The problem of making a trade off between conflicting goals in the light of uncertainties in order for the decision-maker to reach the best decision can be supported by decision support models e.g. classical Bayesian decision analysis, belief networks, multi-criteria decision analysis.
Bayesian Networks is a class of probabilistic models well suited for handling complex interdependent variables. Bayesian Network techniques may be applied for selecting the optimal procedure. A quantitative risk assessment of the various possible procedures is carried out. Each procedure is broken down into a number of consecutive activities. For each activity the main hazards are identified and their consequences and impact on following activities described. The corresponding probability of occurrence is assessed from detailed technical background documentation and during expert workshop sessions. On this basis a risk model for each full procedure is established using Bayesian Networks. Using these models the various procedures may be effectively compared with respect to overall expected risk and costs.
Cost-benefit analysis is a simple and widely used technique for deciding whether to make a change. The values of the benefits of a course of action are added and the costs associated with it are subtracted. Costs are either one-off, or may be ongoing. Benefits are most often received over time. In its simple form, cost-benefit analysis is carried out using only financial costs and financial benefits. A more sophisticated approach is to include a financial value of less intangible costs and benefits like environmental damage and the benefit of quicker and easier travel to work.