A Retrospective, Observational Study of the Adequacy of Elective Loop Stoma Diversion

Diverting stomas are employed for various clinical indications and easier to revert than end stomas. General, plastic, and colorectal surgeons continue to debate whether a diverting loop stoma adequately diverts stool, preventing spillage into the distal limb, which does not receive stool. A retrospective, descriptive study was conducted involving all patients who underwent loop ostomy surgery (colostomy or ileostomy) — defined by current procedural terminology (CPT) codes 44187, 44188, 44153, and 44155–44158 — between April 1, 2002 and October 12, 2010.The purpose of the study was to determine if a diverting loop stoma is completely diverting with no efflux into the distal limb by examining the rate of distal limb contamination. Two surgeons identified patients at the University of Virginia who had loop ostomy surgery and subsequently underwent computed tomography (CT) scan with oral contrast for clinical suspicion of abnormal pathology. A radiologist reviewed these images to evaluate the presence or absence of oral contrast in the ostomy distal limb. Distal progression of oral contrast was deemed adequate if oral contrast was visualized within the ostomy bag or if contrast was intraluminal distal to the loop ostomy. The loop ostomy was considered diverting if oral contrast was only visualized in the ostomy bag with no oral contrast distally. Of the 202 eligible patients, 26 (13%) underwent 41 postoperative CT scans of the abdomen/pelvis. Four (4) were excluded due to inadequate exam or confounding contrast (rectal contrast, contrast retention from preoperative scan). Of the remaining 22 patients with 35 CT scans (median age 54 [range 26–82] years, 10 men, 18 Caucasian, 18 having elective surgery), no patient (0%) had evidence of distal contrast. In this and other studies, loop stomas were found to provide adequate diversion without spillage into the nonfunctional limb for the vast majority of patients and should strongly be considered as the procedure of choice for temporary diversion.

Colon and rectal surgeons commonly create a protective loop ostomy in order to decrease rates of clinically significant anastomotic leaks in colorectal procedures. Improved techniques in colorectal surgery allow for low (below the anterior peritoneal reflection) and ultra low (1.5 cm–2 cm above the dentate line) anastomoses in lieu of abdominoperineal resections.¹ A low anastomosis allows patients to preserve bowel function without the need for a permanent ostomy. Unfortunately, these low anastomoses come with significant leak rates. Numerous descriptive studies and a randomized, controlled prospective study have demonstrated improved outcomes with a temporary defunctioning stoma when performing a low anastomosis, including lower rates of leakage with diverting ostomies in addition to lower rates of re-intervention with loop stomas.^2-6

Initially popularized to avoid ischemia associated with ileal conduits and urostomies due to mesenteric undercutting,⁷ loop ostomies have increased in prevalence in colorectal procedures over the past 25 years for anastomotic protection.^8,9 A literature review demonstrated fecal diversion decreased the risk of anastomotic leaks by diverting stool proximally; furthermore, diversion reduced the need for urgent re-operation when a leak occurs.^9,10 

Loop ostomies are technically easy to create and are described in several retrospective reviews. Traditionally, the ostomy is sited within the rectus abdominus for loop ostomy formation.^8,9 The mesentery of either the ileum or colon is mobilized and the antimesenteric border is grasped with tissue forceps.⁸ An enterotomy then is created, and the proximal end is anchored with mucocutaneous sutures. An enterotomy that encompasses 4/5 of the circumference will allow the distal bowel to “escape” from the skin.⁹ In addition, loop ostomies also are easy to reverse, especially with the advent of bowel staplers. End ostomies require large laparotomy incisions for reversal, whereas loop ostomies can be reversed directly through the ostomy incision, which should decrease morbidity.^7,11

General, plastic, and colorectal surgeons debate the success of fecal diversion with loop stomas and whether they allow fecal contents to spill into the distal limb. Traditional teaching states loop stomas do not divert as adequately as end ostomies.¹² The colorectal surgery division at the University of Virginia (UVA, Charlottesville, VA) has received referrals for conversion of loop ostomies to end ostomies to prevent sacral wound contamination, which in turn would allow for proper wound healing. To support the hypothesis that all loop ostomies are diverting, a retrospective study was conducted to determine if loop ostomies adequately divert stool without distal limb contamination, as shown via oral contrast and computed tomography (CT) of the abdomen/pelvis. The purpose of this retrospective study was to examine the rate of distal limb contamination following elective loop ostomy surgery.

