Abdominal Wall Thickness: Is It Associated with Superficial and Deep Incisional Surgical Site Infection after Colorectal Surgery?

Abstract

Background:

Colorectal surgical procedures have a high rate of surgical site infection (SSI), and obesity has been implicated as a predictor of such infection. We hypothesized that abdominal wall thickness (AWT), as a metric of obesity, would predict postoperative superficial and deep incisional SSI after colorectal surgery, and conducted a study to assess superficial and deep incisional SSI and its relationship to abdominal wall thickness.

Methods:

To measure pre-operative AWT through cross-sectional imaging, and to analyze its relationship to SSI, we conducted a retrospective study at a single academic medical center of patients who had had colorectal resection for any indication in 2008 and 2009.

Results:

We identified 143 patients for inclusion in the study. Superficial or deep incisional SSI occurred in 43 patients (30%). Abdominal wall thickness at the midpoint between the umbilicus and pubis was associated with SSI (OR 1.03; p=0.014). Body-mass index (BMI) was also significantly associated with SSI (OR 1.08; p=0.014). Other significant (p<0.05) predictors of SSI by univariate analysis included a history of soft tissue infection, a surgical wound classification of 3 or 4, and lack of compliance with perioperative antibiotic guidelines. In a multivariable analysis of factors that were statistically significantly associated with SSI in univariate comparisons, lack of appropriate preoperative antibiotic administration independently predicted SSI (OR 4.33; 95% CI 1.08–17.40), but AWT and BMI were not significantly associated with SSI.

Conclusions:

Surgical site infection is common after colorectal surgery. Increased AWT predicts SSI by univariate analysis. Our findings could guide further studies of interventions that may decrease the risk of SSIs in patients with a thick abdominal wall.

Obesity is emphasized commonly as a risk factor for poor health outcomes. In the surgical literature, obesity is investigated as a risk factor for post-operative complications, particularly after abdominal surgery. However, multivariable analytic techniques do not demonstrate consistently an association of obesity with major complications of surgery. Some studies show a relationship of such complications to body mass index (BMI), but others show no clear relationship [1 –3]. One explanation for this discrepancy is variability in the way in which BMI is used in statistical analysis [4]. For example, obesity may be defined as a BMI >25 kg/m² [5,6 ] or >30 kg/m² [7], or may be analyzed in quartiles [8,9]. Another explanation for inconsistent results in studies of BMI and complications of surgery is that overweight or obese patients may be protected from complications through a reduced likelihood of being malnourished. Besides these factors, duration of follow-up leads to variability in the detection of SSI, which in up to 50% of cases may occur after discharge from the hospital [10].

A few authors have suggested recently that the inconsistent findings relating obesity to complications of surgery can be attributed to the choice of BMI as the metric of obesity. Such authors favor exploratory investigation of anthropometric measurements of adiposity, such as waist circumference [11], visceral adiposity [5,12,13], subcutaneous adiposity [14], or waist-to-hip circumference ratios as variables that may be more closely related to the risk of surgical complications than is BMI. In particular, two groups of Japanese authors have suggested that BMI, which tends to be low in the Japanese population, does not capture the impact of abdominal obesity on surgical outcomes, and have found that visceral obesity as measured by cross-sectional imaging is correlated more significantly with complications after abdominal surgery [5,14].

Surgical site infection (SSI) continues to be emphaized as an important healthcare–associated infection (HAI). Some cases may be preventable and some payors, such as Medicare, are unwilling to pay for care related to a SSI. Continued improved understanding of this complication serves several purposes. Colorectal surgical procedures have a high rate of SSI, reported in most studies as occurring in from 10%–20% of such procedures [6,7,15 –17], but some retrospective reviews of cases series at single institutions show higher rates [10,18 –20], and some of them have implicated obesity as a predictor of SSI [10,18,19,21].

Our study was designed to assess superficial and deep incisional SSI and its relationship to abdominal wall thickness (AWT) as a metric of obesity. Our hypothesis was that AWT would predict post-operative superficial and deep incisional SSI after colorectal surgery. Given the focus of the surgical infection literature on BMI rather than on other anthropomorphic measures of obesity, our study was an exploratory analysis of whether AWT should be examined more intensively as a risk factor for SSI in colorectal surgery.

