[Results of needle-biopsy of the pleura (author's transl)].

The technique of needle-biopsy of the pleura started with the Vim-Silvermann needle (often inadequate) and has now spread generally with the greater use of the Harefield-Abrams needle. The overall percentagetrue positive results (tuberculosis, cancer) has, over time, slowly fallen, because of the fall in the number of tuberculous cases and an increase in biopsies with insufficient material (T 1 1964 : 45%, T 2 1979 : 26,5%). In our third study (T 3), we studied in 150 cases selected at random out of the 628 cases studied in T 2. We compared our percentage true positive results with those obtained in T1 (number of biopsies positive for tuberculosis (TB) or cancer (CA) compared with the number of patients suffering from tuberculosis or cancer]. These figures, for percentage true positive results, was 90% for TB 62% for cancer in T 1, and fell to 87% for TB and 53% for cancer in T 3. There were no false positive results. The diagnosis of tuberculosis can, in general, be made with a single biopsy. Diagnosis of cancer requires repeated biopsies. Association of cytology increased the results to 70% (T 1 and T 3). Looking for the tuberculous bacillus from the biopsy material was rewarded in 33 % (T 3). Histological diagnoses of non-specific conditions was possible in 30 % of biopsies, which gave true non-specific results. The technical reliability in T 1 (95% with 4 individuals who carried out the biopsies) fell to 85% in T 3 (57 individuals). This fall was studied and could be explained by: 1) insufficiently repeated biopsies; 2) too great a number of individuals carrying out the biopsies (T 3 : 51 inexperienced individuals out of 57), with numerous cases of insufficient material; 3) the ratio "useful fragments/total fragments", was far too low. This relationship between useful fragments and total fragments is statistically (p less than 0.05) correlated with the experience of the doctor carrying out the procedure. The optimal number of fragments per biopsy is between 2 and 3 : a number greater than this does not improve the results. The later degradation in the diagnostic value of the biopsy by the histologist should be examined : the biopsy should be carried out by an experienced individual, and the biopsy should be read by an experienced histologist. The histologist should be exigent in his requirements, from the doctor carrying out the biopsy, and he should examine all the material brought up in the biopsy.