Background Both early and long-standing rheumatoid arthritis (RA) patients have a low prevalence of work participation, 60% and 40% respectively. Not only is this an economic loss for society, it is also a burden for patients. Studies demonstrate that paid work is associated with better health-related quality of life in patients with RA. TNF blocking agents have a proven effectiveness in the treatment of RA but are expensive. For reasons of cost-effectiveness it is important to know whether the use of TNF blocking agents is associated with increased work participation and to know how it can be predicted.

Objectives The objectives of our study were:

1) To analyze the disease related determinants of work participation at the start of TNF blocking agents

2) To verify whether treatment of RA patients with TNF blocking agents leads to changes in work participation

3) To analyze the determinants for changes in work participation in patients who changed in working status after two years of treatment with TNF blocking agents.

Methods Data were used from 2002-2011 of a large observational cohort of RA patients on biologics (DREAM). Patients were included when starting a TNF blocking agent (t=0) and were <65 years of age at t=2 years. This age was selected because it is the legal retirement age in the Netherlands. Work activity was registered as: paid work, retirement pension, work disability pension or voluntary work. Work participation was defined as having paid work and/or doing voluntary work. Disease related determinants analyzed were: age, gender, disease duration, number of previous DMARDs, rheumatoid factor, presence of erosions, number of co morbidities, HAQ-DI and DAS28. Additional analyzed determinants at 2 years were HAQ-DI response and DAS28 response. To explore influencing determinants, first univariate analyses (excluded if p-value >0.2) were performed followed by multivariate analyses with logistic regression (excluded if p-value >0.1).

Results N=1259 patients started on TNF blocking agents and their work participation status was available. N=537 patients were also <65 years at t=2 years and therefore included. Determinants relating to the work participation at baseline were: having a better HAQ-DI (OR=0.36, p=0.000), not having co morbidities (OR=0.65, p=0.081), and being male (OR=0.62, p=0.042). After two years of TNF blocking therapy 12.1% (n=65) started to work and 13.2% (n=71) lost work. Influencing determinants for starting or stopping work are shown in Table 1.

Conclusions It did appear that in this RA cohort work participation on average did not change during treatment with TNF blocking agents: equal proportions of patients lost or gained work. Work participation increased most in younger, better functioning RA patients without co morbidities.

Disclosure of Interest S. Manders Grant/Research support from: the Dutch affiliation of Abbott Pharmaceuticals, W. Kievit: None Declared, A. Braakman: None Declared, L. Hendriks: None Declared, H. Brus: None Declared, J. Fransen: None Declared, M. van de Laar: None Declared, P. van Riel: None Declared