Colorectal Surgical Site Infections and Their Causative Pathogens: Differences between Left- and Right-Side Resections

Abstract

Background:

Surgical site infections (SSIs), after colorectal resection, pose a significant burden. Recognition of the spectrum of potentially involved pathogens is crucial for determining correct antibiotic prophylaxis. This study aimed to determine whether the distribution of SSI-associated pathogens depends on the location of the colorectal resection.

Methods:

We retrospectively categorized 2713 consecutive colon resections as left- or right-side operations, identified patients having concurrent peritonitis or development of postoperative SSIs and peritonitis, and assigned all subcutaneously and intra-abdominally isolated pathogens according to the location of the resection.

Results:

Gram-positive cocci (especially enterococci) and gram-negative bacilli (especially Pseudomonas aeruginosa) were isolated more frequently from patients in whom SSIs developed after left-side resections than after right-side resections (52.5% vs. 32.6%, p < 0.01 and 15.9% vs. 6.7%, p < 0.01, respectively); enterococci were the causative organisms in a large percentage of SSIs (46.3%). Moreover, intra-abdominal P. aeruginosa and Candida spp. were isolated more frequently during left-side resections than during right-side operations in patients with peritonitis (15.8% vs. 6.3%, p = 0.02 and 14.3% vs. 5.3%, p = 0.02, respectively).

Conclusions:

Our results indicate that differences exist in the distribution of pathogens after left- or right-side colorectal resections. Our data further suggest that gram-positive cocci play an important role in SSIs occurring after colorectal resections; therefore, antibiotic prophylaxis should emphasize their coverage. Further, enterococcal coverage may be especially advantageous during left-side resections.

Despite advances in antisepsis and operating techniques, surgical site infections (SSIs) are still a relevant problem for patients undergoing colorectal surgical procedures. Although SSIs are among the most preventable health-care–associated infections [1], they still represent a significant burden in terms of patient morbidity and death and result in additional costs to health systems and service payors, worldwide [2]. In some studies, SSIs are reported to occur after up to 26% of colorectal surgical procedures [3] and may affect the skin as well as subcutaneous tissue and deeper structures associated with the abdominal cavity [4]. Their occurrence usually leads to increased hospitalization durations and may jeopardize the patient's life [5].

Factors, such as hospitalization and previous antibiotic therapy, influence the microbiology of SSIs [6,7]. Thus, the microbiology of the infection, combined with increasing pathogen resistance, may limit empiric therapy effectiveness. Therefore, knowing the spectrum of potential pathogens is crucial to selecting appropriate treatment. Recently, large register-based studies have demonstrated differences in SSI incidence depending on whether the rectum was included in the resection [8]; higher SSI rates are normally associated with rectal resections than with resections including only the right hemicolon. Conversely, other studies have reported a higher incidence of SSIs in resections involving the small bowel [9].

This study investigated whether the distribution of sub-cutaneous pathogens in patients with SSIs, as well as the abdominal pathogens involved in patients with existing peritonitis, depend on the location of the colorectal resection. We, therefore, retrospectively categorized colorectal resections as either left- or right-side operations and analyzed all the recovered subcutaneous and intra-abdominal pathogens.

Methods

All included colorectal resections (n = 2713) were performed at a district teaching hospital between January 2008 and September 2016. The procedures were categorized as resections within the colorectum (“left hemicolon”) or between the small intestine and colon (“right hemicolon”). Simultaneous resections, involving both sides, were excluded (n = 29); subtotal resections involving ileorectostomies were classified as left hemicolon procedures because the rectum was involved and a stapled, transanal anastomosis was created (n = 123). Isolated colon transversum resections were classified as right hemicolon procedures, because they excluded the distal colon and rectum and hand-sewn anastomoses were created (n = 12). The remaining classifications are shown in Table 1.

Table 1.

Intra- and Post-Operative Characteristics, Depending on Location of the Colorectal Resection

	Left hemicolon		Right hemicolon
	n = 1932	%	n = 751	%	p
Approach
Laparoscopic	977	50.5	322	42.9
Open	848	43.8	264	35.2
Conversion	108	5.6	165	22.0	<0.01
Procedure
Colectomy	65	3.4
Sub-total colectomy	67	3.5
Rectal resection	667	34.5
Left colectomy	140	7.2
Rectosigmoid resection	729	37.7
Abdominoperineal extirpation	61	3.2
Right hemicolectomy			575	76.6
Resection of colon transversum			12	1.6
Ileocecal resection			152	20.2
Restoration of continuity	49	2.5	12	1.6
Other	154	8.0	0
Colostomy	241	12.5	0
Ileostomy	122	6.3	12	1.6
Length of post-operative stay (d)	11 (1–215)		12 (2–345)		0.11

Right-side resections involved hand-sewn, terminolateral or laterolateral anastomoses in both conventional and laparoscopic procedures. If an incision required enlargement to >10 cm, it was considered a conversion. Left-side resections with rectal anastomoses typically involved stapled, transanal anastomoses.