Patients and Methods

The study was approved by the University of Minnesota Institutional Review Board. Patients undergoing elective or emergency colon or rectal resection by colorectal surgeons at the University of Minnesota in 2008 and 2009 were identified through procedure codes 45.7, 45.8, 45.9 of the International Classification of Diseases, Ninth Revision, Clinical Modification. A retrospective review of the patients' medical records was done with a standard data collection form. Inpatient and outpatient electronic medical records and, when available, documents from outside institutions, were reviewed. Pre-operative medications and medical diagnoses were identified through a review of all physician notes. Patients' BMIs were calculated from pre-operative data. Pre-operative cross-sectional imaging studies of the patients' abdomen and pelvis, done within 6 mos of surgery, were examined by a single investigator (M.R.K). Abdominal wall thickness is characterized poorly in the literature. Therefore, four locations at or below the umbilicus were used to measure AWT: In the midline immediately inferior to the umbilicus (AW1); halfway between the umbilicus and the pubis (AW2); at the level of the anterior superior iliac spine (AW3); and at the level of the umbilicus at the lateral border of the rectus abdominis muscle (AW4).

The practice of the surgeons whose patients were included in the study was to administer purgative bowel preparations, without oral antibiotics, on the day before surgery, but documentation of this for each patient was not available for review. Reports of the patients' operations were reviewed to categorize procedures as open, laparoscopic, or laparoscopic converted to open; whether a stoma was created or taken down; and whether the surgery did not create an anastomosis or whether it created multiple anastomoses. Compliance with appropriate pre-operative antibiotic guidelines was defined as the administration within 60 min of the time of surgical incision of cefoxitin, cefotetan, or cefazolin combined with metronidazole, or of ertapenem. Patients who received intravenous (IV) antibiotics to treat infection before surgery (16% of the patients in the study) were excluded from the determination of compliance. Wounds were classified as clean-contaminated (class 2); contaminated, defined as gross spillage or acute inflammation (class 3); or dirty, defined as purulent contamination or a perforated viscus (class 4). If a wound classification was not specified in the patient's chart, an assignment of it was made on the basis of the details of the surgeon's operative report.

Patients were excluded from the study if they had an incision that was left open at the end of the case, if concurrent surgery was done in another body cavity (i.e., thoracic surgery) was involved, if they had a colectomy performed during a re-operation done during the same hospitalization as their original operation, if they has a component separation or other complex wound closure procedure, or if they died within 7 d after undergoing surgery.

The outcome of interest was superficial or deep incisional SSI, as determined from a review of the patients' medical records. Superficial incisional SSI was judged to be present if purulent drainage from a surgical incision was documented; if organisms were isolated from a culture obtained aseptically of fluid or tissue from the incision; if tenderness, localized swelling, redness, or heat was present and the wound was subsequently opened deliberately by the surgeon; or if the diagnosis of a superficial incisional SSI was made by the attending surgeon. Deep incisional SSI was judged to be present if purulent drainage was seen from the deep part of an incision but not from the organ/space; if a deep incision dehisced spontaneously or was opened deliberately when the patient had fever or localized tenderness; if an abscess involving a deep portion of an incision was found on examination during re-operation or by radiology or histopathology; or if the diagnosis of a deep incisional SSI was made by the attending surgeon. Infections were recorded if they developed at any time in the patient's post-operative course, and the patient's outpatient medical record was used to assess for this complication after discharge from the hospital.

Statistical analyses

Statistical analyses were done with SAS version 9.2 (SAS Institute, Cary, NC). The Fisher exact test was used to compare groups, and univariate logistic regression or the t-test was used to evaluate the association of continuous variables with outcome. Logistic regression was used for multivariable analysis. Because BMI and AWT were variables of special interest in this exploratory study, they were investigated in isolation, without other predictors, through multivariable logistic regression. A multivariable analysis was then done with all variables that were statistically significant at a threshold of p<0.05 in univariate comparisons. Because BMI and AWT were correlated, only AWT was entered into the model in the multivariable analysis.

