A review and meta analysis of health state utility values in breast cancer

Background and purpose: Health-related quality of life is an important issue in the treatment of breast cancer, and health-state utilities are essential for cost-utility analysis. This paper identifies and summarises published utilities for common health-related quality of life outcomes for breast cancer, considers the impact of variation in study designs used, and pools utilities for some breast cancer health states.

Data sources and study selection: 13 databases were searched using key words relating to breast cancer and utility measurement. Articles were included if specified empirical methods for deriving utility values were used and details of the method, including number of respondents, were given. Articles were excluded if values were based on expert opinion or were not unique.

Data extraction and synthesis: The authors identified 49 articles which met their inclusion criteria, providing 476 unique utilities for breast cancer health states. Where possible, mean utility estimates were pooled using ordinary least squares with utilities clustered within study group and weighted by both number of respondents and inverse of the variance of each utility. Regressions included controls for disease state, utility assessment method and other features of study design.

Results: Utility values found in the review are summarised for six categories: 1) screening related states, 2) preventative states, 3) adverse events in breast cancer and its treatment, 4) non-specific breast cancer, 5) metastatic breast cancer states and 6) early breast cancer states. Pooled utility values for the latter two categories are estimated, showing base state utility values of between 0.668 and 0.782 for early breast cancer and 0.721 and 0.806 metastatic breast cancer depending upon which model is used. Utilities were found to vary significantly by valuation method, and who conducted the valuation.

Conclusions: A large number of utility values for breast cancer are available in the literature; the states that these refer to are often complex, making pooling of values problematic.