All patients were intra- and post-operatively treated in the same department by the same personnel. The SSI surveillance was performed by the hygiene department and reported to the German National Reference Center for Surveillance of Nosocomial Infections. For all patients with colorectal cancer, simultaneous SSI surveillance was conducted by the regional oncology registrar. For patients (n = 519) included in previous clinical studies, these data were used to verify SSIs and complications [10 –12]. In addition, the electronic data system, including both in- and outpatient data, was screened for wound-related procedures.

All data were merged, and contradictory results checked against the electronic patient data system. Further, all patients with a positive subcutaneous smear, obtained via a revision laparotomy, were counted as having SSIs; this captured as many infections as possible. Moreover, basic patient data, including pre-existing diseases and complications, were exported from the electronic patient data system and analyzed. We also investigated whether additional abdominal procedures were performed within 30 days of the primary operation. A causal differentiation was not performed because of inconsistent coding.

When an SSI was diagnosed, the treating team was urged to obtain a sub-cutaneous swab on the day of diagnosis (BD ESwab™, Copan Italia, Brescia, Italy). During procedures involving patients with existing peritonitis and in patients undergoing re-laparotomies or re-laparoscopies, intra-abdominal swabs were taken routinely, in the same manner. All detected pathogens (n = 4425) were analyzed and assigned according to their intra-abdominal or sub-cutaneous origin. Resistograms were performed for 2399 of these pathogens. Cephalosporins were not included in standard resistogram-patterns for enterococci, however; these were assumed to not be efficacious.

In patients undergoing resections, during the study period, the bowel was prepared with small enemas (Freka Clyss^®, Fresenius Kabi Deutschland, Bad Homburg, Germany) in the evening before and the morning of operation. Oral antibiotic agents were not used at that time.

During elective resections, pre-operative antibiotic prophylaxis involved intravenous ampicillin (2 g) and sulbactam (1 g), with re-administration occurring for long operations; in patients with penicillin allergies, intravenous levofloxacin (500 mg) and metronidazole (500 mg) were administered. Patients, with accompanying peritonitis, undergoing emergency or early-elective operation usually received intravenous ceftriaxone (2 g) and metronidazole (500 mg) therapy. Pre-operative skin preparation involved standardized application of at least three swabs, soaked with disinfectant (Softasept^® N, B. Braun, Melsungen, Germany), to wash the abdominal skin. Post-operatively, primary wound closure was performed: the subcutaneous tissue was first irrigated with plain saline, containing polyhexanide; subcutaneous drains or subcutaneous sutures were not used. Thereafter, the skin was closed using a skin stapler. In SSI cases, the wound was opened later and healing by secondary intention was attempted.

Categoric variables are expressed as percentages, and continuous variables as medians (range). The Pearson chi-squared test was used to evaluate categoric variables; for continuous variables, parametric t tests or non-parametric U tests (Mann-Whitney-Wilcoxon tests) were used. Statistical significance was defined as p < 0.05.

All procedures were performed in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards. The need for written informed consent was waived by the local ethics committee because this retrospective cohort study used only pseudonymized clinical data.

Results

During the study period, 1932 left hemicolon resections (71.2%) and 751 right hemicolon resections (27.7%) were performed. Left-side resections were more common than right-side resections due to patients having indications with attendant peritonitis (29.8% vs. 24.1%, p < 0.01). The SSI incidence after left-side resections (20.4%) was lower than that after right-side resections (25.2%, p < 0.01); patients undergoing left-side resections were also less likely to undergo a second operation (13.9% vs. 16.9%, p = 0.05).

The proportions of male and female patients in the left- and right-side resection groups were equal, but their mean ages differed slightly (Table 2). A history of pulmonary disease was more common in the left-side resection group, but no other differences in pre-existing conditions were observed between the groups. The left-side resection group most commonly underwent successful laparoscopic (50.5%) or open (43.8%) surgical approaches; conversions occurred more often during right-side resections (22.0%). The two groups had similar post-operative lengths of stay (Table 1).

Table 2.

