1) Sampling should take into account knowledge about the target population, the aims and design of the study, and available resources |
2) Sample size estimation is one of the most important parts of study planning and should always be performed |
3) Sample size estimation should be based on the main aim and design of the study. It should also take into account nonresponse and loss to follow-up |
4) When estimating the sample size, one should also consider the possibility of subsample analyses, and adjust the sample size accordingly. For instance, will males and females or different phenotypes of COPD be analysed separately? If so, the sample size should be planned to enable such analyses |