Editorial

Celia Almeida, Paula Braveman, Marthe R Gold, Celia L Szwarcwald, Jose Mendes Ribeiro, Americo Miglionico, John S Millar, Silvia Porto, Nilson do Rosario Costa, Vincente Ortun Rubio, Malcolm Segall, Barbara Starfield, Claudia Travessos, Alicia Uga, Joaquim Valente, Francisco Viacava.

This article will be published in the May 26 issue of The Lancet.

Introduction

The authors of the WHO's World Health Report 20001 have placed on the WHO agenda a commitment to the laudable goals of assessing health systems, monitoring inequalities in health, and achieving equity in health-care financing. Their proposition that health services should be responsive to people's expectations is a welcome one. While these commitments should be sustained, we believe that the approaches taken toward these ends in the World Health Report are seriously flawed. We aim to suggest changes to the approach in the World Health Report to ensure that measurement strategies supporting public health policy throughout the world are scientifically sound, socially responsible, and practical.

Both the conceptual basis and methodological approaches to the World Health Report composite index of health system goal attainment and its individual components, and the indices of health system performance, have major problems. Data needed to calculate four of the five component measures for overall goal attainment were absent for 70-89% of countries, but this was not acknowledged in the report. Because all the measures are new, and imputed values for the 70-89% of countries without data were based on new methods involving multiple non-standard assumptions, readers deserve to know the underlying assumptions, methods, and key limitations, which were not adequately acknowledged. The measures of health inequalities and fair financing do not seem conceptually sound or useful to guide policy; of particular concern are some ethical aspects of the methodology for both these measures, whose implications for social policy are cause for concern. The use of the composite indices for guiding policy is not evident, mainly because of the opacity of the component measures.

In response to criticisms of the report from member states, the WHO Executive Board on Jan 19, 2001, recognised the need to establish a technical consultation process that would obtain input from member states and a small advisory group for the cross-country assessments of health systems (www.who.org, accessed May 15, 2001); we do not know what steps have been taken in that process. The Lancet published an article by Navarro in November, 2000,2 that analysed the World Health Report, focusing mainly on a series of important policy concerns. Little attention was given to methodological discussion. We therefore focus on the methodological and related conceptual issues of the report, in the hope of making an additional, constructive contribution to a thorough process of consultation that must now be opened up by WHO.

Conclusion

The positive contribution of the World Health Report 2000 is its stimulation of fresh thinking about a range of issues relevant to measuring health-system performance. The goals to improve average levels of health as well as distribution of health in populations, and to monitor progress toward these goals, are sound ones. Our comments are offered in the hope that they will help WHO, guided by its member states, to move ahead with an open process of conceptualisation, measurement, and documentation in studying health systems that can serve as a sound basis for policy, planning, and advocacy in the search for health and equity; unfortunately, the World Health Report 2000 does not provide such a basis. As researchers, our recommendations have largely focused on methodological concerns. However, we firmly believe that a strong and sustained response will be needed not only from the research community but from advocates for health and development globally, and particularly from the member states to whom WHO must be accountable. We hope that this paper helps to clarify key concerns on several serious issues related to the methodology of the report. Although we have focused on methodological concerns, these issues are not simply matters of technical and scientific concern, but are profoundly political and likely to have major social consequences.