Search

Eight patients over an 8-year period required operation for spontaneous hemorrhage as a complication of a pancreatic pseudocyst. Three patients presented with abdominal pain or jaundice and bled in hospital while undergoing work-up. Four patients were admitted with upper gastrointestinal bleeding and one with intraperitoneal bleeding. Five patients were managed by pancreatic resection (two of the head and three of the tail) and three were managed by arterial ligation and internal drainage. There was one death (mortality rate, 12.5%). The first four patients in the series had their operations delayed secondary to a perceived need for further work-up or an inability to make an exact diagnosis of the bleeding site. All rebled, necessitating an emergency operation. The last four patients underwent an expedited workup and operation. Successful treatment of bleeding pancreatic pseudocysts requires good surgical judgment, especially when nonoperative methods fail or aren't applicable. The risk of recurrent hemorrhage is high, suggesting the need for immediate intervention once the diagnosis is made. Resection provides definitive control, although selected patients with easily accessible vessels may be managed more conservatively with ligation and drainage.

To assess the relationships between core temperature (T) and other factors relating to operating room (OR) blood loss and mortality following abdominal injury, the records of 122 patients undergoing laparotomy for trauma at Detroit Receiving Hospital over a 1-year period (1989) were reviewed. Most injuries were penetrating (86%) and the mortality rate was 8.2%. Overall, 57 of 122 (47%) had hypothermia (T < or = 35 degrees C) upon arrival in the OR. There was a significant correlation between admission blood pressure and lowest intraoperative temperature (r = 0.60; p < 0.001). Multiple regression analysis revealed that the patient's lowest temperature (p < 0.001) and Trauma Score (TS); p < 0.0015), but not Abdominal Injury Severity Score (AISS) (p = 0.25) correlated with OR blood loss. The 28 patients with high TS (15 or 16) and AISS > or = 9 had significantly less blood loss when the OR temperature was maintained above 35 degrees C versus 33 degrees-35 degrees C (540 +/- 580 mL vs. 1820 +/- 1160 mL; p < 0.003). This suggests that hypothermia may exacerbate OR blood loss independent of degree of physiologic or anatomic injury. Thus hypothermia is common in patients undergoing a laparotomy for trauma. Trauma scores and the presence of shock preoperatively correlate with the development of intraoperative hypothermia. Hypothermic patients with similar injury severity have greater blood loss. Prevention and rapid correction of hypothermia during resuscitation and surgery appear to be extremely important in reducing blood loss in this patient population.

Objective: To evaluate the use of serum lactic acid values to predict flow-dependent increases in oxygen consumption (VO2) in response to increasing oxygen delivery (DO2) after blood transfusion in surgical sepsis., Design: Prospective study., Setting: Tertiary care, trauma center., Patients: Twenty-one patients, postsurgical or posttrauma, judged septic by defined criteria., Interventions: Serum lactic acid concentrations, DO2, and VO2 were measured before and after transfusion therapy., Measurements and Main Results: Overall, the DO2 increased from 532 +/- 146 to 634 +/- 225 (SD) mL/min.m2 (p less than .001), and the VO2 increased from 145 +/- 39 to 160 +/- 56 mL/min.m2 (p = .02). These changes occurred with an Hgb increase from 9.3 +/- 1.1 to 10.7 +/- 1.5 g/dL (p less than .001). The patients were grouped by their pretransfusion serum lactic acid values. In those patients with normal (less than 1.6 mmol/dL) serum lactic acid (n = 10), DO2 increased from 560 +/- 113 to 676 +/- 178 mL/min.m2 (p less than .02), and VO2 increased from 150 +/- 25 to 183 +/- 46 mL/min.m2 (p less than .02). However, in the increased serum lactic acid group (n = 17), VO2 was not significantly changed after transfusion (143 +/- 46 to 146 +/- 58 mL/min.m2) despite increased DO2 (515 +/- 163 to 609 +/- 251 mL/min.m2, p less than .01)., Conclusions: Blood transfusion can be used to augment DO2 and VO2 in septic surgical patients. Increased serum lactic acid values do not predict patients who will respond. The absence of lactic acidosis should not be used in this patient population to justify withholding blood transfusions to improve flow-dependent VO2. Patients who have increased lactate concentrations may have a peripheral oxygen utilization defect that prevents improvement in VO2 with increasing DO2.