Results

The study was done on 143 patients who underwent surgery for colon (21%) or rectal (9%) cancer, diverticular disease (16%), ulcerative colitis (12%), Crohn disease (14%), or other conditions (27%). The mean age of the patients was 54 years (range 18–92 years) and 51.7% were female. A midline incision was used in most (93%) cases. Of the patients with a history of soft tissue infection (11.2%), half had a history of prior SSI after other abdominal operations, 37.5% had infectious complications at the upper or lower extremities ends of their surgical fields (37.5%), and 12.5% had previously had perianal or perineal soft tissue infections. Superficial or deep incisional SSI occurred in 43 (30%) of the 143 patients in the study. Patient characteristics and perioperative factors and their relationship to SSI are described in Tables 1 and 2. Surgical site infections were diagnosed at a median of 8 d postoperatively, and 30% were diagnosed on an outpatient basis. Few (9%) of the patients required return to the operating room for management. A median of 39 d (range 3–240 d) was required for wound healing. Thirteen (9%) of the patients in the study developed an intra-abdominal abscess, and of these patients, 8 (61%) also developed an SSI (p=0.02).

Table 1.

Rates of Surgical Site Infection Stratified by Patient Characteristics (n=143)

	Percent of all patients with characteristic	Percent of SSI in patients with characteristic	Percent of SSI in patients without characteristic	p Value for SSI in group comparison
Indication for surgery
Rectal cancer	9.1	15.4	31.5	NS
Colon cancer	21.0	26.7	31.0	NS
Inflammatory bowel disease	27.3	33.3	28.8	NS
Diverticular disease	16.1	39.1	28.3	NS
Comorbidities
Diabetes	11.2	50	27.6	0.08
History of myocardial infarction	7.0	50	28.6	0.17
Pulmonary disease	9.1	7.7	32.1	0.11
Chronic anticoagulation	7.0	30	30.1	NS
Chronic kidney disease	7.7	18.2	31.1	NS
Prednisone use	14.6	23.8	30.9	NS
Recent immunosuppressant medication	17.5	28	30.5	NS
History of soft tissue infection	11.2	56.2	26.7	0.02
Age (years)	Patients with SSI (mean) 52.6 Patients without SSI (mean) 54.3			NS
Female gender	51.7	36.5	23.2	0.10
Current smoker	22.4	21.9	32.4	NS
Pre-operative weight loss	26.8	39.5	26.9	0.16
Albumin <3.5 g/dL	24.5	25.0	34.1	NS
BMI (in categories)
<25	46.7	25.0	–	NS (Wilcoxon rank sum test)
25–29.9	29.9	29.3	–
30–34.9	12.4	41.2	–
>35	10.9	40.0	–
BMI (kg/m²) (continuous variable)	OR 1.08 (95%CI: 1.02–1.15)			0.014
AW2 (midpoint from umbilicus to pubis)	OR 1.03 (95% CI 1.006–1.05)			0.014
AW3 (at level of ASIS)	OR 1.03 (95% CI 1.007–1.06)			0.013
AW4 (at lateral border of rectus)	OR 1.05 (95% CI 1.01–1.08)			0.012
AW1 (at umbilicus)	OR 1.04 (95% CI 1.006–1.074)			0.02

NS=p>0.20.

AW=abdominal wall level of measurement; BMI=body mass index; SSI=surgical site infection.

Table 2.

Peri-Operative Factors in Risk of Surgical Site Infection

	Percent of all patients with characteristic	Percent SSI in patients with characteristic	Percent SSI in patients without characteristic	p Value for SSI in group comparison
Conditions of surgery
Emergency surgery	11.2	37.5	29.1	NS
Prior abdominal surgery	44.4	36.5	25.3	0.20
Laparoscopic procedure	17.4	36	28.6	NS
Operative duration >180 min	70.9	32	31.2	NS
Wound classification 3 or 4	18.2	50	25.6	0.01
Peri-operative hyperglycemia	16.9	37.5	28.8	NS
Prophylactic antibiotic not administered in a timely manner	17.3	55.6	25.6	0.02
Procedural characteristics
Right-sided colon resection	35.7	29.4	30.4	NS
Stoma formation or takedown	49.6	36.6	23.6	0.10
Stoma takedown	9.8	50	27.9	0.12
More than one anastomosis	9.1	7.7	32.1	0.11

NS=p>0.20.

