FACTORS ASSOCIATED WITH EARLY LOSS TO FOLLOW-UP IN A MULTICENTER LONGITUDINAL RHEUMATOID ARTHRITIS COHORT

Abstract

Background Loss to follow-up can occur with many reasons, which influenced the outcomes of longitudinal cohort studies.

Objectives To compare the characteristics of patients who lost to follow-up and patients who completed the 2 years of follow-up, and examine the associated factors for early loss to follow-up (ELTF).

Methods The study subjects consisted of the KORONA cohort which is a multicenter longitudinal RA cohort. After excluding 1,119 patients who did not complete the 2 years of follow-up but still are actively participating, we divided the patients (n=4,257) into two groups; patients who lost to follow-up within 2 years (ELTF group), and patients who completed the 2 years of follow-up (FU group). Multivariate analysis was performed using the variables significant in univariate analyses and institutional factor. Institutions were divided in 3 groups based on the number of patients enrolled; A (largest) and C (smallest).

Results ELTF group patients (n=1,674, 39%) were older (p=0.04) and less educated (p<0.01). In a multivariate analysis, higher disease activity (OR 1.09, CI 1.02-1.16) and cardiovascular disease (CVD, OR 1.66, CI 1.17-2.34) at enrollment were risk factors for ELTF, whereas longer disease duration (OR 0.98, CI 0.97-0.99) and MTX use (OR 0.59, CI 0.49-0.70) were protective. Patients recruited from institutions of smaller numbers are likely to ELTF (C: OR. 7.65, CI 5.43-10.79, B: OR 3.27, CI 2.73-3.92, for A).

Conclusions RA Patient who have higher disease activity and CVD at enrollment are likely to ELTF, while longer disease duration and MTX use were protective. In addition, institutional factor should be considered as one of main factors for ELTF.