The records of 57 patients presenting with flail chest injury from 1981 through 1987 were reviewed to determine factors affecting morbidity and mortality. Fifteen patients (26%) had 8+ rib fractures with a unilateral flail and seven (12%) had multiple rib fractures with a bilateral flail. Thirty-two (56%) had moderate-severe pulmonary contusions and 44 (77%) required chest tubes for hemo-pneumothorax. Ventilatory assistance was used in 36 (63%). The major factors determining the need for ventilatory assistance were: an ISS greater than or equal to 23, blood transfusions in the first 24 hours, moderate-severe associated injuries (fractures, head injuries or truncal organs requiring operation), and shock on admission (p less than 0.001). An adverse outcome occurred in 15 (28%); nine required ventilatory assistance greater than or equal to 14 days and six died of sepsis with pneumonia. The main factors associated with an adverse outcome were: an ISS greater than or equal to 31 (p less than 0.001), moderate-severe associated injuries (p less than 0.001), and blood transfusions (p less than 0.005). Although the primary determinants of an adverse outcome were the associated injuries and blood loss, a bilateral flail (p less than 0.01) and age greater than or equal to 50 years (p less than 0.02) were contributing factors.

Large soft-tissue defects following close-range shotgun blasts remain a major technical challenge to trauma surgeons. During the period 1980 through 1988, 43 patients who survived greater than 48 hours following this injury were managed in our center. The locations of their soft-tissue defects were: extremity, 22; abdomen/chest, 18; and head/neck, three. All patients underwent immediate surgical exploration and wide debridement of all devitalized tissue along with repair of associated injuries. Management included mandatory frequent dressing changes, debridement, irrigation in the operating room, and the perioperative administration of broad-spectrum antibiotics. Four patients whose abdominal wall defects could not be initially closed had temporary placement of rayon cloth to prevent evisceration. Overall, four patients underwent delayed primary closure, eight were covered with split-thickness skin grafts, nine had closure with myocutaneous flaps, and 19 closed by secondary intent. Two patients, who were transferred to us following initial management, developed wound sepsis due to inadequate debridement and both eventually required amputation as did one patient who developed early myonecrosis following lengthy arterial repair. Frequent operative dressing changes, adequate debridement, and irrigation minimize sepsis following close-range shotgun blasts and should be the treatment of choice for this devastating injury. Techniques of wound closure need to be individualized to the particular situation.

The records of 239 patients surviving more than 24 hours with full-thickness intraperitoneal colonic injuries over a 7-year period were reviewed. During the first 3 years, 29% (31/106) of the patients were managed by primary repair without colostomy. In the next 4 years, almost twice as many patients, 56% (75/133), with similar colonic trauma were treated without fecal diversion (p less than 0.05). Although there was no difference in the mean Trauma Score in the patients with primary repair in the two time periods, the Injury Severity Score (mean +/- sd) in the patients without colostomy in the later periods was significantly higher (17.8 +/- 2.1 vs. 20.2 +/- 5.1) (p less than 0.001). No patient suffered because of the increased incidence of primary repairs. These patients had five abdominal abscesses and only one leak, whereas the patients with colostomy had 15 intraperitoneal abscesses. Because of the safety when primary repair is performed, more liberal use of primary colonic repair following penetrating trauma is warranted.

In the past decade 93 patients with duodenal injury were treated at a trauma center. By chart review, the age, sex, mechanism of injury, time to initial exploration (and the reason for delay), laboratory results, associated injury, extent of duodenal injury, operative repair, use of drains and tube decompression, morbidity, and cause of death were tabulated in order to improve management of these injuries. Of 87 patients surviving until the time of operative repair 73% required no repair (four) or primary closure (59). The remainder had either resection with primary anastomosis (ten), diverticulization (12), or pancreaticoduodenectomy (two). All patients with penetrating trauma were immediately explored. Patients with blunt trauma were explored on the basis of the judgment of house staff and faculty. Overall mortality was 18%. Significant morbidity occurred in 49% of survivors. This urban experience was heavily weighted toward penetrating injury. In this group early death usually resulted from associated vascular injuries. Blunt duodenal injury was less frequently associated with immediate exsanguination. Mortality associated with blunt duodenal injury was usually the result of delayed diagnosis. In blunt duodenal trauma peritoneal lavage is not diagnostic and may often be misleading; in this series 50% of lavages were false negatives. Blunt duodenal trauma, particularly when combined with pancreatic injury or delayed repair, was a lethal combination. A high index of suspicion and aggressive diagnostic evaluation (CT contrast study/amylase) in the emergency department is required in equivocal cases to avoid morbidity and mortality.