Tolerance to and Postoperative Outcomes With Early Oral Feeding Following Elective Bowel Surgery: A Systematic Review

Although practice guidelines recommend resuming oral feeding immediately after gastrointestinal surgery, many practitioners remain reluctant to order early oral feeding (EOF). Therefore, this review aimed to clarify the tolerance to and postoperative outcomes with EOF among patients undergoing bowel surgery. A systematic review of the literature published between January 1990 and July 2022 with the time of oral intake (early or delayed until resolution of ileus) as the exposure variable was conducted using PubMed and Scopus databases. Outcomes of interest included tolerance to EOF and postoperative adverse effects or complications. After screening 1,667 research articles, 18 randomized control trials, six prospective case series, and four cohort studies met our inclusion criteria, collectively representing data from 2,647 patients in eleven countries. These studies indicate that while most patients tolerate EOF, 5-25% may not tolerate EOF until the fourth postoperative day (POD). Moreover, EOF, at best, has no advantage over delayed feeding in terms of vomiting, nausea, nasogastric tube requirement, or other postoperative complications. In addition, early return of bowel function, lower risk of diarrhea, and lower pain score with EOF are inconsistently reported, and shorter hospitalization with EOF may be limited to those who tolerate oral feeding on POD 0 or 1. Nevertheless, shorter hospitalization with EOF could reduce the cost of hospitalization. A substantial number of patients may not be able to tolerate oral feeding after bowel surgery until POD 4, and in patients who tolerate EOF, the only clear benefit is a shorter length of hospitalization.


Introduction And Background
Traditional postoperative care for gastrointestinal (GI) surgeries typically involved withholding oral intake until the resolution of ileus indicated by the passage of flatus or stool.Breaking away from the traditional oral feeding (TOF) practice, the American Society for Enhanced Recovery, Perioperative Quality Initiative (POQI), and the European Society for Clinical Nutrition and Metabolism (ESPEN) recommend resumption of oral feeding -including clear liquids, oral nutritional support, and balanced diet-soon after GI surgery to facilitate post-surgery recovery, shorten the length of hospitalization and reduce postoperative morbidity and mortality [1,2].Resuming oral nutrition immediately after GI surgery is also included as a care element in the Enhanced Recovery After Surgery (ERAS) protocol [3].
These practice guidelines also align with patients' preferences [4].For example, when patients who underwent colon resection were advised to resume a regular diet based on their appetite, 70-80% initiated solid meals by the second postoperative day (POD) [5,6].However, despite the growing scientific consensus and patient preference for early oral feeding (EOF) after GI surgery, many healthcare providers remain reluctant to advise oral intake of fluids or solid meals until the return of bowel function, especially anesthesiologists and general surgeons compared to colorectal surgeons [7][8][9][10][11].The reluctance is not entirely unfounded: Even though most patients prefer solid meals on POD 1 [12] or 2 [6], only a third may be able to tolerate solid meals by POD 2 [5], and half of the patients by POD 4 [11].
The disconnect between practice guidelines and the healthcare provider's reluctance to order EOF after elective GI surgeries warrants a review of evidence to ascertain if patients can tolerate EOF and if it produced improved postoperative outcomes compared to TOF for several reasons.Previous systematic reviews have evaluated postoperative outcomes with EOF in patients undergoing any GI surgery [13,14], only lower GI surgery [15,16], or, more specifically, among patients undergoing colorectal surgery [17,18].Reviews that evaluated the post-surgery outcomes with early feeding to specific surgery sites, such as colon and rectum, were limited by the number of papers meeting the inclusion criteria (n=9 with 879 patients [16], n=7 with 587 patients [17], and n=5 with 985 patients [18]).However, the broad inclusion criteria of some of these reviews led to reporting of the combined outcomes with EOF from several feeding routes -oral, nasojejunal tube, nasoduodenal tube, and jejunostomy [13][14][15] -or with various surgical sites such as upper GI, lower GI, and hepatobiliary surgery [13,14].Also, prior reviews do not make a distinction by surgical approach (laparotomy or laparoscopy) even though the resolution of ileus [19] and tolerance to oral feeding [5] are achieved sooner in patients undergoing laparoscopic GI surgery and are more likely to initiate EOF [8,10] compared to a laparotomy procedure.
Therefore, this study aimed to systematically review the tolerance to and postoperative outcomes with EOF among patients undergoing bowel surgery, representing 58% of all GI surgeries worldwide [20].Moreover, we aimed to separately evaluate the available studies with small or large bowel procedures by surgical approach (laparotomy or laparoscopy).

Review Method
This systematic review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [21].

