32 Intestinal Stomas
Abstract
Abdominal stomas serve a critical function in the management of benign and malignant diseases. Currently, 130,000 patients undergo ostomy surgery in the United States annually. Many patients have major complications from the operation and problems with the stoma. The judicious assessment of the need for the stoma, careful surgical technique, and skilled wound ostomy continence or enterostomal nursing are essential for a satisfactory outcome.
32.1 Introduction
Although the advent of restorative proctocolectomy and stapled low anterior anastomosis has lessened the need for permanent ileostomy and permanent end colostomy, the abdominal stoma still serves a critical function in the management of benign and malignant diseases. Currently, approximately 130,000 patients undergo ostomy surgery in the United States, annually. 1 Many of these patients will have major complications from the operation, and all of them will experience at least some problems, as a result of the stoma. These problems are biopsychosocial in nature and may include surgical complications; peristomal skin complications; and problems with odor, noise, anxiety, depression, sexual dysfunction, and social isolation. 2 , 3 , 4 The judicious assessment of the need for the stoma, careful surgical technique, and skilled wound ostomy continence (WOC) or enterostomal nursing are essential for a satisfactory outcome.
32.2 Ileostomy
32.2.1 History
It was Baum who performed the first ever ileostomy in 1879 on a patient with an obstructing carcinoma of the ascending colon. The patient survived the ileostomy procedure but died from complications of a second operation that involved resecting the primary carcinoma and creating an ileocolic anastomosis. Kraussold, Billroth, Bergman, and Maydl were among the other 19th century surgeons who subsequently performed the ileostomy surgery. Maydl’s patient is generally considered to be the first to survive and fully recover from an ileostomy performed in combination with resection for colon carcinoma. 5
In 1913, Brown 6 reported on 10 patients in whom he constructed an ileostomy. He made the stoma as an end ileostomy, closing off the distal end and bringing 2 in. of the proximal end out through the lower part of the incision. He inserted a catheter into the lumen and allowed it to separate as the surrounding tissue sloughed, and the bowel became granulated, contracted, and eventually formed a mucocutaneous junction with the abdominal wall skin. Brown achieved a favorable result in all of his patients.
The results of ileostomy surgery as reported by subsequent authors were uniformly disappointing. A variety of techniques were introduced, including the method of placing the terminal ileum through a separate right lower quadrant incision as an end ileostomy and various loop ileostomy techniques. The most significant achievement during this period, however, occurred in the late 1920s, when Dr. Alfred Strauss of Chicago performed an ileostomy and later a colectomy for ulcerative colitis on a young chemist named Koenig. Koenig subsequently designed an ileostomy appliance bag, commercially known as the Koenig-Rutzen bag. This appliance was used widely until the introduction of karaya in the 1950s.
In the following years, ileostomy-related fluid and electrolyte problems were identified and the need for earlier operation in patients with severe ulcerative colitis was recognized. The problem of high ileostomy output remained an obstacle to full rehabilitation for many patients until the 1950s, when Warren and McKittrick 7 related the problem of ileostomy dysfunction to partial obstruction of the stoma at the level of the abdominal wall. A few years later, Crile and Turnbull 8 showed that excessive ileostomy fluid loss in the early postoperative period resulted from serositis of the exposed ileal serosal surface. The authors subsequently showed that covering the exposed serosa with an everted layer of mucosa could prevent or minimize ileostomy dysfunction. The simplified full-thickness eversion technique described by Brooke 9 has since gained universal acceptance. Later technical developments include Goligher’s description of the extraperitoneal ileostomy in 1958 10 and the description by Kock et al 11 of the continent ileostomy in the early 1970s.
Two events that were critical to the success of ileostomy surgery and the rehabilitation of the ileostomy patient were the formation of the first ileostomy club at Mount Sinai Hospital in New York in 1951 and the initiation of an educational program for enterostomal therapists (later to be known as enterostomal therapy [ET] nurses and subsequently as wound ostomy continence nurses [WOCN]) by Rupert Turnbull at the Cleveland Clinic in 1961. Today the United Ostomy Associations of America comprises over 300 affiliated support groups. 1 The specialty of WOCN has matured and expanded as well with the opportunity of these specially trained nurses to become board certified. For more than 30 years, over 7,600 dedicated nurses have been certified ( http://www.wocncb.org/ ; accessed December 6, 2016).
32.2.2 Stomal Physiology
Physiologically left-sided colostomy output is very similar to that of normal bowel movements. There are essentially no significant physiologic abnormalities associated with left-sided colostomy; for that reason the physiology segment of this chapter will be dedicated to ileostomy output.
Ileostomy output clearly changes as the patient recovers from the operation and appears to have three distinct phases of adaptation. In the first 3 days, the output is bilious and liquid in nature and increases daily with the maximum output occurring on the third or fourth day. During the second phase, which occurs from the fourth to the sixth day following operation, the output stabilizes, thickens, and even decreases slightly. Phase III, adaptation, occurs from the first week to the eighth week following operation and is associated with a steady decrease in volume and thickening of the stoma output. 12 , 13 After complete adaptation, the output from an end ileostomy created without significant ileal resection stabilizes between 200 and 700 mL/day.
Tang et al 12 studied the ileostomy output of 60 patients who underwent restorative proctectomy with defunctioning ileostomy. By the fourth postoperative day, 65% of patients had a functioning ileostomy. Ileostomy output peaked at the fourth day with a median of 700 mL (range, 10–3,250 mL) per 24-hour period. Output decreased after the fifth day, and by the 10th postoperative day, a median of 300 mL (range, 100–750 mL) per 24-hour period was reached despite normal intake. The authors noted that the critical period for acute dehydration was the third to eighth day and recommended aggressive fluid and electrolyte replacement. They therefore suggested that patients should not be discharged before 9 or 10 days after ileostomy construction. Modern surgical practice in the United States has many of these patients leaving the hospital at 2 to 6 days, stressing the importance of patient education and home health care. Small bowel adaptation following ileostomy creation results in increased reabsorption of water and electrolytes. In a normal individual, between 1,500 and 2,000 mL/day, the terminal ileum exits into the colon. With ileostomy adaptation, 70 to 80% of this output is reabsorbed.
Historically, ileostomies were created without eversion and matured spontaneously in the first 1 to 2 weeks following operation. This was commonly associated with ileostomy dysfunction. In the 1950s, Warren and McKittrick related ileostomy dysfunction to partial obstruction due to stomal edema. 7 This stomal edema was further clarified by Crile and Turnbull 8 when they showed that ileostomy effluent damaged the ileal serosa leading to serositis and subsequent edema and ileostomy obstruction. The primarily “matured” ileostomy described by Brooke has essentially eliminated ileostomy dysfunction. 9
Most patients are able to tolerate an unrestricted diet following postoperative recovery. Dietary changes have little effect on ileostomy output with the exception that fasting can decrease ileostomy output to between 50 and 100 mL/day. Studies have shown that an elemental diet will decrease volume and concentration of digestive enzymes as well as bile acids. 14 Diets high in fat, perhaps due to fat malabsorption and alterations in bile salt circulation, have been shown to increase stoma output to 20% above baseline. 14 , 15 In addition, increases in oral fiber intake greater than 16 g/day have also been shown to increase output, stool frequency, and flatus. 16
Nutrition
Individuals with ileostomies associated with less than 100 cm of resected terminal ileum are able to maintain essentially normal nutrition. Although rare, fat malabsorption, from impaired bile salt absorption or reduced bile salt pool, may cause osmotic diarrhea. Bile salt output may increase either by absorptive inhibition or perhaps by direct secretory stimulation on intestinal mucosa. 14 , 17 Lactase deficiency and/or enterokinase deficiency may be more evident in individuals with an ileostomy. In the absence of significant terminal ileal resection, nutritional consequences of ileostomy are minimal and patients are able to maintain a normal weight and body composition. Resection of greater than 100 cm of terminal ileum or bacterial overgrowth may result in B12 malabsorption, and vitamin B supplementation may be necessary to prevent megaloblastic anemia.
Metabolic Changes
Normal individuals lose between 2 and 10 mEq of sodium in their stool on a daily basis, while the ileostomates lose approximately 60 mEq/day. In the review by Soybel, 13 the daily steady state of fecal losses following ileostomy compared to normal stool were water, 100 to 150 mL versus 650 mL; sodium, 1 to 5 mmol versus 81 mmol; potassium, 5 to 15 mmol versus 6 mmol; and chloride, 1 to 2 mmol versus 34 mmol. Symptomatic salt depletion, however, is rare due to renal compensation. Individuals with ileostomies have been shown to have chronically elevated mineralocorticoid levels that increase water and sodium reabsorption, compensating for the increased losses in the stool. Renal compensation combined with a normal diet makes chronic dehydration and salt depletion a rare circumstance in individuals with well-adapted ileostomies. Additionally, calcium and magnesium levels are unaffected unless extensive terminal ileum has been resected.
Although the sodium concentration of ileostomy fluid rises and falls in relation to total body sodium levels, the usual sodium concentration is about 115 mEq/L. Others have reported the intestinal loss of sodium in patients with a conventional ileostomy to be 62 mmol/24 hours. (For univalent molecules [e.g., Na +, K + ] 1 mEq/L = 1 mmol/L; for divalent molecules [e.g., Ca + +, Mg + + ] 2 mEq/L = 1 mmol/L.) With dehydration, the sodium concentration falls and the potassium level rises; changes in the sodium/potassium ratio reflect the participation of the terminal ileum in sodium conservation during times of salt depletion. Normally, the sodium/potassium ratio is about 12 and rarely rises above 15. 9 The pH is generally on the acidic side, just slightly below 7.
