Required: best value for money

Issues in choosing which health care interventions to provide

Von Guy Hutton / Schweizerisches Tropen- und Public-Health Institut (Swiss TPH)

"To require the health system to obtain the greatest possible level of health from the resources devoted to it, is to ask that it be as cost-effective as it can be. This is the basis for emphasising those interventions that give the greatest value for money, and giving less priority to those that, much as they may help individuals, contribute little per dollar spent to the improvement of the population’s health" (WHO World Health Report 2000, page 52).

In order for health services to provide the best ‘value for money’, often fundamental change is required (2). In order to achieve this, recently many developing country governments have been involved in rationalising the delivery of public health services, consisting of redefining which health interventions are to be provided at different levels of the public health service, integrating vertical programmes back into primary health care programmes, improving quality of care, and exploring alternative health care financing mechanisms. Several countries, for example Mexico, Bangladesh, Columbia, Zambia, and Tanzania have begun implementing "essential" or "basic" health care packages, which specify a list of health care interventions to be carried out at various levels of the health service (3). In general, these health care packages are defined to treat the main burdens of disease using interventions that are cost-effective and affordable. For example, in Bangladesh an Essential Service Package (ESP) has been defined and implemented, consisting of reproductive health, child health, communicable disease control, limited curative care, and behaviour change communication.

There exist two main strategies for maximising the impact of resources spent by the government in the health sector, which ideally should be implemented together.


(1) Improve the mix of health services, focussing on those that have greatest impact at least cost, thus moving resources from cost-ineffective ones to cost-effective ones. This increases ‘allocative’ efficiency. Allocative efficiency can be improved by putting together cost-effectiveness information for a wide range of relevant health interventions - using appropriate cost-effectiveness analysis methodology provided in guidelines (4), (5), (6) - then ranking interventions in terms of cost-effectiveness, before finally building a package of services that jointly have the biggest impact on health status and at least cost. This approach means that the process of priority-setting is explicit, transparency is increased, and it reduces variation in the availability of services in different parts of the same country.

Some published studies report the results of case studies of following such an approach in developing countries (7), (8). However, still considerable uncertainties exist in these studies due to lack of data, cross-setting differences in cost-effectiveness ratios (both within and between countries), and disagreements or weaknesses in cost-effectiveness methodology (9). In addition, sometimes the health sector objectives of maximising efficiency and improving equity come into conflict with each other. For example, given the lack of low cost options for treating those with HIV/AIDS, it may be more cost-effective to invest in health services that target disease that are curable and cost less per patient.

Therefore, considerable work still needs to be done in refining the cost-effectiveness methodology and applying it in a wider range of countries. Future studies may make more use of syntheses of effectiveness reviews, such as provided by the Cochrane collaboration (10), and analysis of generalisability of findings between settings, rather than conducting expensive new research studies (11). Also, further elaboration is required of how the essential package is to be delivered in practice, such as the necessary referral structures, staff supervision, regulation, monitoring and evaluation of services (3). Finally, the limitations of essential packages should be recognised, and they should be implemented within the framework of a comprehensive and system-wide health sector strategy.


(2) Improve the efficiency of health care resources, getting maximum productivity out of them. This increases ‘technical’ efficiency. Technical efficiency can be increased at the operational level by, for example, reducing the time that equipment and rooms/buildings spend out of use, reducing theft and wastage of drugs, and increasing staff motivation to increase patient daily throughput. At the organisational level, different structures and methods of control need to be examined, to decide what mix of hierarchical control, bureaucracy, and market-based interactions are optimal for different settings (1).

In addition to the issues mentioned briefly above, there are issues that cut across allocative and technical efficiency. An important current issue, that of health care financing, affects cost-effectiveness through its impact on health care utilisation patterns. For example, when services are charged for there may be a lower uptake than when services are free, thus changing the cost-effectiveness ratio. Policy makers must also choose their investment strategy, which affects cost-effectiveness as well as technical efficiency. For example, long-term investment (such as in infrastructure) may appear less attractive than short-term investment (such as improving drug supply). However, managers also need to take into account that not investing in health infrastructure in the longer-term will affect the sustainability of health services, and therefore future cost-effectiveness rates and technical efficiency.

