Vogelmeier 2011 [POET].

Methods	Design: randomized, double‐blind, double‐dummy, parallel‐group trial Duration: 1 year (+ 2 week run‐in) Location: 725 centres in 25 countries
Participants	Population: 7376 participants were randomised to tiotropium (3707) and salmeterol (3669) Baseline characteristics Age (mean years): salm 62.8, tio 62.9 % Male: salm 74.9, tio 74.4 % FEV1 predicted: salm 49.4, tio 49.2 Pack‐years (mean): salm 37.8, tio 38.8 Inclusion criteria: at least 40 years of age and had a smoking history of 10 pack‐years or more, a diagnosis of COPD, a forced expiratory volume in 1 second (FEV1) after bronchodilation of <70% of the predicted value, a ratio of FEV1 to forced vital capacity (FVC) of <70%, and a documented history of at least one exacerbation leading to treatment with systemic glucocorticoids or antibiotics or hospitalisation within the previous year Exclusion criteria: significant disease other than COPD; diagnosis of asthma; life‐threatening pulmonary obstruction, or a history of CF; active TB; narrow angle glaucoma; myocardial infarction or hospital admission for heart failure within the year prior to visit 1; cardiac arrhythmia requiring medical or surgical treatment; severe CV disorders; hypersensitivity to components of study drugs; respiratory infection or exacerbation in the 4 weeks prior to visit 1
Interventions	1. Salmeterol 50 bid (LABA) ‐ plus HandiHaler placebo 2. Tiotropium 18 qd (LAMA) ‐ plus pMDI placebo Inhaler device: HandiHaler and pressurised metered dose inhaler (pMDI) Allowed co‐medications: Patients were allowed to continue their usual medications for COPD, except for anticholinergic drugs and long‐acting β2‐agonists, during the double‐blind treatment phase
Outcomes	time to first exacerbation (primary); Secondary and safety end points included time‐to‐event end points, number‐of‐event end points, serious adverse events, and death
Notes	Funding: Boehringer Ingelheim and Pfizer Identifier(s): NCT00563381
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	A randomisation list was generated by the sponsor using a validated system involving a pseudo‐random number generator. Patients were randomized in a 1:1 ratio in blocks of four, with equal allocation of treatment within each block per country site.
Allocation concealment (selection bias)	Low risk	Patients were randomized to treatment via an Interactive Voice Response System (Perceptive Informatics Inc., Berlin, Germany)
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Blinding was maintained by allocation of a dummy placebo MDI to those randomized to the tiotropium arm and a dummy placebo HandiHaler to those in the salmeterol arm. Tiotropium and placebo capsules were identical in size and colour and were therefore indistinguishable.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	A committee assessing cause of death was blind to treatment group. Authors judged that other outcomes were blind also
Incomplete outcome data (attrition bias) All outcomes	Low risk	The efficacy and safety analyses included all the patients who underwent randomisation and who received at least one dose of the study medication. Fewer patients in the tiotropium group than in the salmeterol group withdrew from the study prematurely: 585 patients (15.8%) vs. 648 patients (17.7%) but both were judged to be low over a year and considering imputation of missing values
Selective reporting (reporting bias)	Low risk	Outcomes were well reported in the publications and on clinicaltrials.gov