Samuel D. Smith

Bacterial translocation in trauma patients.

Sepsis and multiple system organ failure (MSOF) are major causes of morbidity and mortality in trauma patients. Bacterial translocation induced by hypotension, endotoxemia, or burns is a reproducible phenomenon in the laboratory. The incidence of bacterial translocation to mesenteric lymph nodes (MLNs) in 29 critically ill patients was evaluated to determine its relationship to subsequent sepsis and MSOF. Bacterial translocation was documented in 3 of 4 patients who underwent laparotomy for gastrointestinal (GI) disease. No trauma patient (25 patients), even at second exploration 3-5 days after injury, had a positive MLN culture. Five patients died; 4 trauma patients, one with GI disease. Forty percent of the trauma patients had major complications, predominantly pulmonary infections with gram-negative bacteria. However, infectious complications and outcome were not related to MLN culture results. The classical progression of bacteria from the gut to the bloodstream via the MLNs may require time and gut mucosal injury. The data suggest that bacterial translocation to the MLNs is not a common occurrence in acutely injured trauma patients.

Fibrin sheath enhances central venous catheter infection

Objective To determine whether fibrin-coated central venous catheters have a higher infection rate, and spawn more septic emboli, than uncoated catheters after exposure to bacteremia. Design Animal study comparing catheter infection and blood cultures of fibrin-coated and uncoated catheters exposed to bacteremia. Setting Animal laboratory. Subjects Adult male Sprague-Dawley rats. Interventions A total of 210 rats had catheters placed with the proximal end buried subcutaneously. Rats were divided into three groups: tail vein bacterial injection on day 0 (no fibrin group) or on day 10 (fibrin group), or no injection/saline injection (control, n = 40). Bacterial injections were 1 × 108 colony forming units of either Staphylococcus epidermidis (n = 100) or Enterobacter cloacae (n = 60). Animals were killed 3 days after injection. Blood cultures were obtained via cardiac puncture, and catheters were removed via the chest. Half of the catheter was rolled onto agar and the other half was placed in trypticase soy broth. Plates and broth were incubated at 37°C for 48 hrs. The presence of >15 colonies on roll plates, or growth in broth, was accepted as a positive sign of infection. Microscopy was performed on day 20-10 catheters. Thirty animals without catheters had bacterial injections and underwent blood culture 3 days after injection. Measurements and Main Results Catheter infection with S. epidermidis occurred in 32% of roll plates and 80% of broth from the fibrin group vs. 4% and 20% from the no fibrin group (p < .01 for each). Catheter infection with E. cloacae occurred in 50% of roll plates and 80% of broth from the fibrin group vs. 0% and 12% from the no fibrin group (p < .01 for each). Positive blood cultures occurred in 47 of 68 animals from the fibrin group vs. 8 of 68 from the no fibrin group (p < .01). Microscopy showed a fibrin sheath on 20 of 20 catheters. Without catheters, 30 of 30 blood cultures were negative. Conclusion The fibrin sheath significantly enhanced catheter-related infection and persistent bacteremia.

Early Video-assisted Thoracic Surgery in the Management of Empyema

Objective. The appropriate timing, as well as the type of intervention, for the treatment of empyema in children is controversial. The advent of video-assisted thoracic surgery (VATS) has changed the way we treat these children. Therefore, we reviewed our experience with the early use of VATS in the treatment of empyema and formulated a treatment algorithm. Methods. We retrospectively reviewed medical records of all patients undergoing VATS for empyema at Arkansas Childrens Hospital from December 1994 to February 1997. All patients were treated by the pediatric surgical service and had the diagnosis of empyema confirmed at surgery. Results are reported as means, unless otherwise noted. Results. Twenty-five children with empyema were treated with VATS during the review period. Their age was 48.3 months, and the duration of symptoms was 7.4 days. All the patients had parapneumonic empyemas and had received preoperative antibiotics for 10.1 days. Preoperative imaging included chest radiography in 25 (100%), ultrasonography in 20 (80%), and computed tomography in 10 (40%). All patients with documented loculated parapneumonic fluid collections underwent VATS within a mean of 2 days of hospitalization. Chest tubes were removed in 3.2 days, resulting in a postoperative length of stay of 4.9 days. Total length of stay was 7.3 days. One patient required conversion to minithoracotomy and required a transfusion. There were no other complications or deaths. Follow-up was available for 22 (88%) children, and there was resolution of symptoms in all children with no recurrences. Conclusions. Earlier intervention with VATS in the treatment of empyema in children is safe and may reduce hospital charges by shortening hospital stay. A treatment algorithm based on early use of VATS is also described.

