There are several scenarios under which clinically effective technologies may be found not to be cost-effective even if they are zero priced. There may be costs associated with delivering the technology which remain even when the price is reduced to zero and these costs alone may outweigh the health benefits achieved. But even in the situation where a clinically effective technology can be acquired and delivered for zero cost, there are scenarios in which that technology may fail to demonstrate cost-effectiveness because it increases other aspects of resource use. We have described four related but different scenarios in which clinically effective treatments result in additional time being spent in health states with high resource use and / or low health-related quality of life either during or after the treatment period.

We have reviewed the methodological literature around the exclusion of unrelated costs from cost-effectiveness analyses to determine whether there is a case for excluding some of the costs incurred in periods of additional survival in the case studies we identified. In the majority of the case studies, the costs incurred during periods of additional survival were related to either the technology being appraised or the condition the technology was intending to treat. These cannot therefore be considered to be unrelated costs.

Whilst we have mainly focused on the issues related to costs incurred in added life years it is also important to consider if the benefits have been properly accounted for in the cost- effectiveness model. In addition we discuss how treatments which are cost-effective in the general population may not be cost-effective in particular groups of patients with high background care costs.