INVESTIGATION FOR MEDIASTINAL DISEASE IN PATIENTS WITH APPARENTLY OPERABLE LUNG-CANCER

Background. The optimal approach to the investigation of mediastinal d isease in patients with apparently operable non-small cell carcinoma o f the lung is controversial. Methods. We conducted a randomized, contr olled trial in thoracic surgery services at mainly academic tertiary a nd secondary care general hospitals. We recruited 685 patients with ap parently operable, suspected or proven, non-small cell carcinoma of th e lung who underwent either mediastinoscopy or computed tomography. De pending on the apparent presence or absence of mediastinal nodes of gr eater than 1 cm, patients undergoing computed tomography either underw ent mediastinoscopy or went directly to thoracotomy. The primary outco me was thoracotomy without cure, defined as resection with recurrence. Secondary outcomes included thoracotomies undertaken in patients with benign disease and costs of the two strategies. Results. The relative risk of thoracotomy without cure in patients in the computed tomograp hy group was 0.95 (95% confidence interval, 0.75 to 1.19). The relativ e risk of thoracotomy without cure or thoracotomy in patients with ben ign disease was 0.88 (95% confidence interval, 0.71 to 1.10). The medi astinoscopy strategy cost $708 more per patient (95% confidence interv al, -$723 to $2,140). Conclusions. The computed tomography strategy is likely to produce the same number of or fewer unnecessary thoracotomi es in comparison with doing mediastinoscopy on all patients, and is al so likely to be as or less expensive.