Unfortunately, some details of the study are vague and potentially misleading. Although the techniques are described in detail, the analyses are not. It is unclear whether the analysis was calculated per patient or per node. A per-node calculation would artificially elevate all of the test performance measures, and would be inappropriate because we care for patients and not individual nodes. The article implies that it addresses the value of EBUS in staging the mediastinum, but the analysis lumps stage N1 and N2 nodes together. This is not appropriate, and is not consistent with how we have approached preoperative staging for lung cancer. If I have added up the numbers correctly, the results of EBUS for mediastinal staging are even better: sensitivity, 100%; specificity, 100%; and false-negative and false-positive result rates of 0. It is not clear why the authors would choose to report on stage N1 and N2,3 nodes combined.