Health variations among breast-cancer patients from different disease states: evidence from China.

BACKGROUND

This study aimed to obtain health utility parameters among Chinese breast cancer patients in different disease states for subsequent health economics model. In addition, we aimed to explore the feasibility of establishing a breast cancer health utility mapping model in China.

METHODS

Multiple patient-reported health attributes were assessed, including quality of life, which was measured by the Functional Assessment of Cancer Therapy-Breast (FACT-B) instrument; health utility and self-rated health, which were measured by the EuroQol-5 Dimension-5 Level (EQ-5D-5L) questionnaire. Multivariate regression models, including a linear regression model, an ordinal logistic regression model and a Tobit model, were employed to analyze health differences among 446 breast cancer patients. Subgroup analyses were performed to examine differences in multiple dimensions of health derived from the FACT-B and EQ-5D-5L instruments. A mapping function was used to estimate health utility from quality of life. Rank correlation analyses were employed to examine the correlation between estimated and observed health utility values.

RESULTS

A total of 446 breast cancer patients with different disease states were analyzed. The health utility values of breast cancer patients in the P state (without cancer recurrence and metastasis), R state (with cancer recurrence within a year), S state (with primary and recurrent breast cancer for the second year and above), and M state (metastatic cancer) were 0.81 (SD ± 0.23), 0.90 (SD ± 0.12), 0.78 (SD ± 0.31), and 0.74 (SD ± 0.27), respectively. There were positive correlations between all scores, including every domain of the FACT-B instrument (p < 0.001). Results from multivariate analysis suggested that patients in the R and M states had lower scores for overall quality of life (R, β = - 9.45, p < 0.01; M, β = - 6.72, p < 0.05). Patients in the M state had lower health utility values than patients in the P state (β = - 0.11, p < 0.05). Estimated health utility values, which were derived from quality of life by using a mapping function, were significantly correlated with directly measured health utility values (p < 0.001).

CONCLUSIONS

We obtained the health utility and health-related quality of life (HRQoL) scores of Chinese breast cancer patients in different disease states. Mapping health utility values from quality of life using four disease states could be feasible in health economic modelling, but the mapping function may need further revision.