Chronic Wound Infection: Bacterial Colonization in the Dermal Pericolostomic Region

Valdemir Salles, MD;1 Sarhan Saad, MD, PhD;2 Delcio Matos, MD, PhD2

The incidence of septic complications following colostomy surgery varies from 0.5% to 25%.1–6 Wound infections appear to be more common following emergency stoma surgery, increasing the occurrence of sepsis in contaminated cases compared to noncontaminated cases. This can be explained by the contamination of the surgical site, starting from contact of the skin tissue with bacterial flora present in the intestinal lumen, which is not noted during clean procedures or those in which the intestine is not manipulated.7


To determine the prevalence and type of bacteria in the peristomal skin of colostomy patients, a prospective study was conducted among patients at the Department of Surgery of the University of Taubaté, Medical School and Federal University of São Paulo Medical School, São Paulo, Brazil, from October 2000 to September 2002. All patients had a temporary colostomy of at least 7 weeks duration located either in the transverse or in the sigmoid colon due to reconstruction of the intestinal transit. They were admitted or excluded from the study depending on the established criteria and after signing a consent form formulated by the Medical Ethics Committee.

Patients with clinical stage I or II, as defined by the American Society of Anesthesiology, and those with colorectal oncological malignancy (Astler-Coller stage B2 colorectal cancer), ie, who were not in the advanced stage of the disease, were admitted to the study. Excluded from the study were patients exhibiting pericolostomic skin disease (stenosis, cellulitis, prolapse, dermatitis, sepsis) or who had undergone chemotherapy and/or radiotherapy within the previous 30 days; with diabetes mellitus, immune-suppressive, and cachectic disorders; with infectious or inflammatory processes in other tissues, and patients being treated or had been treated with antibiotics and anti-inflammatory medications in the previous 30 days.

Based on these criteria, a group of 34 healthy individuals with colostomies and without complications was formed. In order to obtain the test specimen, the patient’s colon was prepared following the pre-surgical procedure. No oral antibiotic preparation was used. Intravenous antibiotic therapy was initiated only after collection of the test specimens. Antiseptic measures were performed in the pericolostomy region using 0.9% saline solution, polyvinylpyrolidone-iodine, and 10% sodium lauryl ether sulfate, for at least 5 minutes. Two biopsies of the cutaneous tissue were taken next to the lower edge of the colostomy, approximately 0.5 cm from the enterocutaneous anastomosis. The material was transported to the Microbiology Laboratory (Stuart Transportation), and was seeded on specific plates for aerobic bacteria (eosin agar methylene blue, blood agar with acid, and chocolate agar). Anaerobic bacteria (blood supplemented agar and bacteroides-bile-esculin agar) was incubated at 37˚C for 72 hours and kept oxygen-free to facilitate identification. Manual biochemical assays were utilized to identify the bacteria (API 20E, Bactray I, II).

Fisher’s Exact Test and linear-by-linear association were applied to examine the association between the presence of microorganisms and the length of time of the colostomy. The rejection level of the null hypothesis was fixed at 0.1 or 10% (α ≤ 0.1) in all tests. Patients were classified into 4 subgroups (comprising 4 weeks): 7.5 to 12 weeks (10 patients), 12.3 to 16 weeks (9 patients), 16.3 to 20 weeks (5 patients), and > 20 weeks (10 patients).


Of the 34 participants, 20 (58.8%) were male and 14 (41.2%) were female. The average age was 51.6 years (range 16–84 years). Using the surgical risk evaluation adopted by the American Society of Anesthesiology, 25 patients (73.5%) were class 1 and 9 (26.5%), class 2.

Among the participants, 16 (47.0%) had a colostomy due to malignant colorectal disease, 13 (81.3%) were classified in stage B2, and 3 (18.7%) in stage B1.

The most frequently cultured bacteria were Escherichia coli (31 of 34 patients, 91.2%). Also noted were Bacteroides spp (13/34, 38.2%), and Peptococcus spp (13/34, 38.2%), Klebsiella spp (11/34, 32.4%), and Bacteroides fragilis (10/34, 29.4% [Table 1]). When bacterial species were compared with consideration to length of time with a colostomy, researchers noted that Escherichia coli was present in all groups, while Bacteroides fragilis and Klebsiella spp were not found. Peptococcus spp increased in the period beyond 20 weeks (Table 2).

According to Fisher’s Exact Test and linear-by-linear association, Bacteroides fragilis (P = 0.021) and Klebsiella spp (P = 0.003) were not found in the > 20-week group, a statistically significant finding. It was observed that Peptococcus spp (P = 0.068) was increasingly present in the > 20-week group; Escherichia coli presence was statistically significant (P < 0.001).


In clinical practice, the authors observe that the rate of infectious complications affecting the pericolostomy region is low, despite the permanent contact of the local skin tissue with the intestinal content.

Several factors can affect the genesis of pericolostomic bacterial colonization, which is facilitated by immediate contact of the local skin area with the enteric content. The impact of the length of time of the colostomy on the bacteria present in the dermis of the pericolostomic region was evident in group 2, where higher values of microorganisms were found compared to the other 3 groups. A reduction of the average values of microorganisms (approximately 250%) occurred in patients in groups 3 and 4 who had a longer colostomy duration (Table 2).

The tissue cultures confirmed what was expected—that the most frequently found bacteria would be Escherichia coli, Bacteroides fragilis, Bacteroides spp, and Peptococcus spp. These organisms are the most common bacteria in fecal material cultures and the main source of bacteria for the colonization of the pericolostomic dermis.8–10 Other bacteria less frequently isolated include Bifidobacterium spp, Staphylococcus epidermidis, Eubacterium spp, Fusobacterium spp, and Pseudomonas aeruginosa. A percentage of these organisms were insufficient to allow for statistical analysis. Experimental data have shown that a level of bacterial growth > 105 organisms per gram of tissue is necessary to cause wound infection.11 These results are the only found on this subject in the literature that employ these methods and define the bacterial flora present in the pericolostomy skin region.

This study was developed with a restricted group of colostomized patients, ie, those who did not present local complications in the ostomy and who were in good health. These factors reduced the risk of local or systemic infection. New studies employing the same methodology, but involving patients with other types of ostomies or even adverse situations, such as pericolostomy complications or systemic involvement, should be developed to extend the knowledge of this surgical condition in a wider group of patients with an ostomy.


The skin of patients with an ostomy is heavily colonized and increases the risk of surgical site infection. Results of this study suggest that enteric bacteria colonize the dermal layer of the abdominal wall in the pericolostomic region, with a sufficient number of microorganisms in tissue to cause local infection. With bacteria in the pericolostomic skin region identified and measured, we can provide colostomized patients adequate treatment in cases of local infection. This is different from what occurs presently, in which antibiotic therapy is used empirically and is often based on the experience of the stomatotherapist or the surgeon, aside from the possibility of reducing the rate of infection with a specific antimicrobial treatment.


Back to top