front |1 |2 |3 |4 |5 |6 |7 |8 |9 |10 |11 |12 |13 |14 |15 |16 |17 |18 |19 |20 |21 |22 |review

Inputs and consequences of a health intervention accrue at different times, especially for chronic disease and population-based programmes to deal with them. In this case we cannot directly compare the inputs of a programme starting today with its consequences which will accrue in thirty years' time. So economists adjust the valuation of such consequences to take account of the difference in time by using a technique called discounting which allows the calculation of the present values of inputs and benefits which accrue in the future. Discounting is mainly based on a time preference which assumes that individuals prefer to forego a part of the benefits if they accrue it now, rather than fully in the uncertain future. The strength of this preference is expressed by the discount rate which is inserted in economic evaluations. The choice of a discount rate and the choice of which items it should be applied to are a matter of intense debate among economists. The comparison between inputs and consequences does not happen in a vacuum, its context influences the comparison itself. For instance, production costs are dependant on numbers of units produced. For example as production rises, costs may decrease if fixed costs (those that do not vary with the production) are divided by a larger number of units and no other investment is necessary. Often the choice is not about whether to carry out a certain intervention or not but about varying the volume of services currently provided or shifting resources between services. The comparison then must be based on inputs necessary at that level of variation and the change in consequences that result from that variation of inputs. This type of logic is called marginal and is at the basis of all the costing procedures which economists use. All comparisons made within the framework of an economic evaluation are based on marginal logic. Evaluations are models which attempt to capture and summarise reality. However, the effects of healthcare are often uncertain and our models tend to be based on real data (epidemiological, clinical or resource data) which often are either incomplete, or of dubious quality or simply not there. Epidemiology, for instance, provides us with an estimate of probabilities (of developing a particular disease, or of effecting a transition from one stage of the disease to the next), or even a range of estimates which is at times wide. Where data is absent or uncertain, the gap may have to be filled by assumptions. In order to deal with uncertainty and carry on with our decision-making process, economic evaluations use a range of techniques, called sensitivity analysis, which repeats the comparison between inputs and consequences varying the assumptions underlying the estimates. In other words it tests the robustness of the conclusions by varying the items around which there is uncertainty.