Purpose: The aim was to determine the responsiveness and minimal important change (MIC) of the questionnaire ProFitMap-neck that measures symptoms and functional limitations in women with neck pain. The same measurement properties were determined for Neck Disability Index (NDI) for comparison purposes.
Methods: Longitudinal data were derived from two randomized controlled trials, including 103 and 120 women with non-specific neck pain, with questionnaire measurements performed before and after interventions. Sensitivity and specificity to discriminate between improved and not or little changed participants, based on categorization of a global rating of change scale (GRCS), were determined for the ProFitMap-neck indices and NDI by using area under receiver operating characteristic curves (AUC). Correlations between the GRCS anchor and change scores of the questionnaires were also used to assess responsiveness. The change score that showed the highest combination of sensitivity and specificity was set for MIC.
Results: The ProFitMap-neck indices showed similar responsiveness as NDI with AUC exceeding 0.70 (Range: ProFitMap-neck, 0.74-0.83; NDI, 0.75-0.86). The MIC in the two samples ranged between 6.6 and 13.6 % for ProFitMap-neck indices and 5.2 and 6.3 % for NDI. Both questionnaires had significant correlations with GRCS (Spearman's rho 0.47-0.72).
Conclusions: Validity of change scores was endorsed for the ProFitMap-neck indices and NDI with adequate ability to discriminate between improved and not or little changed participants. Values of minimal important change were presented.
Keywords: Anchor-based; Discrimination; Physical function; Sensitivity; Specificity; Validity.