Métodos de biópsia pulmonar empregados no Hospital Universitário de Santa Maria: acurácia, sensibilidade e especificidade

Abstract

The use of biopsies to establish a diagnostic is frequently needed during the investigation of lung diseases. To this end, methods such as bronchoscopy, image-guided percutaneous biopsy or even surgical methods may be employed. Nevertheless, it is not always possible to establish a definitive diagnosis even after analysis of the material obtained by biopsies. Explanations for this might be due to the sampling area of the biopsy, the pathological issues, the particularities collecting material method, and the characteristics of the disease itself. Thereby learning the diagnostic accuracy of each biopsy method becomes an essential requirement for a better indication of the procedure. It is critical to understand the risk-benefit inherent in each indicated method. This research aimed to determine the diagnostic accuracy of two methods of obtaining lung biopsies employed in a University Hospital. To reach this objective, the pathological results of lung biopsies performed at the Pathology Department of the University Hospital of Santa Maria S (HUSM), obtained from the Education for Information System (EIS) were analyzed. The following variables were collected: age, sex, date of biopsy, the biopsy method, and their pathological diagnosis. All patients undergoing lung biopsy procedures at HUSM in the period between 2003 and 2013 were considered eligible. In total, 1547 pathological reports were identified as being from lung biopsies of that period. However, 93 reports were found not eligible. After applying the inclusion criteria, 122 pathological reports were the selected for statistical analysis. Of these, 21 were from biopsy procedures guided by computed tomography (CT), 45 from endobronchial biopsy and other 56 from surgical procedures, the latter, regarded as the gold standard for assessment of accuracy. The accuracy of biopsy performed by the endobronchial method and CT-guided were measured by area under the Receiver Operator Characteristic (ROC) curve. The endobronchial biopsy had an area under the curve of 0.645. The sensitivity and specificity calculated to this method were respectively 43% and 85%. The accuracy of CT-guided biopsy assessed by the area the area under the curve was 0.681. For this method, the calculated sensitivity was 61%, and the specificity was 75%. Although the evaluation of the accuracy was obtained solely by the reports of the pathological analysis, the area under the curve was over 0.6 in both methods. In conclusion, both biopsies methods had reasonable accuracy and are therefore important tools in the diagnostic evaluation of lung diseases.