We wish to clarify issues raised by your recent letter regarding the abovementioned paper published in May 2003 in Thorax (2003;58:399–404). Our intention was to notify the editors and to reference the paper by Donohue et al appearing in Chest. We believe this is evidenced by our declaring in the cover letter that the data from one of the two trials were under review. However, we wish to apologise that we did not follow this up and did not remember to notify the Editorial Office about the acceptance of the Donohue manuscript in our response to the reviewers. We wish to assure you that it was not our desire to hide this information and we believe in and support the integrity of the scientific review and publication policies of all peer review journals.

Two separate clinical trials were conducted, as mentioned in the paper. Only the trial by Donohue et al examined 12 hour spirometry, representing the dosing interval of salmeterol, and constituted the first publication. As the St George’s Respiratory Questionnaire and Mahler dyspnoea index were part of both trials, the information was included in the initial report. The absence of including the data that were collected could be perceived negatively. However, as mentioned in the introduction to the paper in Thorax, both protocols stated a priori that they would be combined for health outcome end points such as exacerbations and health resource utilisation, which were the focus of the Thorax publication. In the analysis section in the Thorax paper there is a detailed description of how the health resource utilisation data were analysed, and this is distinct from data presented in the paper by Donohue et al. The data in the results section are presented as combined data and therefore include the contribution from the study by Donohue et al. The health resource data were not analysed or reported for the individual trials and do not appear in the Chest manuscript. The pulmonary function data could only be presented as 3 hour data and were included for completeness as combined data in the Thorax manuscript. The focus of the discussion in the Thorax paper was on exacerbations and health outcome end points.

We sincerely regret our oversight and will ensure that this does not occur again. We thank you for giving us the opportunity to explain this regrettable omission and proclaim the goodness of our intentions.