Abstract

In order to evaluate the role of cytopathologic diagnosis of sputum, bronchial washing and bronchial brushing in the diagnosis of lung cancer, we performed this study. The patients included in this study had undergone sputum, bronchial washing and brushing cytology over the 20-month period of 1985 through 1987. The total number of specimens was 5,495 of 2,242 patients, including 4,830 sputa and 665 bronchial washing and brushings. The average number of sputa and bronchial washings and brushings per case was 2.4 and 1.2 respectively. Among them, about 10% were unsatisfactory specimen, and three-fourths were negative specimens. In sputum cytology, the diagnosis of "atypical cells" was given to 3%, "suspicious for malignancy" was given to 1 %, and "malignancy" was given to 13%. In bronchial washing and brushing cytology, the diagnosis of "atypical cells", "suspicious for malignancy" and malignancy" was given to 6%, 3%, and 20% respectively. The cases diagnosed as "atypical cells" in cytology were actually malignancy in 95% and 84.8% of sputum and bronchial washing and brushings respectively, and the "suspicious for malignancy" were actually malignancy in 100% in both methods. The detection rates of malignancy were 50.4% and 55.2% in sputum and bronchial washing and brushing respectively, and the specificity was 100% in both methods. The accuracy of cell typing was 92% in sputum and 89.7% in bronchial washing and brushing.

• Keywords: Lung carcinoma;Sputum cytology;Bronchial washing;Bronchial brushing;Diagnostic accuracy