Box 1.1. Strategic policy directions adopted by the WHO's global cabinet Box 2.1. Summary of BMJ economic evaluation guidelines (from Drummond and Jefferson 1996) Cost-effectiveness of environmental health interventions v FOREWORD This document has been developed with the aim to provide an overview of currently used methods for economic evaluation and to discuss implications of using these methods for evaluating environmental health interventions. It aims at formulating recommendations for future evaluations in environmental health. The document is intended to contribute to the methodological discussions, and in particular the development of guidelines for evaluation of cost-effectiveness in the framework of WHO's initiatives, and other ongoing work in this area. This work constitutes a background document with preliminary considerations of methods for economic evaluations in environmental health. This document in particular focuses on what is peculiar to environmental health interventions, and therefore how the conduct of economic evaluations may need to be different to other health interventions. The main peculiarity, or difference, is that environmental health is a cross-cutting area. Environmental health interventions may need to be addressed, funded and implemented by various sectors (the health sector, the environment sector, the industrial sector, the transport sector, water or infrastructure services). In return, the benefits from environmental health interventions also accrue to various sectors and sometimes also to fulfil basic needs and increase the comfort or quality of life of the receiver (such as improved water supply). This issue raises the question: "Who will pay for (which part of) the intervention?". In relation to economic evaluation, therefore, it should be decided how the benefits should be accounted for in the cost-effectiveness (or cost-benefit) ratio. Furthermore, environmental health also deals with some exposures with very longterm effects, such as climate change, certain occupational exposures, changes in ecosystems with (short and) long-term effects on health, for example through the change in vector populations. Discounting health, even at very small rates, would make almost any impacts with very long latencies seem negligible, thus raising the question "how can discounting for long-term health effects be compatible with the concepts of prevention and sustainability?"