Patient selection. The institutional review board at the UVA Health System approved this retrospective, observational study with a waiver for written informed consent. The general surgery and colorectal databases were queried for all patients undergoing a colorectal procedure between April 1, 2002 and October 12, 2010. Patients were eligible if they underwent ostomy creation and a subsequent CT scan in the study period. Patients undergoing any colorectal procedure between the specified dates had further chart review performed by a surgery resident and attending surgeon to determine the presence or absence of a loop ostomy (ileostomy or colostomy). Patient name, race, gender, medical record number, procedure type, date of operation, elective or emergent operation, current procedural terminology (CPT) code, and reason for surgery were abstracted and recorded in a separate database for all patients. Ostomy reversal was not abstracted, since it was not relevant for the purposes of the current study. Patients undergoing loop colostomy or ileostomy procedures were defined by CPT codes 44187, 44188, 44153, and 44155–44158. 

Information on type and date of CT scan, if any, was collected along with any recent non-oral contrast administration. CT scans were ordered at the discretion of the treating physician based on clinical suspicion for abnormal pathology or for cancer surveillance. Scans were appropriate for inclusion if they occurred after ostomy creation but before stoma reversal. If multiple scans were ordered for a patient, all eligible scans during the study period were included. Chart review was also essential to determine which patients underwent postoperative CT scans of both the abdomen and pelvis with oral contrast (iohexol). A radiologist independently retrieved and reviewed the scans and corroborated the presence or absence of distal contrast. 

This review was performed independently of the clinically indicated image review and was performed for study purposes only. Technically incomplete studies (ie, imaging of only the abdomen or pelvis, cessation of imaging after the scout acquisition and before cross-sectional imaging, or oral contrast not reaching the level of the ostomy) were excluded. Additionally, patients were excluded if they received rectal contrast or oral contrast within 1 week before creation of the loop ostomy. 

Surgical technique. Standard procedure, diverting loop ileostomies or colostomies were created under the supervision of 1 of 8 attending surgeons. One patient had a loop ileostomy created at another institution that was revised by one of the UVA surgeons during a concurrent operation. In general, patients presenting for elective ostomies undergo ostomy site marking by a specified ostomy team before surgery. Bowel preparation with polyethylene glycol for elective patients before colonic surgery was widely practiced during the study period. Emergent ostomies were placed in a location chosen by the attending surgeon. 

A muscle-splitting incision through rectus abdominus muscle and fascia is made at the preferred site. A loop of small or large bowel then is grasped with Babcock forceps and passed through the incision. Bovie electrocautery is used to create a small enterotomy. Typically, loop ileostomies are created using a Brooke technique, whereas loop colostomies are sutured to the skin. A Brooke ileostomy involves suturing a portion of the ileum to the skin so an inside-out portion of ileum serves as the ostomy.¹³ Surgeons at the authors’ institution classically place a supporting rod underneath the ostomy if any tension is noted during creation of the stoma. Stoma appliances then are fitted in the operating room. 

Examination technique. CT scans were performed at the discretion of the physician with a standardized protocol (single-breath hold, spiral-acquisition, 5-mm reconstructed images). Examinations with and without intravenous contrast medium were included. Oral contrast protocol recommends patients receive one oral dose as soon as the CT order is placed; dose 2 is given 30 minutes after completion of the first dose. The third and final dose of oral contrast is administered 15 minutes before transport to CT. Oral contrast dosing was variable given unpredictable patient wait times for an available scanner. A radiologist reviewed the scans in 3 planes on the hospital imaging system (PACS, Carestream Vue PACS, Rochester, NY). Both diagnostic and cancer surveillance scans were used in the study analysis.

Imaging and data analysis. Each CT was examined to determine the most distal extent of oral contrast progression. Distal progression of oral contrast was deemed adequate if oral contrast was visualized within the ostomy bag or if contrast was intraluminal in bowel distal to the loop ostomy. The loop ostomy was considered diverting if oral contrast was visualized only in the ostomy bag with no oral contrast distally. If oral contrast was visualized in the distal bowel, with or without contrast in the ostomy, the ostomy was considered incompletely diverting.

Data recording and analysis. Data were abstracted from retrospective chart review and maintained in a separate Excel database. Because there is no comparison group, no formal statistical analysis was performed. Descriptive summaries are provided for each variable, including overall number of patients and percent of total.

During the study period, the records of 202 patients who underwent a loop colostomy or ileostomy were identified. Of these, 40 underwent a CT exam of the abdomen or pelvis with a total of 77 radiographic studies. Chart review and analysis of the operative reports excluded 1 patient with 2 CT exams who underwent a divided end ostomy. The radiologist reviewed the remaining CT examinations of 39 patients who underwent 75 studies. Of those, 30 studies were excluded due to technically incomplete examinations. An additional 2 patients were excluded for receiving a water-soluble contrast enema and not oral contrast. Lastly, 5 patients with 6 CT scans were excluded for confounding oral contrast (recent preoperative CT with oral contrast and recent rectal contrast). Ultimately, 22 patients (median age 54 [range 26–82] years) with 35 CT studies had adequate examinations for inclusion: 9 patients (41%) had loop colostomies — 7 patients with descending colostomies underwent 11 CT scans (2 patients with ascending colostomies underwent 1 scan each). The remaining 22 scans were from 13 patients (59%) with loop ileostomies (see Table 1). 

