Article Test

1.1 Introduction

In 2004 we initiated an inception cohort of patients with recent-onset rheumatoid arthritis (RA). From 2008 onward, compliance (C) was assessed at fixed intervals, using a questionnaire (CQ) that additionally investigated 15 predefined motivations for non-persistence with therapy (P), and a visual analogue scale (C-VAS, 0-100mm). Objectives were to examine the correlation between the CQ and the C-VAS and to investigate if the selection of patient-independent motivations for non-P predicted better self-reported C.

1.2 Materials and methods

Up to January 2016, the cohort comprised 180 patients with variable follow-up. Each pre-specified motivation for non-P was classified as patient-dependent or patient-independent by 50 patients randomly selected (≥70% agreement). Descriptive statistics and multiple regression analysis were used. Written informed consent was obtained.

1.3 Results

Length of follow-up from 160 patients for which data were completed and analyzed was 6.7±3.4 years; all the patients scored 1516 pairs of CQ and C-VAS. C-VAS significantly correlated with the CQ, r=0.468, p=0.001. Optimal C-VAS cut-off value to predict CQ-compliance was ≤7.5 mm.

During follow-up, there were 670 CQ scored as with non-P among whom, 654 had at least one motivation for non-P selected; of them, 549 CQ (70.2%) corresponded to non-P patients who selected only patient-independent motivations. The selection of exclusively independent motivations for non-P predicted C-VAS and CQ scores. Also, the selection of exclusively independent motivations for non-P predicted compliance either per VAS (OR: 15.6, 95%CI: 5.4-45.3, p≤0.001) or per CQ (OR: 2.25, 95%CI: 1.062-4.664, p=0.034).

1.4 Conclusions

Patient motivation for non-persistence with medication impacts self-reported C.