Background and aims: Biologics (disease modifying antirheumatic drugs, bDMARD) have been in use in Germany for the treatment of rheumatoid arthritis (RA) since 2001, usually after failure of at least one conventional synthetic (cs)DMARD. We analyzed temporal changes in factors that influence the decision for either a first bDMARD or a further csDMARD.
Material and methods: We analyzed data from 9513 bDMARD-naive RA patients in the German biologics register RABBIT who switched to a new therapy. For three recruitment periods (2001-2003, 2004-2006 and 2009-2015) factors influencing the therapeutic decision were analyzed by means of machine learning methods and logistic regression analysis.
Results: In all recruitment periods the number of previous csDMARDs, high dosages of glucocorticoids (>7.5 mg/day) and a higher DAS28 (>5.1) were significantly associated with the decision for a first bDMARD. Over time, the chance of receiving a bDMARD increased in patients with moderate disease activity, moderate glucocorticoid dosages (5-7.5 mg/day) and those with comorbidities, such as congestive heart failure or prior malignancy. Men had a higher chance of receiving a bDMARD than women only in the first recruitment period. Private health insurance, high education and gainful employment were significantly associated with more frequent prescription of bDMARDs in all recruitment periods.
Discussion: The time-dependent changes in the impact of disease activity, concomitant drugs, gender and comorbidity on the prescription of bDMARDs mirror the increasing therapeutic options and the growing experience in the application of the new substances in patients at higher risk. The influence of demographic and social factors may reflect safety concerns in patients at increased risk of adverse events but also the need to economize drug costs..
Keywords: Cohort Studies; Germany; Model-based boosting; Prescriptions; RABBIT.