Abstract
The accuracy of diagnosis of cell type obtained from sputum cytology, bronchial aspirate, bronchial biopsy, or percutaneous lung biopsy in 161 cases of confirmed primary lung cancer has been examined and the pretreatment histological diagnosis has been compared with the final diagnosis made after surgical resection or necropsy. The yield of positive diagnoses of malignancy obtained by each method of investigation in each cell type showed that cytological examination of sputum was the most accurate method, but a high degree of accuracy was also obtained by bronchoscopic aspiration and bronchial biopsy. Percutaneous lung biopsy was the most effective, but the least accurate, means of obtaining carcinoma cells. The level of diagnostic accuracy was highest in patients with squamous cell carcinoma. Accurate pretreatment diagnosis of patients with adenocarcinoma was particularly difficult, and only 20% of these cases were correctly diagnosed by investigation. Of seven patients with adenocarcinoma and a positive diagnosis of malignancy made on percutaneous lung biopsy, none was correctly diagnosed. The causes of error in diagnosis of cell type of primary lung cancer are discussed.