The STarT Back Screening Tool: The Nepali Translation, Cross-Cultural Adaptation and Measurement Properties in Adults With Non-Specific Low Back Pain.

INTRODUCTION/

OBJECTIVE

The STarT Back Screening Tool (SBST) stratifies low back pain (LBP) patients based on their risk of chronicity to guide treatment accordingly. The absence of its validated Nepali version limits stratified LBP care in Nepal. The study aimed to translate and cross-culturally adapt the SBST into Nepali and evaluate its measurement properties in adults with LBP.

METHODS

The measurement properties of the Nepali SBST were evaluated in 102 Nepali adults with non-specific LBP. We assessed content validity, internal consistency, test-retest reliability, construct and discriminant validity. Item redundancy was evaluated using Cronbach's alpha (α > 0.90), test-retest reliability using Intraclass Correlation Coefficient (ICC2,1) and Cohen's kappa using established cutoffs score for categorising patients into risk groups, construct validity using hypothesis testing (if a minimum of 75% of the hypotheses were supported), and discriminant validity using Area Under the Curve (AUC) with the reference scales administered at baseline.

RESULTS

Cronbach's alpha scores were 0.72 for the overall scale and 0.66 for the psychosocial subscale. Test-retest reliability values were good to excellent with ICC2,1 of 0.94 (95% CI: 0.87-0.97) for the overall scale and 0.87 (95% CI: 0.73-0.94) for the psychosocial subscale and Kappa values of 0.68 (95% CI: 0.43-0.93) for the overall scale and 0.79 (95% CI: 0.52-1.00) for psychosocial subscale. Construct validity was confirmed as 100% of a priori hypotheses were met. Acceptable discriminative validity was observed with reference scales with AUCs (0.75-0.80).

CONCLUSIONS

Nepali SBST demonstrates the reliability and validity of screening for chronicity risk in Nepali adults with LBP. Future studies should evaluate its responsiveness, predictive abilities, and effectiveness in stratifying LBP patients in the Nepalese context.