for and against contracting

For

Allows a greater focus on measurable results

Increases managerial autonomy

Draws on private sector expertise

Increases effectiveness and efficiency through competition. Allows governments to focus on other roles such as planning, standard setting, financing, and regulation

Allows for rapid expansion of health service

Against

Competition may not exist, especially in low income countries where there may be no alternative providers

Contracts may be difficult to specify and monitor

Management costs may wipe out efficiency gains

Contracting may fragment the health system

Governments with weak capacity to deliver services may also be weak in a stewardship role

Plans for the near future include a competitive process for the next rounds of funding through the European Commission and USAID. It remains to be seen whether NGOs will bid for one another's contracts in these subsequent bidding rounds. They may settle into operating in certain areas and be reluctant to move. The desirability of replacing providers once they are in place is also questionable. In such difficult environments, local knowledge and networks may give the incumbent such an advantage that it would be better to accept that the type of contracting being used is more relational (meaning long term and based on trust or dependency) than competitive.10

The long term effects of a competitive process in acting as a spur to efficient service delivery also cannot yet be assessed. In the short term NGOs responded to the competitive process with widely varying estimates of the cost of delivering the basic package of health services. Some may have underestimated costs in an effort to win contracts. Currently, average annual per capita allocations from differing contract rounds range from $2.06 to $4.83. If this variation is reflected in differences in accessibility and quality of services, it is clearly undesirable. On the other hand, costs of delivering services are likely to differ between areas and are currently unknown. Continued competitive bidding, if it took place, would help to assess resource requirements.

Specifying and measuring performance

If contracts are to increase transparency, the quantity and quality of services must be both clearly specified and measurable. NGOs are contracted to deliver a basic package of health services (box 1) but the terms of the contracts vary. Some donors are more focused on inputs (such as numbers of trained staff), some on process indicators (such as utilisation), and some on outputs (such as immunisation rates). They also have different incentives. World Bank contracts, for example, have a performance based element. Four NGOs have recently received bonuses amounting to 1% of their contract price for good performance, which is defined as an increase of at least 10 percentage points above baseline indicators. It will be important to continue to monitor how often payment is made or withheld and the effect that this has on providers' behaviour.

Overall, issues of specification and monitoring seem to be dealt with well—to the extent that service delivery in such settings can be monitored.10 Performance is being taken seriously; one contract with an international NGO has already been terminated for poor performance. For all contracts, progress reports and site visits are part of the monitoring process. In addition, to ensure objective measurement of performance a third party has been contracted to monitor services using household surveys, inspections of facilities, and interviews. This gives detail of volume and processes of service delivery and some measure of access by the community. However, limited sampling means it is less able to reflect on access or health outcomes for the broader community.

The biggest challenge is how to specify contracts to encourage delivery of services to the most remote parts of the population. Although the contracts nominally cover a high proportion of the population, many remain outside the catchment area of any facility. It is currently difficult to specify or monitor the extent to which NGOs extend services into these areas.

Costs and sustainability

Costs associated with monitoring and managing the existing contracts are closely linked to issues of government capacity to carry out stewardship. Currently costs are increased by expatriate technical assistance both to help develop NGO and government capacity and to strengthen contract management—the USAID funded contracting programme has over 20 expatriates in Afghanistan. The Ministry of Public Health has established a specialised unit that manages the World Bank grants and is eventually likely to manage all contracts. The third party evaluator responsible for external monitoring (by household survey and quality assessment visits to facilities) is a further expense—the current contract is worth over $4m until September 2006.

These expenses reflect the costs of a well managed and monitored contracting framework where capacity is weak. They raise issues both of their magnitude in comparison to a government hierarchy for service delivery, and more relevantly, how such costs could be met without continuing substantial donor inputs. Building local capacity to manage this system would reduce costs, but this requires a long term vision of the future model of health care in Afghanistan.

Decentralisation without fragmentation

In settings such as the UK and New Zealand where an existing public sector hierarchy was unbundled to form an internal market, contracting was argued to provide a desirable decentralisation of managerial responsibility. Afghanistan starts from a different perspective. Central authority is limited and in places highly compromised. Despite the existence of a basic package of health services, decentralisation to non-state providers means that fragmentation is virtually inevitable. There is no standardised practice in areas such as user fees, drug procurement systems, and deployment of community health workers.

Variation is not automatically a problem— innovation may lead to advances in service delivery. But it is important to monitor whether differences have implications for equity and efficiency in the longer term. Strong international NGOs may engage in valuable capacity building activities in their area, but a national perspective on such activities may be lost. If these issues are not addressed in the current framework, concerns over the broader reconstruction of the Afghan health system may grow as it attempts to consolidate systems for drug supply and human resources.

What happens to NGOs as they scale-up?

Although the timescale of the current contracting framework is unclear, it is likely to continue in the medium term. If there is a further phase of expansion, the capacity of NGOs to continue to scale-up is unclear. Established NGOs may be unwilling to take on further contracts. The emergence of new, national NGOs may fill this gap, but the extent to which NGOs may overstretch themselves or be unable to sustain quality services if they expand is unknown. NGOs may develop the same weaknesses as government delivery mechanisms if they grow bigger. In addition, some NGOs depend on a few key individuals and a local approach; their effectiveness may be lessened as they grow. Commenting on the performance of national and international NGOs one donor representative stated: “International managers can bring the latest in primary health care while Afghans can bring the knowledge of the area so it's hard to predict who will do better.”

A further issue is the relationship between government and NGOs. The Afghan government is already expressing concern over the role and behaviour of NGOs in the country's reconstruction process.11 Governments in fragile states are often struggling to maintain legitimacy. Delivering health services and controlling health workers are often seen as key government functions, and as a government becomes better established it may wish to resume control of these. Although central Ministry of Public Health staff believe that delivering services through NGOs is a good option for the medium term, our interviews with staff at provincial level suggested some reluctance to accept that NGOs may be there to stay, especially international NGOs. One official stated: “I would prefer the ministry over the NGOs because it is unknown how long they will work in Afghanistan and they may leave. Also the price of NGOs is very high and they take a lot of holidays.”

Conclusion

The millennium development goals cannot be reached without progress in fragile states such as Afghanistan.1 Contracts with NGOs are probably the only way to get systems moving quickly. This pragmatic policy has both benefits and opportunity costs. Issues of NGO capacity and motivation, equity, politics, and the role of government, as well as the danger of bypassing the opportunity for longer term health systems development, make this area an important focus for future research and debate.

Summary points

Fragile states are increasingly contracting out delivery of health services to non-governmental organisations (NGOs)

Afghanistan is the most recent and large scale example of contracting

Use of NGOs enables rapid expansion of health services

Other effects of this valuable new policy approach need to be monitored and evaluated

The appropriate role of government, the capacity and motives of NGOs, and how to limit fragmentation need investigation

We thank Anne Mills and Bayard Roberts for comments on the first draft of this paper and a number of colleagues at LSHTM for comments on the second draft.

Footnotes

  • Contributors and sources NP is a member of the Health Economics and Financing Programme, which is funded by the UK Department for International Development. LS, AW, and ES are members of the Conflict and Health Programme. This discussion draws on previous and ongoing work by all authors on contractual relationships in primary care in developing countries. The Afghan case study was one of 23 projects funded by EC-PREP, a programme of research sponsored by the Department for International Development. ES is the principal investigator of this study. LS and AW are based in Kabul and working on this study. NP, LS, and ES discussed the idea for the paper. NP drafted the first version. LS, AW, and ES wrote sections, edited others, and provided information. NP is guarantor.

  • Competing interests None declared.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
View Abstract