Correlation of patient‐reported routine assessment of patient index data with clinical measures of disease activity in psoriatic arthritis

Abstract Aim A treat‐to‐target strategy is recommended for management of psoriatic arthritis (PsA), although there is lack of agreement regarding the best measure of disease activity to target. Physician assessments included in traditional indices can be complex and time consuming to complete and cannot be readily conducted by telehealth. This study compares the routine assessment of patient index data 3 (RAPID3), an efficient tool comprising patient self‐assessment, with traditional clinician‐led composite measures in the PsA clinic setting. Methods Data were collected prospectively from July 2016 to March 2020 in 2 dedicated PsA clinics in Sydney, Australia. A receiver operating characteristic (ROC) curve was created for comparison of RAPID3 score with composite scores minimal disease activity (MDA), very low disease activity (VLDA) and disease activity in psoriatic arthritis (DAPSA) in low disease activity or remission. Results Ninety‐three patients had simultaneous collection of RAPID3 and MDA measures. Mean (SD) age was 49.9 (13.5) years, 50.5% were male and 23 (24.7%) had erosive disease at baseline. RAPID3 scores ≤3.2 and ≤2.7 (range 0‐30) had high sensitivity and specificity for VLDA and DAPSA remission respectively, with ROC curve area under the curve (95% CI) of 0.94 (0.91‐0.97) and 0.96 (0.93‐0.99). Conclusion RAPID3 has good agreement with physician‐led composite scores of MDA, VLDA and DAPSA, and provides a viable alternative to composite scores. This is particularly helpful in settings that do not allow for clinical examination, for example telehealth.


| INTRODUC TI ON
Psoriatic arthritis (PsA) is associated with peripheral and axial joint involvement and distinctive extra-articular manifestations. 1 It is associated with increased rates of cardiovascular disease, mortality, mental health burden and has a significant impact on quality of life. [2][3][4][5][6] A treat-to-target (T2T) strategy, aiming for early remission or low disease activity, is recommended to reduce symptoms of pain and stiffness, improve joint function and reduce the risk of reversible joint damage. 7,8 Various composite measures to assess PsA disease activity have been developed and validated, although there is no agreement on the optimal measure. [9][10][11] Such measures include minimal disease activity (MDA), very low disease activity (VLDA), disease activity index for psoriatic arthritis (DAPSA), Psoriatic Arthritis Disease Activity Score and Composite Psoriatic Disease Activity Index. [12][13][14][15] Composite measures require multiple clinical and laboratory measures to be recorded, rendering them impractical for busy clinical settings and utilizing time and resources that could be better spent undertaking patient-centered care activities, such as education and shared discussion of treatment options. 16,17 Patient-reported outcome measures (PROMs) offer a potential efficient, patient-led alternative to composite activity measures. If reliable and accurate, they could increase time available in consultation for discussion regarding care and education, thereby improving quality of care. 18 The routine assessment of patient index data 3 (RAPID3) is a PROM initially developed in patients in rheumatoid arthritis. It efficiently provides results comparable to validated disease activity scores in clinical trials and clinical practice. [19][20][21] RAPID3 has been demonstrated to be useful in patients with PsA but has not been extensively studied and there is limited understanding of the expected RAPID3 values to describe low disease activity or disease remission. 22,23 This study aims to assess the ability of RAPID3 to discriminate disease activity in PsA in the clinical practice setting, as compared to assessment by MDA, VLDA and DAPSA, and to determine the respective optimal RAPID3 cut points that correspond to categories of disease activity, as per the composite measures.  24 Scores were recorded on standardized proformas by appropriately trained physicians. For skin assessments, assessors used Psoriasis Area and Severity Index (PASI) and body surface area (BSA) according to standard protocols. 25,26 Achieving MDA requires meeting 5/7 criteria: TJC68 ≤1, SJC66 ≤1,  Area under the curve (AUC) of each ROC curve was calculated and interpreted using the accepted analysis of diagnostic accuracy, with AUC of 0.5 providing no discrimination, 0.8-0.9 demonstrating excellent accuracy and 1 representing perfect accuracy. 34 The optimal cut point was selected by choosing the data point closest to (0,1), representing the optimal trade-off between sensitivity and specificity.

