ABSTRACT
The plausible range of values for the confidence interval of the observed incremental cost-effectiveness ratio (ICER) can lead to different interpretations, especially if the confidence interval covers two or more regions in the cost-effectiveness plane. It is therefore useful to count the proportion of ICERs in each region of the cost-effectiveness plane. When looking at the cost-effectiveness of cancer trials there are several important results that need to be interpreted, both clinically and from a cost-effectiveness perspective. For cost-effectiveness, the primary analysis variable of interest in trials will be the cumulative cost incurred by the patient up to the end of the trial. A key issue present in most cost data is informative censoring due to the lack of a common rate of cost accrual over time among patients. Informative censoring means that the reason for censoring may well be related to treatment resulting in incomplete costs.
