This was a multi-centre randomised controlled trial, carried out in six clinical centres. The duration of the follow-up was for 3 years after randomisation with reviews being carried out at 6 weeks, 3 months, 6 months, 9 months, and 1 year after randomisation and then every 6 months for a total of 3 years. Loss to follow-up was 1 patient in the CT group. Stratification of patients was conducted according to clinical centre and whether their lesions were central (no air between mediastinum and lesion) or non-central. Blocked randomisation was performed in blocks of 4 such that, of every 4 patients, 2 were randomised to CT and 2 to mediastinoscopy. All clinical data collected at each follow-up including clinical notes, reports of investigations, and, where available, autopsy reports were reviewed by an adjudication committee consisting of an internist, two thoracic surgeons and the study statistician.