Percutaneous gastrojejunostomy versus Nissen fundoplication for enteral feeding of the neurologically impaired child with gastroesophageal reflux

To determine the optimal method of providing enteral feeding to neurologically impaired children with gastroesophageal reflux, Nissen fundoplication with simultaneous gastrostomy tube placement (NGT) was compared with anterograde percutaneous gastrojejunostomy (APGJ), a nonsurgical procedure performed under fluoroscopic guidance. The records of 112 neurologically impaired children with gastroesophageal reflux were retrospectively reviewed; 68 had undergone NGT and 44 APGJ. Follow-up data were available for 45 NGT patients (mean age, 6.4 years) and 34 APGJ patients (mean age, 7.9 years). Mean follow-up was 1.8 years in the NGT group and 2.5 years in the APGJ group. Complications resulting from either procedure were classified either as major, which included treatment failures or morbidity resulting in prolonged hospitalization, or as minor, those requiring outpatient treatment only or not directly caused by the procedure. The NGT group had a significantly higher incidence of major complications in comparison with the APGJ group (33.3% vs 11.8%, p < 0.05). Ten patients (22.2%) in the NGT group required reoperation for complications; six required a second NGT for wrap hernia, failure, and continued gastroesophageal reflux. Two patients (5.9%) in the APGJ group required surgery for complications; one of these eventually required an NGT, and the other had an intussusception that necessitated a small-bowel resection. Minor complications were more common in the APGJ group than in the NGT group (44.1% vs 6.6%); the majority of complications were related to the jejunostomy tube. Premature replacement or reinsertion of the jejunostomy tube was necessary in 14 APGJ patients (32%). The mortality rate was 8.8% in the NGT group and 5.9% in the APGJ group (p = not significant). No death occurred within 30 days of either procedure. We conclude that APGJ is a safe alternative method for feeding the neurologically impaired child with gastroesophageal reflux.

Adaptation in short-bowel syndrome: Reassessing old limits☆☆☆

The improving survival of patients with severe short-bowel syndrome along with the advent of successful intestinal transplantation have accentuated the need to answer two questions. (1) Is there an intestinal length below which adaptation to full enteral nutrition can not be expected to occur? (2) How much time is necessary to complete intestinal adaptation? We reviewed the outcome of 21 infants with less than 50 cm of small intestine to answer these questions. Patients were divided into three groups based on intestinal length, regardless of ileocecal valve status: group I, < 10 cm (n = 3); group II, 10 to 30 cm (n = 11); and group III, 30 to 50 cm (n = 7). Data were collected to assess survival, incidence of adaptation, time to adaptation, and causes of mortality. Infants in group I did not achieve intestinal adaptation to full enteral nutrition. One survived and 2 died, one from varicella pneumonia and the other after intestinal transplantation. Eight of the 11 (73%) patients in group II survived and 5 of 8 (63%) survivors achieved full intestinal adaptation after a mean interval of 320 days (range, 148 to 506 days) on parenteral nutrition. Six of the seven patients (86%) in group III survived and all survivors (100%) achieved complete enteral adaptation after an average of 376 days (range, 58 to 727 days). The overall survival was 71% (15/21), but survival in patients with > 10 cm was 78%.(ABSTRACT TRUNCATED AT 250 WORDS)

Selective decontamination in pediatric liver transplants : a randomized prospective study

Although it has been suggested that selective decontamination of the digestive tract (SDD) decreases postoperative aerobic Gram-negative and fungal infections in orthotopic liver transplantation (OLT), no controlled trials exist in pediatric patients. This prospective, randomized controlled study of 36 pediatric OLT patients examines the effect of short-term SDD on postoperative infection and digestive tract flora. Patients were randomized into two groups. The control group received perioperative parenteral antibiotics only. The SDD group received in addition polymyxin E, tobramycin, and amphotericin B enterally and by oropharyngeal swab postoperatively until oral intake was tolerated (6 +/- 4 days). Indications for operation, preoperative status, age, and intensive care unit and hospital length of stay were no different in SDD (n = 18) and control (n = 18) groups. A total of 14 Gram-negative infections (intraabdominal abscess 7, septicemia 5, pneumonia 1, urinary tract 1) developed in the 36 patients studied. Mortality was not significantly different in the two groups. However, there were significantly fewer patients with Gram-negative infections in the SDD group: 3/18 patients (11%) vs. 11/18 patients (50%) in the control group, P < 0.001. There was also significant reduction in aerobic Gram-negative flora in the stool and pharynx in patients receiving SDD. Gram-positive and anaerobic organisms were unaffected. We conclude that short-term postoperative SDD significantly reduces Gram-negative infections in pediatric OLT patients.

Evolution of Staged Versus Primary Closure of Gastroschisis

ObjectiveSince the introduction of a preformed silo to the authors’ practice in 1997, there has been a decrease in primary closure of gastroschisis. To clarify the impact of this change, the authors reviewed their results over the past 10 years. MethodsFrom patient records, the authors abstracted the closure method, mechanical ventilation days, time to full feeds, mechanical and infectious complications, and length of stay. The authors compared groups using the Student t test and the Mann-Whitney test, as appropriate. ResultsBetween 1993 and the present, 124 patients were identified. Between 1993 and 1997, 38 children presented with gastro-schisis. Thirty-two (84.2%) closures were primary and six (18.8%) were staged. After 1997, the authors treated 80 children with gastroschisis. There were 27 (33.8%) primary and 53 (66.2%) staged closures. Six patients with other lethal anomalies were excluded. Length of stay and ventilator days were higher for the staged closure group, but infection and mechanical complications were less common in the staged closure group. The time to full feeds did not differ. ConclusionsA lower incidence of infection and complications related to abdominal compartment syndrome has made staged closure of gastroschisis more common in the authors’ practice. While it has resulted in a longer hospital stay, staged closure decreases the risk of long-term bowel dysfunction and need for reoperation.

