Lessons from the major studies in COPD: problems and pitfalls in translating research evidence into practice.

Translating the growing evidence base on COPD management into practice can be challenging and understanding the strengths and weakness of published studies is crucial. Studies should conform to the standards of CONSORT statement; they should be sufficiently powered, participants should be randomised, there should be assignment concealment, and the outcome measures and analyses should be decided in advance. The interpretation of the results may be affected by age and severity inclusion criteria for the study and the exclusion of patients with co-morbid illnesses. Whether previous medication is continued or stopped can affect the interpretation of the results. Secondary analyses in sub-groups should be viewed with caution unless pre-specified and accommodated in the trial design and power calculations. Real world observational studies may be confounded by non-randomisation of participants but can sometimes yield valuable insights. The way in which the results are presented can influence their interpretation and their magnitude with respect to minimal important differences as well as statistical significance is important. Research help formulate management algorithms but often the questions they address are too specific to allow evidence-based sequencing of therapies.