Eight RCTs (1,868 randomised participants) were included in the review. All trials used adequate sequence generation, addressed incomplete outcome data and were free of selective reporting. Six trials used blinding and reported allocation concealment.
Treatment with tiotropium plus formoterol was associated with significantly greater improvements from baseline in average FEV1 (WMD 105mL, 95% CI 69 to 142; three trials), average FVC (WMD 135mL, 95% CI 96 to 174; two trials) and trough FEV1 (WMD 53mL, 95% CI 30 to 76; four trials) compared with tiotropium alone. There was no significant difference for trough FVC. Mean change in TDI was significantly greater with tiotropium plus formoterol than with formoterol alone (WMD 1.50, 95% CI 1.01 to 1.99; two trials). Patients treated with the two drugs were significantly more likely to achieve a clinically significant change in TDI (OR 2.34, 95% CI 1.58 to 3.46; three trials). Heterogeneity was significant only for change in average FEV1. Adverse events and COPD exacerbations did not differ significantly between groups.
Funnel plots and Egger's test did not suggest publication bias.