Cost data for individual patients included in clinical studies: no amount of statistical analysis can compensate for inadequate costing methods.

This work examines the quality of the cost methods used to derive patient level costs in 45 economic evaluations conducted alongside randomised controlled trials. The perspective of the cost analysis, the methods used to determine quantities and values of resources and how the cost data were reported are examined. The reported costing methods were found to be of poor quality, highlighting the need for greater rigour. Researchers to date appear more concerned with whether cost data have been subjected to the appropriate statistical analysis. For the results of clinical studies to be valid both cost methods and the methods used for the statistical analysis of cost data should be of a high quality. Copyright 2002 John Wiley & Sons, Ltd.