February 1984

Sensitivity, Specificity, and Predictive Values of Closed Pleural Biopsy

Author Affiliations

From the Pulmonary Disease Units, Departments of Medicine, Highland Hospital (Drs Poe and Kallay), St Mary's Hospital (Drs Israel and Greenblatt), and Strong Memorial Hospitals (Drs Utell and Hall), Rochester, NY, and the University of Rochester School of Medicine and Dentistry.

Arch Intern Med. 1984;144(2):325-328. doi:10.1001/archinte.1984.00350140139020

• To determine the clinical value of a nonspecific pleural biopsy specimen and fluid in malignant neoplasm and tuberculosis, we retrospectively reviewed records of all patients with pleural effusions undergoing the procedure at three community hospitals over six years. Two hundred eleven patients underwent biopsies. Adequate tissue was obtained in 207. The results were compared with the ultimate clinical and pathologic outcome by follow-up for 12 to 72 months. The initial procedure was diagnostic of malignant neoplasm in 54 patients and granulomatous disease in ten. A nonspecific or normal result was found in 143 (68%). Malignant neoplasms or tuberculosis was eventually established in 30 and excluded in 101 of the 143 patients. In 12 patients, no diagnosis was made. The procedure's sensitivities were 65% (malignant neoplasm) and 90% (tuberculosis). One false-positive result occurred in a patient with nontuberculous granulomatous pleuritis. The specificity and positive predictive value were 99% and 98%, respectively. The negative predictive value was 77%. Closed pleural biopsy with simultaneous fluid analysis is a valuable diagnostic procedure in community hospital patients, but a nonspecific result does not exclude malignant disease.

(Arch Intern Med 1984;144:325-328)