Baseline Data of All Included Patients, Depending on Location of the Colorectal Resection

	Left hemicolon		Right hemicolon
	n = 1932	%	n = 751	%	p
Age (y)	66.1 ± 13.1		68.0 ± 15.5		<0.01
Gender female	1065	55.1	392	52.2	0.17
Pre-existing disease
Pulmonary	468	24.2	147	19.6	0.01
Cardiovascular (incl. hypertension)	1138	58.9	461	61.4	0.23
Diabetes mellitus	262	13.6	109	14.5	0.52
Hepatic	120	6.2	39	5.2	0.32
Renal	195	10.1	80	10.7	0.67
Obesity					0.69
I°	214	11.1	92	12.3
II°	63	3.3	20	2.7
III°	20	1.0	9	1.2
Reason for resection					<0.01
Malignoma	703	36.4	399	53.1
Colorectal cancer	626	32.4	367	48.9
Stadium 0	8	1.3	0	0.0
Stadium I	146	23.3	81	22.1
Stadium II	205	32.7	115	31.3
Stadium III	163	26.2	93	25.3
Stadium IV	103	16.5	77	21.0
Stadium X			1	0.3
Benign neoplasia	65	3.4	122	16.3
Diverticular disease	834	43.2	19	2.5
Acute appendicitis	1	0.1	60	8.0
IBD	22	1.1	33	4.4
Fistula	26	1.4	3	0.4
Ileus	52	2.7	28	3,7
Other	229	12.9	87	11.6

IBD = inflammatory bowel disease.

Peritonitis, present at the time of the primary resection, existed in 757 patients (28.2%) because of their causative diseases. In 298 (39.4%) of these patients, one or more intra-abdominal pathogens were isolated. The most frequently isolated bacteria were Enterobacteriaceae (64.1%) and anaerobic rods (61.8%). Gram-negative bacilli (e.g., Pseudomonas aeruginosa) and Candida spp. were found most often during left-side resections (Table 3). Among these 757 patients, SSIs developed in 223 (29.5%).

Table 3.

Spectrum of Pathogens Isolated from Intra-Abdominal Swabs of Patients Who Have Undergone Colorectal Resections because of an Indication with Attendant Peritonitis

	Left hemicolon		Right hemicolon
	n = 203	%	n = 95	%	p
Gram-positive cocci	123	60.6	57	60.0	0.92
Gram-positive cocci, excl. enterococci	60	29.6	34	35.8	0.28
MSSA	2	1.0	4	4.2
MRSA	1	0.5	1	1.1
CNS	12	5.9	6	6.3
Streptococcus spp.	46	22.7	26	27.4
Enterococci	75	37.0	27	28.4	0.15
Enterococcus faecalis	35	17.2	13	13.7
Enterococcus faecium	31	15.3	9	9.5
Enterococcus spp.	14	6.9	7	7.4
VRE	2	1.0	0

Anaerobic gram-positive bacteria	15	7.4	9	9.5	0.54
Clostridium perfringens	3	1.5	0	-
Clostridium spp.	11	5.4	6	6.3
Peptostreptococcus spp.	1	0.5	3	3.2
Gram-positive bacilli	5	2.5	2	2.1	0.85
Gram-negative bacilli, enterobacteriaceae	126	62.1	65	68.4	0.29
Escherichia coli	110	54.2	53	55.8
Klebsiella pneumonia	10	4.9	6	6.3
Klebsiella oxytoca	3	1.5	2	2.1
Enterobacter aerogenes	2	1.0	1	1.1
Enterobacter cloacae	3	1.5	4	4.2
Proteus spp.	5	2.5	1	1.1
Proteus mirabilis	4	2.0	4	4.2
Citrobacter freundii	5	2.5	4	4.2
Morganella morgagnii	7	3.5	3	3.2
ESBL	5	2.5	5	2.5
Gram-negative bacilli, non-fermenting	32	15.8	6	6.3	0.02
Pseudomonas aeruginosa	32	15.8	6	6.3
Anaerobic gram-negative bacilli	124	61.1	60	63.2	0.73
Bacteroides fragilis	50	24.6	15	15.8
Bacteroides (not fragilis)	82	40.4	47	49.5
Prevotella spp.	10	4.9	4	4.2
Other bacteria	14	6.9	6	6.3	0.85
Molds	29	14.3	5	5.3	0.02
Candida (not albicans)	12	5.9	2	2.1
Candida albicans	17	8.4	4	4.2

MSSA = methicillin-sensitive Staphylococcus aureus; MRSA = methicillin-resistant Staphylococcus aureus; CNS = coagulase-negative staphylococci; VRE = vancomycin-resistant Enterococcus faecium; ESBL, extended-spectrum beta-lactamases.

Proportion of patients with at least one recovered isolate of the corresponding pathogen, relative to the total number of patients from which bacterial species were isolated.

