Design: Systematic review and meta-analysis of randomised trials
Population: 5262 patients with rheumatoid arthritis in sustained control
Primary outcome: Disease flare
Effect: Tapering RR 1.56 (95% CI 1.26 to 1.98); withdrawal RR 2.23 (1.83 to 2.94)

A meta-analysis of randomised trials — 27 publications and 5262 patients — examined what happens when DMARDs are tapered or withdrawn in rheumatoid arthritis with sustained disease control.

Tapering increased flare risk by about half, driven mainly by biologic tapering and mainly in the first nine months; beyond nine months the excess was not significant. Withdrawal more than doubled flare risk, and the excess persisted at every follow-up point. Both were associated with radiographic progression beyond nine months — weakly for tapering, strongly for withdrawal. Neither reduced adverse events.

This supports current practice in which dose reduction is an option for patients in sustained remission who want it, but stopping is generally avoided. Many patients who flare regain control on restarting, though this analysis did not assess that. In India, cost often drives biologic interruption; where that is unavoidable, patients need a clear plan for early review and restarting.

Discuss tapering only with patients in sustained remission, and explain the higher flare risk in the first months.
Avoid stopping DMARDs completely in controlled rheumatoid arthritis where possible.
Taper gradually and review within about three months, when flares were most frequent.
Agree a flare plan with the patient before tapering, including how to restart quickly.
Do not taper expecting fewer side effects; adverse events were not reduced.

Why it matters

Patients often ask to stop treatment once well, and this gives clear numbers for that conversation.

The statistics, in plain English

A risk ratio of 2.23 for withdrawal means flares were more than twice as common as with continued treatment. The tapering risk ratio of 1.56 (95% CI 1.26 to 1.98) is smaller and concentrated early.