Sensitivity of Preference-Based Quality-of-Life Measures for Economic Evaluations in Early-Stage Melanoma

Abstract

The diagnosis of a life-threatening disease like melanoma can affect all aspects of a person's life, including health-related quality of life (HRQOL) and psychological aspects of melanoma such as fear of cancer recurrence (FCR). Economic evaluations of psychological interventions require preference-based (utility) instruments that are sensitive to changes in well-being and HRQOL; however, very few studies have evaluated the sensitivity of these instruments when used for people with melanoma. To compare utility scores from the multiple-attribute instrument Assessment of Quality of Life-8-Dimension Scale (AQoL-8D) with the mapped utility scores of the Functional Assessment of Cancer Therapy-Melanoma (FACT-M) and to investigate the sensitivity of both instruments in identifying the influence of FCR on HRQOL. This assessment of data from a randomized clinical trial of a psychoeducational intervention to reduce FCR, conducted at 3 high-risk melanoma clinics in Australia, evaluated 164 patients with early-stage melanoma and a high risk of developing a second primary melanoma. The FACT-M and AQoL-8D were used to assess HRQOL and FCR among the study participants. Concurrent validity was assessed by comparing the total and subdomain scores of the 2 instruments, and the strength of associations was assessed using Pearson correlation coefficient. Convergent validity was assessed by comparing participants' HRQOL, demographic, and clinical characteristics using the χ2 test and F statistic. Both the FACT-M and AQoL-8D utilities were regressed on FCR Inventory (FCRI) severity scores to estimate the effect of elevated FCR on HRQOL. A total of 164 participants completed the baseline questionnaires, but only 163 met all inclusion criteria and underwent the full analysis: 72 were women; 91 were men; and mean (SD) age was 58.2 (12.1) years. Both the AQoL-8D and FACT-M instruments showed good concurrent validity and could differentiate between relevant subgroups including level of FCRI severity. The AQoL-8D and FACT-M utilities were strongly correlated (r2 = 0.57). Respondents had a mean (SD) AQoL-8D utility of 0.77 (0.2), and a mean (SD) FACT-M utility score of 0.76 (0.07). High levels of FCRI severity were associated with a decrease in utility of 0.12 (95% CI, -0.19 to -0.05) as measured by AQoL-8D, and a decrease of 0.03 (95% CI, -0.05 to -0.01) as measured by the FACT-M. For economic evaluations of psychological interventions in melanoma, the AQoL-8D and FACT-M are valid measures of utility; however, the AQoL-8D demonstrates greater sensitivity to FCRI severity. Our results suggest a significant association between FCR and HRQOL.