VALIDATION DATA

The validation study of SPAKE included patients recruited from 13 rheumatology centers in France and one private practice center. Eight centers tested reproducibility, and six centers tested sensitivity to change by including patients scheduled for an educational session after completing the questionnaire. Inclusion criteria were patients aged ≥18 years, diagnosed with axSpA, fulfilling the 2009 Assessment of SpondyloArthritis international Society (ASAS) classification criteria, and able to complete a French-language questionnaire. This study reports the development and the validation of the SPAKE, showing good acceptability with no missing responses, good internal and external consistency, good reproducibility, and a high sensitivity to change assessed before and after PE sessions.

The validation of SPAKE showed good internal consistency, with a Kuder-Richardson α coefficient of 0.85 for the long-form and 0.79 for the short-form. The item-retest correlation for the long-form ranged from 0.14 to 0.50, with the highest correlations found in the knowledge of disease and pharmacological treatment domains, and the lowest correlation found in the comorbidity domain. The correlation coefficient between the long-form and short-form scores was excellent at 0.98. 

Reproducibility of SPAKE was assessed in 61 subjects, with good Lin concordance correlation coefficients of 0.81 (95% CI 0.72-0.89) for the long-form and 0.80 (95% CI 0.70-0.88) for the short-form. Better reproducibility was found for knowledge of disease and treatments, while self-care and adaptive skills had lower reproducibility. Of the 42 questions, 28 had a concordance rate of >80%.

The SPAKE questionnaire was developed in three steps. Step 1 involved extracting knowledge items from existing questionnaires and using a Delphi process to narrow down the list of potentially useful items. Participants included rheumatologists, HCPs, and patients from 13 multidisciplinary teams in France. Step 2 involved a final Delphi round to select essential and useful items. The first version of the questionnaire was developed by 2 rheumatologists and 1 rheumatology nurse, based on the results of the final Delphi round. The questionnaire was designed for patients with axSpA-predominant features, with or without associated features of peripheral SpA, PsO, or IBD. The questionnaire was later reviewed and reformulated by 2 rheumatologists, 2 rheumatology nurses, and 1 patient research partner to ensure its ease of understanding and relevance to the Delphi results.

The questionnaire underwent linguistic validation and cognitive debriefing by 10 patients. The completion time was noted. The questionnaire was then reviewed by the investigating centers to obtain the final version.

In a multivariate analysis, factors independently associated with better knowledge levels were sex (female vs male), grade level (> high school vs ≤ high school), how well informed the patient was according to the opinion of the rheumatologist or nurse, and patient education sessions (yes vs no). There was no correlation found with age or disease duration.