The patients' mean BMI was 27.1 (SEM±5.9) and their median BMI was 25.7. When treated as a continuous variable, BMI was significantly associated with SSI, with an OR of 1.08 (95% CI 1.02–1.15). The mean AWT halfway between the umbilicus and pubic bone (AW2) was 22.9 mm. The median AWT was 17.6 mm. Abdominal wall thickness, a wound classification of 3 or 4, a history of soft-tissue infection, and deviation from pre-operative antibiotics were significantly associated with SSI (Tables 1 and 2).

Of patients with an AWT exceeding 1 SD above the mean (>3.6 cm), 43% had an SSI (Table 3). Pre-operative cross-sectional imaging data were available for the measurement of AWT in 110 patients, of whom 35 (32%) developed an SSI. The availability of pre-operative imaging data was not associated with the incidence of SSI (24.2% for patients having imaging data vs. 31.5% for those not having such data, p=0.51). Abdominal wall thickness was significantly correlated (p<0.0001) with BMI (Pearson correlation coefficients: AW1: 0.71; AW2: 0.75; AW3: 0.71; AW4: 0.72). The distribution of values of BMI and AWT (AW2) is plotted in Figure 1.

FIG. 1.

Scatter plot of values of body-mass index (BMI) (abscissa) vs. values of abdominal-wall thickness (AWT) (ordinate) of patients in study.

Table 3.

Association between Categories of Abdominal-Wall Thickness and Surgical Site Infection (n=110)

Abdominal wall thickness (AW2)	0–0.9 cm (<25th percentile AWT)	0.9–1.8 cm (25th–50th percentile AWT)	1.8–3.6 cm (50th–75th percentile AWT)	>3.6 cm (>75th percentile AWT)
SSI	22% (6/27)	25% (7/21)	37% (10/27)	43% (12/28)

AW=abdominal wall thickness category; SSI=surgical site infection.

Patient and procedural factors not associated with SSI were age, gender, pre-operative indication for surgery, emergency surgery, prior abdominal surgery, serum albumin concentration, use of prednisone, diabetes mellitus, smoking, pulmonary disease, weight loss, laparoscopic surgery, creation or take-down of a stoma, duration of surgery >180 min, and perioperative hyperglycemia.

In a multivariable analysis done only with BMI and AWT as predictors of superficial or deep incisional SSI, AWT was a significant predictor of SSI whereas BMI was not (Table 4). A multivariable analysis of all predictors of SSI that were statistically significant (p<0.05) on univariate analysis showed that a lack of appropriately administered prophylactic antibiotic with regard to type and timing predicted SSI independently (OR 4.33; 95% CI 1.08–17.4). The other variables examined in the study, including AWT, were not statistically significant predictors of SSI (Table 5). The c-statistic for this model was 0.765.

Table 4.

Body-Mass Index and Abdominal Wall Thickness as Predictors of Surgical Site Infection in a Multivariable Analysis

	OR (95% CI)	p Value
BMI	0.98 (0.88–1.09)	0.70
AWT halfway between umbilicus and pubis [AW2]	1.04 (1.00–1.08)	0.04

AW=level of measurement of abdominal wall thickness; AWT=abdominal wall thickness; BMI=body mass index; CI=confidence interval; OR=odds ratio.

Table 5.

Significant Predictors of Surgical Site Infection on Multivariable Analysis

Predictor	Odds Ratio	95% CI
Abdominal wall thickness (AW2)	1.026	0.995–1.06
Wound classification 3 or 4	2.80	0.74–10.6
History of soft tissue infection	2.29	0.55–9.6
Prophylactic antibiotic not administered in a timely manner	4.33	1.08–17.4

c-statistic=0.763.

AW2=level of measurement of abdominal wall thickness; CI=confidence interval.

Discussion

In this retrospective review of patients undergoing colorectal surgery at a single institution, AWT, as measured in a preoperative cross-sectional imaging study, was associated with superficial or deep incisional SSI by univariate analysis. In determining AWT, the point halfway between the umbilicus and pubic bone was selected for measurement of the distance between the skin and fascia. For each 1 mm increase in AWT, the odds of SSI increased by 3%; for each 10 mm increase in AWT, the odds of SSI increased by 1.3 (1.03) [10].

