Validation of the Hindi version of the Multidimensional Fatigue Inventory-20 (MFI-20) in Indian cancer patients.

OBJECTIVE: The present study was designed to validate the Hindi version of the Multidimensional Fatigue Inventory-20 (MFI-20) in Indian oncology population.
METHODS: The original English version of the MFI-20 was translated into Hindi (hMFI-20) using the translation and back translation processes. The hMFI-20 was administered to 200 cancer patients. The item analysis for hMFI-20 was carried out using the corrected item-total correlation. The confirmatory factor analysis (CFA) was employed to test whether the original factor structure of MFI-20 is confirmed for the hMFI-20. Further, convergent and discriminant validities were also tested. The reliability of the hMFI-20 was evaluated by computing composite reliability and Cronbach's α coefficient.
RESULTS: Corrected item-total correlation value for each of the items of hMFI-20 was greater than 0.6. Results of the CFA (comparative fit indices (CFI) = 0.91, root mean squared residual (RMR) = 0.04, root mean square error of approximation (RMSEA) = 0.028, and χ (2) = 45.68, p > 0.05) indicated that the five-factor model provided a good fit to the data. The findings indicated that hMFI-20 has a good convergent (composite reliability (CR) >0.7; average variance extracted value (AVE) >0.5) and discriminant (maximum shared variance (MSV) < AVE; average shared variance (ASV) < AVE; square root of AVE > inter-factor correlations) validities. The Cronbach's α coefficient for the total hMFI-20 was 0.8 and was more than 0.7 for each of the five factors.
CONCLUSIONS: We conclude that the hMFI-20 has a high internal consistency and reasonable construct validity. Therefore, the hMFI-20 is a reliable and valid tool to assess the multidimensional fatigue in Indian oncology population. However, we recommend further validation of hMFI-20 in population of cancer patients of different linguistic settings and regions of India.