Clinimetric properties of the ASAS health index in a cohort of Italian patients with axial spondyloarthritis

Marco Di Carlo, Valentina Lato, Marina Carotti, Fausto Salaffi

The impact of axial spondyloarthritis (axSpA) is considerable in many aspects of the life. Over the last decades, many efforts have been conducted to develop useful tools for the evaluation of disease activity. However, since the development of Assessment of SpondyloArthritis international Society Health Index (ASAS HI), no specific freely questionnaire to describe the overall picture of impairments, limitations and restrictions in activities or social partecipation were available. The aims of this study were to test the feasibility, reliability, and construct validity of the ASAS HI, in order to compare its clinimetric properties with the current available measures of disease activity, functional limitation and health status assessments in patients with axSpA. A cohort of 140 consecutive axSpA has been the object of study. The feasibility has been determined by the percentage of patients who were able to complete the questionnaire by themselves and by the time employed to fill the ASAS HI. The reliability has been evaluated performing a test-retest of the questionnaire within a week. The construct validity was examined in three ways. First, we examined construct convergent validity by correlating the scores of the ASAS HI with the Ankylosing Spondylitis Disease Activity Score (ASDAS)-CRP/ESR, the Simplified Ankylosing Spondylitis Disease Activity Score (SASDAS), the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), the Bath Ankylosing Spondylitis Metrology Index (BASMI), the Bath Ankylosing Spondylitis Functional Index (BASFI), the Ankylosing Spondylitis Quality of Life scale (ASQoL) and the EuroQoL Five Dimensional Questionnaire (EQ-5D). Secondly, we have created patient groups based on the patients' activity ranks (ASDAS-CRP and SASDAS categorisation) within the cohort to assess discriminative accuracy. Additionally, to distinguish patients with active and non-active disease and to assess their respective cut-off points values, the receiver operating characteristic (ROC) curve analysis was used. Thirdly, we analyzed the contribution of demographic (age, sex, and disease duration) and clinical variables (number of comorbidity and disease activity by ASAS-CRP) to the attainment of an ASAS HI condition by stepwise logistic regression. The mean time to complete the ASAS HI was 1.92 ± 0.76 min. Overall, the ASAS HI questionnaire was correctly completed by the majority of the patients (99,2 %). Coefficients of agreement between ASAS HI scores on first and second administrations were excellent and all items showed very good agreement (ICC = 0.976; range 0.966 to 0.982). The ASAS HI was correlated significantly with all other comparator scores (p <0.0001). The highest correlations were seen with ASQoL (rho 0.784; p <0.0001), BASFI (rho 0.671; p <0.0001) and SASDAS (rho 0.640; p <0.0003). On categorizing patients into different cut-off point of disease activity, with respect to the both ASDAS-CRP and SASDAS, ASAS HI scores were highly significantly different between the four categories (p <0.0001). An ASAS HI value of 4.0 resulted the cut-off with the highest combination of sensitivity and specificity (82.6 % and 86.3 %, respectively) to define the inactive disease. In the logistic regression model, high disease activity measured by ASDAS-CRP (coefficient 2.39; p <0.0001), was the only independent variable associated with ASAS HI. The results reported in this study confirm the feasibility, reliability and validity of the ASAS HI in Italian patients with axSpA. Even if ASAS HI is not a disease activity index, of particular interest appears the cut-off value of 4.0, under which could be defined the inactive disease. This value could represent an easily applicable starting point in daily clinical practice.