Early Closure of Temporary Loop Ileostomies: A Systematic Review

A temporary loop ileostomy is a common surgical procedure to protect colorectal anastomoses. The aim of this systematic review was to determine whether early closure of a defunctioning loop ileostomy (<2 weeks from index operation) is safe and reduces stoma-related morbidity.

 A systematic literature search was conducted using Ovid MEDLINE, EMBASE, Cochrane Collaboration, and the Cumulative Index to Nursing and Allied Health (CINAHL®) databases to identify all publications from January 1996 to March 2014 that reported the outcomes of early ileostomy closure.The following search terms (and their variations) were used as both medical subject headings (MeSH terms) and text words: ileostomy, surgical stoma, stoma, early, reversal, closure. No language restrictions were applied. The main outcomes of interest were stoma-related complications and postclosure complications. Studies that included pediatric patients (<18 years of age), small cohorts (<10 participants), case reports, conference abstracts, reviews, and letters; studies involving defunctioning colostomies or other types of small bowel stomas; and studies where results from closure of an ileostomy at >14 days could not be separated from early closure results were excluded. Where multiple studies were reported by the same institution and/or authors, only the most recent was included. This search strategy identified 4 studies (2 retrospective case series, 1 prospective nonrandomized study, and 1 randomized controlled trial), yielding a pooled population of 142 patients, ages 18–89 years old. Three studies reported indication for ileostomy; colorectal cancer accounted for 96 patients (78%). Time to ileostomy closure ranged from 8–14 days. No reported deaths were related to ileostomy closure. Wound infections were reported in 3 studies and were the most common complications, affecting 24 patients (19.8%). Of the 2 studies that reported ileostomy-related complications, 4 patients (3.6%) experienced a stoma-related complication before closure. Ileus or small bowel obstruction (SBO) occurred in 7 patients (4.9%). Compared to traditionally timed closure (8–12 weeks), reported stoma-related complication rates were lower in patients undergoing early closure. Both mortality and ileus/SBO rates also compare favorably with traditionally timed closure; however, wound infection rates appear to be increased. Additional studies to accurately define which individuals stand to benefit from early closure, as well as to further evaluate the impact of early ileostomy closure on quality of life and health care costs, are warranted.

Anastomotic leaks after colorectal surgery are common, with reported incidence following low anterior resections between 3% and 26%.^1-4 Furthermore, large prospective and retrospective series have shown these leaks to be associated with major adverse events, with mortality rates between 6% and 22%.^3,5 To mitigate the consequences of anastomotic leaks, surgeons often construct a loop ileostomy at the time of surgery. Randomized controlled trials and meta-analyses have shown defunctioning loop ileostomies are effective in reducing both the rate of symptomatic anastomotic leaks as well as the need for reoperation in such cases^6-8 and therefore recommended in all patients with high-risk anastomoses.^7,8

Although a temporary loop ileostomy can provide a protective benefit, it is not without risk. Stoma-related morbidity due to parastomal hernias, small bowel obstruction (SBO), or dehydration from high stoma output are common. Three retrospective studies from Germany,⁹ The Netherlands,¹⁰ and the United States¹¹ involving 120, 39, and 154 ileostomy patients, respectively, reported rates of loop ileostomy-related morbidity ranging from 13.3% to 74%. Furthermore, a nonrandomized, prospective study by Silva et al¹² demonstrated loop ileostomies also have a significant negative impact on quality of life, sexual relationships, and activities of daily living.

The majority of patients undergoing colorectal surgery will recover without anastomotic complications.^1-4,13 In these patients, the formation of a diverting loop ileostomy serves only as an unnecessary source of morbidity. A recent review¹⁴ demonstrated more than 90% of patients undergoing colorectal surgery in their cohort derived no direct benefit from their defunctioning stoma. Unfortunately, the identification of patients that stand to benefit from fecal diversion cannot be accurately predicted preoperatively.

