Tech Coloproctol (2005) 9:206–208 DOI 10.1007/s10151-005-0228-z O R I G I N A L A RT I C L E

G. Lahat • H. Tulchinsky • G. Goldman • J.M. Klauzner • M. Rabau

Wound infection after ileostomy closure: a prospective randomized study comparing primary vs. delayed primary closure techniques

Received: 22 November 2004 / Accepted: 18 May 2005 / Published online: 21 November 2005

Abstract Background Closure of ileostomy is considered a contaminated operation. The infection rate of the stoma wound is ≥30%. Several ileostomy-closure techniques intended to reduce the high rate of infection have been described in the literature. Among them, delayed primary closure of the stoma wound is a commonly used method that was reported to reduce the infection rate according to several retrospective studies. We therefore conducted the first prospective randomized trial comparing primary with delayed primary closure of a stoma wound. Methods During 2003, 40 patients were admitted to our ward for closure of ileostomy. The ileostomies were taken down by the same team using the same surgical technique except for the technique of wound closure. We randomly divided the patients into two groups. In Group 1 (n=20), the wound was left open for delayed primary closure and not closed until postoperative day 4. In Group 2, the wound was primarily closed at the end of

G. Lahat • H. Tulchinsky () • G. Goldman • J.M. Klauzner M. Rabau Proctology Unit, Division of Surgery B Tel Aviv Sourasky Medical Center Sackler School of Medicine Tel Aviv University, Israel E-mail: [email protected]

the procedure. Results The total wound infection rate was relatively low (15%). Infection occurred more frequently (4 cases, 20%) in Group 1 than in Group 2 (2 cases, 10%). The length of hospital stay was similar for both groups. Conclusions In this first prospective comparison of two techniques during ileostomy take down, primary closure unexpectedly produced less wound infection than delayed primary closure. Key words Ileostomy • Wound infection • Closure

Introduction Loop ileostomy is commonly used to temporarily defunction distal colorectal, coloanal or ileoanal anastomoses. It is easy to construct, offers excellent fecal diversion and is easy to handle. Gastrointestinal continuity is renewed after a period of approximately 3 months. Its closure is associated with a considerable complication rate. The commonest complication is wound infection, ranging from 2% to 40%, with an average of approximately 10% [1-5]. Several closure techniques for stoma wound have been described: delayed primary closure (DPC) of the skin is commonly used in dirty operations, to reduce the wound infection rate [6, 7]. Some retrospective studies have shown a lower rate of wound infection with DPC than with primary closure (PC) of the stoma wound [8]. We therefore carried out a prospective study to compare PC to DPC of ileostomy wound.

Patients and methods All consecutive patients that underwent closure of loop ileostomy during 2003 were enrolled. Patients who had a previous complicated attempt of ileostomy closure were excluded.

G. Lahat et al.: Wound infection after ileostomy closure Ethical approval was not requested since both methods of wound closure during ileostomy takedown are established, commonly used techniques. Elective takedown was performed after a minimum of 2 months from the primary procedure. Closure of stoma was preceded by contrast radiographic examination to evaluate distal bowel integrity. No preoperative bowel preparation was undertaken. Preoperatively, a first-generation cephalosporin was given intravenously. The same surgical team, using the same technique, reversed all ileostomies. Surgical technique included a circumferential incision around the ileostomy, release of the ileal loop down into the abdomen, and stapled side to side (functional end to end) anastomosis with resection of the ileostomy end. Closure of the abdominal wall was performed with continuous 0/0 PDS (Polydioxanone) sutures. Patients were randomized (each consecutive case referred alternately without any selection) to one of two strategies of wound management: PC or DPC. For PC, the skin was closed with skin staples; no topical antibiotic solution was used during the procedure. For DPC, wounds were packed with saline-soaked gauze and were not manipulated until postoperative day 3, at which time the dressing was changed using sterile technique and the wound was evaluated for closure. If the wound appeared pristine, showing no drainage, skin was closed with nylon sutures the next day. Otherwise the skin was left open until suitable for closure. Wounds were observed twice daily until discharge and subsequently at the outpatient clinic at 1 and 2 weeks. Wound infection was determined when purulent drainage was observed. Wound infection was suspected based on local inflammatory signs or systemic signs such as fever and tachycardia. Possibly infected wounds were evaluated by one of three colorectal surgeons, were observed closely and opened if purulent drainage, increasing erythema, or induration developed.

