Pleural effusions: is thoracoscopy a reliable investigation? A retrospective review.

In this paper, we consider the results of thoracoscopy in a busy thoracic unit where the referring physicians place their greatest emphasis upon simple standard investigation of pleural disease. Between 1985 and 1989 620 patients with a pleural effusion of unknown aetiology were referred to our thoracic medical unit. Initial investigations included aspiration of pleural fluid for cytology and culture, and blind pleural biopsy for histological examination. Recourse to thoracoscopy was only taken in the absence of a diagnosis or non-resolution of the patients symptoms and signs. Of these 620 patients only 48 (8%) remained without a diagnosis and were referred for thoracoscopy. Histological assessment of biopsies obtained at thoracoscopy revealed malignancy in 24 patients (50%) and benign conditions in 16 patients (33%). In eight patients (17%) no conclusive diagnosis was established; in this latter group, six patients continued with their symptoms and further invasive investigations revealed malignancy. In this setting where thoracoscopy was used as a last resort, the sensitivity for thoracoscopy was 83% and the specificity was 100% with a predictive value of a negative result being 25%. In conclusion, from our experience, the majority of pleural disease may be diagnosed using simple techniques but thoracoscopy can be very helpful in the more complex cases. Moreover, inconclusive histology following thoracoscopy is an indication for further investigation if the condition does not improve.