The study included an approximately equal number of men and women. The majority of patients were Caucasian (82%), and most stomas were created under elective conditions (82%). Seven (7) patients with a colostomy had a sigmoid colostomy; 2 patients had a right colostomy (see Table 1). 

None (0) of the 22 patients undergoing 35 CT scans of the abdomen/pelvis with oral iohexol contrast administration was identified as having contrast in bowel distal to the level of the loop ostomy, regardless of reason for ostomy formation. 

Some colorectal surgeons have received referrals, particularly from plastic surgeons, requesting conversion of a loop to an end ostomy. The referrals frequently come with complaints of anal or rectal leakage, even with a diverting loop stoma. The concern is that the discharge may actually be nondiverted stool, causing contamination of wounds or flaps created in previous procedures. It is known that patients with diverting ostomies, even end stomas, continue to pass mucous discharge, which occurs even in the presence of true fecal diversion.¹⁴ Additionally, the drainage may result from diversion colitis, a term coined by David Glotzer,¹⁵ who first described this phenomenon in 1981 based on a 10-patient sample with inflammatory findings on proctoscopy. A systematic review by Kabir et al¹⁶ of 22 case series, 9 case reports, 2 retrospective, and 2 prospective studies qualitatively demonstrated diversion colitis occurs in colonic segments distal to the diversion and is thought to be a nonspecific inflammation occurring in up to 33% of patients with a diverting ostomy. Diversion colitis can be asymptomatic or can cause symptoms ranging from pelvic/abdominal pain and tenesmus in up to 15% of patients, bloody discharge or mucous discharge in up to 40%, and anorectal pain. 

Few studies have evaluated fecal diversion with loop ostomies. Winslet et al⁸ prospectively evaluated the effectiveness of the defunctioning capacity of the loop stoma by administering radioactive dye orally followed by rectal lavage with a Foley catheter to assess the presence of dye. The authors found a 99% effectiveness of complete diversion in 18 patients. Similarly, in a 10-patient prospective cohort Schofield et al¹⁷ were unable to detect any radiolabeled dye in the distal limb, concluding loop stomas are adequately diverting. Morris et al¹² prospectively assessed diversion through administration of oral barium in 23 patients with loop colostomies followed by mechanical bowel preparation to determine if barium passed to the distal limb; similar to the current study, no patient was found to have barium in the distal portion of bowel.

In this study, CT scans of the abdomen/pelvis were evaluated in 22 patients who had previously undergone diverting stoma and the presence of contrast distal to the ostomy was used as an indicator of distal spillage of stool. The current study demonstrates no patient with either loop ileostomy or loop colostomy had evidence of contrast distal to the anastomosis, regardless of reason for operation. Patients who had distal contrast received oral or rectal contrast shortly before the CT in question and were therefore appropriately excluded from analysis. 

 This retrospective, observational study has important limitations. The CT scans excluded for confounding oral contrast may have had contrast in the distal limb from post-ostomy formation iohexal. Although this is a possibility, it seems unlikely because no other patients had distal contrast in the nonfunctional limb. Additionally, the current study included a small sample size of 22 patients, which may limit the ability to identify distal contrast. Furthermore, newer imaging techniques may have greater sensitivity to detect distal contrast or radioactive tracer. The timing between administration of iohexol and initiation of CT imaging is unknown. Lastly, as a single institution, retrospective study, the generalizability to other patient cohorts is limited. 

The results from this and previous research confirm that loop stomas are adequately diverting. Given the ease of reversal, they serve as a preferred method for temporary diversion.

This research was funded by the National Institutes of Health Grant #Surgical Oncology T32 CA163177. The funding was used to provide salary support for the first author. The funder had no role in study design, data collection, or manuscript drafting.

Dr. Shah is a general surgery resident, University of Virginia, Charlottesville, VA. Dr. Mauro is an internventional radiology fellow, Mount Sinai Hospital, New York, NY. Dr. Friel and Dr. Hedrick are colorectal surgeons, University of Virginia.

Please address correspondence to: Puja M. Shah, MD, University of Virginia, Department of Surgery, PO Box 800679, Charlottesville, VA 22908-0679; email: ps5au@virginia.edu.