Search Strategy
Our search strategy covered studies indexed in PubMed and Scopus databases between 01 January 1990 and 31 July 2022 to capture the most recent trends (last searched: 05 November 2022) using the following search term: (postoperativ*) AND (rectum OR anus OR colon OR intestine OR bowel) AND (surger* OR colestom* OR resection* OR abdominoperineal) AND (clear fluid* OR oral feed* OR solid feed* OR eat OR eating).The reference lists of earlier systematic reviews and meta-analyses on postoperative feeding practices were also screened.

Study Selection Criteria
Randomized control trials (RCT), prospective case series (PCS), and cohort studies published in English with online access to the full-text article meeting the following criteria were considered for inclusion: a. Participants ≥18 years had undergone elective bowel surgery primarily to treat an underlying bowel condition.
b. Time to first oral feed (liquid, semi-solid, or solid) or EOF (as defined by the study authors) after surgery was used as exposure variables.c.Reported tolerance to EOF, adverse effects (e.g., nausea and vomiting), or postoperative complication (including mortality).
Studies with patients who had undergone emergency bowel surgery, palliative bowel surgery, or required prolonged postoperative enteral nutrition through, for instance, a feeding jejunostomy were excluded.Finally, studies reporting the effect of EOF as a part of multimodal recovery programs such as ERAS and "fast-track surgeries" were excluded unless the experimental and control groups varied only by feeding timeline and all other modalities were equally applied to both groups.
Case series and cohort studies were differentiated using the distinction criteria proposed by Mathes and Pieper [22].

Data Extraction and Analysis
Data from eligible studies were tabulated using a predefined form which included study characteristics (author, year of publication, country), population characteristics (sample size, age with standard deviation or range, and gender distribution), surgery characteristics (indication, location, and approach), and the outcomes of interests (tolerance to EOF, postoperative adverse effects, and complications).If the country of the study was not explicitly mentioned, the country of the primary author's affiliated institution was used.

Results
Our literature search strategy yielded 1,667 papers, of which 809 were removed during prescreening (Figure 1).The title/abstract of the remaining 858 records was screened with full-text screening for 312 records.Finally, 25 articles met the inclusion criteria.In addition, another three articles were identified through the reference lists of previous systematic reviews and meta-analyses.The detailed flow of the study selection process is described in Figure 1.

Study Design, Surgical Approach, and Surgery Location of the Included Studies
Eighteen out of the 28 included studies were RCTs (64.28%) which collectively included 1,183 patients from Australia [23], Brazil [24,25], China [26], Egypt [27], India [28], Iran [29,30], Italy [31,32], Spain [33,34], Turkey [35], and USA [36][37][38][39][40].The average age of the participants in most RCTs was 50 years or older.In ten RCTs, patients underwent a laparotomy procedure, out of which six specifically involved the large bowel, one involved the small bowel, and three studies did not separately report data from patients who underwent small or large bowel procedures (Table 1  Half of the remaining eight RCTs reported combined data from patients who had undergone either laparotomy or laparoscopy procedures (Table 2), while the other half did not specify the surgical approach (Table 3).In both RCT subgroups, three specifically involved the large bowel, and one involved both small and large bowel procedures (Tables 2-3).We did not identify any RCT that specifically reported the outcomes of EOF after laparoscopic bowel surgery during our study period.Additionally, we identified six PCS [41][42][43][44][45][46] reporting tolerance to and/or postoperative outcomes with EOF, all of which involved the large bowel surgery (Surgery approach: Laparotomy = 3, Laparoscopy = 1, Laparotomy or laparoscopy = 1, and Unspecified = 1; Table 4).Collectively these studies included data from 750 patients from four countries.Finally, four cohort studies [47][48][49][50] met our inclusion criteria and were included in this review; all four involved large bowel surgery (Surgery approach: Laparotomy = 1, Laparotomy or laparoscopy = 3; Table 5).

Author
Collectively, the cohort studies included data from 714 patients from three countries.The occurrence of postoperative nausea, vomiting, diarrhea, abdominal distension, NGT reinsertion or other complications was similar between the two groups.However, the median length of hospitalization was shorter in the EOF group (9 days vs. 12 days; p = 0.01).

Author
Retrospective cohort studies

Post-surgery Diet Resumption Plan of the Included Studies
EOF in 15 RCTs was initiated with a liquid diet on POD 0 or 1, transitioning to a semi-solid or regular diet irrespective of the passage of flatus or stool.However, three studies initiated patients on a regular diet after 8-24 hours of bowel rest [26,39,40].The control group in all RCTs followed the TOF regime.
There was more heterogeneity with the initiation of oral feeding among the PCS and cohort studies.For example, one PCS initiated EOF on POD 0 [44], another at POD 1 [43], two on POD 2 [41,42], and another two included patient groups who started EOF on POD 1, 2, or 3 [45,46].Similarly, in one cohort study, EOF was started on POD 0 or 1 [50], two initiated feeding on POD 1 [48,49], and one on POD 2 [47].In all PCS and cohort studies included in this review, EOF was initiated with a liquid diet and the subsequent introduction of a regular diet.