Currently, gastrointestinal continuity is usually restored after a period of 8–12 weeks.¹⁵ Closing the stoma before that time is thought to be associated with marked peristomal and intraperitoneal adhesions, making surgery more hazardous^9,16; however, the optimal time for reversing an ileostomy remains unknown, and a high percentage of patients will experience stoma-related complications during this period. One proposed method of minimizing stoma-related morbidity and mortality while providing the protective benefits of fecal diversion is to close the stoma within 2 weeks of the index operation before adhesions have developed. This time period is thought to provide the minimal delay to closure while still allowing adequate time for anastomotic healing. 

A systematic review of the literature was performed to assess the effectiveness and safety of early closure of a defunctioning loop ileostomy and its effect in minimizing stoma-related morbidity.

A systematic review was conducted in accordance with the code of practice as set out in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement.¹⁷

Literature search. A systematic literature search of the Ovid MEDLINE, EMBASE, Cochrane Collaboration, and Cumulative Index to Nursing and Allied Health (CINAHL®) databases was performed of publications from January 1996 to March 2014 (inclusive) to reflect contemporary surgical practice. Search terms included [ileostomy OR surgical stoma OR stoma OR Ileost*.mp.] AND [early OR earl*.mp.] AND [revers*.mp.] OR ([closure* OR close* OR closing]).mp.]. Searches were restricted to studies involving humans. No language limits were applied.

Searches were concomitantly performed by 2 of the authors; all results were screened to identify reports on early ileostomy closure. Full-text papers of selected studies then were further evaluated by the authors independent of one another to identify those eligible for inclusion. Final decision regarding eligibility was determined by group consensus; any conflicts were adjudicated by the senior author.

In addition, all eligible articles were hand-searched for potentially relevant references not identified during the initial literature search.

Inclusion and exclusion criteria. All published studies reporting early ileostomy closure (defined as <14 days from the index operation in which the ileostomy was formed) were included. Studies had to report at least 1 of the outcome measures detailed to follow.

Studies that involved pediatric patients (<18 years of age), a small cohort (<10 participants), case reports, conference abstracts, reviews, and letters; studies concerning defunctioning colostomies or other types of small bowel stomas; and studies where results from closure of the ileostomy in >14 days could not be separated from early closure were excluded. Where multiple studies were reported by the same institution and/or authors, only the most recent was included.

Outcomes of interest and definitions.

Primary outcomes. Primary outcomes included stoma-related complications and postoperative complications such as death, return to surgical theater, SBO/ileus, anastomotic leak, fistula, abscess, and wound infection.

Secondary outcomes. Secondary outcomes examined patient-related data (demographics, timing of ileostomy closure, indication for ileostomy, and index operation) and operative and hospital-related outcomes (operative time, length of stay post-ileostomy closure, and total length of stay).

Data extraction and analysis. Data were extracted by 1 of the researchers into standardized data extraction tables and checked by a second researcher. Included studies were assigned a level-of-evidence grade according to the Oxford Centre for Evidence-based Medicine Levels of Evidence¹⁸ (see Figure 1). Descriptive statistics were applied. Data were expressed as a median and range between parentheses unless otherwise stated. Data were pooled from the included studies and presented as totals and percentages.

Details of the initial search results and refined inclusions are presented in Figure 2. Of the 1,591 articles identified based on the above search strategy, 18 full-text articles were reviewed. Of these, 14 were excluded because they had a cohort of <10 patients,^16,19 the early ileostomy closure group was >14 days from index operation,^20-23 there were insufficient details to extract or calculate the necessary data from the published results,^24-28 the study pertained to early reversal of a colostomy (not an ileostomy),²⁹ or because they included reversal of jejunostomy or colostomy within the ileostomy cohort.^30-32

The remaining 4 studies ultimately analyzed, yielding 142 patients, included 2 retrospective case series^33,34 (Oxford level of evidence 4), 1 prospective nonrandomized study³⁵ (Oxford level of evidence 4), and 1 randomized controlled trial³⁶ (RCT) (Oxford level of evidence 1b). Characteristics of the included studies are outlined in Table 1.

