On the use of risk and decision analysis to support decision-making

Risk analysis is always part of a decision context. While it is essential that risk analysts understand the decisional context of their work and results, this is often not very clear in risk analysis and engineering texts. Important knowledge for risk analysts are what tools are available for decision analysis and what the methodological and practical strengths and limitations of alternative approaches are. In the present paper, we discuss and give recommendations on the choice of approach to decision problems involving risk and uncertainty. Using an example from investment in risk reducing equipment on offshore helicopters, we illustrate how information from risk analysis is used to reach a decision, combined with cost-effectiveness indices, cost/benefit calculations and expected utility optimisation. Methodological differences are discussed. It is argued that, for risk related organisational decisions, presentations of consequence and uncertainties from the risk analysis should be highlighted, rather than synthesized measures of utility gains and losses. Results from decision analysis-cost/benefit as well as expected utility-are thus seen as secondary information and strictly as decision aid, rather than methods for prescription of choice. In a model for decision-making, this requires an explicit element of review and judgement by the decision maker. Cost/benefit calculations can provide results that are comprehendible, but will often cause disagreement, about the procedure and reference for the valuation of consequences. We argue that expected utility, although theoretically a more well founded approach, is seen to be of limited practical use. (C) 2002 Elsevier Science Ltd. All rights reserved.