Hueniken K, et al. Cancer 2020.
BACKGROUND: The treatment of head and neck cancer (HNC) may cause significant financial toxicity to patients. Herein, the authors have presented the development and validation of the Financial Index of Toxicity (FIT) instrument.
METHODS: Items were generated using literature review and were based on expert opinion. In item reduction, items with factor loadings of a magnitude <0.3 in exploratory factor analysis and inverse correlations (r < 0) in test-retest analysis were eliminated. Retained items constituted the FIT. Reliability tests included internal consistency (Cronbach α) and test-retest reliability (intraclass correlation). Validity was tested using the Spearman rho by comparing FIT scores with baseline income, posttreatment lost income, and the Financial Concerns subscale of the Social Difficulties Inventory. Responsiveness analysis compared change in income and change in FIT between 12 and 24 months.
RESULTS: A total of 14 items were generated and subsequently reduced to 9 items comprising 3 domains identified on exploratory factor analysis: financial stress, financial strain, and lost productivity. The FIT was administered to 430 patients with HNC at 12 to 24 months after treatment. Internal consistency was good (α = .77). Test-retest reliability was satisfactory (intraclass correlation, 0.70). Concurrent validation demonstrated mild to strong correlations between the FIT and Social Difficulties Inventory Money Matters subscale (Spearman rho, 0.26-0.61; P < .05). FIT scores were found to be inversely correlated with baseline household income (Spearman rho, -0.34; P < .001) and positively correlated with lost income (Spearman rho, 0.24; P < .001). Change in income was negatively correlated with change in FIT over time (Spearman rho, -0.25; P = .04).
CONCLUSIONS: The 9-item FIT demonstrated internal and test-retest reliability as well as concurrent and construct validity. Prospective testing in patients with HNC who were treated at other facilities is needed to further establish its responsiveness and generalizability.