Feed Abdominal Surgery Patients to Improve Outcomes

Reference: Herbert G, Perry R, Andersen HK, Atkinson C, Penfold C, Lewis SJ, Ness AR, Thomas S. Early enteral nutrition within 24 hours of lower gastrointestinal surgery versus later commencement for length of hospital stay and postoperative complications. Cochrane Database Syst Rev. 2019;7(7):CD004080. doi:10.1002/14651858.CD004080.pub4

Rationale: Historically, patients undergoing lower GI surgery were starved, with food by mouth forbidden for 6 hours to 12 hours preoperatively or several days postoperatively until peristalsis resumed. Recent evidence that these periods of starvation are not needed, and may in fact adversely affect patient outcomes, led to the current systematic review of these effects.

Objective: Conduct a systematic review of studies exploring the effects of early postoperative enteral feeding on subsequent length of hospital stay (LOS), anastomotic dehiscence, wound infection, pneumonia, and abdominal abscesses.

Methods: Cochrane procedures were used to search the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, LILACS, ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry Platform from inception through November 15, 2017, for unique data from randomized clinical trials (RCTs) on adults undergoing lower GI surgery and exploring the effects of oral or tube enteral feeding started during the first 24 hours postoperatively as compared with the traditional approach in which oral intake or tube feeding is avoided before the return of bowel function. Studies on parenteral feeding or individuals younger than 18 years of age or comparing only different forms of enteral feeding were excluded. Two authors blinded to each other’s evaluations successively coded titles for exclusion and then coded full-text for inclusion, referring to colleagues for translation or resolution if their decisions differed. Primary outcomes reported were LOS or standardized wound infections, intra-abdominal abscesses, anastomotic leakage or dehiscence, or pneumonia during the hospital stay. Secondary outcomes were standardized quality of life, mortality, or nausea/vomiting during the hospital stay. The 95% CI was used for statistical significance of weighted mean differences for the continuous variable LOS and for relative risks of the dichotomous variables studied. Heterogeneity was assessed using forest plots with the χ² statistic with P < .01. The I² statistic was applied to test whether statistical heterogeneity contributed to intervention effects. Trial sequential analyses were performed to test whether cumulative Z curves crossed the boundaries of benefit, harm, or futility, with alpha set at 0.033 for LOS, allowing for multiple outcome comparisons. Sensitivity analysis of LOS was conducted without studies at high risk of bias due to 2 or more of the following: nonrandom assignment, lack of blinding, incomplete outcome data, or selective outcome reporting. 

Results: Among 4018 records identified, 17 RCTs including 1437 participants qualified for inclusion. The overall effect of any form of oral or tube feeding started within 24 hours after GI surgery was to reduce hospital LOS by 1.95 days (P < .002), with no significant effect on wound infection, anastomotic leakage, dehiscence, intra-abdominal abscess, pneumonia, nausea, vomiting, or mortality. These results were confirmed by trial sequential analysis and remained true across international borders or for subset analyses involving (a) only oral feeding or oral plus tube feeding or (b) surgeries proximal or distal to the ligaments of Treitz, or when (c) more biased studies were excluded from the analyses.

Authors’ Conclusions: This review provides sufficient evidence that early postoperative enteral oral or tube feeding reduces hospital LOS. Evidence was inconclusive for all other outcomes; additional research is necessary before a conclusion can be justified.

Reference: Burden S, Billson HA, Lal S, Owen KA, Muneer A. Perioperative nutrition for the treatment of bladder cancer by radical cystectomy. Cochrane Database Syst Rev. 2019;5(5):CD010127. doi:10.1002/14651858.CD010127.pub2

Rationale: Patients with muscle- invasive urothelial bladder cancer usually undergo RC, with a long, often complicated hospital stay and recovery period. This may be improved by better nutrition, which in this study resulted in reduced recovery time and complications for those with other forms of cancer. 

Objective: Systematically review RCT literature exploring the effects on postoperative outcomes of perioperative nutritional interventions for patients undergoing RC for bladder cancer. 

Methods: The authors searched AMED, CINAHL, EMBASE, MEDLINE, and Evidence Based Medicine Reviews clinical trial registries from inception through June 2018, as well as meeting abstracts through June 2017, for RCTs including at least 65% of adult patients who underwent RC for bladder cancer and who received any standardized perioperative nutritional intervention compared with any control. Two reviewers successively evaluated the titles and then the full-text version of potentially qualifying articles measuring the primary outcome (LOS or complications) or secondary outcome (mortality within 90 days after surgery). After determining GRADE ratings of study quality, clinically similar studies were pooled for meta-analyses of quantitative data using a random-effects model with P ≤ .05 indicating statistical significance.

Results: Of 5514 studies examined, 8 RCTs including 500 patients met the inclusion criteria. One RCT of 157 participants reported more complications within 30 days after surgery for those receiving postoperative parenteral nutrition compared with those receiving oral nutrition with intravenous lactated Ringer solution (P = .01). No other comparison was statistically significant, although in one small RCT, postoperative immune-oriented oral nutrition resulted in marginally fewer complications than standard oral nutrition within the first 90 days after RC surgery (P = .10).

Authors’ Conclusions: Postoperative parenteral nutrition may increase the likelihood of complications compared with oral feeding within the first 30 days after RC. Immune-oriented nutrition merits further research. 

For those whose patients cannot wait for perfect evidence, there is low-quality evidence of shorter hospital stays for patients fed orally or by tube beginning on the first day after lower GI surgery.⁶ Patients receiving parenteral nutrition experienced more complications than those receiving oral nutrition during the first 30 days after RC.⁷ Similarly, early postoperative oral feeding led to shorter hospital LOS with fewer complications than the traditional nil-by-mouth approach, as reported in a systematic review of 7 RCTs (n = 583 patients) following colorectal surgery (P < .04).⁸

Until one understands how Cochrane reviewers are trained to downgrade evidence quality for small or biased studies with statistical heterogeneity arising from nonstandardized procedures or outcomes reporting, it may seem irrational that so many RCTs provided only low-quality support for oral feeding after lower abdominal surgery. Clinical study quality can be improved by conducting double-blind studies using clearly described techniques for avoiding bias in patient selection and allocation, study performance, or outcomes measurement and by reporting all valid, reliable standardized outcome measures⁹ documented on all participants in the study. Evidence can be further strengthened by reducing unidentified sources of intervention or patient variability that expand data heterogeneity, thereby weakening statistical significance. Systematic review conclusions are only as strong as the studies on which they are based.