Re-laparotomy or re-laparoscopy was performed in 396 patients (14.8%) (Table 4). The reasons for the repeat surgical procedures included anastomotic leakage, persisting peritonitis, post-operative ileus, or intra-abdominal hematoma. Pathogens were isolated in 282 cases (71.2%); the most commonly recovered bacteria were enterococci (51.4%) and Enterobacteriaceae (50.7%). Although the presence of enterococci was higher than that noted in patients undergoing primary procedures, the frequency of isolating anaerobic species was lower. Differences were observed between the left- and right-side resection groups relative to the frequency of P. aeruginosa isolation; a trend was also seen for differences in the frequency of gram-positive cocci isolation, excluding enterococci. Both were more frequently isolated from patients undergoing left-side colon resection.

Table 4.

Spectrum of Pathogens Isolated from Intra-Abdominal Swabs of Patients Undergoing Subsequent Re-Laparotomy/-Scopy after Colorectal Resections

	Left hemicolon		Right hemicolon
	n = 194	%	n = 88	%	p
Gram-positive cocci	144	74.2	57	64.8	0.10
Gram-positive cocci, excl. enterococci	73	37.6	23	26.1	0.06
MSSA	17	8.8	1	1.1
MRSA	2	1.0	1	1.1
CNS	45	23.2	15	17.1
Streptococcus spp.	17	8.8	11	12.5
Enterococci	98	50.5	47	53.4	0.65
Enterococcus faecalis	45	23.2	20	22.7
Enterococcus faecium	60	30.9	32	36.4
Enterococcus spp.	10	5.2	2	2.3
VRE	1	0.5	3	3.4
Anaerobic gram-positive bacteria	13	6.7	2	2.3	0.12
Clostridium perfringens	2	1.0	1	1.1
Clostridium spp.	8	4.1	2	2.3
Peptostreptococcus spp.	4	2.1	1	1.1
Gram-positive bacilli	5	2.6	3	3.4	0.70
Propionibacterium spp.	2	1.0	0	-
Corynebacterium spp.	3	1.6	3	3.4
Gram-negative bacilli, Enterobacteriaceae	96	49.5	47	53.4	0.54
Escherichia coli	70	36.1	38	43.2
Klebsiella pneumoniae	12	6.2	3	3.4
Klebsiella oxytoca	6	3.1	2	2.3
Enterobacter spp.	1	0.5	1	1.1
Enterobacter cloacae	10	5.2	1	1.1
Proteus spp.	3	1.6	1	1.1
Proteus mirabilis	4	2.1	2	2.3
Citrobacter freundii	4	2.1	0	-
Citrobacter spp.	5	2.6	3	3.4
Morganella morgagnii	7	3.6	1	1.1
ESBL	9	4.6	0
Gram-negative bacilli, non-fermenting	24	12.4	1	1.1	<0.01
Pseudomonas aeruginosa	24	12.4	1	1.1
Anaerobic gram-negative bacteria	69	35.6	35	39.8	0.50
Bacteroides fragilis	29	15.0	13	14.8
Bacteroides (not fragilis)	47	24.2	23	26.1
Prevotella spp.	7	3.6	2	2.3
Other bacteria	11	5.7	10	11.4	0.09
Molds	44	22.7	18	20.5	0.68
Candida (not albicans)	19	9.8	9	10.2
Candida albicans	29	15.0	12	13.6

Proportion of patients with at least one recovered isolate of the corresponding pathogen, relative to the total number of patients from which bacterial species were isolated.

Among the 396 patients undergoing re-operations, in 60 cases (15.2%), pathogens were detected during both the primary and subsequent procedures. Identical species were detected during both procedures in 55 patients (91.7%), with the most frequently recovered pathogens being enterococci (15 cases), Enterobacteriaceae (16 cases), and molds (9 cases).

Overall, SSIs occurred in 583 patients (21.7%); in 430 of these (73.8%), one or more bacterial species were isolated from sub-cutaneous swabs. Gram-positive cocci, excluding enterococci, were found in 49.1% of cases, and enterococci and Enterobacteriaceae were found in 46.3% and 42.6%, respectively. The most frequently isolated organisms were Escherichia coli (31.2%), Enterococcus faecalis (27.8%), and E. faecium (23.7%). Patients with SSIs after left-side resections had higher proportions of gram-positive cocci, enterococci, and gram-negative bacilli, such as P. aeruginosa and Stenotrophomonas maltophilia (Table 5). Among the 583 patients with SSI, in 72 cases (12.3%), pathogens were detected during both the primary operation and the occurrence of SSI. Identical species were found in 54 patients (75.0%), with the most frequently recovered pathogens being enterococci (19 cases), Enterobacteriaceae (16 cases), and molds (six cases).

Table 5.