| RE SULTS
One hundred and three patients were enrolled, and data were collected from July 2016 to March 2020. Nine patients were excluded due to missing data and 1 patient later withdrew consent to complete required PROMs. Simultaneous RAPID3 and MDA/VLDA data were available for 336 clinic visits from 93 patients (44 and 49 from Concord and Liverpool Hospitals, respectively) and simultaneous RAPID3 and DAPSA data were available for 85 patients over 290 clinic visits. The median (interquartile range) number of visits per patient was 3 (1)(2)(3)(4)(5). At baseline, mean (±SD) age was 49.9 (±13.5) years with 47 (50.5%) male patients ( Table 1). Twenty-three (24.7%) patients had erosive disease.
The mean duration of PsA was 10.4 years. Thirty-three (35.5%) patients were taking methotrexate at baseline and 41 (44.1%) were taking a biologic or targeted synthetic disease-modifying agent ( Table 2).  Table 3. The optimal RAPID3 cut points were  Table 3. The generated ROC curves are displayed in Figures 1 and 2.
The relationships between patients in a state of MDA and DAPSA-LDA and the optimal RAPID3 cut points of 6 and 10 respectively are demonstrated in Figure 3. The majority of patient visits in which patients were in MDA were associated with a RAPID3 score ≤6 ( Figure 3A). In 67 patient visits, scores were concordant scores (ie patients both in MDA and achieving RAPID3 score ≤6). In 14 patient visits, patients were in MDA but had a RAPID3 score exceeding 6 and in 31 patient visits, patients had a RAPID3 score ≤6 but patients did not meet MDA criteria. Not depicted in the Venn diagram are the 224 patient visits where patients did not meet MDA and had RAPID3 >6, that is concordant scores for patients with greater than minimal disease activity. Similar results are noted for the relationship between patients with a RAPID3 score ≤6 and in DAPSA-LDA ( Figure 3B).
The distribution of RAPID3 scores for patients in MDA and for patients not in MDA is represented in the supplementary data and demonstrates the ability of RAPID3 scores ≤6 to identify patients who are likely in a state of minimal disease activity ( Figure S1).
Patients who are not in a state of MDA are more likely to have a RAPID3 score >6 ( Figure S2). Figure S3 shows the range of RAPID3 and DAPSA scores across all patient visits and this scatter plot illustrates the linear relationship between RAPID3 scores and DAPSA scores for each individual visit.

| DISCUSS ION
In this study, self-reported RAPID3 scores from patients attend- The high degree of concordance of RAPID3 scores using the above cut points with MDA and DAPSA-LDA is visualized in the Venn diagrams in Figure 3 and scatter plots in Figures S1-S3.
This study provides a novel understanding of how RAPID3 could be used to estimate disease activity in PsA by defining the cut points for assessments of disease activity based on 2 composite scores.
RAPID3 cut points used in RA are ≤3 for remission, 3.1-6.0 for low disease activity, 6. Abbreviations: AUC, area under the curve; RAPID3, routine assessment of patient index data 3; MDA, minimal disease activity; VLDA, very low disease activity; DAPSA-LDA, disease activity in psoriatic arthritis -remission/low disease activity; DAPSA-REM, disease activity in psoriatic arthritis -remission.

TA B L E 3
Receiver operating characteristic curve output F I G U R E 1 Receiver operator characteristic (ROC) curve for minimal disease activity (MDA)/very low disease activity (VLDA) and routine assessment of patient index (RAPID3). The ROC curves show the ability of RAPID3 to identify patients who meet MDA activity criteria ( Figure 1A) and those who meet VLDA criteria ( Figure 1B) 36,37 Second, the application of PROMs in settings independent of traditional appointments presents a promising opportunity to empower patients. In a recent pilot study employing online self-monitoring in patients with inflammatory arthritis, patients reported increased knowledge and awareness of their disease, with some noting earlier self-recognition of disease flare. 38 Self-monitoring presents a potential counter to key challenges identified by patients with PsA and psoriasis, notably fear of deterioration, lack of control and disempowerment by lack of personalized care. 39 The use of PROMs also presents a feasible and efficient alternate in assessment of disease activity to facilitate the T2T model. Despite the strong evidence for its use, T2T has low uptake in the clinical setting due to various barriers including time limitations and need for training. 17 The application of PROMs electronically, as discussed above, presents an opportunity for remote disease assessment to be incorporated in the T2T model. 40 The use of remote PROMs in assessment of disease activity has never been more compelling, given the Extending the application of RAPID3 to also include the symptom checklist in the MDHAQ provides opportunity for clinicians to ex- Qualitative analysis of the patient experience with RAPID3 in the PsA setting would further inform its inclusion into regular clinical care.
In conclusion, RAPID3 is able to discriminate between levels of disease activity in PsA. This study supports the use of a RAPID3 score cut point of ≤3 for remission in PsA. RAPID3 offers a potential solution for assessment of disease activity in situations where clinical assessment is not readily possible, for example with telehealth consultations.

ACK N OWLED G EM ENTS
Open access publishing facilitated by The University of Sydney, as part of the Wiley -The University of Sydney agreement via the Council of Australian University Librarians.

CO N FLI C T O F I NTE R E S T
The authors do not have financial interests that could create a potential conflict of interest or the appearance of a conflict of interest with regard to the work.

F I G U R E 3
Venn diagram for the relationship of routine assessment of patient index (RAPID3) with minimal disease activity (MDA) and disease activity in psoriatic arthritis with low disease activity (DAPSA-LDA). The Venn diagram in Figure 3A demonstrates the relationship between the number of patient visits in which patients met MDA criteria (light gray; Figure 3A) and the number of patient visits in which patients reported a RAPID3 score ≤6 (dark gray; Figure 3A). In 67 patient visits, patients were in MDA and scored RAPID3 ≤6. In 14 patient visits, patients were in MDA but scored RAPID3 >6. Figure 3B demonstrated the relationship between patients who meet DAPSA-LDA criteria (light gray; Figure 3B) and the number of patients who report a RAPID3 score ≤10 (dark gray; Figure 3B). In 92 patient visits, patients were in DAPSA-LDA and scored RAPID3≤10. In 18 patient visits, patients were in DAPSA-LDA but scored RAPID3 >10 O RCI D