Development and validation of a screening tool for SPondyloArthritis Screening in Sub-Saharan Africa: SpASSS questionnaire

Abstract

Abstract Objective To develop and validate a screening tool to identify patients with a high likelihood for Spondyloarthritis (SpA) in the Democratic Republic of the Congo (DR Congo). Methods The development of the SpA Screening questionnaire in Sub Saharian Africa (SpASSS) questionnaire followed 3 steps: The item generation was carried out by a systematic literature review according to the PRISMA guidelines on the clinical manifestations of SpA, interviewing clinical experts and the classification criteria for Spondyloarthritis. The candidate questions were tested in a population of 50 consecutive patients with confirmed diagnosis of spondyloarthritis, in a control population of rheumatic disease excluding SpA and in a group of 200 non-rheumatic participants, randomly chosen in the general population for question reduction and validation. Descriptive statistical analyses were performed to assess socio-demographic characteristics and response distribution for each item. Their diagnostic performance was investigated using ROC curves. For validation, principal component analysis was performed using factor analysis. Referral strategy score for SpA was determined by adjusted Cronbach’s alpha coefficient. Results Mean ± SD age of SpA cases was 41.8 ± 14.4 years, 56% were men compared to diseased controls 60.0 ± 12.5 years, 28.7% men (p < 0.001). 14/20 items showed a statistically significant difference (p < 0.05) between SpA cases and control groups. All items were factorable and 6 components were identified. Only the two first components (C1 with 8 items, C2 with 3 items) showed a significant threshold for reliability in detection of suspected SpA with a Cronbach's alpha of 0.830 and 0.708. All validated items of these two components showed the global reliability threshold with α-adjusted Cronbach calculated at 66.9%. The performance for correctly screening SpA was demonstrated with an area under the curve of 0.938 (0.884–0.991) and 0.794 (0.728–0.861) for C1 and C2 respectively. Conclusions This validation and item reduction of the SpASSS questionnaire for SpA might identify patients to refer for case ascertainment and will help conducting future epidemiological and clinical studies in the DR Congo. Strengths and limitations of this study • To the best of our knowledge, this is the first study in Sub-Saharan Africa based on local data to develop a screening tool for SpA in the population for epidemiological and clinical use. • Referral strategies based on context-specific data are necessary to provide accurate case definition and epidemiological data, thus reducing methodological bias. • In the SpA group, no discrimination was made regarding SpA subtypes, disease duration, activity and severity.