Measuring indirect costs: is there a problem?

It is generally accepted that, in order to appraise from a societal perspective, indirect costs should be included in economic evaluations. What is less generally accepted, however, is the method by which such indirect costs should be calculated. Different methods and assumptions can, in theory, produce different results. Previous studies have commented on this variability, and most suggest a need for consensus. This having been said, no previous study appears to have demonstrated that variability in method is actually a practical problem, in the sense that the use of different costing methods would lead to different policy conclusions. In this paper, we examine this issue with respect to a specific intervention, namely, paediatric cochlear implantation (PCI). Based on questionnaire data, we estimate the indirect costs of PCI using a variety of methods. Thereafter, we integrate these indirect costs into a cost-utility analysis of PCI, and demonstrate that the variability in methods can significantly affect the outcome of a cost-effectiveness study. Therefore, in this case at least, the measurement of indirect cost is indeed a problem.