Management of blunt and penetrating external esophageal trauma.

The records of 26 patients with external blunt or penetrating esophageal trauma were reviewed to determine clinical features and results of therapy. Twenty-one injuries (four blunt, 17 penetrating) were to the cervical esophagus, and five to the thoracic esophagus. Major physical signs included subcutaneous air, neck hematoma, and blood in the nasogastric tube. Helpful roentgenographic findings were cervical and/or mediastinal air, mediastinal widening, pleural effusion, and pneumothorax (15%). Nine of 12 (75%) contrast studies and five of six (83%) esophagoscopies were positive. Twenty-four patients had associated injuries, the most common of which was tracheal (14 patients) (64%). All patients were managed by prompt surgical exploration, primary closure, and drainage. There were three early deaths. Thirteen patients had postoperative complications, four of which were esophageal leaks. Two of the leaks caused mediastinitis, pleural sepsis, and led to death. They were not treated by early esophageal exclusion or excision. There were no significant strictures or esophageal sequelae in the other patients. It is concluded that early primary closure and drainage results in a relatively high incidence of survival. If a thoracic esophageal leak occurs, aggressive management of prompt esophageal exclusion or excision is necessary to control sepsis and improve survival.

Objective evaluation of the efficacy of various venous cannulas.

Six venous cannulas (USCI No. 32, USCI No. 40, USCI No. 44, Sarns No. 40, Sarns two-stage cavoatrial, and Ferguson Argyle No. 40) were tested for efficiency of venous flow during cardiopulmonary bypass, with and without aortic cross-clamping. Each cannula was tested six times in dog models (twice in each of three dogs) and the data were averaged. The tip of the cavoatrial Sarns catheter was positioned as recommended. Two No. 32 USCI caval cannulas were placed either with or without caval snaring. The other cannulas were placed in the right atrium. Arterial flow was constant at 80 ml/kg/min, and aortic pressure averaged 84 mm Hg. Central venous pressure and the right ventricle excluded from the pulmonary artery so that blood which was not drained via the venous cannula was measured. Each single atrial cannula decompressed the right atrium (right atrial pressure 0 to 1 mm Hg) better than two caval cannulas (right atrial pressure 0 to 2 mm Hg). Caval snaring did not empty the right atrium (right atrial pressure 2 to 5 mm Hg) as well as any single cannula. The cavoatrial Sarns catheter did not empty the right atrium as well (right atrial pressure 2 to 4 mm Hg), but drained blood from the inferior vena cava (central venous pressure 0 mm Hg) better than other atrial cannulas (central venous pressure 1 to 5 mm Hg). All of the atrial cannulas performed equally well. The right ventricular vent flow varied between 63 and 70 ml/min with each of the atrial cannulas but rose to 85 ml/min with the cavoatrial Sarns catheter and 190 and 74 ml/min during caval cannulation with and without snaring, respectively. Aortic cross-clamping eliminated coronary sinus flow and decreased right ventricular vent flow. Therefore, a single atrial cannula is more efficient in draining blood from the right side of the heart than are two caval or a cavoatrial cannula. This advantage is negated by aortic cross-clamping.

Diagnosis and management of major tracheobronchial injuries.

From 1968 to 1978, 14 patients were treated for major tracheal or bronchial injury. Five injuries resulted from blunt trauma and nine from penetrating injury. Of the 5 patients with injury due to blunt trauma, three had avulsions of the right main bronchus from the trachea. In 2 of them, the injury was associated with stellate tears of the distal trachea and bronchus. The simple avulsion was repaired by a primary anastomosis of the right main bronchus to the distal trachea. For the other 2 patients, treatment consisted of right pneumonectomy. The remaining 2 patients in this group had complete transection of the trachea and underwent primary repair. Of the 9 patients with a penetrating injury, 4 had lacerations of the cervical trachea which were treated with neck exploration and tracheostomy. Three patients with partial transections of the cervical or upper mediastinal trachea were treated by primary closure. The other 2 patients had gunshot wounds to the distal right lateral trachea, which were treated by right thoracotomy and primary closure. There were no deaths, and the subsequent course was generally good in all patients.