The Fisher exact test was used to test whether lung point, a specific feature of pneumothorax at PU, was more prevalent in pneumothoraces detected only by PU than in pneumothoraces detected by both CXR and PU. To maximize power, the analysis was made PU-wise but yielded a nonsignificant difference. Therefore, analyzing on the basis of the patient would not have altered the conclusion. Our results show that in this population, more pneumothoraces were detected by PU than by CXR (McNemar test, P < .006), and this was especially true for evaluation at 6 h after clamping. Based on this significant difference, we believe that the study had sufficient power on which to base our conclusion.