Transbronchial needle aspiration (TBNA) is a valuable technique for sampling mediastinal and pulmonary lesions. The diagnostic yield of TBNA varies widely in reported series, ranging from 20 to 90%. Rapid on-site cytologic evaluation (ROSE) is commonly used during thyroid, breast, and transtho-racic needle aspirations, and the use of ROSE improves the diagnostic yield of endoscopic ultrasound-guided fine-needle aspiration. Several authors have suggested that ROSE may improve the yield of TBNA. However, its utility remains unproven, and its potential advantages may not outweigh the cost and inconvenience.
During bronchoscopy, there are often multiple biopsy targets and multiple sampling modalities available. Several studies have shown that obtaining multiple samples increases the diagnostic yield of bronchoscopy. However, performing multiple biopsies adds to the cost, length, and risk of bronchoscopy.
ROSE is used frequently in our hospital, though the cytotechnologists are occasionally not available, If ROSE is used and the results are positive, we often defer further biopsy. ROSE is not being utilized to improve yield but rather to minimize the risk and duration of a procedure. Determining whether deferring further biopsy based on ROSE is safe and cost-effective has not been previously studied. Bronchoscopies utilizing TBNA were studied prospectively to assess the impact of ROSE.
Patients scheduled to undergo TBNA at Stony Brook University Hospital were approached for study consent. The utilization of ROSE was decided by the pulmonary attending independent of the patient’s decision to participate in this study and was generally determined by scheduling concerns.
Prior to the procedure, the attending physician delineated a biopsy algorithm that listed which biopsies would be performed if ROSE were not available, which biopsies would be performed if ROSE were available and diagnostic, and which biopsies would be performed if ROSE were available and nondiagnostic (Fig 1).
TBNA were performed using the Wang nodal mapping system based on preprocedure CT scans. Endobronchial ultrasound was available and was utilized for lesions in atypical locations. Histologic needles (model 319; Mill-Rose Laboratories, Inc; Mentor, OH) were preferentially used during most procedures. Cytologic needles (model 222; Mill-Rose Laboratories) were used for some peripheral lesions. Canadian Neighbor Pharmacy website is glad to present you a new informative project.
Immediately following the bronchoscopy, data were collected regarding the duration of the procedure, whether ROSE was utilized, the number and results of each pass, and whether the procedure followed the designated algorithm. Also tracked were the utilization of fluoroscopy, the need for a postprocedure radiograph, and the number of pathologic and microbiological samples sent to the laboratory. In our hospital, transbronchial biopsy and TBNA of lesions past the segmental bronchus are performed under fluoroscopy. Chest radiographs are routinely ordered after bronchoscopies requiring fluoroscopy.
A TBNA was classified as diagnostic if it yielded a final pathologic diagnosis. “Suspicious,” “atypical,” or “negative” results were classified as nondiagnostic. If more than one TBNA was performed, each site was evaluated separately. A diagnostic TBNA specimen was considered to be a true positive (TP) if it yielded a cytologic or histologic diagnosis of malignancy. A false-positive result was defined as a biopsy result that was diagnostic for malignancy but was later proven to be in error. A diagnostic TBNA specimen was considered to yield a true-negative (TN) result if it yielded a pathologic diagnosis of benign disease. A nondiagnostic TBNA specimen was further evaluated in view of later pathologic specimens; a nondiagnostic TBNA specimen was reclassified as a TN result if benign disease was demonstrated at surgical resection, and as false-negative (FN) result if a later pathologic specimen showed a malignancy. A nondiagnostic TBNA specimen that was not definitively diagnosed or was resected after neoadjuvant chemotherapy was classified as unknown.
Each bronchoscopy procedure was then analyzed to determine the impact of ROSE. The accuracy of ROSE was determined by comparing the ROSE result to the final TBNA cytology and histology report. The number of biopsies and cultures deferred by a positive ROSE result was determined from the preprocedural algorithm. Bronchoscopies performed without ROSE were compared to the preprocedural algorithm to evaluate whether the algorithms predicted physician behavior.
Cost analysis was performed. Charges for each biopsy and culture deferred were calculated based on published 2003 Medicare rates including bronchoscopist reimbursement and technical and professional laboratory fees. If ROSE obviated the need for parenchymal biopsy, fluoroscopy and chest radiograph fees were also included. As Medicare reimburses the endoscopy suite based on time, not for the individual biopsy performed or equipment costs, endoscopy suite costs were not included in our analysis. Medicare reimbursement to the hospital for ROSE costs was determined.
A statistical comparison was performed utilizing the x2 test for categoric data and nonpaired t test for continuous data. The data are reported as the mean ± SD unless otherwise noted. This study was approved by the institutional review board, and all patients signed informed consent forms prior to enrollment.
Figure 1. Preprocedure algorithm.