Measurement Properties of the Spinal Function Sort in Patients with Sub-acute Whiplash-Associated Disorders

Abstract

PURPOSE: To extensively analyze the measurement properties the Spinal Function Sort (SFS) in patients with sub-acute whiplash-associated disorders (WAD). METHODS: Three-hundred-two patients with WAD were recruited from an outpatient work rehabilitation center. Internal consistency was assessed by Cronbach's α. Construct validity was tested based on eight a priori hypotheses. Structural validity was measured with principal component analysis (PCA). Test-retest reliability and agreement was evaluated in a sub sample (n = 32) using intraclass correlation coefficient (ICC) and limits of agreement (LoA). The predictive validity of SFS for future work status at 1, 3, 6, and 12 months follow-up was determined by area under the curve (AUC) of receiver operating characteristics. Non-return to work (N-RTW) was defined with two cut-off points: workcapacity <50 and <100 %. RESULTS: N-RTW decreased from 50%, 1 month follow-up, to 14%, 12 months follow-up. Cronbach's α was 0.98, PCA revealed evidence for unidimensionality. ICC was 0.86, LoA was ±33 points. Seven out of eight hypotheses for construct validity were not rejected. AUC reduced with a longer follow-up from 0.71 for 1 month to 0.61 at 12 months, for cut-off point <50%. For cut-off point <100% these values were 0.71 and 0.59. CONCLUSION: In patients with sub-acute WAD test-retest reliability, internal consistency, construct- and structural validity of the SFS were adequate. LoA were substantial. Sensitivity to accurately predict N-RTW was poor. The predictive validity of the SFS for N-RTW of patients with sub-acute WAD from an outpatient work rehabilitation setting was only sufficient for the short term (1 month).