The bacteriologic composition of the ileostomy fluid is different from that of normal small or large intestine. Gorbach et al 18 showed that the total number of bacteria in ileostomy effluent is approximately 80 times greater than that in the normal ileum. The ileostomy effluent contains a 100-fold increase in the number of aerobes, a 2,500-fold increase in coliform bacteria, and increases in total anaerobes compared to the normal ileum.
The single most important factor influencing ileostomy output is the amount of uninjured intestine proximally. 13 Increased body mass is associated with increased output. Water intake plays virtually no role. Elemental diets are associated with decreased volume, while high fat content is associated with higher losses. Increases in fiber content increase ileostomy output by as much as 20 to 25% when dietary bran supplementation exceeds 16 g/day. Losses of nutrients such as carbohydrates and protein are also increased with diets high in fiber or fat. Overproduction of gastric acid contributes to the volume of ileostomy output. Ileostomy output decreases when patients receive antisecretory agents such as omeprazole, but such benefits are reaped only with patients with large ileostomy output (22.61 L/day) or those who have had significant amounts of small bowel resected.
Intestinal transit does slow in individuals with end ileostomy. Using radioisotopes, Soper et al demonstrated that gastric emptying was normal. However, oral stomal transit time was significantly increased compared to normal controls (348 vs. 243 minutes). 19 Similar findings were observed in a study of transit times in individuals undergoing proctocolectomy and ileostomy. 20 Similarly, Bruewer et al demonstrated, by lactose breath test, that oral pouch transit time was prolonged in individuals undergoing proctocolectomy and ileal pouch–anal anastomosis. 21 Adaptation occurs over a period greater than 1 year and the mechanism remains unknown. Theories include epithelial hypertrophy with an increased absorptive surface. 22 In addition, an inverse relationship between nutrition and electrolyte absorption and intestinal transit time has been identified. 23
Bacterial flora in the ileostomy approaches that of the colon. Gorbach et al found an 80-fold increase in total number of organisms in the terminal ileum. 18 Specifically coliforms were 2,500 times more common than normal ileal fluid. However, ileostomy bacterial content was still considerably less than that of normal stool. Bacteroides fragilis, a normal inhabitant of colonic stool, was rarely found in ileostomy output.
Individuals with inflammatory bowel disease, particularly associated with ileal resection and/or ileostomy, are at an increased risk for urinary stone formation. The incidence in these individuals ranges from 3 to 13%, whereas it is 4% in the normal population. 24 Uric acid stones are rare in the normal population, but comprise 60% of stones in individuals with an ileostomy. 25 Low urinary volume and pH combined with increased concentrations of calcium and oxalate are thought to be responsible for this phenomenon. 26 In addition, decreased urinary volume and decreased urinary pH facilitate stone precipitation. 26 Prophylaxis and treatment consist of increasing daily fluid intake to increase urinary volume and urinary pH.
The association between ileostomies and gallstone formation has not been well established. Individuals with well-adapted ileostomies are found to have bile acid secretions similar to those with an intact colon. 27 , 28 However, significant terminal ileal resection or extensive terminal ileal disease does affect the enterohepatic circulation. Bile absorption and/or depletion may occur, resulting in increased saturation of bile that leads to bile salt precipitation and subsequent stone formation. Despite this, Ritchie found no increase in incidence in cholecystectomy rates in ileostomy patients compared with that in the normal population. 29 However, Kurchin et al recommended prophylactic cholecystectomy in female individuals undergoing proctocolectomy or small intestinal resection for inflammatory bowel disease because of a threefold increase in asymptomatic gallstones. 30
Ileostomy Dysfunction
Significant diarrhea may occur in individuals with ileostomies for many reasons. It develops more rapidly and has greater physiologic consequences than when it occurs in individuals with a normal colon. Diarrhea may result from significant ileal resection, partial small obstruction, bacteria overgrowth, recurrent or persistent regional enteritis, or infections. In most cases, treatment is similar to those given to individuals with an intact colon. Specific attention must be paid to fluid and electrolyte repletion, however, because these individuals are at significant risk for dehydration and metabolic abnormalities.
Kusuhara et al 31 found that the administration of a somatostatin analog, SMS 201–995 (100 mg thrice a day for 5 days), reduced the daily output of proximal ileostomies from 997 to 736 g along with a decrease in daily sodium and chloride excretion. The authors suggested a possible role for this agent in the management of a proximal ileostomy.
Bacterial overgrowth may develop when stasis occurs in the small bowel and can lead to deconjugation of bile salts and subsequent osmotic diarrhea as well as B12 malabsorption. Unconjugated bile salts impair sugar, water, sodium, and potassium reabsorption in the small intestine. 32 In addition, anaerobic bacteria bind to the B12-intrinsic factor complex inhibiting reabsorption. Treatment is directed at eliminating the cause of stagnation in the small intestine and, if appropriate, antibiotics are used to decrease bacterial load.
32.2.3 Types of Ileostomies
End Ileostomy
An end ileostomy is indicated after a total proctocolectomy for inflammatory bowel disease or familial polyposis, and in some cases after more limited resection for other colonic diseases. The procedure is rarely performed today without concomitant resection. It is potentially reversible when it has been performed with a partial or total colectomy (when the hazards of primary anastomosis between the ileum and colon or rectum were considered to be unacceptably high), such as in cases of ischemic infarction, peritonitis, or severe nutritional deficiency.
When the ileostomy is performed for benign inflammatory conditions, it is desirable to preserve as much ileum as possible; consequently, the ileum should be divided within 1 to 2 cm of the ileocecal valve. If the resection is performed for benign conditions, it is possible to preserve the ileocolic artery, although collateral flow from the next proximal mesenteric vessel is sufficient to nourish the entire distal ileum.
Once the blood supply and mesentery have been taken, the ileum is divided between either staples or intestinal clamps. The small mesenteric vessels and fat are trimmed proximally for a distance of 5 to 6 cm. A 1-cm strip of mesentery is left attached to the terminal bowel in order to prevent ischemia. A 2.5- to 3-cm circumferential incision is made in the skin around the previously marked site in the right lower quadrant (▶ Fig. 32.1a). Care should be exercised not to remove excess subcutaneous fat because its presence is beneficial in the support of the ileostomy appliance. 33 The fascia is exposed, and a vertical incision is made through the fascia (▶ Fig. 32.1b). The rectus muscle is spread apart with a hemostat or scissors, and the incision is carefully extended through the posterior rectus sheath and peritoneum, with electrocautery or scissors (▶ Fig. 32.1c). The opening is extended to produce a defect of 3 to 4 cm that corresponds to the width of the tips of the index and middle fingers of a surgeon with average-size hands.
The ileum is drawn through the circular incision in the right lower quadrant so that the ileal mesentery is oriented cephalad, toward the diaphragm (▶ Fig. 32.1d). Exteriorization of 5 to 6 cm of ileum should be accomplished (▶ Fig. 32.1e). The mesenteric defect may be closed by suturing the cut edge of the mesentery to the anterior abdominal wall 2 cm lateral to the midline incision from the ileostomy site to the ligamentum teres (▶ Fig. 32.1f). The primary Brooke-type maturation of the ileostomy can either be performed at this point or after the abdominal wall has been closed. The advantage of waiting until the abdomen is closed is that it prevents the spilling of ileostomy contents into the abdominal cavity and onto the wound. The advantage of maturing the stoma immediately is that it allows the surgeon to make adjustments, as needed, if inspection of the ileostomy suggests that its viability is tenuous. The mucosal surface should always be pink and should bleed freely at its edges. The primary maturation is performed by placing interrupted three-point buttressing sutures of absorbable suture between the tip of the ileum, the seromuscular layer of the ileum at the level of the fascia, and the deep dermis (▶ Fig. 32.1g). Interrupted sutures connecting the full-thickness bowel wall directly to the dermis of the skin are then added to secure the ileostomy (▶ Fig. 32.1h). The ileostomy should protrude at least 2 cm beyond the abdominal wall (▶ Fig. 32.1i). An appliance bag is fitted immediately in the operating room.
The Brooke ileostomy is now the most widely used technique for the construction of an end ileostomy. In most patients, it is relatively simple to perform and problems with retraction, serositis, or obstruction are uncommon. In some patients, however, the terminal ileal segment may be so edematous that full-thickness eversion using the Brooke technique is not possible. In these instances, the Turnbull technique may be useful. 8 This involves resecting the distal serosa and muscularis and everting the mucosa over the proximal ileum.
The Guy rope suture technique is an alternative method of creating eversion in a thickened terminal ileum without the need for seromuscular resection. With this technique, three permanent 3–0 absorbable sutures are placed between the distal edge of the ileum and the full thickness of the skin. In addition, three temporary internal Guy traction sutures are placed 1.5 to 2 cm apart inside the ileal lumen. Upward traction on the Guy sutures with simultaneous outward traction on the permanent external sutures helps evert the thickened segment without damage or tearing of the mucosa. 34 Carlsen and Bergan 35 reviewed the complications of 358 end ileostomies: 224 were primary constructions, 96 were reconstructed by laparotomy, and 38 were local reconstruction. Only two ileostomies were primarily located on the left side. The mean length was 5 cm. The authors performed 11.6% of reoperations after primary ileostomy and 7.3 and 7.9% of reoperations after reconstruction by laparotomy and local approach, respectively. There were 12.9 and 8.7% reoperations after emergency and elective primary operations, respectively. Closing the lateral gutter or fixation of ileum to the rectus fascia did not significantly influence the number of reoperations. Postoperative discoloration of the ileostomy was not predictive of ileostomy dysfunction. Stenosis of the ileostomy, peristomal fistulas, and peristomal dermatitis were seen in 23 (10.3%), 21 (9.4%), and 18 (8%) of the patients after primary ileostomies, respectively. Patients with Crohn’s disease had significantly more of these problems than patients with ulcerative colitis. Only a few patients had retraction of the ileostomy (2.7%), stomal prolapse (1.8%), or parastomal herniation (1.8%). Women had significantly more parastomal herniation than men; otherwise, there were no differences between the sexes.