All anastomoses were investigated radiologically before ileostomy closure to ensure anastomotic integrity. Patients noted as having anastomotic complications before closure did not have their ileostomy reversed early. Patients also were excluded from early closure if their recovery from their index operation was complicated by 1 or more of the following: sepsis or active infection,^35,36 organ failure,^34,36 poor postoperative recovery or condition (not further defined),^33,35 or prolonged postoperative ileus.³⁴

Patient characteristics. Patient characteristics are summarized in Table 2. A total of 142 patients underwent early closure of their ileostomy. Early closure population size ranged from 11 to 90 patients. Patient age ranged from 18–89 years old. The indication for ileostomy was reported in 3 studies^34-36; colorectal cancer was the most common indication for the index operation (78%).

Operative and hospital outcomes. The timing of ileostomy closure ranged from 8–14 days after index operation. Operative time was reported in 3 studies^34-36; median times ranged from 35–94 minutes. The median total length of stay was 1535  to 16 days,^34-36 and 1 study reported a mean total length of stay of 14 days.³³

Postoperative complications. Complications following early reversal of an ileostomy are outlined in Table 3. All studies reported 1 or more major postoperative complication but no deaths related to ileostomy closure. Complications necessitating reoperation occurred in 9 out of 142 patients (5.4%), with indications including SBO (n =1),³³ anastomotic leak (n = 3),³⁶ hemorrhage (n = 1),³⁶ rectovaginal fistula (n = 2),³⁶ and anastomotic stenosis (n = 2).³⁶

Wound infections were reported in 3 studies^32,34,35 and were the most common complication, affecting 26 out of 131 patients (19.8%). Four out of 112 patients (3.6%) experienced a stoma-related complication before closure, including obstruction or high ileostomy output. Five patients developed an enterocutaneous fistula from the ileostomy closure site, all managed nonoperatively. SBO or ileus occurred in 4.9% of patients. Before ileostomy closure, 6 patients had a false negative contrast study (missed anastomotic leak) and 3 subsequently required further intervention.

Loop ileostomies are created to protect distal anastomoses and reduce the impact of anastomotic complications. However, for a defunctioning ileostomy to have clinical utility, ileostomy-related complications (including those related to ileostomy closure) must not exceed the risk of anastomotic dehiscence. Despite the use of adhesional barriers³⁷ and laparoscopic techniques³⁸ during ileostomy closure demonstrating some reduction in perioperative complications, stoma-related morbidity remains significant for many patients.⁹

This study reviewed the literature to determine the effectiveness and safety of early closure of a defunctioning loop ileostomy and its effect in minimizing stoma-related morbidity. Overall, early closure is safe and feasible in patients without anastomotic or postoperative complications; when compared to a recent systematic review³⁹ evaluating 6,107 patients undergoing traditionally timed ileostomy closure, early closure is not associated with increased mortality or major complications. Elderly and medically comorbid patients were included in all studies evaluated, with 75% of all patients undergoing rectal resection with loop ileostomy in the RCT³⁶ qualifying for early closure.This suggests a significant number of patients may be candidates for early closure. However, further classification of eligible populations is required because inclusion and exclusion criteria often were poorly defined in the current studies.

Ileostomies usually are closed between 8 and 12 weeks.¹⁵ This delay is thought to not only allow sufficient time for patients to fully recover from their index operation, but also to reduce bowel friability and adhesions, enabling an easier ileostomy reversal.^9,16,28 However, the operating time for early ileostomy closure was comparable to a published series³⁹ on delayed ileostomy closure, suggesting a period of delay may not significantly improve ease of operation.

