Measurement properties of the Functional Assessment of Cancer Therapy-Breast may depend upon the education levels of patients

Abstract

ABSTRACT Background: Education level may vary in older cancer patients. This study aims to compare the measurement properties of the Functional Assessment of Cancer Therapy-Breast (FACT-B) when self-administered by breast cancer patients with various education levels. Methods: An observational study of 244 Singaporean breast cancer patients who self-administered the instrument. FACT-General and FACT-B scores were assessed for the discriminatory ability, responsiveness to change and test–retest reliability. We hypothesized that patients with better performance status would result in higher FACT-General and FACT-B scores. Regression models were constructed, and relative precisions were also examined. Results: The mean baseline FACT-General and FACT-B scores monotonically decreased with performance status in the higher but not lower educated group (relative precision = 1.13 and 1.08, respectively). Stronger responsiveness to deterioration in quality of life were found in the higher (each p < 0.001) than lower educated group (p = 0.146 and 0.245). Larger intra-class correlation coefficient of the FACT scores (each p < 0.05) and smaller variability in the Bland-Altman plots indicated better test–retest reliability in higher educated group. Conclusion: Measurement properties may be reduced when the instrument is self-administered by lower educated cancer patients.