Loop Ileostomy
The loop ileostomy is used to protect an ileoanal or high-risk colonic anastomosis. Although a loop transverse colostomy had traditionally been used to protect a low-lying colorectal anastomosis, Williams et al 36 recommended the loop ileostomy whenever a stoma is needed to divert stool from the distal colorectum. The authors compared the loop ileostomy to the loop colostomy and found that the ileostomy had significantly less odor, required significantly fewer appliance changes, and had fewer overall problems than the colostomy. Others have advocated loop ileostomy as a superior method for diverting the bowel to cover a colonic or rectal anastomosis and found septic complications following ileostomy closure less frequent than those following colostomy closure. However, postoperative bowel obstruction and other serious complications such as severe dehydration requiring antidiarrheal medication and hospitalization, skin breakdown, and even cholelithiasis may occur, rarely. 37
Winslet et al 38 assessed the defunctioning efficiency of the loop ileostomy using a radioisotope and dye technique. The median defunctioning capacity in patients without episodes of fecal discharge per rectum (n = 18) was 99% and was not affected by body position or the formation of a dependent stoma. In four patients who passed fecal material per rectum but who had no stomal retraction, the median defunctioning efficiency was 99%, and continued fecal discharge was considered to be due to mucopurulent secretion from active distal disease. In four patients who passed fecal material per rectum and also had a retracted stoma, the defunctioning efficiency was significantly reduced (median, 85%), owing to the overspill into the distal limb. Chen and Stuart 39 reviewed the choice of a defunctioning stoma in restorative resection for colorectal carcinoma. The authors found the morbidity of stoma construction and closure comparable, but favored loop ileostomy because it is generally easier to manage.
Loop ileostomies are most commonly created at the time of laparotomy and bowel resection; therefore, exposure is generally through a midline incision. A segment of ileum several centimeters proximal to the ileocecal valve (which reaches the predetermined stoma site without tension; ▶ Fig. 32.2a) is chosen and encircled with a Penrose drain or umbilical tape or grasped with a Babcock clamp. The distal limb is tagged with a suture or scored to ensure correct orientation. A disc of skin is excised and a defect in the abdominal wall is created similar to or slightly larger than that of an end ileostomy. The loop is then passed through the abdominal wall with care to avoid twisting. The suture marking the distal limb should be clearly visible. The loop is matured by making an incision across the distal aspect of ileum from one mesenteric border to the other at skin level (▶ Fig. 32.2b). This allows the functional limb to be the larger of the two stomas. The functional, or proximal, limb will assume a cephalad position, and the defunctionalized loop will assume a caudad position. The distal loop is sutured to the dermis of the skin with 3–0 absorbable suture. The cephalad, proximal end is then everted to assume the appearance of an end ileostomy by placing three-point anchoring sutures through the ileostomy tip, the ileum at the fascia level, and the dermal layer of the skin (▶ Fig. 32.2c). Interrupted sutures securing the remainder of the afferent opening to the skin level are placed to complete the stoma maturation (▶ Fig. 32.2d). In most circumstances, a stomal support rod is not necessary.
Some authors 36 , 40 , 41 prefer to rotate the loop 180 degrees so that the afferent loop is in the caudad position and the defunctionalized efferent limb is cephalad to allow for maximal fecal diversion when the patient is upright. However, this increases the risk of bowel obstruction and does not improve diversion rates. 42
Prasad et al 43 described a rodless end-loop ileostomy procedure in which the ileum is divided between surgical staples after only a short segment of mesentery has been taken and the stapled proximal end is brought out and matured as a Brooke ileostomy. The antimesenteric corner of the distal stapled end is brought out inferiorly, and a small distal ileostomy opening is created and matured primarily at the skin level (▶ Fig. 32.3). Sitzmann 44 described a similar technique in which the staple line of the distal limb is left closed and the limb is tacked to the afferent limb at the level of the fascia.
Loop-end ileostomies may be created in any situation where a loop ileostomy is indicated. They may be particularly helpful in individuals with thick abdominal walls or short ileal mesenteries. In addition, if there is a significant likelihood that the “temporary” ileostomy may become permanent, an end-loop stoma may make long-term management easier for the ostomate. The end-loop ileostomy is constructed by dividing the ileum with staples and by elevating the ileum at a point where the bowel is redundant enough to permit elevation above the abdominal wall. This ileostomy is matured in the same way as a conventional incontinuity loop ileostomy (▶ Fig. 32.4). 35
32.2.4 Loop Ileostomy Takedown
Before closing a diverting ileostomy, it is important to ensure that there is no distal stricture as may be seen in patients with a diverting ileostomy constructed at the time of an ileal-pouch anal reservoir. Takedown of a loop ileostomy generally can be performed through a peristomal incision and usually does not necessitate a formal laparotomy. In some patients, particularly in those who are obese or have unusually dense adhesions, it may be necessary to open the previous abdominal incision and mobilize additional small bowel in order to safely perform the procedure.
An incision is made at the mucocutaneous junction and extended down to the fascia and the peritoneum by sharp dissection, with meticulous care taken to avoid bowel injury. Once the proximal and distal limbs of the loop have been freed down to the peritoneal level, the remaining peritoneal attachments to the abdominal wall can be divided by blunt or sharp dissection. The ileal segment has been adequately mobilized when it can be elevated from the incision and the wound edges are clear to the peritoneum (▶ Fig. 32.5a,b). The surgeon then has three options of closure: anterior ileal wall closure, stomal resection with end-to-end anastomosis, and functional end-to-end stapled anastomosis.
The choice of anterior wall closure as opposed to resection and anastomosis is one that can be made at the time of surgery. It is often feasible to simply trim the stomal margin and then transversely close the bowel in either one or two layers or with staples (▶ Fig. 32.5c,d). If induration or distortion of the lumen is present, however, it is safe to resect the edges of both limbs and perform a primary end-to-end anastomosis using healthy ileal tissue proximally and distally (▶ Fig. 32.5e–i).
Kestenberg and Becker 45 proposed the technique of stapled functional end-to-end anastomosis because of the dissatisfaction they experienced with long operating times, anastomotic edema, stricture, and partial obstruction following end-to-end anastomosis. A simplified method involves mobilizing both proximal and distal limbs of the ileostomy and inserting a linear cutting stapler in both limbs of the ileostomy and firing the linear stapler. The residual opening can be closed in two layers with sutures or with a stapler (▶ Fig. 32.6).
van de Pavoordt et al 46 reviewed their experience with closure of loop ileostomies. Ninety-three percent of the stoma closures were done through simple transverse incisions. The overall complication rate was 17%. Of the early postoperative complications (13%), the major complication was small bowel obstruction, commonly in patients in whom the stoma was protecting a pelvic ileal reservoir. Abdominal septic complications occurred in 1% of patients. The wound infection rate after healing by both secondary intention and primary skin closure was 3%. Only one incisional hernia was observed in the late postoperative period; in three patients, a posterior rectus sheath defect at the stoma site was found incidentally at laparotomy, without clinical evidence of an incisional hernia.
An additional adjunct to ileostomy closure is the application of sodium hyaluronate/carboxymethyl cellulose (SH/CMC) sheets (Seprafilm Genzyme, Cambridge, MA) during creation of the ileostomy. Bertoni et al 47 retrospectively reviewed a series of ileostomy closures and compared 146 patients in which SH/CMC was used to 147 patients without SH/CMC. The SH/CMC patients had a shorter closure time. Other studies showed that use of SH/CMC resulted in lower perioperative complications and facilitation of early stomal closure. 48 , 49
32.3 Colostomy
32.3.1 History
In 1776, Pillore, a French surgeon, created a cutaneous cecostomy for a wine merchant suffering from obstructing rectal carcinoma. Although the patient died 2 weeks later from a small bowel perforation induced by forced catharsis, the cecostomy was the first recorded instance of a successful colonic stoma being performed. 50 At autopsy, the cecostomy was healthy. A number of sporadic reports of colostomies followed, and these generally involved construction of a loop colostomy for obstruction in adults or imperforate anus in infants. In 1793, Duret, a naval surgeon from Brent, performed the first successful left iliac colostomy for a case of imperforate anus in a 3-day-old child. The patient lived up to the age of 45 years. In 1839, Ammusat, a Parisian surgeon, reported 29 cases of colostomy, with all of them sited in the lumbar area. Twenty-one of these cases involved infants with an imperforate anus; of these, four survived.
In 1884, Maydl introduced the technique of loop colostomy supported by a rod, when he suggested using a goose quill to support the loop against the abdominal wall. In 1881, Schitininger described the creation of an end sigmoid colostomy and an oversewn distal stump. This operation was the forerunner of the Hartmann procedure described in 1923, which consisted of sigmoid resection with end colostomy and an oversewn rectal stump. 50 The operation of colostomy for obstruction came into widespread use in the 20th century.