As expected, AWT and BMI are correlated significantly. However, the scatter plot (Fig. 1) of AWT versus BMI shows that some patients have a thick abdominal wall despite having a BMI <30 (non-obese). In our multivariable analysis done with BMI and AWT (Table 4), AWT significantly predicted SSI, suggesting that it may be a more a specific predictor of SSI than is BMI. Furthermore, the OR for BMI in this analysis changed direction from its value of 1.08 in univariate analysis to 0.98 in multivariable analysis. Although this is not statistically significant, its interpretation is that after the incorporation of AWT in assessing the risk of SSI, an increased BMI may be protective against SSI.

In the multivariable analysis incorporating all significant predictors of SSI found by univariate analysis, AWT was not a significant predictor of SSI. Therefore, in our study, AWT was not an independent risk factor for SSI as compared with other clinical factors. The relatively small sample size of our study and the heterogeneity of the patient population limit the power of this analysis. Arguably, the relationship of AWT with SSI on univariate analysis remains meaningful, in that it is not readily apparent how other factors in the analysis could confound the association between AWT and SSI. It would not be expected, for example, that patients with thicker abdominal walls would be less likely to receive appropriate antibiotic prophylaxis for surgery or would be more likely than others to undergo surgery for conditions involving a greater degree of wound contamination. Abdominal wall thickness, an easily measured and clinically accessible variable, merits further investigation.

In addition to answering the primary question to which it was addressed, the study also found that a history of soft-tissue infection, either as a prior SSI after abdominal surgery or as an infection at another site, is correlated with post-operative SSI in univariate analysis, and is worth further investigation in this regard.

Prior studies have found BMI to be an insensitive measure in understanding the relationship of obesity to complications of surgery. Tsujinaka et al. [5] studied visceral and subcutaneous fat in Japanese patients who had laparoscopic sigmoid colectomy for cancer, defining obesity as a BMI>25 or visceral fat as >130 cm² as measured with pre-operative computed tomography (CT) (51% of the patients were obese by this definition). Surgical site infection was more frequent in patients defined as obese by the area of their visceral fat on CT, but not in patients defined as obese by their BMI. Tokunaga et al. [13] examined 135 patients who had gastrectomy and found that visceral fat was associated with post-operative intra-abdominal infection, mortality, and length of stay. Fujii et al. [14] found that a thickness of subcutaneous fat of >2 cm was associated with incisional SSI with an OR of 2.81 in a multivariable analysis of 152 patients who underwent elective colorectal surgery. In the United States, two studies at the Memorial Sloan-Kettering Cancer Center have examined the distribution of intra-abdominal fat. Visceral fat was associated with post-operative complications and with pancreatic fistula [22], and with mortality, complications, and length of stay in patients undergoing liver resection [12].

Extravisceral or extra-abdominal fat may be more pertinent to infection occurring outside than inside the peritoneal or pelvic cavity. Superficial SSIs after colorectal surgery may in large part be related to local factors such as microscopic or macroscopic contamination with fecal flora in the settings of dead space, necrotic fat, and varying degrees of wound vascularization. Interventions that address local wound factors, such as wound protectors, loose closure of sutures, healing by secondary intention, the use of intra-incisional antibiotics, debridement, and the evacuation of necrotic fat before skin closure, or subcutaneous drainage may be useful in patients with a thick abdominal wall if the latter is associated with infection.

Our study is limited by its retrospective nature. We provide no data about whether AWT as determined with CT is equivalent to what AWT would measure in the operating room. We also do not account for other factors that may affect the local wound environment, such as the length of an incision; degree of wound contamination (although this is indirectly assessed through wound classification); or devitalization of skin, adipose tissue, or fascia. Some of these variables may explain the relationship between AWT and wound infection.

Upon reviewing our findings in the present study, we believe that patients who have increased subcutaneous fat probably experience a substantial proportion of wound infections after colorectal surgery, but this was not confirmed in a multivariable analysis. To resolve this issue, we recommend further investigation with a larger, more homogeneous (e.g., elective cases only) sample of patients.

Footnotes

Acknowledgments

Dr. Kwaan conceived and designed the study; Drs. Kwaan and Sirany acquired the study data; and Drs. Kwaan, Sirany, Madoff, and Rothenberger analyzed and interpreted the data and wrote this paper.

