P. De Negri IRCCS H “Case Sollievo della Sofferenza” S. Giovanni Rotondo (FG), Italy G. Ivani “Regina Margherita” Children’s Hospital Turin, Italy S. Eksborg Hospital Pharmacy Karolinska Hospital Stockholm, Sweden References 1. De Negri P, Ivani G, Visconti C, et al. The dose-response relationship for clonidine added to a postoperative continuous epidural infusion of ropivacaine in children. Anesth Analg 2001;93:71– 6. 2. Ghignone M, Noe C, Calvillo O, Quintin L. Anesthesia for ophthalmic surgery in the elderly: the effects of clonidine on intraocular pressure, perioperative hemodynamics, and anesthetic requirement. Anesthesiology 1988;68:707–16. 3. Ivani G, Bergendahl HT, Lampugnani E, et al. Plasma levels of clonidine following epidural bolus injection in children. Acta Anaesthesiol Scand 1998;42:306 –311. [Please note printing error in this publication; correct epidural clonidine dose is 2 g/kg -1 h -1 .] DOI: 10.1213/01.ANE.0000022681.46533.3B Beware of Gastric Tube! To the Editor: We would like to introduce a rare case of a patient experiencing trouble with a gastric tube. A 66-year-old man classified as ASA physical status one was scheduled for distal gastrectomy and Billroth’s operation. The surgery was performed under general anesthesia with epidural analgesia without any events. A post- operative chest radiograph showed that a gastric tube (Salem sump 18F) was located at the place of esophagogastric anasto- mosis, forming a curve (Fig. 1). The tip of the tube was approx- imately 50 cm from the nasal foramen. Considering that it was too deep, we pulled the tube out about 3 cm. Six hours after the operation in the PACU, the total blood loss exceeded 1500 mL from the drain tube, and hypotension of 70 – 80/40 –50 mmHg also continued. Despite transfusion of 4 units of blood, hemoglo- bin and hematocrit remained 9g/dL and 27%, respectively, and bleeding from the drain tube was expected to continue. When the second operation for hemostasis was performed, we found that the gastric tube was unintentionally sutured (stapled) at the esophagogastric anastomosis during the first operation. The fo- cus of the hemorrhage was from the transverse mesocolon artery, which was injured when we pulled out the gastric tube. The tip of the gastric tube section is shown in Figure 2. Checking out the tip of the gastric tube to determine whether it is sutured may be one of the most important duties for anesthesiolo- gists during gastrectomy. Beware of the gastric tube, otherwise you may have a bitter experience. Yoo J. Kondo, MD Toshiyuki Okutomi, MD Sumio Hoka, MD Department of Anesthesiology Kitasato University School of Medicine Sagamihara Kanagawa, Japan DOI: 10.1213/01.ANE.0000022685.46533.2E Diagnosis of Postpneumonectomy Bronchopleural Fistula Using Inhalation of Oxygen or Nitrous Oxide To the Editor: The postoperative development of a bronchopleural fistula (BPF) is a serious complication of pneumonectomy. Diagnosis can present problems when the bronchial disruption is limited. However, the early detection of this complication is critical for optimal manage- ment of patients. A few publications recommend nuclear medicine techniques such as inhalation of Xe-133 or scintigraphy with Tc-99m DTPA for the diagnosis of postpneumonectomy BPF (1). We report the use of a simple technique for the early detection of a postpneu- monectomy BPF by measuring O 2 and N 2 O concentrations in the pneumonectomy cavity at baseline and after the patient inhaled gas mixtures enriched with O 2 and/or N 2 O. The technique is performed in the immediate postoperative pe- riod when a chest drain is in place and involves sampling of gas from the postpneumonectomy cavity. The drain is connected to a chest drainage unit (Pleur-Evac). The sterile sampling tube of an anesthetic gas analyser (Capnomac Ultima, Datex, Helsinki, Fin- land) is connected to the chest tube through a Y connector or an intravenous catheter introduced under aseptic condition in the lu- men of the chest tube. When the patient is spontaneously breathing room air, the concentration of oxygen in the sampled gas is approx- imately 20%. The informed patient subsequently inhales a mixture of N 2 O (50%) and O 2 (50%) via a tight fitting face mask for a period of 4 to 5 min. In 10 patients where no BPF was present, no N 2 O was detected in the gas sampled from the chest tube during the 4-min inhalation. During this short inhalation period, we did not observe any nitrous oxide-induced excitement. In the presence of a BPF, the N 2 O concentration will rapidly increase, reaching a concentration between 30 and 50% after 60 to 90 s of inhalation of the O 2 /N 2 O mixture. The test has been successfully applied in the case of a 68 year-old man who underwent right pneumonectomy associated with the Figure 1. Postoperative chest radiograph showing gastric tube. Figure 2. Tip of the gastric tube. 1122 LETTERS TO THE EDITOR ANESTH ANALG 2002;95:1119 –28