Results A total of 40 patients who required closure of loop ileostomy were enrolled (Table 1). Indications for construction of loop ileostomy were restorative proctocolectomy in 23 patients (18 ulcerative colitis, 1 Crohn’s disease and 4 familial adenomatous polyposis), very low anterior resection for rectal cancer (15 patients), and perforated diverticulitis with rectosigmoid resection and primary anastomosis (2 patients). The patients were randomized, 20 per group, to receive either primary closure (PC) or delayed primary closure (DPC) strategy of wound management. Average length of time for reversal was 92 days (range, 64 days to 6 months) since the first operation: 14 patients (70%) in PC group and 15 (75%) in DPC group had ileostomy reversal within 90 days of the first operation. The groups were comparable concerning sex, age, original operation, risk factors for wound infection and duration between ileostomy construction and takedown (Table 1). Wound infection occurred overall in 6 patients (15%), including 2 patients in PC group (10%) and 4 in DPC group

207 Table 1 Characteristics of 40 patients who underwent closure of loop ileostomy, by type of wound management strategy DPC (n=20)

PC (n=20)

Men, n (%)

11 (55)

8 (40)

yearsa

45.5 (18–80)

47.5 (21–73)

91 (67–103)

93 (65–180)

18 18 12 11 11

10 17 12 10 11

18 18 11 15 13 10 17

19 17 10 17 11 17 16

Age,

Time from first operation, daysa Underlying disease, n Ulcerative colitis Rectal cancer FAP Crohn’s disease Acute diverticulitis Risk factors for infection, n Malignancy Pelvic radiation Diabetes mellitus Steroid treatment Cardiovascular disease Smoking habit Patients with ≥1 risk factor

a Values are mean (range) FAP, Familial Adenomatous Polyposis

(20%). All patients had local signs; 5 had superficial wound infection and 1 had deep wound infection. Five patients were diagnosed during postoperative hospitalization and one was readmitted 9 days after operation. Three patients were febrile, and two had elevated white blood cell counts. There were 3 wound cultures positive for Pseudomonas aeruginosa, E. coli and Staphylococcus aureous. One wound in the DPC group was not closed on postoperative day 4 due to excessive drainage and was left for secondary closure. Overall, 16 patients in DPC group (80%) and 18 patients in PC group (90%) had normal healing (Table 2). Two patients, one from each group, had an anastomotic (ileal) microleak and presented with fever, abdominal pain and local peritoneal signs. The diagnosis was confirmed by abdominal computed tomography; both patients were managed conservatively as there was no evidence of wound infection. Two wounds from the PC group that were suspected to be infected were opened with no evidence of purulent discharge and negative cultures; therefore they were not considered to be infected. Table 2 Outcome of loop ileostomy closure, by type of wound management strategy

Wound infection, n (%) Normal healing, n (%) Secondary healing, n (%)

DPC (n=20)

PC (n=20)

4 (20) 16 (80) 4 (20)

2 (10) 18 (90) 2 (10)

208

Average length of hospitalization for the DPC group was 6.5 (range, 5–11) days; for the PC group it was 5.9 (range; 5–13) days. Average length of hospital stay for patients with infected wounds was 9.4 (range, 8–13) days compared with 6.2 (range, 5–8) days for patients with normal healing wounds.

G. Lahat et al.: Wound infection after ileostomy closure

sion, our study, although consisting of a small number of patients, did not find any advantage to DPC concerning wound sepsis. A larger sample should be studied in order to gain statistical significance.

