Context: Fatigue is the most common under-recognized symptom in cancer. Administering fatigue tools in multi-lingual and multi-literate populations may affect the quality and accuracy of the data collected as they rely on language to elicit responses.
Aim: The aim of the study is to develop and validate a tool to assess fatigue in cancer patients using response formats that are not language-dependent.
Methods: The content validity of the tool was established using the Delphi procedure and was field tested with 102 cancer patients. Test-retest reliability of the tool was tested with 55 cancer patients and 47 healthy individuals. Convergent, concurrent, and discriminant validity and internal consistency were established with 374 cancer patients, 202 survivors, and 75 healthy controls.
Statistical analysis: Qualitative analyses, descriptive statistics, product-moment correlation, analysis of variance, Cronbach's α coefficient, and exploratory factor analysis were conducted.
Results: The Cronbach's alpha of the SAFE in cancer patients and healthy individuals was .86 and .92, and their test-retest reliability ranged from .44 to .83. SAFE correlated significantly with measures of quality of life (QOL) (r = -0.54, p < .01), anxiety (r = 0.54, p < .01), depression (r = 0.5, p < .01), and sleep (r = 0.52, p < .01). The tool was able to distinguish between cancer patients, survivors, and healthy controls (p < .05). Two factors emerged namely "Fatigue Extent and impact" and "General fatigue" contributing to 52% of the variance in fatigue.
Conclusion: A symbolic tool using visual response formats to assess fatigue and its impact in cancer patients was developed and standardized with good reliability and construct, concurrent, and discriminant validity established.
Keywords: Cancer; Fatigue tool; Measurement; Standardization; Symbolic response formats; Validity.