Evaluation of the spinal pain score in AS--a psychometric analysis

To evaluate the spinal pain score (SpiPS), a semi-objective instrument for measuring disease activity in AS.

The SpiPS is based on the physical examination of the spine and was evaluated according to the outcome measures in rheumatology filter in 659 AS patients from different cohorts including two interventional trials. Aspects of truth, discrimination and feasibility were assessed.

The SpiPS significantly distinguished between patients with high and low disease activity. The correlation of the SpiPS with the disease activity index BASDAI was relatively weak in the entire cohort of AS patients (r = 0.36) but moderate to good in AS patients with short disease duration (r = 0.66). Correlations with the functional index BASFI (r = 0.38), patient global (r = 0.18), physician global (r = 0.51) and BASMI (r = 0.36) were weak to moderate in the entire cohort. Logistic regression revealed SpiPS and patient global to be independently associated with disease activity (BASDAI) after adjustments for age, gender, disease duration, CRP and HLA-B27. Sensitivity to change expressed as effect size (ES) was 0.82 and 1.58, respectively, and highly comparable with that of BASDAI in the two interventional trials. Reproducibility of the SpiPS was good (intra-observer variability r = 0.93, inter-observer variability r = 0.79).

The SpiPS is a measure of disease activity in AS and is sensitive to change. However, disease domains other than disease activity are also captured by the SpiPS, and our analysis failed to demonstrate any additional value of the SpiPS over existing instruments.