Similarly, there appears to be a limitation in the evidence for the increasingly widespread incorporation of ultrasonography for the rapid bedside diagnosis of pneumothorax. In this issue of CHEST (see page 859), Ding and colleagues8 provide a comprehensive look at the available evidence (prospective cohort studies, n =14; retrospective studies, n =2; study type not reported, n =4) with, to my knowledge, the first published meta-analysis assessing the diagnostic value of ultrasonography compared with chest radiography for pneumothorax. The findings may pleasantly surprise some readers. Across multiple practice settings and physician types using ultrasonography, the pooled sensitivity for ultrasonography (0.88) was higher compared with that of chest radiography (0.52); the specificity was similar (0.99 and 1.00, respectively). Importantly, the pooled sensitivity and specificity for nonradiologists, including pulmonary, critical care, and emergency physicians, are impressive (0.89 and 0.99, respectively). Similarly, high diagnostic accuracy of ultrasonography by nonradiologists (respiratory physicians) for pleural effusions has been reported in a recent prospective evaluation that included 960 ultrasound scans.9 However, despite the apparent higher sensitivity for ultrasound detection of pneumothorax compared with chest radiography, there is no significant difference in the summary receiver operating characteristic between ultrasonography and chest radiography.8 At a minimum, these data likely can be interpreted as ultrasonography being as accurate as chest radiography for the detection of a pneumothorax, but the authors note an important caveat: Operator skill is central to success. The current data used for the meta-analysis do not show any difference in the ultrasonography diagnostic signs of a pneumothorax. However, based on their extensive ultrasonography experience, the authors assert that among the signs of pneumothorax on ultrasonography, only the absence of both the lung sliding sign and the comet tail sign safely lead to a diagnosis of pneumothorax, and they note that ultrasonography-based diagnosis of pneumothorax often is a “rule out” test. More study is needed to support the authors’ assertions.