Miles 51 has been credited for the 1908 description of the operation of end sigmoid colostomy and abdominoperineal resection. In fact, C.H. Mayo had previously described the technique in 1904. This operation, with the loop colostomy that preceded it, ushered in a new surgical era. The end sigmoid colostomy and loop diverting colostomy as described by Miles, Maydl, and others have now been performed by thousands of surgeons on hundreds of thousands of patients in this century with relatively few technical modifications. In the last two decades, the need for end sigmoid colostomy has diminished as surgical stapling techniques have made low anterior anastomosis more feasible. 52
32.3.2 Fecal Diversion
Although the standard method of diverting the fecal stream away from the left colon has traditionally been the transverse loop colostomy, the indications for fecal diversion as well as the choice of a loop colostomy as the best means to achieve this have become controversial. The most common indications for diversion in the past have been obstruction or inflammation of the left colon, trauma to the colon, anterior resection of the rectum, and perineal sepsis. The increasing popularity of Hartmann’s procedure with end colostomy for diverticulitis or obstructing carcinoma, the more aggressive use of primary resection and anastomosis in some of these cases, and the increasing reliance on primary closure for many colon injuries have no doubt decreased the popularity of loop colostomy in the emergency setting.
Fielding et al 53 prospectively studied more than 2,000 patients having elective colorectal anastomoses and found that 15.8% received a synchronous covering stoma to protect the anastomosis. No differences in mortality were observed between patients who had a covering stoma and those who did not, although significant differences in surgical practice were identified in different surgeons. The authors concluded that the covering diverting stoma was not necessary for surgeons who were able to keep their patients’ leak rates at an acceptably low level.
Although many patients with colon and rectal injuries, particularly those with multiple abdominal injuries, need some type of fecal diversion, surgeons have increasingly recognized that the isolated and uncomplicated colonic injury treated promptly can be managed successfully with primary closure. 54
The methods of diversion not involving resection include the “medical colostomy,” ileostomy, cecostomy, and diverting colostomy. The concept of a “medical colostomy” using a chemically defined diet was recommended by Gordon for such anorectal procedures as sphincteroplasty, postanal repair, skin grafting following excision of a neoplasm or hidradenitis suppurativa, and repair of rectovaginal fistula. 55 It puts the gastrointestinal tract at rest so that the repaired area does not have to be challenged with bowel movements. Today it is not necessary for most anal sphincter and rectovaginal fistula repairs, but may be beneficial in the treatment of extrasphincteric fistulas. Robertson 56 achieved similar benefits with an elemental diet in 16 patients with inflammatory bowel disease, anorectal sphincter surgery, or wounds that had to be protected from fecal soiling. None of the patients on the elemental diet had bowel movements within 3 days of operation, and most had bowel movements only at 5- to 6-day intervals. Nutritional support, either by total parenteral nutrition or the elemental diet, is a useful alternative in select patients who need a brief period of gastrointestinal tract rest but do not otherwise need total prolonged fecal diversion.
The value of the loop ileostomy was discussed earlier and should be considered as an alternative to loop colostomy, in most patients. Both the loop ileostomy and loop colostomy, however, can pose management problems for the patient, and neither is an ideal stoma for long-term use. The comparative study by Williams et al 36 showed that both loop ileostomy and loop transverse colostomy successfully defunctionalized the distal bowel. The ileostomy had fewer peristomal problems and fewer management problems than the colostomy, although dehydration from diarrhea or high ileostomy output could be a potential problem with the ileostomy, particularly in elderly patients.
Corman 57 stated that “cecostomy could be removed from our surgical armamentarium with very little consequence to health care delivery.” Others 58 , 59 have been more supportive of the cecostomy as a method of decompressing select patients without subjecting them to the morbidity of a loop colostomy. Attention has focused on less invasive ways of performing cecostomy, including the computed tomography (CT)-guided percutaneous method 60 and the percutaneous endoscopic method for nonobstructing colonic dilatation. 61
Winkler and Volpe 62 reviewed loop transverse colostomy among 29 patients in a community hospital and found that 5 patients died, 8 had complications, and only 18 patients underwent colostomy closure. The authors concluded that the first stage in the management of these patients should be resection of the diseased segment and that the colostomy should be end bearing and placed as close to the disease process as possible. Hopkins 63 supported the concept that temporary loop colostomy may not be “temporary” in many patients by reporting that colostomy closure was not performed in 19 of 45 patients (42%) having transverse colostomy for malignancy. Abrams et al 64 similarly found that colostomy closure was not performed in 156 of 248 patients (63%) having transverse and left colostomies. The difficulties in managing a loop transverse colostomy compared to an end sigmoid colostomy are so substantial that the surgeon must always consider that the loop colostomy may ultimately subject the patients to a less-than-ideal colostomy for life. Although much of the controversy on indications and methods for fecal diversion involves personal preferences on the part of individual authors, some general principle should be applicable for most patients. The principle of resecting primary disease and performing an end-bearing stoma in preference to a loop colostomy or other proximal diverting method is a sound principle that should be adhered to whenever possible. For other patients needing fecal diversion, the loop colostomy or loop ileostomy can be performed. In deciding whether to perform a fecal diversion for patients undergoing anterior resection or for patients with trauma, the surgeon must weigh the risks and consequences of anastomotic or colonic closure leakage against the morbidity and mortality of later colostomy takedown. These risks vary among patients and also among surgeons. The medical colostomy is a useful means of avoiding a surgical stoma in patients who need only short-term diversion to protect perineal wounds or perineal incisions. Tube cecostomy is of limited value and should be used only for decompression and not for fecal diversion.
Cecostomy
Since the turn of the century, tube cecostomy has been advocated mainly as a decompression procedure for acute colonic obstruction and as a safety valve for colonic resection. It has also been suggested for perforation or impending perforation of the cecum, for cecal volvulus, adynamic ileus of the colon, and toxic dilatation in inflammatory bowel disease. The procedure is best performed as a tube cecostomy rather than a primary stoma. When a primary stoma is needed, it is better to perform a loop ileostomy or a loop transverse colostomy, as both are more diverting and are easier for the patients to manage.
The virtues of tube cecostomy, as cited by its proponents, are that it is easy to perform, can be carried out under local anesthesia, the cecum is the portion of colon most likely to perforate, it does not interfere with subsequent resection of the left side of the colon, and it may eventually close spontaneously. In cases of extreme obesity of the abdominal wall and when gross distention of the bowel has “used up” the leaves of the transverse mesocolon, transverse colostomy may be impossible. Under these circumstances, cecostomy may be the only method of relieving obstruction.
The use of cecostomy has frequently been criticized and has lost favor with many surgeons. Detractors point out that it provides less effective decompression than does a transverse colostomy and that it does not allow complete diversion.
Tube cecostomy can be performed as a primary operation through a right lower quadrant (McBurney) or a right lateral transverse incision. Once the peritoneum is opened, the distended cecum is easily delivered through this incision. The serosal surface should be secured to the peritoneal surface with interrupted absorbable sutures. Two pursestring sutures of nonabsorbable material such as 3–0 silk are then placed around the apex of the cecum, and a no. 30-Fr bladder catheter or no. 36 to 40 Pezzer catheter is inserted inside the pursestring sutures. The two pursestring sutures are tied, and the tube is secured in place. It is best to deliver the catheter through a stab wound either cephalad or caudad to this incision so that the incision can be closed primarily. The tube can be removed after 7 to 10 days, and the cecocutaneous fistula should be closed spontaneously.
Rosenberg and Gordon 65 conducted a retrospective review of 59 tube cecostomies to evaluate operative indications, outcome, and associated morbidity. Tube cecostomy was performed as a complementary procedure in 81.4% of cases; in the other 18.6%, it represented either the only operative intervention or the initial stage of a two-stage procedure. Complications included local infection in 32% of cases, pericatheter leak in 25%, skin excoriation in 24%, and pain in 12%. Catheters remained in place for an average of 14 days, but function was adequate in only 40% of cases. Cecal drainage persisted from 24 hours to 90 days after the tube was removed.
Two additional procedures were required to close persistent cecal fistulas. In addition, other reported complications included wound dehiscence, recurrence of a fistula after closure of the cecostomy, subcutaneous emphysema, prolapse of the omentum, stitch granuloma, evisceration, retraction, and hernia formation at the cecostomy site. Aseptic decompression of a distended cecum can be difficult and peritonitis and poor tube function after cecostomy have been responsible for a substantial number of deaths. Tube cecostomy fails to decompress the bowel adequately in as many as 50% of cases. The authors concluded that the high morbidity associated with this procedure militates against its use. Benacci and Wolff 59 were more encouraging about cecostomy. In a retrospective chart review of 67 patients who had catheter tube cecostomy placement, clinical indications for tube cecostomy were colonic pseudoobstruction (39%), distal colonic obstruction (16%), cecal perforation (15%), cecal volvulus (13%), preanastomotic decompression (12%), and miscellaneous (5%). Operation was emergent in 64% of patients and elective in 36% of patients. Tube cecostomy was the primary procedure in 70% of patients and complementary in 36% of patients. Minor complications were seen in 45%, including pericatheter leak (15%), superficial wound infection (12%), tube occlusion (7%), skin excoriation (4%), premature tube dislodgement (4%), colocutaneous fistula (3%), and ventral hernia (12%). No patients required reoperation for tube-related morbidity. The authors concluded that catheter tube cecostomy is of therapeutic value in select clinical situations.
