| Sign In to gain access to subscriptions and/or personal tools. |
DOI: 10.1177/0148607107031005410
Feeding the Open Abdomen  
From the * Department of Biostatistics, Section of
Surgical Sciences, and Correspondence: Bryan Collier, DO, CNSP, FACS, Trauma Patient Care Center, Vanderbilt University Medical Center, 404 Medical Arts Building, 1211 21st Avenue, Nashville, TN 37212. Electronic mail may be sent to bryan.collier{at}vanderbilt.edu.
Background: The purpose of this study was to determine if early
enteral nutrition improves outcome for trauma patients with an open abdomen
(OA). Methods: Retrospective review was used to identify 78 patients
who required an OA for The use of damage-control celiotomy has increased the survival of trauma patients in extremis, whereas the early use of decompressive celiotomy has greatly reduced mortality associated with abdominal compartment syndrome (ACS).1–5 These management strategies, while saving numerous lives, have created an entirely new set of clinical challenges and have developed care of the open abdomen into a subspecialty within trauma and emergency surgery. Controversies regarding the care of these patients have centered on the optimal timing and method of abdominal wall closure, as well as the general approach to nutrition and route of delivery. Timing the closure of these patients is dictated by the patient's physiology (acidosis, hypothermia, coagulopathy), the mechanical effect of a distended bowel, and the clinical plans for unaddressed injuries or concerns regarding the integrity of surgical interventions.6,7 During this time period, ranging from hours to weeks, the trauma patient undergoes the ebb and flow of a systemic inflammatory response syndrome (SIRS) and is at high risk for developing complications stemming from protein-calorie malnutrition. This includes poor healing, infections, and an ongoing SIRS response.8
Trauma patients without an open abdomen have improved outcome with the
initiation of early enteral
feedings.9–15
In addition, early critical care nutrition studies have demonstrated an
increase of morbidity and mortality once an
After approval from the Vanderbilt University institutional review board, a retrospective review of all trauma admissions from January 1, 1995, to December 31, 2002, was performed to identify patients who underwent exploratory celiotomy and subsequently required an open abdomen either as part of a planned staged damage-control celiotomy (DCC) or after the development of an ACS. The Trauma Registry of the American College of Surgeons (TRACS) and hospital billing database were used to identify patients 15 years or older who required a celiotomy and subsequent open abdomen for  4 days and survived
the hospital stay (Figure
1). DCC was defined as a staged approach initiated to control exsanguination and contamination in a critically injured patient, and was followed by abdominal packing and temporary abdominal closure.4–6,19,20 The patient was then transferred to the trauma intensive care unit to address physiologic parameters, including acidemia, coagulopathy, and hypothermia. Once these parameters were normalized, the patient underwent reexploration for removal of the packs, definitive surgery was completed, and fascial closure was attempted. ACS was defined as elevated intra-abdominal pressures associated with renal dysfunction, cardiopulmonary compromise, or hemodynamic instability.21–25 Once ACS was diagnosed, abdominal decompression was performed and temporary closure provided. Data were obtained from TRACS, manual and computerized chart review, and the hospital billing database. Demographic data including age; sex; race; mechanism of injury; injury severity score (ISS); abbreviated injury scores (AIS) of the abdomen, pelvic, and thoracic regions; initial base deficit and lactate; initial blood product administration; and hospital and intensive care unit length of stay (LOS). Additionally, nutrition variables were evaluated, and included the presence of enteral nutrition, rate of enteral feedings, and percent of goal with enteral feedings. Infectious outcomes included ventilator-associated pneumonia (VAP), empyema, bloodstream infections (BSI), surgical site infections (SSI), wound infections, and urinary tract infections (UTI). Infectious complications were defined by the National Nosocomial Infections Surveillance System (NNIS). Noninfectious outcomes included fistula formation, timing of closure, and hospital charges. Timing of closure evaluated whether the patient's fascia was closed primarily within 8 days of initial celiotomy.26 Hospital charges included only those involving the initial hospitalization. All patients identified through inclusion criteria and having nutrition data available formed our study population. These 78 patients were then grouped according to whether enteral feedings were initiated within 4 hospital days (EEN) or after >4 hospital days (LEN) postceliotomy. EEN patients received enteral feedings at the time of fascial closure or while the abdomen was open. Outcomes related to EEN vs LEN included infectious complications, fistula formation, early (<8 days open) fascial closure, and primary hospitalization charges.
Continuous demographic variables were presented as mean ± SD.
Wilcoxon tests were used to compare continuous variables, and Pearson
During the 7-year period, 22,281 patients were admitted to the adult trauma service. There were a total of 78 patients who met study criteria (Figure 1). Mean time for the initiation of feedings for the EEN group was 3.0 ± 1 day postceliotomy. Mean time for the initiation of feedings for the LEN group was at least 7.9 ± 3.9 days postceliotomy. Fifteen patients were not fed enterally during any of the 14 postoperative days evaluated.
