Psychometric evaluation of the Chinese CancerSupportSourceTM-Caregiver among family caregivers to colorectal cancer patients using CTT and Rasch analyses

Family caregivers of colorectal cancer (CRC) patients often experience significant distress, yet no validated instrument exists in China to screen for this distress. This study aimed to translate the CancerSupportSource™-Caregiver (CSS-Caregiver) scale into Chinese and evaluate its psychometric properties among CRC family caregivers. A cross-sectional study was conducted with 340 family caregivers of CRC patients from three tertiary hospitals in Guangzhou. The Chinese CSS-Caregiver was translated and culturally adapted. Its psychometric properties were assessed using both Classical Test Theory and Rasch analysis. Reliability was evaluated via internal consistency and test-retest reliability. Validity was examined through confirmatory factor analysis (CFA), item-total correlations, and convergent validity. Measurement invariance was tested using multi-group CFA and Differential Item Functioning (DIF). The optimal cut-off score was determined via ROC analysis. The Chinese CSS-Caregiver demonstrated excellent internal consistency (Cronbach’s α = 0.902) and test-retest reliability (ICC = 0.935). CFA supported the original five-factor structure with good model fit (χ²/df = 2.064, CFI = 0.956, RMSEA = 0.073). Rasch analysis indicated generally satisfactory item fit and person separation. The scale exhibited strict measurement invariance across sex, age, and residence, with negligible DIF. Strong convergent validity was demonstrated by correlations with the criterion measures (r = 0.81 ~ 0.85). An optimal cut-off score of 18.5 was established. The Chinese CSS-Caregiver is a valid and reliable instrument for assessing distress among family caregivers of CRC patients in China, facilitating early identification and timely psychosocial support.