The overall reported mortality rate associated with early ileostomy closure was 0%, which compares favorably with a published series³⁹ on delayed closure that reports mortality rates of between 0% and 6.9%. Major complications, such as enterocutaneous fistulas and intra-abdominal abscesses,  result in substantial morbidity following ileostomy closure. Reported enterocutaneous fistula rates range from 0% as reported in the retrospective case series of 96 patients by Rullier et al⁴⁰ to 8.6% reported in the loop ileostomy arm of the RCT by Gooszen et al⁴¹(N = 29).  In addition, the 2 retrospective case series by Kaiser et al⁴² (N = 56) and Gunnarson et al⁴³  (N = 143) report rates of intra-abdominal abscesses following traditionally timed closure of 0% to 1.4%, respectively. In the current pooled analysis, enterocutaneous fistulas and intra-abdominal abscesses occurred in 3.5% and 1% of patients, respectively, rates that can be considered comparable with reported rates following delayed closure.

SBO and ileus often are reported as the most common complications following ileostomy closure.^28,39 However, there is considerable heterogeneity in the definitions of both, leading to considerable overlap in incidence. For the purposes of this review, these complications were pooled together to allow for indirect comparison with the current literature. In this review, the incidence of SBO/ileus following early ileostomy closure appears to be reduced when compared to contemporary literature,³⁹ occurring in only 4.9% of patients. This finding in part may be due to the problems of the definitions highlighted previously; however, in the single RCT³⁶ included in this review, SBO also was found to be significantly less common in the early closure (3%) compared to late closure (16%) group (P = 0.002), supporting the pooled data results. Furthermore, a recent analysis¹⁹ of 134 patients reported patients with a longer time interval before ileostomy closure had a significantly higher incidence of postoperative nausea, vomiting, and need for nasogastric tube insertion. In an in vitro study evaluating changes that occur in the distal limb of an ileostomy, Williams et al⁴⁴ found significant decreases in ileal smooth muscle contractility following a period of fecal diversion.; therefore, it was proposed that such changes may result in reduced compliance and contractility of the defunctioned bowel segment in vivo, leading to the obstructive symptoms commonly observed following ileostomy closure.

Wound infection was the most common complication found in this review, with an incidence of 19.8%. This is considerably higher when compared to the pooled analysis by Chow et al,³⁹ who reported a rate of 5%. Several factors may account for this difference. First, this could be a result of how wound infections are defined, because definitions of wound infection are known to vary widely between studies,  and despite the existence of specific standardized criteria for defining surgical site infections, none of the studies evaluated in this review utilized such definitions for reporting wound complications. Second, differences in skin closure technique have been demonstrated to have an impact on wound infection rate. A recent systematic review⁴⁵ evaluating the method of skin closure after loop ileostomy reversal found circular or purse-string closure was associated with the lowest risk of wound infection. In this review,^33-36 no studies standardized patients to purse-string closure because the majority of closures were left to surgeon preference. As such, purse-string closure could be added to a standardized protocol following early ileostomy reversal in order to reduce the incidence of this complication.

Another consideration is the relative immunosuppression known to occur following major abdominal surgery, predisposing patients to infection. Early ileostomy closure may occur during the postoperative period where immunity has not fully recovered. Improving postoperative outcomes for these patients poses a more difficult task. However, the development of enhanced recovery programs with the routine use of epidural anesthesia is thought to be beneficial in attenuating stress-induced immunosuppression in surgical patients, as shown in the pilot RCT by Ahlers et al,⁴⁶ which demonstrated a significant reduction in immunological alterations in patients receiving epidural anesthesia.

Perhaps the most important finding was the overall reduction in stoma-related morbidity during the presence of the ileostomy. According to the literature, stoma-related morbidity ranges anywhere from 13.3% to 74%,^9-11 and a large retrospective review47 of 603 patients with loop ileostomies found as many as 16.9% of these patients will require subsequent readmission. This poses a substantial burden to the patient’s quality of life as well as increased associated health care costs. In the current review, only 3.6% of patients undergoing early closure experienced a stoma-related complication, which is intuitive given the substantial decrease in time with a stoma. This reduction in stoma complications (and time having a stoma) is likely to have a positive impact on patient quality of life and also may lead to health care savings; therefore, further studies to accurately assess these outcomes are justified.

