Validity and Reliability of the Thai Version of the Barthel Index Self-Report (BI-SR-TH) for People with Spinal Cord Injury

OBJECTIVE This study aimed to cross-culturally translate and evaluate the psychometric properties of the Thai version of the Barthel Index Self-Report (BI-SR-TH) for Thai people with spinal cord injury (SCI). METHODS A descriptive cross-sectional validation and reliability study was conducted at the Maharaj Nakorn Chiang Mai Hospital, involving 109 Thai people aged 18 or older with either traumatic or non-traumatic SCI. The English postal version of the Barthel Index was translated into Thai following standard cross-cultural procedures, including forward and backward translation. Participants completed both the BI-SR-TH and the Thai EuroQoL (EQ-5D-5L) questionnaires via self-report and clinician assessment using the Barthel Index. The BI-SR-TH was re-administered to the participants one week later. Psychometric evaluation included assessments of internal consistency (Cronbach’s alpha), concurrent validity (Spearman’s correlation with clinician-rated BI), construct validity (convergent and discriminant validity assessed by correlations with relevant and distinct EQ-5D-5L domains), and test-retest reliability (intraclass correlation coefficient, ICC). The agreement between the BI-SR-TH and BI scores was assessed using ICC and the Bland-Altmand Altman analysis. Item-level agreement was assessed using the Kappa statistic. RESULTS The BI-SR-TH demonstrated good internal consistency (Cronbach’s alpha = 0.836) and excellent concurrent validity with a clinician-rated BI (r = 0.884). In terms of construct validity, the BI-SR-TH showed strong negative correlations with relevant EQ-5D-5L domains, e.g., mobility, self-care, and usual activities (r = –0.67 to –0.81, p < 0.01), demonstrating convergent validity. In contrast, correlations with unrelated domains such as anxiety and pain were weak or nonsignificant, indicating discriminant validity. The total score agreement between BI-SR-TH and the clinician BI was high (ICC = 0.916), and Bland-Altman analysis indicated a small mean difference. Most BI-SR-TH items showed good agreement with clinician ratings (Kappa 0.61–0.72) with the exception of the bowels and bladder items which showed only fair agreement (Kappa = 0.34 and 0.35, respectively). Test-retest reliability was also high (ICC = 0.924). CONCLUSIONS The BI-SR-TH is a valid and reliable tool for assessing functional independence in Thai people with SCI. However, the bowels and bladder items require further refinement for this population KEYWORDS Barthel Index, self report, functional status, spinal cord injury, reliability and validity