We appreciate the highly pertinent questions and comments by Drs Aro and Ronkainen 1 concerning our observations made on disease evolution in patients with gastrooesophageal reflux disease under routine care, over a follow-up period of 5 years.As to the number of endoscopists from 1253 participating centres who enrolled a total of 6215 patients with erosive reflux disease (ERD) and non-erosive reflux disease (NERD), we cannot provide the exact figure. However, the majority of participating centres were each invited to include four consecutive patients with GERD, and most of these were out-patient services led by one individual specialist performing the endoscopy.All participating endoscopists were educated in the adoption of the Los Angeles classification for the description of oesophageal erosions as well as in the accurate endoscopic description of any detectable columnar lined epithelium (Barrett's oesophagus). This was the best we could do, considering the pragmatic nature of the study that intended to follow-up patients with ERD and NERD under routine medical care. Inter-observer validation was impossible to do, but the standard of education provided and training in endoscopy is high in our country.Concerning the healing rate of 74% of patients with Barrett's oesophagus (BO) on regular PPI medication, as pointed out by Dr Aro and Dr Ronkainen, intended healing of erosions, if they were present, or complete symptom relief in patients who had no erosions. We have indicated with an asterisk in table 1 of the manuscript 2 that healing referred to erosions in the group with ERD and complete symptom relief in the group with NERD in the initial healing phase on PPI, 3 and this does also apply to the subset of 240 patients with additional BO at baseline, but does not imply regression of BO. Thank you, for allowing us to make this clarification.Patients were also allowed to be included if they were found to have BO (either suspected endoscopically or confirmed by histology) at the index endoscopy, and they were assigned to one of the two categories ERD or NERD. Patients with BO at baseline have been excluded from the analysis in this follow-up study 2 as their evolution will be addressed in a separate report. Concerning progression of BO during follow-up, in spite of the regular use of PPI, may result from inadequate dosing of PPI in the specific subset of patients. Beyond this hypothesis a definitive explanation cannot be given due to the observational nature of our study.The definition of BO in the follow-up cohort was either by endoscopy alone, if the detection of columnar lined epithelium was not corroborated by histology, or was 'confirmed' if combined with positive histology. According to the proposal of the Montreal classification of gastro-oesophageal reflux disease, we considered the histology to be positive for BO if the columnar lined epithelium contained either specialized intestinal metaplasia (IM) or gastric metaplasia. 4 An important finding in this context was that in the newly developed BO segments ...