A protective diversionary cecostomy or transverse colostomy is now rarely warranted for left-sided colon procedures as this can be safely performed without diversion. A review of contemporary literature suggests that reasonable current indications for tube cecostomy include volvulus, pseudoobstruction (if an attempt at colonoscopic decompression has failed), cecal perforation associated with an obstructed left-sided lesion, and to relieve obstruction in morbidly obese patients when transverse colostomy is technically impossible.
Loop Transverse Colostomy
In the past, the loop transverse colostomy was a commonly used method to achieve temporary or short-term fecal diversion, whether necessitated by obstruction, inflammation, trauma, low colorectal anastomoses, or perineal wounds. Fontes et al 66 studied the ability of the emergency loop colostomy to achieve total fecal diversion in 62 patients. The authors found that fecal diversion was virtually complete in all patients for the first 3 months but that the diversion became incomplete in about 15% of the patients in subsequent months, probably because of recession and retraction of the stoma. Schofield et al 67 found that they could achieve total fecal diversion from a right transverse colostomy by rotating the stoma counterclockwise for 90 degrees so that proximal limb was in the most dependent position caudad to the distal limb. Morris and Rayburn 68 studied 23 patients with loop colostomies by giving them barium by mouth. In no patient was barium found in the distal limb, indicating full diversion. At least 17 other individual techniques have been recorded in the literature to ensure complete fecal diversion from a loop colostomy. 66 Most of these manipulations appear to be unnecessary, as a traditional loop colostomy should be able to achieve complete fecal diversion for at least a few months postoperatively. Complete diversions should rarely, if ever, be necessary for a longer period of time.
Loop Colostomy Construction
Although a seemingly endless number of techniques have been described for construction of a loop colostomy, three generic methods are used most often. The first involves construction of the loop colostomy over a fascial bridge without the use of a supporting rod or prosthetic bridge material. The second method involves the use of a rod or bridge, and the third method involves creating an end loop or divided type of colostomy. The dissatisfaction with the postoperative management problems caused by loops and bridges has no doubt led to the proliferation of techniques.
In addition to the three types of loop colostomies, a fourth type of diverting temporary colostomy has been described by Turnbull et al 69 as part of an operation for the treatment of toxic megacolon complicating ulcerative colitis. That procedure consisted of a skin-level (trephine) colostomy along with a loop ileostomy as the preliminary operative treatment of critically ill patients with toxic megacolon. This procedure is infrequently performed today, and a skin-level temporary colostomy with its attendant management problems is rarely indicated. However, a trefine stoma may be useful in selected cases of carcinomatous where it may be the only option for fecal diversion.
Loop Colostomy over Fascial Bridge
A transverse incision is made in the right upper quadrant at the previously marked stoma site. The incision is extended across the rectus fascia anteriorly and the rectus muscle fibers are separated. If the patient does not have a concomitant lower transverse incision, rectus can be divided between clamps and ligated to allow for a larger opening. If, however, a simultaneous lower transverse incision through the rectus muscle below the umbilicus has been made, then it is best not to devascularize rectus muscle in the right upper quadrant because of the risk of leaving an ischemic muscle segment between the two transverse incisions. The peritoneum is incised either sharply or with a fine hemostat, and the peritoneal cavity is entered.
The right transverse colon should be grasped with a Babcock forceps and elevated out of the incision, with care taken to identify the taeniae and confirm that the segment is colon. The incision may have to be enlarged if there is too much tension on the bowel as it leaves the abdominal wall. On occasion, it may be necessary to extend the incision and take down the hepatic flexure of the right colon to allow the right transverse colon enough mobility to be delivered through the incision, but this maneuver is usually not necessary.
Once the transverse colon has been elevated from the abdomen, the gastrocolic omentum adherent to the superior surface of the colon is cleaned away, and a short segment of mesocolon (< 5 cm) is taken between clamps and ligated with sutures. The fascial bridge is defined by grasping the fascia and peritoneum together with either an Allis or Kocher clamp. Two or three sutures of heavy nonabsorbable material such as braided nylon or slowly absorbable material are placed through both the fascia and peritoneum of the superior lip of the incision, then through the window created between the colon and the mesocolon, and then through the fascia and peritoneum of the inferior lip on the opposite side. After all sutures have been placed, they are tied on the same side as the incision. The fascia should be felt carefully to make sure it is not constricting the lumen or blood supply of either loop. If the fascial or peritoneal openings are tight, they should be incised either medially or laterally but not along the fascial bridge. If the abdominal wall incision appears too large on either side, it can be partially closed with an interrupted nonabsorbable or slowly absorbable suture.
An incision is made across the apex of the colon. A primary maturation is performed with a 3–0 absorbable material such as Vicryl or chromic catgut. The residual lateral or medial extension of the incision can be closed with a running subcuticular absorbable suture (▶ Fig. 32.7).
Loop Transverse Colostomy over Rod
This procedure is identical to that described for the fascial bridge technique, except that the colostomy is supported through the mesocolic defect by a plastic rod, bridge, or other type of supporting material (▶ Fig. 32.8). Regardless of type, the supporting structure is removed approximately 3 to 10 days after surgery, at which time the colostomy is adherent enough to the incision that it should not recede. Schofield et al 67 described the operation of dependent proximal loop colostomy. This stoma avoids the complications of prolapse and paracolostomy hernia and is easy to close. It defunctions the distal colon as evidenced by the study of 10 patients given a radioactive tracer, chromium 51, by mouth and none spilled over into the distal loop.
Divided-Loop Colostomy
The divided-loop colostomy is analogous to the divided-loop ileostomy described by Prasad et al. 70 In this technique, the transverse colon is divided between surgical staples. The proximal end is brought out as an end stoma through the incision, and the tip of the distal limb is brought out alongside the proximal stoma. The corner is opened and matured as a small distal stoma (▶ Fig. 32.9). The end-loop colostomy eliminates the risk of distal segment prolapse and creates a stoma easier to care for postoperatively.
Unti et al 71 conducted a retrospective review of their 7-year experience in 229 patients with end-loop colostomies, 72 ileocolostomies, 70 and ileostomies. 23 A total of 30 stoma-related complications were observed in 27 stomas, for an overall complication rate of 13.1%. The most common complications were skin excoriation secondary to leakage (3.5%), retraction (3.5%), partial necrosis (2.6%), and peristomal sepsis (1.8%). Mucocutaneous separation, prolapse, and stenosis were each seen in less than 1% of patients. No cases of stomal herniation, obstruction, or hemorrhage were encountered. Twelve deaths occurred, but none was attributed to stoma-related complications.
Hidden Colostomy
Sometimes a patient is found to have an unresectable carcinoma of the colon or rectum in association with extensive metastases or carcinomatosis. If the primary lesion is not causing complete or partial obstruction, the surgeon must decide between doing nothing and risking that the patient will need a second operation later for obstruction or creating a colostomy at the first operation, knowing that the patient may never need it. The hidden colostomy can be a useful technique in this setting.
The colostomy is performed as a loop over a fascial bridge proximal to the carcinoma, which is usually in the rectum. The colon is not elevated above the skin level but is instead buried in the subcutaneous fat where it can easily be identified later. Then if obstruction occurs, the colostomy can be opened at the bedside with the patient under local anesthesia. The resulting stoma is not anatomically ideal, but the patient is usually terminal or near terminal when it is performed, and overall it serves the patient well.
Loop Colostomy Closure
The decision to perform a temporary loop colostomy must always involve some consideration of the morbidity, the mortality, and the cost of colostomy closure. Parks and Hastings 73 reported that complications related to colostomy takedown decreased if colostomy closure was put off for 90 days or more. Williams et al 74 analyzed the morbidity, mortality, and cost of colostomy closure in patients having a colostomy constructed for either trauma or nontraumatic diseases and found no significant differences in complication rates, although the authors noted that the colostomy was less likely to be closed if it was performed for nontraumatic disease. Thal and Yeary 75 reported no deaths and a relatively low complication rate of only 10.2% among 137 patients having a colostomy closed following diverting colostomy for trauma. They emphasized the importance of meticulous surgical technique and attention to detail during the procedure.
Altomare et al 37 reviewed the closure of 87 protective colostomies and identified hypoalbuminemia and interval to closure as risk factors. Operative mortality was 4.6%, major complications developed in 13% (fecal fistula, sepsis, intestinal obstruction, and myocardial insufficiency), and minor complications were recorded in 29% of patients. Colostomy closures within 30 days had a mortality rate of 12.5%, a major complication rate of 25%, and a minor complication rate of 50%. For closures performed between 31 and 90 days, rates were 5.2, 10.5, and 36.8%, respectively. For closure after 90 days, rates were 0, 9.3, and 9.3%, respectively. Their review of the literature, which included 26 reports, revealed a range of death rates from 0 to 4.6%, fecal fistula from 2 to 43%, and total complications from 5 to 61%. The authors emphasized that these data underscore the very significant morbidity and mortality associated with colostomy closure, and therefore the same skill and meticulous approach are required for this operation as for any major procedure performed on the colon.
Velmahos et al 76 reported conflicting recommendations from randomized trials of trauma patients in which they compared early (within 15 days of initial operation) and late (> 90 days after the initial operation) closure. The authors found no significant difference in morbidity between the two groups, with an overall complication rate of 26.3%. Technically, the early closure of colostomies was far easier than late closure, required significantly less operating time, and resulted in less intraoperative blood loss. The closure of end colostomies was more time consuming, both early and late, and caused more bleeding. Total hospitalization was marginally shorter overall for early closure, but late closure of end colostomies resulted in prolonged hospitalization. The authors recommend early closure of colostomies and the use of loop colostomies whenever possible as both are safe and beneficial for patients with colonic injury after trauma. Contraindications to early closure include a nonhealing distal bowel, persistent would sepsis, and persistent postoperative instability.