Spectrum of Pathogens Isolated from Sub-Cutaneous Swabs of All Patients with Surgical Site Infections after Colorectal Resections

	Left hemicolon		Right hemicolon
	n = 295	%	n = 135	%	p
Gram-positive cocci	240	81.4	94	69.6	<0.01
Gram-positive cocci, excl. enterococci	148	50.2	63	46.7	0.50
MSSA	23	7.8	10	7.4
MRSA	9	3.1	1	0.7
CNS	113	38.3	50	37.0
Streptococcus spp.	31	10.5	9	6.7
Enterococci	155	52.5	44	32.6	<0.01
Enterococcus faecalis	82	27.8	21	15.6
Enterococcus faecium	78	26.4	24	17.8
Enterococcus spp.	14	4.8	1	0.8
VRE	0		1	0.8
Anaerobic gram-positive bacteria	13	4.4	2	1.5	0.12
Clostridium perfringens	3	1.0	0	-
Clostridium spp.	5	1.7	2	1.5
Peptostreptococcus spp.	4	1.4	0	-
Gram-positive bacilli	6	2.0	2	1.5	0.69
Gram-negative bacilli, Enterobacteriaceae	126	42.7	57	42.2	0.92
Escherichia coli	92	31.2	45	33.3
Klebsiella pneumonia	12	4.1	6	4.4
Klebsiella oxytoca	11	3.7	1	0.7
Enterobacter spp.	5	1.7	2	1.5
Enterobacter cloacae	9	3.1	5	3.7
Proteus spp.	4	1.4	1	0.7
Proteus mirabilis	9	3.1	2	1.5
Citrobacter freundii	7	2.4	1	0.7
Morganella morgagnii	12	4.1	3	2.2
ESBL	16	5.4	7	5.2
Gram-negative bacilli, non-fermenting	47	15.9	9	6.7	<0.01
Pseudomonas aeruginosa	43	14.6	8	5.9
Stenotrophomonas maltophilia	4	1.4	1	0.7
Anaerobic gram-negative bacteria	89	30.2	45	33.3	0.51
Bacteroides fragilis	49	16.6	32	23.7
Bacteroides (not fragilis)	42	14.2	18	13.3
Prevotella spp.	8	2.7	1	0.7
Other bacteria	20	6.8	6	4.4	0.35
Molds	41	13.9	13	9.6	0.22
Candida (not albicans)	19	6.4	6	4.4
Candida albicans	26	8.8	9	6.8

MSSA = methicillin-sensitive Staphylococcus aureus; MRSA = methicillin-resistant Staphylococcus aureus; CNS = coagulase-negative staphylococci; VRE, = vancomycin-resistant Enterococcus faecium; ESBL = extended-spectrum beta-lactamases.

Proportion of patients with at least one recovered isolate of the corresponding pathogen, relative to the total number of patients from which bacterial species were isolated.

Further, we investigated the number of recovered pathogens that were resistant to commonly used antibiotic agents. Table 6 shows the proportion of patients with at least one recovered isolate that was resistant to the indicated antibiotic agent, relative to the total number of patients from which bacterial species were isolated and for which antibiograms were available. Resistances against levofloxacin were found more commonly after left-side resections than after right-side resections among patients with pre-existing peritonitis, and among patients in whom SSIs developed after the primary resection; ceftriaxone resistance was more common after left-side resections in patients with SSIs. Resistances against ampicillin and sulbactam were found more commonly after left-side resections among patients with pre-existing peritonitis. Antibiotic resistant bacteria were detected more often after re-operations and in patients with SSIs than in patients undergoing primary resections.

Table 6.

Proportion of Patients with at Least One Resistant Bacteria against Selected Antibiotic Agents

	Left hemicolon		Right hemicolon
Resistance against	n	%	n	%	p
Primary resection, attendant peritonitis (intra-abdominal swab)
Ceftriaxone	85 of 157	54.1	34 of 72	47.2	0.33
Ampicillin/sulbactam	86 of 153	56.2	30 of 72	41.7	0.04
Levofloxacin	62 of 158	39.2	18 of 72	25.0	0.04
Meropenem	6 of 137	4.4	2 of 64	3.1	0.67
Re-laparotomy/-scopy (intra-abdominal swab)
Ceftriaxone	111 of 152	73.0	49 of 67	73.1	0.99
Ampicillin/sulbactam	102 of 150	68.0	41 of 65	63.1	0.48
Levofloxacin	96 of 160	60.0	35 of 67	52.2	0.28
Meropenem	18 of 111	16.2	5 of 45	11.1	0.42
Surgical-site infection (subcutaneous swab)
Ceftriaxone	185 of 230	80.4	63 of 102	61.8	<0.01
Ampicillin/sulbactam	166 of 227	73.1	68 of 100	69.0	0.34
Levofloxacin	153 of 244	62.7	52 of 107	48.6	0.01
Meropenem	36 of 164	22.0	12 of 75	16.0	0.29

Proportion of patients with at least one recovered isolate that was resistant to the indicated antibiotic agent, relative to the total number of patients from which bacterial species were isolated and for which antibiograms were available.

