Our study has several potential limitations, the first being the use of symptom diaries by control patients. This may increase awareness of symptoms and could therefore influence behaviour towards changing symptoms (e.g. seeking earlier contactwith health care providers).We are aware that this may lead to an underestimation of the effect size, although in the
COPE-II study a similar approach did not confound results in this regard. Secondly we are aware of the risk of contamination bias. Case managers and physicians will therefore repeatedly be instructed not to give control patients any information regarding the content of the study action plans throughout the
trial. A final point of consideration is the frequent use of COPD actions plans in the source population. This might decrease the effect size because control patients will also be guided to act in case of respiratory symptoms change. However, as said previously, we strongly feel that the use of solely symptom-based COPD action plans can lead to initiation of incorrect actions and/or delay of proper treatment in COPD patients with comorbidities and therefore hypothesise benefits in the intervention
group; we also wish to ensure that harm does not come to either group.