Fifteen RCTs were included (with at least 4,008 patients, range 20 to 1,049). Eight trials evaluated combination disease-modifying antirheumatic drugs; eight trials evaluated tumour necrosis factor inhibitors/methotrexate (one trial included both combination groups). Follow-up ranged between 12 and 24 months. The average Jadad score was 3.93 (range 2 to 5).
Pooled analysis showed that there were significantly more withdrawals due to toxicity in the combination arms compared with methotrexate alone (15 RCTs). For combination disease-modifying antirheumatic drugs, the odds ratio was 2.33 (95% CI 1.43 to 3.77) and for tumour necrosis factor inhibitors/methotrexate, the odds ratio was 1.44 (95% CI 1.05 to 1.98). The authors stated there was no statistical heterogeneity. The only statistically significant impact of combination treatments on adverse events were an increase in serious adverse events with tumour necrosis factor inhibitors/methotrexate (OR 1.66, 95% CI 1.26 to 2.20; four RCTs) and nausea with combination disease-modifying antirheumatic drugs (OR 2.06, 95% CI 1.30 to 3.27; three RCTs). There were discrepancies between results reported in the tables and text.
Sensitivity analysis showed that patients receiving combination disease-modifying antirheumatic drugs had significantly higher withdrawals due to toxicity at 24 months (OR 2.44, 95% CI 1.28 to 4.66; three RCTs), and had significantly more episodes of nausea during the first 12 months (OR 3.48, 95% CI 2.00 to 6.08).