Discussion

The results obtained in our study provide insight into the microbiology of colorectal-derived infections. After analyzing data from a large, consecutive series of resections, our data demonstrate that differences exist in the distribution of pathogens, depending on whether the ileum was involved in the corresponding operation. Previous reports suggest some differences exist in the species isolated from the upper gastrointestinal tract compared with the lower gastrointestinal tract [13], reflecting differences in the native flora of each location [14]. We were interested specifically in microbiologic differences subsequent to resections involving the left or right hemicolons, however, excluding and including the native flora of the ileum.

The rationale for the observed differences includes the reported elevation of SSI rates associated with small bowel resections [9]. The anastomosis technique used differed between the two groups, however; right-side resections are usually performed with an open, hand-sutured technique in both laparoscopic and conventional surgical procedures, whereas left-side resections are performed using stapled, closed transections and stapled, transanal anastomoses. Therefore, the potential period for contamination is significantly longer during the former procedure.

The two surgical groups (left- and right-side resections), in the present study, did not differ in sex distribution, pre-existing disease burden (except pulmonary diseases), and post-operative lengths of stay. The statistically significant difference in mean age (1.9 years) between the groups was not clinically relevant. Right-side colon resections, however, demonstrated more frequent conversions to conventional, open approaches than did left-side conversions. This difference may be explained by the more difficult radical oncologic resections required for patients with right colon cancer; left-side resections included sigmoid resections that were standardized to be initiated laparoscopically.

Among the intra-abdominal pathogens isolated from patients with ongoing peritonitis during their primary surgical procedures, aerobic and anaerobic gram-negative bacilli were found most commonly. Differences between the right- and left-side resection groups included the isolation of yeasts and gram-negative bacilli, such as P. aeruginosa, not derived from the gastrointestinal tract; both pathogen types were isolated more frequently from patients undergoing left-side operations.

Previous reports indicated that the isolation of Candida spp. is associated with esophageal and stomach procedures, anaerobic antibiotic coverage, and lack of intra-abdominal focus control [15,16]. In those studies, Candida spp. was found to be more frequently associated with surgical procedures involving the upper gastrointestinal tract, as opposed to the lower tract. In our study, Candida spp. accounted for 11.4% of the isolates and was notably isolated with peritonitis associated with left-side colon surgery. In subsequent procedures, the recovery of Candida spp. was even more frequent (22.0%).

To date, empiric coverage of molds is recommended only in patients with hospital-acquired peritonitis after recurrent bowel perforations or perforations of the upper gastrointestinal tract [17]. Based on our data, empiric coverage of Candida spp. might also be considered, for example, after perforation of the left hemicolon necessitating re-laparotomy. The same applies for empiric coverage of Pseudomonas spp.

The SSI incidence, in the present study, seems rather high (21.7%), compared with register-based studies [8]. It is similar to that reported in other studies focused on the spectrum of bacteria involved in SSIs, however [18]. Following the criteria of the National Reference Center for the Surveillance of Nosocomial Infections, the rates in our population are significantly lower. For conventional and laparoscopic procedures, the SSI incidences for colon resections were, respectively, 11.9% and 14.8%, and 6.7% and 4.1% for rectal resections. These differences to our study are because of, for example, the end of the surveillance if a subsequent re-operation is performed and the exclusion of multi-visceral resections, both of which are situations of high-risk for SSIs. As we aimed to encompass the full spectrum of pathogens and not to report comparable SSI rates, we included all SSI cases.

The SSI rate for subcutaneously detected pathogens was higher after right-side resections than after left-side procedures. This contradicts previously published register data and is difficult to explain because the bacterial concentration in the rectum should be 10-fold higher than on the right side [8,19 –21]. An earlier trial involving a similar population, however, with 30-day follow-up results and SSI frequencies as the primary endpoints, showed results similar to those presented here [11]. The difference from the expected results might be related to the high frequency of intended laparoscopic right-side resections (64.9%) involving extracorporeal, hand-sewn anastomoses and the high risk of wound contamination.

Consistent with the results of other studies, gram-positive species, such as staphylococci, streptococci, and enterococci, were the most commonly isolated pathogens from patients with SSIs, followed by aerobic and anaerobic gram-negative bacilli [13,22]. The microbiology of SSIs after left-side resections showed some significant differences compared with that of SSIs after right-side resections. Gram-positive cocci, especially enterococci, were found more often after left-sided resections, as were gram-negative bacilli (P. aeruginosa and S. maltophilia).