The loop colostomy can be closed by any of the following three methods: mobilization of the colostomy with closure of the anterior wall of the colon, mobilization with functional end-to-end surgical staple closure, and resection of the colostomy with primary end-to-end anastomosis. The three techniques are analogous to those used for loop ileostomy takedown. If the colon is satisfactorily mobilized proximally and distally; if the colon edges are satisfactorily debrided so that sutures or staples are placed only through healthy, noninflamed tissue; if the blood supply to both ends of the colon is satisfactory; and if there is no tension on the suture line, then all three methods should give comparably good results.
32.3.3 End Sigmoid Colostomy
End sigmoid colostomy may be temporary or permanent. It is performed as a permanent procedure following abdominoperineal resection of the rectum for malignant disease, and it is also indicated as a definitive stoma for incontinent patients who are not suitable candidates for other procedures. End sigmoid colostomy may be performed as a temporary stoma following resection of the rectosigmoid for benign or malignant disease, and it may serve as a means of achieving temporary diversion for other conditions such as radiation proctitis. The end colostomy allows the patient to take advantage of the absorptive capacity of the proximal colon, making it the easiest type of intestinal stoma for the patient to manage.
A key technical consideration in end sigmoid colostomy construction involves deciding whether to fashion the colostomy as an extraperitoneal or intraperitoneal structure. Goligher 10 first described the extraperitoneal approach after observing that patients with both ileostomy and colostomy seemed especially predisposed to postoperative intestinal obstruction. The extraperitoneal colostomy was designed to prevent obstruction from occurring so frequently. Subsequently, Whittaker and Goligher 77 reviewed their experience with 251 patients and found that patients having an extraperitoneal iliac colostomy did not have a lower incidence of mechanical obstruction, but they did have a significantly lower incidence of pericolostomy herniation, prolapse, and recession compared to patients having an intraperitoneal colostomy.
The extraperitoneal colostomy is a good option for patients needing a permanent iliac colostomy. Extraperitoneal colostomy is also good for patients with ascites. The leaking of fluid is minimized or eliminated. For those patients needing only a temporary stoma, the intraperitoneal method is preferable, because takedown of the colostomy and subsequent reanastomosis are less difficult when the proximal colon has been left in an intraperitoneal position.
Technique of Extraperitoneal Colostomy Construction
The technique of extraperitoneal end sigmoid colostomy construction is the same whether it is performed as part of an abdominoperineal resection or as a permanent end colostomy for benign disease. Either a lower midline or transverse incision is suitable once it has been determined that the preselected stoma site will not be too close to either incision. A minimum of 3 cm and preferably 5 to 8 cm should separate the edge of the stoma from the abdominal incision.
Once the abdomen is opened, the sigmoid colon is mobilized by incising the lateral peritoneal reflection along the white line of Toldt, carefully identifying the left ureter as it crosses the left common iliac artery. The intraperitoneal rectosigmoid is then resected and the rectal stump, if retained, closed with staples. Otherwise, a complete proctectomy is performed as described in Chapter 22 for resection of rectal carcinoma. The colostomy site is then created by excising a disk of skin and separating subcutaneous fat down to the level of the fascia. It is important that a layer of fat be retained between the abdominal incision and the colostomy site so that the two wounds are not in direct communication at the subcutaneous level. Goligher 10 originally recommended that the extraperitoneal colostomy be placed a little more laterally so that the rectus muscle is not included. Subsequent experience has shown that the extraperitoneal tunnel can be readily extended into the rectus sheath; this allows the stoma to be positioned medially through the rectus muscle. Although either method is acceptable, the more medial placement of the stoma is preferable because the rectus muscle adds extra support to the stoma. Sjödahl et al 78 reported a dramatic reduction in parastomal hernia development when a permanent stoma was brought out through the rectus muscle. In a review of 130 patients, the prevalence of parastomal hernia was 2.8% when the stoma was constructed through the rectus muscle versus 21.6% when brought out lateral to the rectus muscle. In contrast, Ortiz et al 79 found no correlation between the presence of parastomal hernia and the position of the stoma in the abdominal wall.
The left lateral peritoneal leaf is then grasped with a forceps, and an extraperitoneal tunnel is created by a combination of sharp and blunt dissection to connect with the colostomy wound in the abdominal wall. The tunnel is gently developed by finger dissection and the colon delivered through the tunnel so that 3 to 5 cm of colon comfortably protrudes above the skin level. The colon is then secured to the fascia with interrupted or running sutures of 3–0 polyglycolic acid. The pelvic peritoneum should be secured over the colon, and, if the procedure is performed as part of an abdominoperineal resection, the process of pelvic reperitonealization is completed at this time.
Primary maturation of the colostomy is performed using interrupted 3–0 absorbable sutures with the matured colostomy fashioned to protrude 1 to 2 cm above the abdominal wall. Maturation is accomplished by placing four anchoring stitches through the full thickness of the colon, then the colonic wall below the skin level, then the dermis. Interrupted simple sutures are placed between each of the four quadrants to complete the maturation (▶ Fig. 32.10).
The intraperitoneal colostomy differs only in that the colon is delivered through the rectus muscle, and the defect in the left lateral peritoneal gutter may be closed with interrupted slowly absorbable sutures. However, most surgeons currently leave this space open.
Several techniques of end colostomy maturation using stapling devices have been described using either the linear skin stapler 80 or the circular stapling device. 81 Neither of these methods allows the surgeon the opportunity to optimize stomal contour and stomal protrusion for the individual abdominal wall. In addition, the use of staplers adds significant cost to the procedure.
Porter et al 82 reported on 126 patients who underwent 130 end colostomies, 44 for benign, and 86 for malignant disease, and were followed up for an average of 35 months. The left or sigmoid colon was used in 99 and the transverse colon in 31 patients. Stomas were made electively in 98 patients and urgently in 32 patients. Seventy-six stomas were brought out through the incision and 54 from separate sites. There were 69 complications in 55 patients (44%), including 11 strictures, 9 wound infections, 14 hernias, 9 small bowel obstructions, 4 prolapses, 2 abscesses, 1 peristomal fistula, 17 skin erosions, and 2 poor stoma locations. Fifteen complications required reoperation. Five of these procedures included stoma revision. Total numbers of complications were not related to the stoma site, the disease process, the urgency of the procedure, or the segment of colon used. Wound infections, however, were increased in urgently made stomas. The incidence of hernia was equivalent in stomas brought out through the incision or at a separate site. Forty-one patients (30%) had 43 colostomies closed at an average of 3.5 months after creation. Thirteen patients had 14 complications (five wound infections, six hernias, two small bowel obstructions, and one rectovaginal fistula). One patient died. Four patients required reoperation. There were no anastomotic leaks. Complications were equivalent in Hartmann’s closures and transverse colostomy closures. Complications were similar in stomas created for carcinoma and those created for diverticular disease.
32.3.4 Continent Colostomies
Numerous devices have been introduced for use with an end sigmoid colostomy to help the patient achieve greater control of both gas and stool output from the stoma. Interest in magnetic devices was stimulated by the work of Feustel and Hennig, 83 who described a magnetic ring system in the early 1970s. The system consists of a magnetic ring that is implanted subcutaneously, and the colonic stoma is brought through the ring. An external magnetic cap fitted over the stoma mechanically prevents the expulsion of stool. A disposable charcoal filter disk on the undersurface of the cap permits gas to escape.
Khubchandani et al, 84 using careful patient selection methods, reported favorable results among 14 patients. Husemann and Hager 85 reported the results of 240 patients who underwent implantation of the Erlangen magnetic device. Forty-five rings were explored because of infection, pressure necrosis, parastomal hernia, invagination, prolapse, and stenosis. Although continence was obtained in 68% of patients, only 43% of surviving patients still use the system. Reasons given for not using the cap were pain and weight. Although the magnetic ring device has been successful in some patients, it has two main disadvantages. It requires a magnetic material to be permanently implanted in the abdominal wall and placement must be nearly perfect to allow for satisfactory function. For these reasons, the device has failed to gain widespread popularity.
In Prager’s preliminary experience 86 with a two-part silicone device, three of five patients were able to achieve continence by inserting a specially designed inflatable plug into the stoma. The Conseal colostomy system (Coloplast, Inc., Tampa, FL) is composed of an open-cell polyurethane foam with a water-repellent cover. The cover contains a charcoal filter that enables intestinal gas to pass without odor while preventing the passage of fecal material. The Conseal plug expands to form a bell-shaped plug when it is inserted into the stoma (▶ Fig. 32.11). Clague and Heald 87 evaluated the Conseal system in a multicenter trial of 100 patients and found that it was successful in approximately one-third of the patients. Patients who irrigated their colostomy tended to retain the plug longer than those who used the natural method. Codina Cazador et al 88 conducted another multicenter trial in which 43 patients were evaluated. No complications arose, and complete fecal continence was obtained in 71%. This continence device did not achieve its original promise.
A number of methods have been used to achieve continence without using external devices, the most promising being the intestinal smooth muscle graft. Schmidt 89 reported an 80% success rate with this procedure among 500 patients. The procedure consists of a free transplant of a 10- to 15-cm segment of colonic smooth muscle that is sutured to become ruffled into a muscle mass of about half that length. The graft is wrapped around the bowel and sutured to produce continence.
