Longitudinal study of clinical prognostic factors in patients with early rheumatoid arthritis: the PREDICT study.

Aim: To assess the association between baseline clinical prognostic factors and subsequent Disease Activity Score of 28 joints (DAS28) remission in early rheumatoid arthritis (RA).

Methods: Data were collected using point of care clinical software from participating rheumatology centres. Patients aged 18 years or over whose date of clinical onset of RA was within the previous 12-24 months, who had at least 6 months of follow-up data and a DAS28-ESR (erythrocyte sedimentation rate) score recorded between 12 and 24 months from first being seen for RA were included. Data collected included baseline demographics, mode of disease onset, pattern of joint involvement at onset, smoking status, DAS28, rheumatoid factor (RF), anti-citrullinated peptide antibodies (ACPA), time from symptom onset to presentation and disease activity at baseline. Univariate and multivariate logistic regression of DAS28-ESR remission between 12 and 24 months after first assessment were performed.

Results: Data from 1017 patients were analyzed: 70% female; mean age 60 years (SD: 14.7); 70% RF-positive, 58% ACPA-positive. The strongest age and sex adjusted baseline predictors of DAS28-ESR remission at 12-24 months were remission at baseline (odds ratio [OR]: 4.49, 95% CI: 2.17-9.29, P < 0.001), being male (OR: 2.42, 95% CI: 1.46-4.01, P < 0.001), abstaining from alcohol (P < 0.001) and being lower weight (OR: 0.98, 95% CI: 0.97-1.00, P = 0.015). There was no statistically significant association between joint onset patterns, mode of onset, RF, ACPA or smoking status.

Conclusion: In this observational study, patients with early RA at risk of not achieving remission include those with high disease activity at baseline, women, those who drink alcohol and those with higher body weight.