Study objective: To describe the diagnostic efficacy, morbidity, and patient outcome of thoracoscopy; to quantify the direct impact of thoracoscopy on clinical management; and to determine preoperative variables associated with finding malignancy at thoracoscopy to aid patient selection.
Design: Retrospective chart review of consecutive cases of thoracoscopy for pleural disease.
Setting: Single tertiary medical center.
Patients: One hundred eighty-two consecutive patients who underwent thoracoscopy for pleural disease over a 5-year period (from 1987 through 1992).
Measurements and results: Final diagnoses were 98 (54%) malignant, 58 (32%) benign, and 26 (14%) idiopathic. Thoracoscopy had a diagnostic sensitivity of 95% for malignancy and 100% for benign disease. Malignancy was shown by thoracoscopy in 27 of 41 (66%) patients who had a preoperative nondiagnostic closed pleural biopsy, and in 24 of 35 (69%) patients who had at least 2 preoperative negative pleural cytologic specimens. Chart review by preestablished criteria showed information obtained from thoracoscopy directly influenced treatment in 155 (85%) patients. Thirty-seven (20%) patients, however, had at least one perioperative complication (15% major, 8% minor). Ten (6%) patients died during the same hospitalization in which a thoracoscopy was performed, although none died within 48 h. There was one thoracoscopy-related death. Sixty-two (34%) patients died within 6 months of thoracoscopy (death by all causes). Forty-seven (48%) patients who had intrathoracic malignancy present at thoracoscopy died within 6 months. Patients found to have malignant pleural disease by thoracoscopy were more likely to have a preoperative history of a malignancy (p = 0.001). Age more than 50 years was associated with finding malignancy at thoracoscopy (p = 0.04). A combined lymphocytic and hemorrhagic effusion was associated with malignancy (p = 0.004). Preoperative pleural data showed that idiopathic effusions had a significantly lower median lactate dehydrogenase (LDH) value (192, which was normal) compared with malignant or benign effusions.
Conclusions: (1) Thoracoscopy increases yield for malignant and benign disease when thoracentesis and closed pleural biopsy are nondiagnostic. (2) Thoracoscopy directly affects clinical management in 85% of patients. (3) Significant complications can occur in patients receiving tertiary care. (4) For the evaluation of suspected malignant pleural disease, thoracoscopy has its greatest diagnostic yield in older patients who have a history of malignancy and who present with a lymphocytic, hemorrhagic, high LDH effusion.