| Abstract |
Background To assess the 10-item Spine Functional Index (SFI-10) clinimetric properties in a general musculoskeletal disorder (MSD) spine population. Ascertain the psychometric characteristics’ consistency with the developmental study findings for structural and criterion validity, internal consistency, and floor/ceiling effect; establish the longitudinal characteristics for test–retest reliability, responsiveness, construct validity, and error scores; and clarify practical characteristics of readability, missing responses, and time/errors for completion/scoring related to administrative burden. Methods A longitudinal study of deidentified spine MSD patients (n = 1317, 53.4% female, age = 18-91yrs, av = 49.5 ± 16.4yrs; neck = 36.5%, mid-back = 8.4%, low back = 56.0%, multi-site = 0.3%) who completed the SFI-10, the Patient Specific Functional Scale (PSFS), and Numerical Rating Scales for Global-function (G-NRS) and Pain (P-NRS). Structural validity used factor analysis, exploratory (EFA) and confirmatory (CFA), plus Rasch analysis. Criterion validity used Spearman's correlation coefficient (r) between the SFI-10 and criteria (PSFS, G-NRS and P-PRS) scores, and construct validity (n = 91, known-groups independent t-test). Internal consistency used Cronbach's alpha (α) and floor/ceiling effects were determined. Subgroups determined reliability (n = 104, intraclass correlation coefficient, ICC2.1); error (n = 171) through the standard error of measurement (SEM) and minimum detectable change (MDC90). Responsiveness (n = 171) was calculated using effect-size (ES), standard response mean (SRM), and area under the curve (AUC); and interpretability through the minimal clinically important difference (MCID). Practicality (n = 16) clarified missing responses, readability, and time/errors for completion/scoring. Results The SFI-10’s structural validity was unequivocally one-dimensional from EFA and verified by CFA with acceptable fit-indices (chi-square/df = 2.88, CFI = 0.981, TLI = 0.975, RMSEA = 0.061), and supported by Rasch analysis (PSR = 0.79, Infit = 0.678–1.216, Outfit = 0.604–1.279, Item-difficulties = -1,215–2.488). Criterion validity varied from high (G-NRS, r = 0.60) and moderate (PSFS, r = 0.43) to low-inverse (P-NRS, r = -0.24). Internal consistency was strong (α = 0.84) and no floor/ceiling effects were present. Reliability was excellent (ICC2.1 = 0.97), responsiveness substantial (ES = 1.54; SRM = 1.64; AUC = 0.89), and measurement error robust (SEM = 3.84; MDC90 = 8.98%, MDIC = 10.4%), with construct validity confirmed (p < 0.001). Practicality showed no missing responses, completion/scoring errors < 1%, excellent readability (Grade = 5.1, Ease = 74.1%), short completion (39.2 ± 10.3 s) and scoring times (8.5 ± 1.8 s). Conclusions The SFI-10 demonstrates sound measurement properties in a general physiotherapy outpatient MSD spine population for both psychometric and practical characteristics. Further investigation in culturally diverse settings that include both inpatients and community settings with whole-spine and regional-spine criteria is required. |