32.3.5 End Colostomy Takedown
End colostomy takedown with intraperitoneal anastomosis involves mobilizing proximal and distal colonic segments and creating an end-to-end anastomosis using either staples or sutures. The previous abdominal incision, either vertical or transverse, needs to be reopened, and often, extensive adhesions need to be taken down. The colostomy is mobilized by making an incision at the mucocutaneous junction and then dissecting down to the fascia and peritoneum until the colon is free. The mucous fistula or distal rectosigmoid is also mobilized, and the anastomosis can usually be performed without difficulty.
Before the advent of circular staplers, end sigmoid colostomy takedown with extraperitoneal anastomosis to a low-lying rectal stump was often an exceptionally difficult and hazardous procedure. This procedure has been greatly simplified by the introduction of the newer circular stapling devices. After the proximal sigmoid has been mobilized and prepared for anastomosis, the distal stump needs to be dissected only enough to define the apex and free the colon from adhesions that will inhibit smooth placement of the stapling device. The stapler is inserted per anum, and the device is opened so that the trocar can penetrate the apex of the rectal stump and bring the barrel of the device through that hole. A pursestring suture is placed on the proximal sigmoid and tied around the anvil shaft. This shaft is reengaged into the central shaft of the cartridge, and the anvil is approximated to the cartridge. The stapler is fired and then removed. The procedure should carry minimal morbidity as long as the apex of the rectal stump has been carefully defined and a clean circular anastomosis has been obtained. This particular operation lends itself well to a laparoscopic approach. In the event that considerable residual sigmoid colon remains, the circular stapler may not reach the apex of the sigmoid stump. Also there may be kinking of rectum in the hollow of the sacrum. In either event, an additional portion of rectosigmoid will require resection.
The purported ease and safety of closure of loop colostomy are cited as reasons for avoiding end colostomy, with or without resection. Data comparing the complications of loop colostomy closure and end colostomy takedown and anastomosis are sparse. Mileski et al 90 analyzed data from 93 consecutive colostomy closures, 62 of which were loop and 31 were end colostomies. The two groups were comparable with respect to age, underlying disease, and risk factors, such as coronary artery disease, diabetes, hypertension, steroid dependence, hypoalbuminemia, and smoking. Closure of end colostomies took longer and was associated with more blood loss than closure of loop colostomies. However, the mortality rates for closure of loop (4.8%) and end (3.2%) colostomies were not significantly different. The complication rates were identical (16%). Although none of the other risk factors was associated with increased rates of mortality or morbidity, the detrimental effects of steroid dependence and preoperative hypoalbuminemia were striking. Four deaths and 60% of the complications occurred in patients with steroid dependence or hypoalbuminemia, or both. The rates of wound infection after primary or secondary closure of the stoma site were not significantly different. The authors concluded that loop colostomy closure is not associated with fewer complications than closure of end colostomy, even though the latter takes longer to perform and is more difficult. Primary closure of the stomal site is safe and reduces the length of hospital stay.
Mosdell and Doberneck 91 reviewed the charts of 59 patients undergoing Hartmann’s procedure and 43 patients undergoing ostomy closure after divided colostomy, loop colostomy, or divided ileostomy-colostomy. Ostomy closure after Hartmann’s procedure was accomplished in 46 patients. These 46 patients (group I) were compared with the 43 patients having ostomy closure following divided colostomy, loop colostomy, or divided ileostomy-colostomy (group II). No deaths occurred in either group. The morbidity rate was 30% in group I and 19% in group II. Major complications involved wound, lung, small bowel, and colonic anastomoses. Anastomotic stricture rate was 9% in group I and 5% in group II. Small bowel and anastomotic complications in both groups occurred only when ostomy closure was performed after a delay of less than 6 months after ostomy construction. Stricture occurred only after end-to-end colocolostomy and coloproctostomy and did not occur after ileostomy or ileoproctostomy. All strictures were successfully treated by reoperation. Anastomotic leak and pelvic abscess did not occur in either group. The authors concluded that ostomy closure after Hartmann’s procedure may be more difficult and time consuming than ostomy closure after loop colostomy, divided colostomy, or divided ileostomy-colostomy, but that ostomy closure after Hartmann’s procedure does not result in a higher morbidity rate. The authors advise an interval of 6 months between ostomy construction and ostomy closure and submit that all patients whose general condition permits reoperation may safely undergo ostomy closure. Pearce et al 92 reviewed the outcome of 145 patients undergoing Hartmann’s resection. The mortality rate of the primary procedure was 8%. Eighty patients also underwent reanastomosis. The interval between the primary and secondary procedures was found to be the most important risk factor. Six of 12 patients had clinical evidence of a leak when this interval was less than 3 months, compared with 7 of 28 patients for 3 to 6 months, and none of 40 when the second operation was delayed for greater than 6 months. All deaths (three patients) and clinical septicemia (four patients) occurred in the two “early” groups. All colovaginal fistulas (three patients) and strictures (three patients) were associated with stapled anastomoses. No association was found between the complication rate following reanastomosis and the initial pathology or experience of the surgeon undertaking the secondary operation.
Keck et al 93 reviewed the case records of 111 patients undergoing Hartmann’s procedure mostly for advanced carcinoma and complicated diverticular disease. Of 96 patients who survived, 50 (52%) underwent reversal. Of those with diverticular disease, 40 of 48 (83%) underwent reversal. Mortality for Hartmann’s reversal was 2%, anastomotic leak rate was 4%, and overall complication rate was 26%. Early reversal was performed in 13 patients and late reversal in 37 patients. There was no difference in these groups in mortality, morbidity, or anastomotic leakage. However, bed stay was longer in the early group and graded operative difficulty greater. In particular, cases in which adhesion density was most severe and in which accidental enterotomy occurred were more common in the early group.
Livingston et al 94 questioned whether the risks after colostomy closure were exaggerated. The authors reviewed 121 patients who underwent colostomy closure for trauma. There was no mortality and a 4.9% incidence of major morbidity. Although there was no apparent relationship between the interval between colostomy creation and closure, three of the six major complications occurred in patients whose colostomies were closed soon after complicated initial injuries. The authors recommended that if the primary operation is complicated by intra-abdominal sepsis or major wound problems, 6 months should ensue before attempting closure. Long-term follow-up of these patients (mean, 39 months) disclosed a low incidence of late complications secondary to colostomy closure. Although the trend toward the increased use of primary repair of colon injuries in selected patients is supported, the authors’ study indicates that the risk of colostomy closure has been exaggerated and should not be a factor in the decision to create a colostomy after colon trauma.
Khoury et al 95 retrospectively reviewed the records of 46 patients who underwent colostomy closure. Patients ranged in age from 24 to 87 years, and 54% were women. Stomas had been created during emergency operations in 87%. Most operations (54%) were performed for complications of acute diverticulitis. Of the 46 procedures, 40 (87%) were end colostomies and 6 were loop colostomies. Stomas were closed at a range of 11 to 1,357 days after creation (mean, 207 days; median, 116 days). Inpatient complications occurred in 15% of patients, including congestive heart failure (2%), cerebrovascular accident (4%), pneumonia (2%), enterocutaneous fistula (2%), and pulmonary embolus with death (2%). The most common long-term complication was midline wound hernia, which occurred in 10% of surviving patients. Overall complications occurred in 24%. The authors concluded that colostomy closure is a major operation.
32.4 Complications
Colostomy complications are similar to those associated with ileostomies. The most frequently encountered problems include ischemia, recession, prolapse, and parastomal hernia. Allen-Mersh and Thomson 96 reviewed 156 operations on 23 patients over a 3-year period for correction of late colostomy complications and found that problems included stenosis in 65 patients, paracolostomy hernia in 42, and prolapse in 16.
Leenen and Kuypers 97 studied 266 patients with 345 stomas on the small and large bowel to determine possible etiologic factors for stomal complications. The overall complication rate for creating a stoma was 36%. No differences in overall complication rate were encountered when comparing acute and elective management; however, high-output stomas and necrosis were encountered more often in the acutely managed group. Preoperative contamination was followed more often by stomal retraction. Septic events, however, occurred less frequently than in the noncontaminated procedures. Moderate obesity had no significant influence on the outcome of the procedure. Adipose patients had a statistically significant greater number of necroses. The outcome of stoma surgery was greatly influenced by bowel quality. Crohn’s disease and bowel ischemia were encountered in 50% of patients with stoma complications. In ischemic disease, significantly more necrosis was found. Retraction of the stoma occurred more often in patients with Crohn’s disease. Patients with chronic ulcerative colitis did not have a higher complication rate.
Although complications of ileostomy are common, they can be minimized by careful surgical techniques and good ET nursing. Carlstedt et al 98 reviewed the ileostomy complications in 203 patients operated on with proctocolectomy and ileostomy for ulcerative colitis and Crohn’s disease. The crude rate of ileostomy complications necessitating reconstruction was 34% and was significantly higher in patients with Crohn’s disease compared with patients with ulcerative colitis. The cumulative rate of surgical revision after 8 years was 75% in the former group and 44% in the latter. Ileostomy stenosis and sliding recession were the two most common indications for reconstruction. Eighty-three percent of the revisions were performed as local procedures, making a formal laparotomy unnecessary. Causative factors such as surgical technique, length of concomitant ileal resection, and postoperative weight gain were analyzed for possible influence on the rate of reconstruction, but no significant association was identified.