Author Disclosure Statement

No competing financial interests exist.

References

Blee

, Belzer

, Lambert

. Obesity: Is there an increase in perioperative complications in those undergoing elective colon and rectal resection for carcinoma? Am Surg, 2002; 58:163–166.

Dindo

, Muller

, Weber

, Clavien

P-A

. Obesity in general elective surgery. Lancet, 2003; 361:2032–2035.

Ballian

, Yamane

, Leverson

et al. Body mass index does not affect postoperative morbidity and oncologic outcomes of total mesorectal excision for rectal adenocarcinoma. Ann Surg Oncol, 2010; 17:1606–1613.

Gendall

, Raniga

, Kennedy

, Frizelle

. The impact of obesity on outcome after major colorectal surgery. Dis Colon Rectum, 2007; 50:2223–2237.

Tsujinaka

, Konishi

, Kawamura

et al. Visceral obesity predicts surgical outcomes after laparoscopic colectomy for sigmoid colon cancer. Dis Colon Rectum, 2008; 51:1757–1767.

Degrate

, Garancini

, Misani

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

Wick

, Vogel

, Church

et al.

Surgical site infections in a “high outlier” institution: are colorectal surgeons to blame?

Dis Colon Rectum, 2009; 52:374–379.

Merkow

, Bilimoria

, McCarter

, Bentrem

. Effect of body mass index on short-term outcomes after colectomy for cancer. J Am Coll Surg, 2009; 208:53–61.

Tanner

, Khan

, Aplin

et al. Post-discharge surveillance to identify colorectal surgical site infection rates and related costs. J Hosp Infect, 2009; 72:243–250.

10.

Smith

, Bohl

, McElearney

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

11.

Balentine

, Wilks

, Robinson

et al. Obesity increases wound complications in rectal cancer surgery. J Surg Res, 2010; 163:35–39.

12.

Morris

, Tuorto

, Gonen

et al. Simple measurement of intraabdominal fat for abdominal surgery outcome prediction. Arch Surg, 2010; 145:1069–1073.

13.

Tokunaga

, Hiki

, Fukunaga

et al. Effect of individual fat areas on early surgical outcomes after open gastrectomy for gastric cancer. Br J Surg, 2009; 96:496–500.

14.

Fujii

, Tsutsumi

, Matsumoto

et al. Thickness of subcutaneous fat as a strong risk factor for wound infections in elective colorectal surgery: Impact of prediction using preoperative CT. Dig Surg, 2010; 27:331–335.

15.

Pastor

, Artinyan

, Varma

et al. An increase in compliance with the surgical care improvement project measures does not prevent surgical site infection in colorectal surgery. Dis Colon Rectum, 2010; 53:24–30.

16.

Hennessey

, Burke

, Ni-Dhonochu

et al. Preoperative hypoalbuminemia is an independent risk factor for the development of surgical site infection following gastrointestinal surgery. Ann Surg, 2010; 252:325–329.

17.

Govinda

, Kasuya

, Bala

et al. Early postoperative subcutaneous tissue oxygen predicts surgical site infection. Anesth Analg, 2010; 111:946–952.

18.

Blumetti

, Luu

, Sarosi

et al.

Surgical site infections after colorectal surgery: do risk factors vary depending on the type of infection considered?

Surgery, 2007; 142:704–711.

19.

Balentine

, Robinson

, Marshall

et al. Waist circumference predicts increased complications in rectal cancer surgery. J Gastrointest Surg, 2010; 14:1669–1679.

20.

, Barie

, Stein

et al. Antibiotic regimen and the timing of propylaxis are important for reducing surgical site infection after elective abdominal colorectal surgery. Surg Infect, 2011; 12:255–260.

21.

Pastor

, Baek

J-H

, Varma

et al. Validation of the risk index category as a predictor of surgical site infection in elective colorectal surgery. Dis Colon Rectum, 2010; 53:721–727.

22.

House

, Fong

, Arnaoutakis

et al. Preoperative predictors for complications after pancreaticoduodenectomy: Impact of BMI and body fat distribution. J Gastrointest Surg, 2008; 12:270–278.