FRI0554 DEVELOPMENT OF A DISEASE ACTIVITY INDEX FOR RA PATIENTS USING HANDSCAN.

Abstract

Background:Disease activity in RA patients is usually measured by DAS28,1a composite measure consisting of 28 swollen and/or tender joint counts (SJC28/TJC28), an acute phase reactant (APR, e.g. ESR/CRP) and patient’s general health typically using a visual analogue scale (VAS). Particularly assessment of joint counts is time consuming, requires a trained health professional and its inter-observer variety is high. The HandScan is developed to measure inflammation in hand joints using optical spectral transmission (OST, score 0 to 66) without taking time of a health professional.2The correlation between DAS28 and a single measurement of OST is moderate.3We hypothesised that a composite measure including OST (representing joint inflammation), VAS and APR would lead to an appropriate disease activity index.Objectives:To establish a method for assessing disease activity in RA patients using HandScan +/- other disease activity parameters, with cut-offs for remission and low disease activity (LDA).Methods:RA patients, visiting the outpatient clinic of Máxima Medical Center Eindhoven, were eligible for inclusion. Inclusion criteria were; (1) RA according to classification criteria, (2) at least one HandScan and DAS28 measurement performed at the same visit, and (3) aged ≥18. Data was extracted from medical records. A random sample of 2/3 of included patients was used as development cohort, the remaining 1/3 was used as validation cohort. In the development cohort, linear regression analyses were performed to create a formula for an OST index (DASost). In these, DAS28 was the outcome variable and, OST, ESR and VAS were predictors. Also other parameters were tested in the model to see if they increased the fit of the model or modified the association between OST and DAS28. A final model was derived, based on statistical significance of predictors and improvement of model fit (adjusted R-square). Agreement of DAS28 with DASost was tested with the random one-way intraclass correlation coefficient (ICC). DAS28 based remission and LDA were calculated for DASost using the established DAS28 cut-offs (i.e. DASost<2.6 and DASost<3.2). A cut-off for DASost for Boolean remission was defined using receiver operating characteristic (ROC) curves and Youden’s index. In the validation cohort, diagnostic values were calculated for DASost using the cut-offs as defined above.Results:Data of 3358 observations within 1505 unique RA patients were extracted. Patients’ demographic and clinical data are shown in Table 1. The formula for DASost derived in the development cohort was: -0.44 + (OST*0.03) + (male*-0.11) + (LN(ESR)*0.77) + (VAS*0.03). The optimal cut-off on DASost found for Boolean remission was 2.2. The ICC was 0.88 (95%CI 0.87 - 0.89). The explained variance of DASost in the validation cohort was 78%. Diagnostic accuracy of DASost in the validation cohort for DAS28 based remission, LDA and Boolean remission are shown in Table 2.Table 1.Patients’ demographics and clinical dataPatient demographicsNumber of patients1505 (100%)Females976 (65%)Age (year)65.5 (12.1)duration of RA (year)11.5 (8.3)Seropositivity1068 (71%)Clinical dataNumber of observations3358DAS282.5 (1.3)ESR (mm/hr)9.0 (5.0 – 21.0)SJC0 (0 – 2)TJC0 (0 – 2)VAS30.0 (10.0 – 50.0)OST12.6 (5.0)Data are n (%), mean (SD), or median (IQR).Table 2.Diagnostic values of DASostDisease activity stateAU ROCSensitivitySpecificityPPVNPVaccuracyRemission0.9389%83%88%84%86%LDA0.9291%67%89%73%85%Boolean remission0.8791%95%39%98%94%AU ROC= area under the receiver operating characteristic curve, PPV= positive predictive value, NPV= negative predictive value.Conclusion:The HandScan could be used as a tool to quickly assess disease activity in RA patients, if OST is combined with other disease activity parameters into an index.