Five RCTs (n=6,522) were included in the meta-analysis. Trial quality was considered satisfactory. All studies reported adequate sequence generation or allocation concealment, double-blinding and clearly reported loss to follow-up (where necessary). Methods of adverse event monitoring was reported in all trials.
The tiotropium mist inhaler was significantly associated with an increased risk of mortality (RR 1.52, 95% CI 1.06 to 2.16; five trials); this result was driven by results from three 12-month trials. The tiotropium inhaler was associated with significantly increased mortality risks at doses of 5µg (RR 1.46, 95% CI 1.01 to 2.10; five trials) and and 10µg (RR 2.15, 95% CI 1.03 to 4.51; four trials). There was no evidence of statistical heterogeneity in any of the analyses.
Sensitivity analyses did not materially alter the main findings. The annualized NNT was 124 (one excess death likely from every 124 patients treated with the 5µg tiotropium inhaler over one year).
Risk of cardiovascular death was significantly increased following use of the inhaler (RR 2.05, 95% CI 1.06 to 3.99, I2 = 0%; four trials).