References Discussion Loop ileostomies are commonly used among colorectal surgeons because they are simple to construct and to take down. Wound sepsis following skin closure during closure of a stoma would appear almost inevitable because the wound is contaminated [9]. It is a surgical principle that contaminated wounds be left open to heal by second intention. DPC of dirty wounds was popularized during World War I, as described by Hepburn in 1919. Several prospective studies comparing DPC to PC of dirty wounds have shown controversial results [10–13]. Still, DPC is a common technique used for dirty wounds, assuming that it lowers rates of infection. The complications and the financial impact of wound infection are significant. Wound infection is a significant contributor for dehiscence and herniation as previously reported [14–16]. It is also the cause for increased costs, associated with the longer hospital stay [17, 18]. The issue of wound infection following closure of a stoma was evaluated in a few retrospective studies reporting high rates of infection [1–3]. Terry et al. [19] found that closed wounds had a much higher infection rate (14.2%) than wounds left partly open (4%). DPC is an acceptable and widely used method for the contaminated stoma wound. Hackam and Rotstein [8] found that PC of stoma wound was associated with a markedly increased wound infection rate (41%) compared with DPC or secondary wound closure (15%). Our study is the first to prospectively randomize two skin closure techniques for closure of stoma wounds. We expected better results for the DPC group as reported in previous retrospective series. However, DPC of the stoma wound showed no benefit. On the contrary, wound infection rate was twice that of the PC group. Our results are statistically insignificant due to the small number of patients enrolled, but they still raise a doubt regarding the advantage of DPC vs. PC on the stoma wound during ileostomy takedown. Equal wound infection rates of the two closure techniques were noticed in three prospective randomized studies dealing with skin closure during dirty operations [14–16]. There was no significant difference in length of hospital stay between the two groups. As expected, hospitalization was longer for patients with infected wounds. One limitation of the study is the small number of patients enrolled, making it difficult to conclude which method is superior, based on statistical power. In conclu-

1. Metcalf AM, Dosois RR, Beart RW et al (1986) Temporary ileostomy for ileal pouch anal anastomosis. Function and complications. Dis Colon Rectum 29:300–303 2. Van de Pavoordt HD, Fazio VW, Jagelman DG et al (1987) The outcome of loop ileostomy closure in 293 cases. Int J Colorectal Dis 2:214–217 3. Feinberg SM, McLeod RS, Cohen Z (1987) Complications of loop ileostomy. Am J Surg 153:102–107 4. Senapati A, Nicholls RJ, Ritchie JK et al (1993) Temporary loop ileostomy for restorative proctocolectomy. Br J Surg 80:628–630 5. Khoo RE, Cohen MM, Chapman GM et al (1994) Loop ileostomy for temporary fecal diversion. Am J Surg 167:519–522 6. Cohn SM, Gianotti G, Ong AW et al (2001) Prospective randomized trial of two wound management strategies for dirty abdominal wounds. Ann Surg 233:409–413 7. Varrier ED, Bossart KJ, Heer FW (1979) Reduction of infection rates in abdominal incisions by delayed wound closure techniques. Am J Surg 138:22–28 8. Hackam DJ, Rotstein OD (1995) Stoma closure and wound infection: an evaluation of risk factors. Can J Surg 38:144–148 9. Samhouri F, Grodinsky C (1979) The morbidity and mortality of colostomy closure. Dis Colon Rectum 22:312–314 10. Tsang TM, Tam PKH, Saing H (1992) Delayed primary wound closure using skin tapes for advanced appendicitis in children. Arch Surg 127:451–453 11. Pettigrew R (1981) Delayed primary wound closure in gangrenous and perforated appendicitis. Br J Surg 68:635–638 12. Andersen B, Bendtsen A, Holbraad L (1972) Wound infections after appendectomy. Acta Chir Scand 138:531–536 13. Stephen MC, Giovanni G, Adrian W et al (2001) Prospective randomized trial of two wound management strategies for dirty abdominal wounds. Ann Surg 233:409–413 14. Riou JPA, Cohen JR, Johnson H (1992) Factors influencing wound dehiscence. Am J Surg 163:324–330 15. Bucknall TE, Cox PJ, Ellis H (1982) Burst abdomen and incisional hernia: a prospective study of 1129 major laparotomies. Br Med J 284:931–933 16. Irvin TT, Stoddard CJ, Greaney MG et al (1977) Abdominal wound healing: a prospective clinical study. Br Med J 2:351–352 17. Shulkin DJ, Kinosian B, Glick H et al (1993) The economic impact of infections. Arch Surg 128:449–452 18. Poulsen KB, Bremmelgaard A, Sorensen AI et al (1994) Estimated costs of postoperative wound infections. Epidemiol Infect 113:283–295 19. Terry P, Hain JM, Ramirez JJ et al (1999) Techniques and complications of ileostomy takedown. Am J Surg 177:463–466