Background/Purpose: The Psoriatic Arthritis Disease Activity Score (PASDAS) is a newly developed composite disease activity measure that summarizes psoriatic arthritis (PsA) disease activity with a score ranging from 0-10. PASDAS captures articular and extra-articular manifestations of the disease and the impact of the disease on the patient via the following variables: swollen and tender joints, dactylitis, Leeds enthesitis index, C-reactive protein, physician and patient global disease activity, and the physical component summary score (PCS) of the medical outcomes survey Short Form 36 (SF-36). A limitation of PASDAS is that the score depends on the patient completing the SF-36, which requires significant time to complete. A shorter 12-question subset of SF-36, the SF-12, with the same range of values, agrees well with the SF-36 in many patient populations. The current objective was to measure the agreement between PASDAS calculated using the standard scoring formula and a modified PASDAS (mPASDAS) calculated by replacing the SF-36-PCS with SF-12-PCS in the scoring formula.

Methods: 100 patients meeting CASPAR criteria for PsA attending a PsA clinic for follow-up visits were consecutively recruited in June and July 2015. All variables required to calculate PASDAS were collected and PASDAS was calculated for each patient. The 12 item responses for SF-12 were extracted from the SF-36 questionnaires. The mPASDAS was subsequently calculated based on the PASDAS scoring formula where SF-36 –PCS was replaced by SF-12 – PCS. A Bland-Altman plot of the mean differences in scores calculated for PASDAS and mPASDAS measured agreement between the two sets of scores. The misclassification of patients based on disease activity as measured with mPASDAS compared to the classification based on the original PASDAS was also determined.

Results: An analysis of 100 patients [53% male, mean (SD) age 57.3 (11.9) years, mean (SD) disease duration 16.9 (11.7) years] revealed that the mean (SD) PASDAS was 3.29 (1.39) and the mean (SD) mPASDAS was 3.24 (1.27). The Bland-Altman plot produced a mean difference (95%CI) between mPASDAS and PASDAS of -0.05 (-0.07, -0.03). The lower limit of agreement was -0.24 (95%CI -0.21, -0.28) and the upper limit was 0.14 (95%CI 0.10, 0.17). The validity of the limits of agreement was supported by a number of normality tests indicating normally distributed differences. No relationship between the differences and the mean values of the scores exist indicating these limits of agreement are valid across the full range of the measures. Discrepancies were seen in 5 out of the 100 cases resulting in a misclassification rate of 5%. PASDAS classified 33 cases as low disease, 42 cases as moderate disease, and 25 cases as high disease. Using the same cutoff scores, mPASDAS classified 34 cases as low disease, 43 cases as moderate disease, and 23 cases a high disease.

Conclusion: The mPASDAS and PASDAS show strong agreement without clinically significant systematic bias in mPASDAS scores. In clinical settings, the mPASDAS may replace PASDAS in disease activity assessment given the strong agreement, low misclassification rate and significantly reduced patient questionnaire burden.

« Back to 2016 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/a-modification-of-the-psoriatic-arthritis-disease-activity-score-mpasdas-using-sf-12-as-a-measure-of-quality-of-life/