tially hypercoagulable state), and all were taking an- ticoagulants. The concerns for our population of men were not just for myocardial infarction but also for cerebrovascular complications. We accept that our data may not meet the standards for making conclu- sions on the basis of statistically significant differences; however, the small numbers of cardiovascular and cerebrovascular events observed is of clinical relevance and perhaps allow some reassurance to physicians that the careful cessation of anticoagulants in the periop- erative period for TURP appears to carry an acceptable risk. As urologic surgeons, it is appropriate to consult relevant specialist colleagues when managing signifi- cant co-morbidity before surgery. In our experience, our colleagues are pragmatic in their acceptance that antico- agulation should cease during the perioperative period rather than their involvement being what will prevent a cardiovascular or cerebrovascular event. We are in agree- ment with our colleagues that it is a case of assisting with optimization of medical comorbidity, defining the proto- cols for cessation and recommencement of anticoagu- lants, and being available to assist should complications arise. The differences between early and delayed hemor- rhage in both groups of patients was not statistically significant as outlined in the results and discussion of our manuscript, but we would accept that with much larger numbers, it would be possible that this differ- ence could reach statistical significance. Even if it did, would it be of clinical significance? In the context of treating patients with significant comorbidity, our opinion is that such a finding may not be of clinical significance. None of our patients were on 5--reductase inhibitors (5ARI). Until very recently, the use of this class of drug was limited in Australia. During the study period, the only 5ARI available in Australia was finasteride. In the absence of reimbursement and competitive pricing, the personal cost of the medication was prohibitive, in excess of $100 USD per month, and public hospital patients were those who could least afford such medications. Mariolyn D. Raj, M.B.B.S. Henry H. Woo, M.B.B.S., F.R.A.C.S.(Urol) Department of Urology Westmead Hospital Sydney, Australia Reference 1. Eagle KA, Rihal CS, Mickel MC, et al. Cardiac risk of noncardiac surgery: influence of coronary disease and type of surgery in 3368 operations. CASS investigators and university of Michigan heart care program. Coronary artery surgery study. Circulation. 1997;96:1882-1887. Re: Bjurlin et al.: Is Routine Postoperative Chest Radiography Needed After Percutaneous Nephrolithotomy? (Urology 2012;79:791-795) TO THE EDITOR: We read with interest the study on the role of routine chest radiography (CXR) after percutaneous nephro- lithotomy (PCNL) through supracostal access (n = 46) as well as subcostal access (n = 168). The authors con- clude that selective CXR is advisable when clinically indicated, such as for the suspicion of pleural or pulmo- nary injury intraoperatively, patient oxygen desaturation, respiratory difficulty, or decreased breath sounds on phys- ical examination in the postoperative period and routine CXR is not recommended. There are certain issues in the article with which we do not agree. First, the authors recommend that all punc- tures should be done in full expiratory phase of respira- tion. Although this may be an important step if the puncture is likely to be supracostal, for stones where infracostal puncture seems feasible this maneuver will likely lead to all punctures becoming supracostal. Second, the authors use intraoperative fluoroscopy of the lungs to detect evidence of hydropneumothorax. We do not agree with this step because this will unnecessarily expose the patient to radiation and, moreover, fluoroscopy is not very sensitive. In a recent multicentric study 1 involving 5803 pa- tients, hydrothorax developed in 1.8% (n = 104) and supracostal puncture was needed in 16.7% of the pa- tients. Supracostal puncture is a known risk factor for development of hydropneumothorax. However, not all patients with supracostal puncture develop this com- plication. The factors predicting the development of this complication are largely unknown. In our experi- ence, the chances of postoperative hydropneumothorax in supracostal puncture are higher if the operative pro- cedure is prolonged (1 hour) and if the nephroscopy sheath is misplaced during the procedure. In 2 small series, the incidence of hydropneumothorax with subcostal puncture was found to be 0.5% and 2.6%. 2,3 In our experience, the chances of this compli- cation with the subcostal approach is negligible. We agree with the authors that postoperative CXR is justified in patients with intraoperative respiratory compromise (increase in airway resistance and fall in oxygen saturation) or in patients who develop postop- erative clinical features such as dyspnea, tachypnea, chest pain, or decreased air entry. However, we feel that postoperative CXR should also be done if the surgery was prolonged or if the Amplatz sheath was displaced during surgery only while using the supra- costal approach. 228 UROLOGY 80 (1), 2012