Abstract

The role of fine needle aspiration biopsy has been assessed prospectively in the diagnosis of discrete lung shadows. A questionnaire was completed before each of 100 biopsies (in 97 patients) to determine the clinician's pretest diagnosis and the likelihood of malignancy. The latter estimates were combined with the previously established sensitivity (71%) and specificity (100%) of the procedure for diagnosing malignancy in the unit to allow calculation in each case of the change in certainty of malignancy as a result of the investigation. Among the 100 biopsies there were 73 true positive and 13 true negative results. There were no false positive results but there were 14 false negatives (cases where malignancy was later proved but where the biopsy did not show unequivocal evidence of malignancy). Among the 27 negative biopsy results the clinician had estimated the likelihood of malignancy as 80% or more in 13 cases. In 11 of these 13 patients the eventual diagnosis proved to be a malignant tumour; on the other hand, six of the 10 patients given a less than 50% chance of malignancy had a benign outcome. A positive biopsy result was therefore quantitively of greatest value when the prior estimate of malignancy was low. In the case of the false negative results the prior probability of malignancy was usually sufficiently high to merit further investigation. It is estimated that the procedure led to the avoidance of thoracotomy in up to 14 of 97 patients.