Tolerance to EOF
Thirteen out of the 18 RCTs reported tolerance to EOF on different PODs.Among patients who underwent open large bowel surgery, tolerance to a regular diet was reported in 85.9% of patients on POD 1 [35] and 75% to 95% between POD 2 and 3 [23,27,33] (Table 1).Lucha et al. (2005) [39] reported that patients who underwent open large bowel surgery tolerated a regular diet on an average of 4.7 days on EOF modality compared to 5.5 days of TOF patients, albeit non-significantly.Similarly, high tolerance to EOF was reported among patients who underwent an open small bowel procedure [30] or an unspecified open bowel procedure [28,31,36], with patients in the EOF groups tolerating regular diet between POD 2 and 3 compared to POD 5 to 7 in TOF groups [28,36] (Table 1).
Consistent with these observations, RCTs that reported combined outcomes from patients who underwent either laparotomy or laparoscopy bowel procedures [24,25,34] (Table 2) or did not specify the surgery approach [29,37] (Table 3) also suggest that most patients tolerate EOF with liquid and subsequent solid diet.It is noteworthy that Ortiz et al. (1996) [34] showed a comparable tolerance to EOF (~80% on POD 1) among patients who underwent laparotomy or laparoscopy colorectal surgery (Table 2).
Although RCTs investigating the tolerance to oral feeding on POD 0 are scarce, one PCS suggests that patients who tolerate oral feeding on POD 0 might have better postoperative outcomes in terms of return of bowel function, earlier tolerance to a solid diet, in-hospital complications, or the length of hospitalization [44].However, all patients in this PCS [44] were operated on with laparoscopy, and there is no direct evidence that patients who undergo an open procedure and tolerate oral feeding on POD 0 also have better postoperative outcomes.
The amount of blood loss [46], male gender [41], and type of operation (total abdominal colectomy or total proctocolectomy) [41] have been identified as factors determining early tolerance to oral feeding.Moreover, the type of surgery may also predict a shorter hospital stay among EOF patients [43].However, age, comorbidities, operative time, and additional surgical procedures were not associated with early tolerance to oral feeding or better postoperative outcomes [41,46].