Khoo et al 99 prospectively assessed the morbidity of creating and closing loop ileostomies in a consecutive series of 203 patients having an ileoanal pouch procedure. There was one death as a result of liver failure. One patient developed a persistent pouch-vaginal fistula that resulted in pouch excision. The remaining 201 patients had their ileostomy closed at a mean time of 10 weeks after the primary procedure. Only 7% needed reoperation to correct ileostomy-related problems. After ileostomy closure, complications were noted in only 2% of patients. Stothert et al 100 reviewed the outcome of 49 high-risk patients having 51 stomas created as an emergency measure and found a morbidity of more than 50%. Feinberg et al 101 found complications referable either to a temporary loop ileostomy or its closure in 69 of 117 patients undergoing loop ileostomy in association with the pelvic reservoir procedure and ileoanal anastomosis. Grobler et al, 102 in a randomized trial of loop ileostomy in restorative proctocolectomy, found that 52% of patients developed ileostomy-related complications. On the other hand, Wexner et al 103 found that loop ileostomy was a safe option for fecal diversion in a study of 83 patients who required temporary fecal diversion after either ileoanal or low colorectal anastomosis (72 patients), for perianal Crohn’s disease (5 patients), or for other reasons (6 patients). All loop ileostomies were supported with a rod, and fecal diversion was maintained for a mean of 10 weeks. Sixty-seven patients had reestablishment of intestinal continuity. Stoma closure was effected through a parastomal incision in 64 patients; in 3, a laparotomy was required. The closure was stapled side to side in 49 patients, while a hand-sewn anastomosis was performed in the other 18 patients. All skin wounds were left open. Nine patients (10.8%) developed 10 complications. Four patients developed dehydration and electrolyte abnormalities secondary to high stoma output, two had anastomotic leaks that spontaneously healed following conservative management, one patient developed a superficial wound infection that spontaneously drained and one patient developed a partial small bowel obstruction that resolved without operation after a 4-day hospitalization. One stoma retracted after supporting rod removal and prompted premature closure. There were no instances of stomal ischemia, hemorrhage, prolapse, or mortality in this series.
Leong et al 104 conducted an actuarial analysis of complications of 150 permanent end ileostomies constructed over a 10-year period. By 20 years, the incidence of stomal complications approached 76% in patients operated on for ulcerative colitis and 59% in those with Crohn’s disease. Revisional surgery rates were higher in patients with ulcerative colitis than in those with Crohn’s disease (28 vs. 16%). Complications encountered were skin problems (cumulative probability, 34%), intestinal obstruction (23%), retraction (17%), parastomal herniation (16%), fistula or suppuration (12%), prolapse (11%), stenosis (5%), and necrosis (1%). Closure of the lateral space did not reduce the probability of developing intestinal obstruction (18% at 20 years in those with closure vs. 3% in those without). Fixation of the mesentery did not reduce the probability of developing prolapse of the ileostomy (11% in those with fixation vs. none in those without). The incidence of parastomal herniation was not reduced by siting through the rectus abdominis (21% in those sited through the body of the rectus abdominis vs. 7% in those sited through the oblique muscles). Some of the surgical dogmas relating to ileostomy construction are not supported by the results of this study.
In a review by Senapati et al, 105 of 310 patients who underwent restorative proctocolectomy, 296 had a covering ileostomy and 14 did not. The stoma had been closed in 88.9% at a median interval from formation of 12 weeks. Ileostomy-related complications before closure occurred in 5.7%. Laparotomy for obstruction due to the ileostomy was required in 2.4%. Retraction requiring revision occurred in 1.0%, an abscess behind the stoma in 0.3%, and miscellaneous appliance problems in 2.4%. Following closure, 22.4% developed an ileostomy-related complication. There were 30 cases of small bowel obstruction, treated conservatively in 19 (7.2%) and by laparotomy in 11 (4.2%). Peritonitis requiring laparotomy occurred in 1.1%, and 0.8% developed an enterocutaneous fistula. There were 5.3% wound infections and 6.1% other miscellaneous problems. Significant complications associated with a temporary ileostomy were less frequent in this series than in some other reports. Obstruction was the most common complication and fistula was rare.
In a review of the complications of temporary loop ileostomy by Kaidar-Person et al, 106 factors suggested to predispose to stoma complications were high body mass index, inflammatory bowel disease, use of steroids and immunosuppressive therapy, diabetes, old age, emergency operation, operative technique, and surgeon experience. Preoperative evaluation and proper marking of the stoma site before an operation were associated with fewer adverse outcomes. In the review of 14 publications, the reported morbidity ranged from 3 to 100% (with many in the 35–45% range). Major complications included stenosis, small bowel obstruction, retraction, necrosis, prolapse, stricture, fistula, and parastomal hernia. Small bowel obstruction ranged from 0 to 17% (mostly in the 3–8% range), high output 1 to 72% (mostly in the 3–5% range), appliance leakage 7 to 68% (mostly in the 17–38% range), skin irritation 2% to 41% (mostly in the 4–7% range), and reoperation rate 1% to 9%. Minor complications included dermatitis, electrolyte imbalance, and dehydration from high stoma output. The latter may often necessitate early closure of the stoma. Although not a complication, nonclosure of a temporary ileostomy ranged from 0 to 19%.
32.4.1 Bowel Obstruction
Small intestinal obstruction is a relatively common occurrence among ileostomy patients. It is usually caused by extrinsic compression from adhesions or an internal hernia, intraluminal compression from impacted boluses of food, or recurrent Crohn’s disease.
Although food bolus obstruction is usually self-limited, it is clinically indistinguishable from complete extrinsic small intestinal obstruction. Food bolus obstructions tend to be complete and present as an impaction at a point of angulation or stricture at the fascial level. Since the impaction may occur after the ingestion of exotic or high-fiber foods, such as coconut, the patient often relates the onset of pain to a time just after eating these foods. The patient frequently has a history of similar episodes.
The onset of pain and obstructive symptoms tends to be sudden, and cessation of ileostomy output usually follows. With a partial high-grade obstruction, the patient describes passing a large volume of watery effluent from the stoma and may confuse this event with gastroenteritis. Whether the obstruction is complete or partial, significant dehydration can result from associated nausea and vomiting, as well as from the secretion of fluid into the dilated small bowel proximal to the obstruction. Toxicity is generally not present, and perforation and peritonitis are uncommon.
Management of food bolus obstruction should include fluid and electrolyte restoration, nasogastric suction, and close observation. Gentle irrigation of the ileostomy with a soft, well-lubricated rubber catheter, using 4 oz of either saline or liquid glycerine often dislodges the impaction and relieves the obstruction. Alternatively, lavage of the ileostomy with up to 500 mL of saline may be tried using an enema bag, a cone-tip irrigator, and an irrigation sleeve. Operation is rarely needed.
Mechanical extrinsic obstruction is less common but potentially more serious than intraluminal obstruction. Patients who have had operations for inflammatory bowel disease have a greater predilection for obstruction than those who have undergone abdominal operations and bowel resection for other disease. Hughes et al 107 found a 9.1% incidence of small bowel obstruction requiring operation among 463 patients having resectional therapy for inflammatory bowel disease compared to a 2.3% incidence of operation among 2,474 patients having resection for colon and rectal neoplasms. Two-thirds of the obstructions in this series were related to adhesions, and the other third was related to the stoma. Fasth and Hultén 41 noted a lower incidence of obstruction, approximately 3%, for patients with loop ileostomies.
Intestinal obstruction has been reported as an especially frequent complication following restorative proctocolectomy for inflammatory bowel disease. The obstructive events are related either to the temporary ileostomy or to the adhesions. Obstructive phenomena have been reported in up to 43.5% of patients having such restorative procedures. 108 Francois et al 109 reviewed the Mayo Clinic experience with small bowel obstruction complicating ileal pouch–anal anastomosis and found that 17% of their patients developed small bowel obstruction; 7.5% of these patients required operation. In this series, obstruction was more likely to occur in patients who had a temporary Brooke ileostomy (12.5%) than those who had a loop ileostomy (4.6%). Although the loop ileostomy with an open mesenteric defect laterally appears especially vulnerable to obstruction, it is clear that patients having either a Brooke end ileostomy or loop ileostomy are at risk.
Anderson et al 110 described a simple technique to minimize the risk of volvulus following establishment of a loop ileostomy. A broad antimesenteric attachment of the seromuscular layer of ileum to the parietes at the stoma site is achieved with absorbable sutures, creating a broader fulcrum. After adopting this technique, no complications were noted in 30 patients followed up for a minimum of 4 years.
The patient with mechanical small bowel obstruction is often indistinguishable from the patient with food bolus obstruction. The initial management is the same, consisting of fluid and electrolyte resuscitation, nasogastric suction, and a brief trial of irrigation or lavage of the stoma. If pain persists despite nasogastric tube decompression and lavage of the stoma, then early operation is the safest course. If, however, the patient is comfortable with the nasogastric tube in place and if the abdomen is soft and free of signs of peritonitis, then the patient may be observed for a period of 24 to 48 hours. Immediate operation must be considered if the patient develops increasing pain, increased distention, leukocytosis, or fever. For all patients with an uncertain obstruction and unsatisfactory course, early operation is often the safest approach. Although Francois et al 109 emphasized that necrosis from small bowel obstruction complicating the loop or Brooke ileostomy was rare in their series, the consequences of additional small bowel loss from infarction in a patient who already has an ileostomy are significant and the need for resection must be avoided.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