Choices must also be made between investing in labour (such as training) and in capital (equipment, buildings), which affects service delivery. Clearly not all of these aspects can be taken into account in determining cost-effectiveness, and therefore information on relevant aspects of financing and expenditure should be provided alongside the cost-effectiveness ratio.

It is clear from the above discussion that to plan which health care interventions to provide, and the service delivery method, requires significant amounts of data and technical support, which are often not available in sufficient quantity in most developing countries. However, as the capacity of both international and national level institutions develops, critical issues are now coming more into focus in the minds of key policy makers and organisations. In the recent shake up of the World Health Organization, a new cluster "Evidence and Information for Policy" has been put in place. Within this cluster, a unit "Choosing health interventions: effectiveness, quality, cost and ethics" has been given the responsibility of, among other things, defining what interventions and mixes of interventions are most appropriate in different groups of countries.

Also, computer technology is being applied to health sector planning, with software being developed to increase the capacity of local decision makers to make optimal decisions. For example, the W.H.O. department "Organisation of Health Service Delivery" is building a comprehensive health sector planning software package, which is currently being field tested, and will allow staff at country level to plan health care activities more rationally.

In conclusion, choosing a package of health services for a population requires a range of data as well as the necessary technical expertise to know how to use the data; this data needs to be updated regularly to ensure adjustments made to health systems reflect the changing conditions. Clearly, decisions about which health services to provide require more than just a ranking of cost-effectiveness ratios (which are lacking in most developing countries, anyway), but a full consideration of the demand for services, external costs and benefits of interventions, the distribution of the costs and benefits (e.g. whether pro-poor or not), and the potential for services to be provided privately (12).

Guy Hutton, MSc, PhD, works as Health Economist for the Swiss Centre for International Health in the Swiss Tropical Institute, Basel. guy.hutton@unibas.ch

1. World Health Organization. World Health Report 2000. Health systems: Improving performance.

2. C Murray, J Kreuser, W Whang. Cost-effectiveness analysis and policy choices: Investing in health systems. Bulletin of the World Health Organization 1994; 72(4):663-674.

3. E Tarimo (1997). Essential health services packages: Use, abuse and future directions. Division of Analysis, Research and Assessment. World Health Organization. ARA Paper Number 15. WHO/ARA/CC/97.7.

4. M Drummond, T Jefferson (1996). "Guidelines for authors and peer reviewers of economic sumissions to the BMJ." British Medical Journal 313 (3 August): 275-283.

5. C Murray, D Evans, et al. (2000). "Development of WHO guidelines on generalised cost-effectiveness analysis." Health Economics 9(3): 235-252.

6. L Brenzel. Selecting an essential package of health services using cost-effectiveness analysis: A manual for professionals in developing countries. World Bank / Data for Decision Making Project, 1993.

7. P Jha, O Bangoura, K Ranson. The cost-effectiveness of forty health interventions in Guinea. Health Policy and Planning 1998; 13(3): 249-262.

8. J Bobadilla, P Cowley, P Musgrove, H Saxenian. Design, content and financing of an essential national package of health services. Bulletin of the World Health Organization 1994; 72(4):653-662.

9. M Palman, H Bekedam, L Hawken, D Nyheim. A critical review of priority setting in the health sector: the methodology of the 1993 World Development Report. Health Policy and Planning 1998; 13(1): 13-31.

10. Visit the Cochrane library at their website: http://wwwsom.fmc.flinders.edu.au/FUSA/COCHRANE/cochrane/cdsr.htm

11. G Hutton. Can the costs of the World Health Organization antenatal care programme be predicted in developing countries? PhD Thesis, London School of Hygiene and Tropical Medicine, London University, 2000.

12. See Figure 3.2 on page 55 of The World Health Report 2000 (see note 1 above).