The authors concluded that enteral nutrition after gastrointestinal surgery was associated with a significant reduction in any complication, any infectious complication, anastomotic leaks, intra-abdominal abscesses, and length of hospital stay. Whilst the authors' conclusions reflected the evidence presented, the under-reporting of how trials were selected and quality assessed means that caution is warranted when judging the reliability of this review.

To compare the effectiveness of enteral nutrition versus parenteral nutrition in patients following gastrointestinal surgery.

MEDLINE, Web of Science, and the Cochrane Library were searched for studies published in English from 1974 to 2006. Search terms were reported. Bibliographies were scanned for further studies. Unpublished data were retrieved from the included trials.

Randomised controlled trials (RCTs) that compared enteral nutrition with parenteral nutrition in adult patients following elective gastrointestinal surgery were eligible for inclusion in the review. Enteral nutrition was defined as the delivery of any nutrient (solid or liquid) that passed through the digestive tract, irrespective of other standard nutritional care or tube feeding. Parenteral nutrition was defined as nutritional fluids containing a minimum of glucose and amino acids delivered through the central or peripheral venous system. The following were excluded: trials that evaluated only nutritional or physiological outcomes; trials of home-based parenteral nutrition; trials that included patients undergoing transplantation surgery, chemotherapy, or radiotherapy; trials of critically-ill patients; and trials located in developing countries.

The primary outcomes were the number of patients with any complication, any infectious complication and mortality. Secondary outcomes were the number of patients with wound infections or dehiscence, anastomotic leaks, intra-abdominal abscesses, pneumonia, respiratory failure, urinary tract infections, renal failure, adverse effects, and length of hospital stay.

In all included trials, enteral nutrition commenced within six to 24 hours after surgery. The included patients underwent a variety of upper- and lower-gastrointestinal surgeries (including oesophageal, gastric, pancreatic, colorectal, small/large bowel, and hepatic surgery). Most of the included trials were of patients with malignant, or malignant and benign conditions (where reported). Enteral nutrition was delivered mainly through a naso-jejunal or catheter jejunostomy. Control patients received total parenteral nutrition or peripheral parenteral nutrition. Five trials were conducted in the UK.

The authors did not state how many reviewers performed the study selection.

Trial quality was assessed on fulfilment of the following components: adequate allocation concealment (yes, no, or unclear), double blinding of outcome assessment (yes or no), and handling of withdrawals and drop-outs (whether intention-to-treat analysis was used).

The authors did not state how many reviewers carried out the quality assessment.

Data were extracted to enable the calculation of relative risks (RR), and mean differences (MD), along with 95% confidence intervals (CI). Investigators were contacted for additional information or missing data (where necessary).

Two reviewers independently carried out the data extraction, with any disagreements resolved by discussion.

Relative risks, weighted mean differences (WMDs), and 95% confidence intervals were pooled in fixed-effect meta-analyses, using inverse-variance weighting. Heterogeneity was assessed with the Q-test, and I² statistic. I² values indicated low (I²=25%), moderate (I²=50%), or high (I²=75%) levels of heterogeneity in the evidence. Where statistical heterogeneity was present, the DerSimonian and Laird random-effects model was applied. Possible sources of variation were explored in random-effects meta-regression.

Sensitivity analyses were performed for the primary outcomes, based on disease pathology (malignant or non-malignant); nutritional status (malnourished or malnourished plus well-nourished); enteral nutrition administration route (immunonutrient or standard nutrient); type of parenteral nutrition (total parenteral nutrition or peripheral parenteral nutrition); administration of parenteral lipid; and location (Europe, or North America). Further analyses were conducted to explore the impact of methodological quality components.

Publication bias was assessed visually using a funnel plot and statistically using Egger's test and Begg's test.

Twenty-nine RCTs were included in the review (n=2,552 patients). Allocation concealment was reported in eight trials; one trial used double-blinding; and 15 trials used intention-to-treat analysis. There was no evidence for publication bias, except for the outcome of wound infection.

Enteral nutrition was found to have significantly positive effects on reducing any complication (RR 0.85, 95% CI 0.74 to 0.99; 13 trials; I²=22%) and any infectious complication (RR 0.69, 95% CI 0.56 to 0.86; 13 trials; I²=10%). There was no significant effect on mortality (15 trials).

Enteral nutrition was associated with positive effects on the reduction in anastomotic leaks (RR 0.67, 95% CI 0.47 to 0.95; 17 trials) and intra-abdominal abscesses (RR 0.63, 95% CI 0.41 to 0.95; 16 trials), with no evidence of heterogeneity. An increased incidence of vomiting was associated with enteral nutrition (RR 1.61, 95% CI 1.25 to 2.09; seven trials; I² = 0%) than with parenteral nutrition. Hospital stay was reduced as a result of enteral nutrition compared with parenteral nutrition (WMD -0.81, 95% CI -1.25 to -0.38; I²=66%).

Sensitivity analyses showed that enteral nutrition produced statistically significant reductions in any complications where trials included malignant diseases, malnourished patients, standard nutrient, total parenteral nutrition, parenteral nutrition with parenteral lipid, and a European country. This direction of effect was reflected for enteral nutrition in reducing any infectious complications, with the addition of immunonutrient and tube feeding as statistically significant moderators. There was no significant association between any variable and mortality. Double-blinded outcome assessment was the only methodological domain to influence any complication (RR 2.16, 95% CI 1.01 to 4.62), and any infectious complication (RR 5.04, 95% CI 1.22 to 20.80).

Enteral nutrition after gastrointestinal surgery was associated with a significant reduction in the incidence of any complication, any infectious complication, anastomotic leaks, intra-abdominal abscesses, and length of stay in hospital.

The review question was clear and was supported by potentially reproducible inclusion critieria for all aspects apart from outcomes, which appeared to be loosely defined. The search strategy included some relevant data sources. Unpublished data were included in the review. The language restriction to English papers might mean that relevant studies were missed, and language bias was a potential threat. The review process was poorly-reported, with only the data extraction process indicating that attempts were made to minimise error and bias.

Appropriate quality assessment components were applied to the included trials, and the results of this assessment were developed in the analysis. Trial characteristics were presented in sufficient detail. The chosen method of synthesis was appropriate. An extensive exploration of trial variation was carried out.

The authors' conclusions reflected the evidence presented. The under-reporting of how trials were selected and quality assessed means that some caution is warranted when judging the reliability of this review.

Practice: The authors stated that enteral nutrition initiated within 24 hours is recommended in patients following gastrointestinal surgery.

Research: The authors did not state any implications for research.

None.

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.