22. There are a variety of packages available which take the analysis of risk using probability distributions a step further. In general, computer simulations start by generating a series of simple profiles to fit a number of defined cases for each risk (usually three): the worst case (maximum impact), the most likely case (expected impact) and the best case (minimum impact). A limited number of standard profiles is usually available within the software and in some programmes the input can be in the form of a curve chosen from a standard set.
23. Monte Carlo simulation is widely available in the form of a number of software packages for this sort of analysis, although alternative statistical methods are equally acceptable. In this method, a random value of probability and its associated value of consequence is selected from the sample profile for each risk in turn and these are then combined to give a total value for the overall project. This procedure is repeated for a large number of iterations. The more iterations, the better the accuracy.
24. Such an analysis is a very powerful method of assessing risk and risk transfer. However, there may not be sufficient data to determine the probability distributions easily and therefore managers will have to generate a lot of the data based upon a "common sense" analysis of possible outcomes. Ultimately it is better for the NHS body to use robust assumptions and unsophisticated techniques than vice versa as the output of any analysis can only be as reliable as the input.