Dataset from Combined Bronchodilators in Chronic Obstructive Pulmonary Disease and the Risk of Adverse Cardio-pulmonary Events: A Population-based Observational Study

Background: Recent observational studies have reported possible arrhythmogenic effects with long-acting beta-agonists (LABA), while the long-acting anticholinergic tiotropium has been associated with cardiovascular and cerebrovascular events. Finally, pneumonia was the object of a recent signal in trials of LABAs submitted for marketing approval. Aim: To assess the potential cardio-pulmonary risk arising from the concurrent use of two long-acting bronchodilators as well as from monotherapy use of each of the long-acting bronchodilators. Methods: A series of population-based cohort studies, using both cohort and nested case-control analyses will be conducted using data from the United Kingdom's Clinical Practice Research Datalink (CPRD). The base cohort will consist of new users of long-acting bronchodilators from Jan 2002 until Aug 2012, age >= 55 with chronic obstructive pulmonary disease (COPD) and at least two years of baseline medical history information. The high-dimensional propensity score technique will be used to match new users of each long-acting bronchodilator and new users of two bronchodilators with comparable subjects from the base cohort, with one-year follow-up for outcomes of acute myocardial infarction, stroke, heart failure, arrhythmia and community acquired pneumonia. Data will be analysed using time-dependent Cox proportional hazard regression models and conditional logistic regression models.