Study objectives: To evaluate factors affecting the diagnostic yield of flexible fiberoptic bronchoscopy in evaluating solitary pulmonary nodules (SPNs).
Design: Retrospective analysis of bronchoscopies performed over a 4-year period.
Setting: A tertiary teaching hospital.
Patients: One hundred seventy-seven patients with pulmonary nodules without endobronchial lesions who underwent bronchoscopy with brushing, washing, and transbronchial biopsy.
Results: There were 151 malignant and 26 benign lesions. The diagnostic accuracy of bronchoscopy in malignant and benign lesions were 64% (97 of 151) and 35% (9 of 26), respectively. The yield of bronchoscopy was directly related to lesion size (p < 0.001, chi(2)). When lesions were grouped according to distance from the hilum, yields of bronchoscopy in central, intermediate, and peripherally located lesions were 82, 61, and 53%, respectively (p = 0.05, chi(2)). When we stratified distance from the hilum by lesion size, the difference in yield was not significant. However, lesions </= 2 cm had a diagnostic yield of 14% (2 of 14) when located in the peripheral third vs 31% (5 of 16) when located in the inner two thirds of the lung. There was a trend toward higher combined diagnostic yield in right middle and lingular lobes when compared to all other segments (p = 0.09, chi(2)). Transbronchial biopsy, washing, and brushing were complementary in improving the yield of bronchoscopy.
Conclusions: Size is the strongest determinant of diagnostic yield in bronchoscopy when evaluating SPNs. The yield of bronchoscopy is particularly low in lesions </= 2 cm that are located in the outer third of the lung. Thus, alternative diagnostic approaches may be preferable in this situation.