Discussion
The currently available evidence suggests that although many patients may be able to tolerate oral feeding by POD 2 or 3, about 5-25% of patients do not tolerate oral feeding until POD 4, which is also typically when the first flatus and the resolution of ileus occurs, and patients in the TOF begin to tolerate oral feeding [23,25,29,30,35,36,39].Moreover, the benefits of EOF on postoperative outcomes in patients undergoing bowel surgery may be marginal compared to TOF.For instance, EOF, at best, has no advantage over TOF in terms of vomiting, nausea, and NGT requirement and, at worst, may increase vomiting and nausea.Furthermore, all studies consistently show similar postoperative complication rates with EOF and TOF.However, the evidence for early return of bowel function, lower risk of postoperative diarrhea, and lower pain score with EOF are inconsistent, and shorter hospital stays with EOF may be limited to patients who can tolerate oral feeding on POD 0 or 1.Finally, there is some evidence to indicate that the type of surgery and blood loss during surgery may be a determinant of early tolerance to oral feeding, and it remains plausible that tolerating oral feeding early and improved postoperative outcomes, if any, may both be downstream of the type and events during surgery.
Notably, 79% of the included studies reported patient outcomes with EOF following large bowel surgery, and the literature is less extensive with small bowel procedures.Similarly, laparotomy was the surgical approach in 46% of the studies, while the rest combined patients who underwent laparotomy or laparoscopy procedures.Only two studies specifically reported patient outcomes with laparoscopy, one of which compared patient outcomes with EOF following laparotomy or laparoscopy procedures.Therefore, there is insufficient literature to conclude whether EOF has comparable outcomes following laparotomy or laparoscopy bowel surgery.
Nonetheless, the findings of this study are consistent with earlier systematic reviews.For example, McAlee et al. and Zhuang et al. showed that EOF in patients who underwent elective colorectal surgery was associated with earlier resolution of postoperative ileus and hospital discharge without a statistically significant increase in nausea, vomiting, or reinsertion of an NGT even though this conclusion was based on the review of only five [18] and seven [17] studies respectively.A more exhaustive Cochrane review compared early nutrition, orally or through any kind of tube feeding, with traditional nutrition of nil-bymouth until ileus resolution among patients undergoing any type of lower GI surgery also noted no differences in adverse events and complication rates, and although early feeding was associated with shorter hospitalization, there was considerable heterogeneity in reporting [15].Nevertheless, a shorter hospital stay in EOF patients could significantly lower hospitalization costs [51].
Our study design allowed the inclusion of more studies than previous systematic reviews in this domain, which enabled a more thorough comparative analysis.However, our literature review was limited by the paucity of available information on several aspects of EOF after bowel surgery.First, only a few studies initiated EOF with solid or regular diets.In 90% of the included studies, feeding was initiated conservatively with a clear liquid diet, even though earlier studies indicate similar outcomes with the early introduction of liquid or solid diets after GI surgery.For instance, in an RCT where one group was initiated on an oral liquid diet and the other on a regular diet soon after the removal of the NGT, there was no significant difference in the occurrence of vomiting, abdominal distention, intestinal obstruction, acute gastric dilatation, or the overall complication rate [52].More recently, a low-residue diet has been shown to reduce adverse effects and complication rates after colorectal surgery compared to a clear liquid diet [53].However, initiating early feeding with a liquid diet may also have merits.For example, early intake of warm water was associated with early flatulence among patients who underwent laparoscopic cholecystectomy and were randomized to drink warm water in the fourth postoperative hour or nil by mouth until the eighth postoperative hour when patients in both groups initiated oral fluids and soft food [54].
Second, while the type of surgery is associated with tolerance to EOF, only two studies investigated such an association.Third, we could not ascertain if the responses to EOF varied by indication for the surgery; most studies presented data from cohorts of patients with a mixed diagnosis, and when responses to EOF were reported in patients with a specific condition, it was usually colorectal cancer.If EOF is tolerated or produces a superior outcome compared to TOF in patients with, for instance, Crohn's disease, diverticulitis, or inflammatory bowel disease is currently unknown.Finally, there was a notable absence of studies comparing EOF and TOF in patients who underwent bowel surgery with laparotomy or complete laparoscopy.In the one study we identified, there was no difference in tolerance to EOF, adverse effects, or complications between patients who had undergone laparoscopic or open bowel surgery [34].

Conclusions
The current evidence indicates that < 25% of patients may not be able to tolerate EOF after bowel surgery until POD 4.However, patients who tolerate EOF may have a shorter length of hospitalization which may be associated with a lower cost of hospitalization.Therefore it is imperative to investigate factors associated with tolerance to EOF.A handful of available studies have identified blood loss, gender, and type of procedure but not age, comorbidities, operative time, or additional surgical procedures as factors determining early tolerance to oral feeding.Also, most studies report similar adverse effects and postoperative complication rates with EOF and TOF, and healthcare professionals must weigh the benefits of EOF with medical resource availability and/or utilization for implementing EOF.However, it is important to note that the differences in the studies' design, setting, and population may not permit generalization.Therefore, a meta-analysis of this study would provide greater insight into the results of this study.Further studies should be conducted with identical baseline conditions to verify the results of this study.

FIGURE 1 :
FIGURE 1: PRISMA flow diagram showing results of the study selection process.PRISMA: Preferred Reporting Items for Systematic Reviews and Meta-Analyses ).

TABLE 1 : Randomized controlled trials reporting tolerance to and/or postoperative outcomes with early oral feeding (EOF) following elective open bowel surgery.
TOF = traditional oral feeding; NGT = nasogastric tube; POD = postoperative day.$Gender distribution not reported.
groups.The time to the first bowel movement was 5.4 days with a 26.6% complication rate in the laparoscopy group and 5.5 days with a 13.3% complication rate in the laparotomy group (p=NS for both).

TABLE 2 : Randomized controlled trials reporting tolerance to and/or postoperative outcomes with early oral feeding (EOF) following elective open or laparoscopic bowel surgery.
TOF = traditional oral feeding; NGT = nasogastric tube; POD = postoperative day

, Type of Feeding Tolerance to Year Country Participants Surgery Procedure EOF Post-operative Complications/Predictors
The amount of estimated blood loss was the 2023 Mvoula et al.Cureus 15(8): e42943.DOI 10.7759/cureus.42943onlyvariable that was significantly associated with successful early oral postoperative feeding, while a null association was observed with age, sex, ASA score, epidural analgesia use, operating time, side of surgery, previous abdominal surgery, additional procedures performed, colloid volume, oral intake with initial feeding, day of initial feeding, and crystalloid fluid requirements.