OBJECTIVE: To systematically identify utility weights for health states in breast cancer.
METHODS: Searches were performed of electronic databases (PubMed, EMBASE and the Cochrane Library, including DARE, NHS EED and HTA databases) and internet resources for the period 1990 to date. Sources potentially containing relevant information were retrieved and reviewed.
RESULTS: Fifty-nine studies were identified as potentially containing utility weights for breast cancer health states. These were assessed for methodological compliance with the NICE reference case, leading to the exclusion of 50 noncompliant studies. Within the remaining nine studies there was wide variability between both alternative NICE compliant estimates for similar health states, and the health states defined. In some cases estimates for poor health states (for example metastatic disease) were higher than those for good health states (for example remission). For some health states (notably terminal disease) there are no estimates available based on NICE recommendations.
CONCLUSION: A review of utility weights for breast cancer health states has revealed high levels of uncertainty within the identified estimates. Despite the quantity of information available there is no universally accepted set of health states covering the whole of the disease pathway for breast cancer, which has lead to the development of numerous utility estimates for numerous health states. The review also highlighted that the majority of studies undertaken (85% of potential studies identified) do not conform to the methodological standards stipulated in the NICE reference case. This is predominantly due to the elicitation method used. For some health states there are no estimates available based on these recommendations. It is a difficult task to identify a coherent set of health state utilities covering the entire disease pathway in breast cancer using previously published data and which conform to NICE standards.