Table I shows when enteral
nutrition was provided over the 14 postoperative days for all 78 patients. On
postoperative day 1, 88% had no nutrition provided, enterally or parenterally.
Two patients (3%) had enteral nutrition provided. Patients were given
increasing amounts of tube feedings through the postoperative days. The
enteral feedings may have been supplemented with PN. Excluding the first
postoperative day as a small group representative (n = 2) and an outlier, the
percent goal obtained increased from
The mean age of the entire group was 36.5 ± 16.5 years. Sixty-eight percent were men, 71% were white, and 65% had blunt trauma. Patient demographics, ISS, AIS (abdominal, pelvic, and thoracic regions), initial base deficit and lactate levels, and length of hospital and intensive care unit stays are also represented (Table II). Both groups were similar regarding demographics and injury profile. Only thoracic AISs differed between the groups, with the EEN group demonstrating greater injury. The EEN group trended toward worse initial base deficit and higher lactate, as well as initial transfusion requirements (Table II).
Seventy-eight patients underwent at least 4 hospital days of open abdomen care. Fifty-one (65%) patients required DCC and 27 (35%) patients had ACS. Of these, 15 patients had primary ACS, with significant intra-abdominal injury as the cause. Twelve patients had secondary ACS or no associated intra-abdominal injury. Table III demonstrates rates of infectious complications. Both groups demonstrated similar rates of VAP, empyema, BSI, SSI, wound infections, wound cellulitis, and UTI. Table IV demonstrates rates of noninfectious outcomes. The EEN group had a higher rate of early primary fascia closure compared with the LEN group (74% vs 49%; p = .02). In addition, the EEN group had a lower fistula rate (9% vs 26%; p = .05) and lower total hospital charges ($172,283 ± $118,010 vs $223,349 ± $138,324; p = .04) by more than $50,000.
In the critically ill or severely injured patient, aggressive nutrition support and early administration of enteral nutrition is widely advocated. It has clearly been shown to improve outcomes related to infectious complications.9–11,14,27 The reasoning can be summarized simply: to prevent acute calorie malnutrition, modulate the immune response, and promote gastrointestinal structure and function.28 Both concepts of open abdomen care and early enteral feeding are well accepted. However, it is unclear according to the literature whether those with an open abdomen secondary to DCC or ACS can or should be fed enterally. Patients requiring an open abdomen may not tolerate enteral nutrition. Generalized gastrointestinal dysfunction could induce gastroparesis or an ileus. In addition, the edematous bowel, which is often associated with an open abdomen, may dilate further with the introduction of enteral nutrition. This may make closure of the fascia more difficult. Previous work from our institution has demonstrated that open-abdomen patients whose fascia is closed within 8 days have fewer infections and fistulae.26 The current study was undertaken to identify complications (both infectious and noninfectious) of enteral nutrition administration in patients with an open abdomen. In our review of 78 patients with an open abdomen for at least 4 hospital days, initiation of enteral nutrition within 4 days was associated with better outcomes. With early enteral nutrition, patients were more likely to achieve fascial closure by postoperative day 8, were less likely to form fistulae, and had less overall hospital charges. Infectious complications were similar between those fed early vs those who had delayed enteral feedings. Literature supporting the enteral feeding of open-abdomen patients is sparse. Tsuei et al,29 in a small retrospective study, described the safety of enteral feedings with patients who had an open abdomen. In fact, the patients received on average 77% of nutrition goal via the enteral route. Complications such as diarrhea (42%) and reflux of tube feedings (36%) were reported, with no noted adverse outcome. Time to abdominal closure and infectious or noninfectious complications were not noted. More recently, Cheatham et al30 described the increased insensible loss of nitrogen/protein in patients with the open abdomen. In view of these findings, early enteral nutrition was attempted. Postpyloric feedings were present in 75%, and approximately 800 kcal per day with 30 g of protein were provided over the first 5 days. Though patient outcomes were not described, the author suggests that open-abdomen enteral feedings are not only plausible but perhaps beneficial. Evidence supports that bowel wall edema and impaired gut function are still present even when the fascia is closed. Cothren et al31 described the feasibility of early enteral feedings immediately after definitive abdominal closure. Subsequently, with protocol-driven administration of enteral feedings, 92% of the patients attained goal feedings in 3 days. Five patients did tolerate a low rate (15 mL/h) with an open abdomen: 2 had a Bogata bag and 3 underwent split-thickness skin grafting. While commonly practiced, withholding enteral nutrition in patients with an open abdomen is not supported in the literature. Although specific studies supporting feeding the open abdomen are lacking, several concepts can be supported when enteral feedings are provided in this critically ill population. First, enteral feedings provide a hyperemic response not only in the healthy individual but also in the shock model, both hemorrhagic and septic.32–38 With improved blood supply to the intestines and the liver, less inflammation at the local level occurs and increased venous return is seen. Enteral feedings may even diminish pressor requirements during the shock state by increasing intestinal blood flow.39 In theory, more arterial supply and venous drainage from the abdominal contents provide an environment of reduced swelling and smaller structures, thereby allowing for safe fascial closure. Typically, it is the swollen abdominal contents that do not allow one to close the fascia primarily in the days after the initial insult.
