The FACT-melanoma quality-of-life instrument: Comparison of a five-point and four-point response scale using the Rasch measurement model

摘要

The FACT-Melanoma (FACT-M) is one of only two validated quality-of-life instruments designed specifically for use in patients with melanoma. The instrument incorporates FACT-G, followed by a set of questionnaire items that are specific to melanoma; all items are scored on a five-point response scale. The primary aim of this study was to evaluate the five-point response format of the FACT-M for goodness of fit to the Rasch measurement model, and to investigate whether rescoring the instrument using a four-point response format improved the psychometric properties. Two data sets of similar patient sample sizes (n=127 and 123) were used to test the reliability and validity of the generic instrument (FACT-G) to measure quality of life for patients with melanoma. The Additional Concerns and Melanoma Surgery subscales were subjected to a more detailed analysis using a combination of confirmatory factor analysis and Rasch analysis techniques. The Rasch model fit of the FACT-M was improved by the use of a four-point response format, together with the deletion of three items. Principal components analysis suggested that two melanoma-specific subscales existed within the Additional Concerns subscale and each could be reduced to seven items, respectively, with improved goodness of fit to the Rasch model. The FACT-M instrument showed improved fit to the Rasch model specifications when the items adopted a four-point response format. These results point to possible improvements in the content and structure of the FACT-M for use in future melanoma clinical trials. However, further study should be conducted with larger samples, selected by disease and treatment status.