The higher proportion of patients in whom enterococcal infections developed after left-side resections (52.5% vs. 32.6%) could be because of the pre-operative antibiotic treatment of patients with complicated diverticulitis of the sigmoid colon. These patients are usually treated pre-operatively with cephalosporins and undergo the surgical procedure a short while later. Cephalosporins are not active against enterococci and may favor their appearance [23], coinciding with the higher proportion of ceftriaxone resistance observed after left-side resections in patients with SSIs.

Because enterococci-associated SSIs were common in both resection groups, this may suggest that coverage of these pathogens with antibiotic prophylaxis should be recommended. Thus far, however, this type of coverage is not recommended in the existing guidelines [24]. The high rate of E. faecium involvement makes providing appropriate antibiotic therapy even more difficult because only a few antibiotic agents have sufficient efficacy against this species. In addition, gram-negative bacilli, such as pseudomonads, are likewise not covered by commonly used antibiotic agents. If risk factors for Pseudomonas-associated SSIs exist, these need to be investigated in future studies to identify patients who might benefit from anti-pseudomonal prophylaxis. Nevertheless, the isolated pathogens were somewhat similar after either right- or left-side resections, justifying the use of a similar antibiotic prophylaxis regimen for both circumstances [24].

The marked diversity of pathogens potentially involved in these infections highlights the risk of inappropriate empiric therapy, which occurs in approximately 15% of intra-abdominal infections [25 –27]. This, in turn, could lead to increased death, often because of infections with antibiotic-resistant gram-negative, beta-lactamase–producing bacteria. In particular, the prevalence of AmpC-type beta-lactamases increases after exposure to cephalosporins, which are generally used in therapy and prophylaxis [23]. In addition, there are progressively more extended-spectrum beta-lactamase–producing Enterobacteriaceae causing community-acquired peritonitis [28]. Other possible causes are beta-lactam- or vancomycin-resistant enterococci or the appearance of molds [25].

Recently, published guidelines have provided recommendations regarding coverage of these species that are dependent on the patient's risk category [17]. To establish the most suitable empiric antibiotic treatment, knowing the specific spectrum of pathogens associated with each procedure is important. The most important factor contributing to the development of post-operative peritonitis involving multi-resistant bacterial species, however, is the application of broad-spectrum antibiotic agents at the time of the initial operation [6]. Therefore, well-differentiated and appropriate use of antibiotic agents is mandatory.

The main limitation of this study is that we did not distinguish emergency surgical procedures from elective operation. Further, we were unable to collect certain treatment variables, such as the type and length of pre-operative outpatient antibiotic treatment. Therefore, we could not prove a correlation between bacterial isolates and pre-operative infection treatments. To better distinguish possible differences in the detected spectrum potentially depending on pre-treatment, additional surveys that consider the different causes of resections are required.

Conclusion

The pathogens isolated in conjunction with subcutaneous SSIs, as well as in peritonitis occurring during primary or secondary surgical procedures, showed some differences when comparing left- and right-side colon resections. Our data suggest that gram-positive cocci play an important role in SSIs; therefore, antibiotic prophylaxis in colorectal resections should emphasize their coverage. Enterococci and Pseudomonas spp. were found more frequently in SSIs developing after left-side resections, which involved the colon and excluded the ileum. P. aeruginosa and Candida spp. were isolated more frequently from patients with pre-existing peritonitis undergoing left-side operations. Enterococci were often found in SSIs (>32%), after both left- and right-side colon resections; their coverage could be advantageous, especially in patients requiring left-side resections. The information obtained from this study allows a better understanding of the etiology of surgical infections in patients undergoing colorectal resections and may have therapeutic implications.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Harbarth

, Sax

, Gastmeier

. The preventable proportion of nosocomial infections: An overview of published reports. J Hosp Infect, 2003; 54:258–266.

Allegranzi

, Bagheri Nejad

, Combescure

, et al. Burden of endemic health-care-associated infection in developing countries: Systematic review and meta-analysis. Lancet, 2011; 377:228–241.

Smith

, Bohl

, McElearney

, et al. Wound infection after elective colorectal resection. Ann Surg, 2004; 239:599–605.

Horan

, Gaynes

, Martone

, et al. CDC definitions of nosocomial surgical site infections, 1992: A modification of CDC definitions of surgical wound infections. Infect Control Hosp Epidemiol, 1992; 13:606–608.

Fry

. The economic costs of surgical site infection. Surg Infect, 2002; 3(Suppl 1):S37–S43.

Augustin

, Kermarrec

, Muller-Serieys

, et al. Risk factors for multidrug resistant bacteria and optimization of empirical antibiotic therapy in postoperative peritonitis. Crit Care, 2010; 14:R20.