Enteral feedings also provide anti-inflammatory or immunomodulation to the
small bowel. Provision of In addition, it is not uncommon for the open-abdomen population to be very ill, especially during the acute setting. In fact, the deadly triad (acidosis, coagulopathy, hypothermia), coupled with pressor requirement, is frequently seen. Many clinicians still withhold enteral feedings, with fear of small bowel necrosis. This phenomenon rarely occurs, and there are only case reports with the administration of early enteral nutrition.44,45 Frequent bedside visualization of the bowel (edema, peristalsis, etc) is the best method not only to assess tolerance of enteral nutrition but also viability. Finally, a continuous glucose source is a key component to aggressive glycemic control. In the landmark study by Van den Berge et al,46 insulin administered via continuous infusion coupled with a continuous glucose source provided not only a lowering of glucose levels but also infectious complications and mortality. Though this glucose source was IV initially, early conversion to enteral feedings occurred the subsequent day as tolerated. The limitations of this study stem from its retrospective study design. This type of investigation has obvious shortcomings, including variability of care. Outcomes were affected by initial and ongoing resuscitation needs, timing of dressing changes, and method of fascial closures. In addition, the initiation and administration of nutrition were surgeon specific, consistent with the lack of evidence-based guidelines on the subject. Our groups were defined by the timing of enteral nutrition. However, the reason for this decision could not be attributed to any identifiable factor. In fact, 15 patients did not have any enteral nutrition initiated within the 14-day data-collection period. Also, a selection bias may exist where patients who had early initiation of enteral nutrition were less edematous and more likely to have earlier fascial closure and subsequently less fistula formation. This same patient population may also have less risk (from a genetic disposition) of an ongoing inflammatory response, sepsis, and mortality.47 Finally, the constraints of a retrospective review and lack of comprehensive nutrition data support led to the exclusion of 54 open-abdomen patients. Despite these exclusions and potential selection bias, our study group represents the largest cohort of open abdomens and nutrition support, with outcome measures described.
Using early enteral nutrition during the care of DCC or ACS is feasible and safe, and is associated with improved outcomes. Earlier fascial closure and reduced fistula formation lead to decreased initial hospital charges and avoid later hospital admissions and surgical procedures, which contribute to the morbidity of an open abdomen. To achieve these outcomes, however, nasoenteric access should be placed during the first or second celiotomy, and feedings initiated soon thereafter. Bedside evaluation with hemodynamic monitoring, bowel inspection, and overall clinical improvement can determine tolerance of enteral feedings. By combining the tenets of surgery caring for the open abdomen, fundamentals of intensive care, and aggressive early enteral nutrition, the complications of ACS and DCC should decrease and the success of managing this difficult patient population will become common. However, these results should be validated with a large prospective randomized trial comparing enteral nutrition and PN, combined with strict glucose control and fluid management. Given the limited number of patients even a busy level I trauma center could provide, a multi-institutional trial would be required.
Presented at Clinical Nutrition Week (American Society for Parenteral and Enteral Nutrition), January 28–31, 2007, Phoenix, Arizona; and the World Congress Abdominal Compartment Syndrome, March 22–24, 2007, Antwerp, Belgium. Received for publication March 9, 2007. Accepted for publication May 4, 2007.
  CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati What's this?
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
 |
Sign In to gain access to subscriptions and/or personal tools.
4 days
within celiotomy; and late enteral feeding (LEN; >4 days). Outcomes
included infectious complications, early closure of the abdominal cavity
(<8 days from original celiotomy), and fistula formation. Results:
Fifty-three of 78 (68%) patients were men, with a mean age of 35 years; 74%
had blunt trauma. Forty-three of 78 (55%) patients had EEN, whereas 35 of 78
(45%) had LEN. There was no difference with respect to demographics, injury
severity, or infectious complication rates. Thirty-two of 43 (74%) patients
with EEN had early closure of the abdominal cavity, whereas 17 of 35 (49%)
patients with late feeding had early closure (p = .02). Four of 43
(9%) patients with EEN demonstrated fistula formation, whereas 9 of 35 (26%)
patients with late feeding formed fistulae (p = .05). The EEN group
had lower hospital charges (p = .04) by more than $50,000.
Conclusions: EEN in the OA was associated with (1) earlier primary
abdominal closure, (2) lower fistula rate, (3) lower hospital charges. 
10,000 caloric deficit
occurs.
Top
MATERIALS AND METHODS
RESULTS
2 tests were used for categorical data. Statistical
significance was defined as p