Colorectal cancer was the most common indication for the index operation in this review, representing 78% of all cases. An important consideration for these patients specifically is the administration of adjuvant chemotherapy. A 2005 review⁴⁸ evaluating adjuvant chemotherapy in colorectal cancer demonstrated systemic therapies significantly improve outcomes and can increase the likelihood of cure in patients with stage III colorectal cancer by 30%. Unfortunately, as 2 recent retrospective studies^9,49 have shown, patients receiving adjuvant chemotherapy often have their loop ileostomy closure significantly delayed and face postoperative complications twice that of patients that do not undergo adjuvant therapy. It is unclear whether this is due to the effects adjuvant therapy have on the patient’s general condition or because patients requiring adjuvant therapy usually have more extensive cancer, making them more prone to complications, as proposed by the retrospective study by Bakx et al.⁵⁰

It has been suggested early ileostomy closure could be utilized to reduce the uncertainty of timing of closure and decrease postoperative complications after reversal. A recent systematic review and meta-analysis⁵¹ has shown maximal benefit is achieved if adjuvant chemotherapy is initiated within the first 8 weeks after surgery. However, clinicians would first need to ensure complications after early ileostomy closure do not delay the time to initiation of adjuvant chemotherapy. Anastomotic leaks were missed in 6 patients undergoing early ileostomy closure. This is consistent with the false negative rates reported in the literature⁵²; in the RCT, no statistically significant differences in incidence between early and late closure groups were noted. However, in the context of administering adjuvant chemotherapy major complications such as missed leaks could delay time to treatment. It is not clear from this review if early closure would result in such delays; further studies are required to evaluate the advantages and disadvantages of early ileostomy closure in relation to adjuvant chemotherapy.

The conclusions that can be drawn from this review are limited by the quality and quantity of available evidence. Only 1 RCT has been performed evaluating early ileostomy closure. This provided the largest patient contribution to the pooled analysis, with smaller retrospective and prospective studies contributing the balance. These smaller studies are prone to selection bias, reducing the validity of their reported controls and precluding a further pooled analysis of well-matched control groups. This necessitated indirect comparisons between the pooled early closure outcomes and the current literature evaluating traditionally timed closure. Essential data were not reported in all studies (such as follow-up time, patient comorbidities, indication for stoma, and wound infection rates) and considerable heterogeneity existed between studies in terms of defining patient selection and outcomes. Therefore, the results of this systematic review should be interpreted with these limitations in mind.

This systematic review identified 4 studies (2 retrospective case series, 1 prospective non-randomized study, and 1 RCT), yielding a pooled population of 142 patients undergoing early ileostomy closure. From the data available, early ileostomy closure is possible in select patients with some significant advantages. This review demonstrated patients having early ileostomy closure experience a reduced incidence of postoperative SBO/ileus and stoma-related complications, with these complications occurring in only 4.9% and 3.6% of patients, respectively. However, these advantages must be weighed against an increased rate of wound infections, with wound infections occurring in 19.8% of early closure patients. Additional large RCTs are required to accurately define which individuals stand to benefit from early closure, as well as to further evaluate the impact of early closure on quality of life, health care costs, and the administration and complications of adjuvant chemotherapy.

Dr. Robertson, Dr. Puckett, and Dr. Vather are general surgery trainees and researchers, and Dr, Jaung is a researcher, Department of Surgery, University of Auckland, New Zealand. Dr. Bissett is a colorectal surgeon and Head, Department of Surgery, Auckland City Hospital; and Associate Professor of Surgery, University of Auckland. Please address correspondence to: Jason P. Robertson, MBChB, University of Auckland, Department of Surgery, Private Bag 92019 Auckland Mail Centre 1142 New Zealand; email: jrobertson171@gmail.com.