Montravers

, Lepape

, Dubreuil

, et al. Clinical and microbiological profiles of community-acquired and nosocomial intra-abdominal infections: Results of the French prospective, observational EBIIA study. J Antimicrob Chemother, 2009; 63:785–794.

Murray

, Pasam

, Estrada

, Kiran

. Risk of surgical site infection varies based on location of disease and segment of colorectal resection for cancer. Dis Colon Rectum, 2016; 59:493–500.

Leaper

, van Goor

, Reilly

, et al. Surgical site infection—a European perspective of incidence and economic burden. Int Wound J, 2004; 1:247–223.

10.

Pochhammer

, Weller

, Schäffer

. Polihexanide for prevention of wound infection in surgery. Is the contact time essential? POLIS-trial: A historic controlled, clinical pilot trial. Wound Med, 2016; 14:19–24.

11.

Pochhammer

, Zacheja

, Schaffer

. Subcutaneous application of gentamicin collagen implants as prophylaxis of surgical site infections in laparoscopic colorectal surgery: A randomized, double-blinded, three-arm trial. Langenbecks Arch Surg, 2015; 400:1–8.

12.

Diener

, Knebel

, Kieser

, et al. Effectiveness of triclosan-coated PDS Plus versus uncoated PDS II sutures for prevention of surgical site infection after abdominal wall closure: The randomised controlled PROUD trial. Lancet, 2014; 384:142–152.

13.

Munez

, Ramos

, Espejo

, et al. [Microbiology of surgical site infections in abdominal tract surgery patients]. (Spa) Cir Esp, 2011; 89:606–612.

14.

Qadan

, Cheadle

. Common microbial pathogens in surgical practice. Surg Clin North Am, 2009; 89:295–310, vii.

15.

Swenson

, Metzger

, Hedrick

, et al. Choosing antibiotics for intra-abdominal infections: What do we mean by “high risk”?. Surg Infect, 2009; 10:29–39.

16.

Montravers

, Dupont

, Gauzit

, et al. Candida as a risk factor for mortality in peritonitis. Crit Care Med, 2006; 34:646–652.

17.

Mazuski

, Tessier

, May

, et al. The Surgical Infection Society Revised Guidelines on the Management of Intra-Abdominal Infection. Surg Infect, 2017; 18:1–76.

18.

Kalakouti

, Simillis

, Pellino

, et al. Characteristics of surgical site infection following colorectal surgery in a tertiary center: Extended-spectrum beta-lactamase-producing bacteria culprits in disease. Wounds, 2017; 29:108–113.

19.

Gomila

, Carratala

, Camprubi

, et al. Risk factors and outcomes of organ-space surgical site infections after elective colon and rectal surgery. Antimicrob Resist Infect Control, 2017; 6:40.

20.

Goto

, Hasegawa

, Hata

, et al. Differences in surgical site infection between laparoscopic colon and rectal surgeries: Sub-analysis of a multicenter randomized controlled trial (Japan-Multinational Trial Organization PREV 07-01). Int J Colorectal Dis, 2016; 31:1775–1784.

21.

Degrate

, Garancini

, Misani

, et al. Right colon, left colon, and rectal surgeries are not similar for surgical site infection development. Analysis of 277 elective and urgent colorectal resections. Int J Colorectal Dis, 2011; 26:61–69.

22.

Schnuriger

, Inaba

, Eberle

, et al. Microbiological profile and antimicrobial susceptibility in surgical site infections following hollow viscus injury. J Gastrointest Surg, 2010; 14:1304–1310.

23.

Dahms

, Johnson

, Statz

, et al. Third-generation cephalosporins and vancomycin as risk factors for postoperative vancomycin-resistant enterococcus infection. Arch Surg, 1998; 133:1343–1346.

24.

Ban

, Minei

, Laronga

, et al. American College of Surgeons and Surgical Infection Society: Surgical Site Infection Guidelines, 2016 Update. J Am Coll Surg, 2017; 224:59–74.

25.

Ibrahim

, Sherman

, Ward

, et al. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest, 2000; 118:146–155.

26.

Bare

, Castells

, Garcia

, et al. Importance of appropriateness of empiric antibiotic therapy on clinical outcomes in intra-abdominal infections. Int J Technol Assess Health Care, 2006; 22:242–248.

27.

Sturkenboom

, Goettsch

, Picelli

, et al. Inappropriate initial treatment of secondary intra-abdominal infections leads to increased risk of clinical failure and costs. Br J Clin Pharmacol, 2005; 60:438–443.

28.

Canton

, Novais

, Valverde

, et al. Prevalence and spread of extended-spectrum beta-lactamase-producing Enterobacteriaceae in Europe. Clin Microbiol Infect, 2008; 14(Suppl 1):144–153.