Delayed gastric emptying in neurologically impaired children with gastroesophageal reflux: The role of pyloroplasty

The presence of delayed gastric emptying in neurologically impaired children with gastroesophageal reflux has led to controversy regarding appropriate surgical management. The authors reviewed the charts of neurologically impaired children requiring fundoplication to answer two questions: (1) is pyloroplasty needed in addition to fundoplication for delayed gastric emptying? and (2) Does delayed gastric emptying influence the morbidity associated with fundoplication? To answer the first question, 40 neurologically impaired children with delayed gastric emptying undergoing fundoplication were divided into two groups: Nissen and pyloroplasty (n = 21) and Nissen only (n = 19). The Nissen and pyloroplasty group had significantly more postoperative complications (23.8% v 5.0%) and took longer to reach full feeding (14.6 v 3.9) days. There were no differences in the incidence of recurrent symptoms, readmissions, or reoperations. To answer the second question, 58 neurologically impaired children undergoing fundoplication were grouped based on gastric emptying scan results: normal gastric emptying (> 32% in 1 hour) (n = 29) and delayed gastric emptying (n = 29). There were no differences in postoperative feeding tolerance, postoperative complications, recurrent symptoms, readmissions, or reoperations between the two groups. Delayed gastric emptying in neurologically impaired children with gastroesophageal reflux did not increase postoperative morbidity after fundoplication, and the addition of a pyloroplasty to fundoplication provided no additional benefit. The authors conclude that the procedure of choice for neurologically impaired children with gastroesophageal reflux is a fundoplication without pyloroplasty, regardless of the degree of delay in gastric emptying.

Throwing out the "grade" book: management of isolated spleen and liver injury based on hemodynamic status

PURPOSE Current organizational guidelines for the management of isolated spleen and liver injuries are based on injury grade. We propose that management based on hemodynamic status is safe in children and results in decreased length of stay (LOS) and resource use compared to current grade-based guidelines. METHODS Patients with spleen or liver injuries for a 5-year period were identified using our institutional trauma registry. All patients were managed using a pathway based on hemodynamic status. Charts were reviewed for demographics, mechanism, hematrocrit values, transfusion requirement, imaging, injury grade, LOS, and outcome. Exclusion criteria included penetrating mechanism, associated injuries altering LOS or ambulation status, combined spleen/liver injury, initial operative management or death. Statistical comparison was performed using Students t test; P < .05 is significant. RESULTS One hundred one patients (50 spleen, 51 liver) meeting inclusion criteria were identified. Average actual LOS for all patients was 1.9 days vs 3.2 projected days based on American Pediatric Surgical Association guidelines (P < .0001). Actual vs projected LOS for grades III to V was 2.5 vs 4.3 days (P < .0001). All patients returned to full activity without complication. CONCLUSIONS Isolated blunt spleen and liver injuries, regardless of grade, can be safely managed using a pathway based on hemodynamic status, resulting in decreased LOS and resource use compared to current guidelines.

Prospective results of a standardized algorithm based on hemodynamic status for managing pediatric solid organ injury

BACKGROUND Controversy surrounds the need for ICU admission, prolonged bed rest, and the duration of activity restrictions for children sustaining blunt trauma. Adult literature supports management based on hemodynamic status, not CT grade. STUDY DESIGN A 3-year prospective study of a standardized management algorithm for hemodynamically normal pediatric patients with blunt liver or spleen injury was performed. Patient selection was based on vital signs, irrespective of injury grade on CT. Patients requiring ICU admission for nonliver or nonspleen injury were excluded. Patients were admitted to a surgical ward with serial hematocrit levels. Discharge occurred 48 hours postinjury if patients had no abdominal tenderness, tolerated a regular diet, and had a stable hematocrit. Patients were allowed noncontact activity, including school, after discharge. Patients were followed up at 1 month with ultrasonographic imaging. RESULTS Eighty-nine patients sustained blunt liver or spleen injury. Forty-five patients were excluded for other injuries (Glasgow Coma Scale < 13, 32 of 45); the remaining 44 patients had a mean age of 8.9 years (range 2 to 17 years), Injury Severity Score 10.6 (range 4 to 33), liver grade 2.1, and splenic injury grade 2.3. Mechanisms of injury were predominately motor vehicle collisions (59%). All patients were managed nonoperatively without transfusion; 43 of 44 patients completed the algorithm. Mean observation was 55.2 +/- 12.3 hours. One-month followup occurred in 33 of 44 patients, with one complication detected and no delayed bleeding. CONCLUSION Management of pediatric solid organ injury should be guided by hemodynamic status and not injury grade on CT. Hemodynamically normal children can be safely managed without intensive care monitoring, do not need prolonged hospitalization, and can resume school on discharge.