Leaving the Skin Incision Open May Not Be as Beneficial as We Have Been Taught

Abstract

Background:

Currently, various methods of skin closures are used in contaminated and dirty abdominal wounds without solid, evidence-based guidance. This study investigates whether closure methods affect surgical site infection (SSI) and other incisional complications. We hypothesize that open management of the skin would have the lowest complications, including SSI.

Patients and Methods:

Patients age ≥18 who underwent trauma laparotomy (TL) or damage control laparotomy (DCL) from 2008–2013 and had class III/IV wounds were included. Demographic, injury, treatment, and outcome variables were compared based on skin closure methods: Primary closure, intermittently stapled with wicks, or open management. Subgroup analyses for TL, DCL, and high-risk patients with stomach, colon, or rectal injuries were performed. Bivariable and multivariable logistic regression (MLR) analyses were performed to identify risk factors for superficial/deep SSI and surgical incision complications.

Results:

A total of 348 patients were included. The median age was 47 years; 14% were female; 21% had blunt injuries. Overall SSI was highest for open incisions (p < 0.05), but there was no difference in superficial/deep SSI. Primary closures healed a median of 20 days, compared with 68 and 71 days for the intermittently stapled and open groups, respectively (p < 0.001). Primary closure in TL and high-risk patients also had the lowest SSI rates (all p < 0.05), but there were no differences in superficial/deep SSI in any subgroup. In TL patients, diabetes mellitus and colon injuries were independently associated with the development of superficial/deep SSI and surgical incision complications; however, skin closure method was not.

Conclusion:

In class III and IV wounds, primary closure was associated with the lowest SSI, shortest length of stay and healing time. Method of skin closure, however, did not have an independent effect on the development of superficial/deep SSI or surgical incision complications. These suggest that primary skin closure in contaminated and dirty abdominal wounds may be performed more safely than commonly perceived.

Surgical site infections (SSI) are one the most common causes of health-care–associated infections (HAI) and are estimated to account for 21.8% of HAI in the United States [1]. The Centers for Disease Control and Prevention (CDC) National Healthcare Safety Network (NHSN) is the most widely used surveillance system for SSI and other health-care–acquired infections in the United States [2]. The CDC categorizes SSI as superficial incisional, deep incisional, or organ space infection [3]. For SSI to occur, microbial contamination of the surgical site is needed. It has been shown that a bacterial load of greater than 10⁵ per gram of tissue significantly increases the risk of SSI, and the amount is even less in the setting of foreign material. The source of bacteria is mostly patient endogenous flora from skin, mucous membrane, or viscera. Alternatively, when a breach in sterile technique occurs, exogenous sources of SSI can originate from surgical personnel, operating suite, and surgical equipment [4].

The most recent report from the NHSN demonstrated a 17% decrease in rates of SSI from 2008 to 2014 [2]. While significant prevention efforts have no doubt contributed to this reduction, SSI still remains a significant source of patient death and healthcare costs [5]. Studies have shown that SSIs increase incidence of incisional hernias to as high as 45% [6], facial dehiscence to 25%, and can lead to necrotizing fasciitis [7,8]. Surgical site infections can also increase the average hospital length of stay (LOS) by 11 days, and add $20,000 to healthcare costs per occurrence [9 –11].

Historically, open management of contaminated surgical sites has been used to decrease SSI since the work of Hunter in the 1790s [12] and was shown to be superior to primary closure by Kern in 1806 [13]. Since then, healing by secondary intention and delayed primary closure (DPC) have been studied both nationally and internationally, and have been demonstrated to be effective in lowering SSI [14 –18]. A recent meta-analysis of eight randomized controlled trials, however, did not show definitive evidence supporting DPC's ability to reduce SSI [19].These conflicting data mirror the wide variety of closure methods currently used to manage class III and IV wounds [3]. In addition, there are very limited recent data that compare SSI rates between complete skin closure, loose skin closure, and no closure; and this is even scarcer in the trauma literature [20].

This study compares surgical incision complication rates and specifically SSI based on three skin closure methods: Primary closure, intermittent staple with wicks in between staples, and open management. We hypothesized that leaving the incision open would be associated with the lowest SSI and surgical incision complication rates, but would take significantly longer to heal.

Patients and Methods

This study was conducted at MetroHealth Medical Center, a regional level I trauma center. The institutional trauma registry was queried for patients age 18 or older who underwent trauma laparotomy (TL) or damage control laparotomy (DCL) from January 1, 2008 through December 31, 2013, and had class III or IV wounds as defined by the CDC [4].

Patient factors collected included age, gender, body mass index (BMI), active infection on arrival to the trauma bay (referred to as concurrent infection), pre-operative albumin level, and co-morbidities including steroid use, immunocompromised illness, blood alcohol level >0.08 on arrival to the trauma bay, ascites, cigarette use, chronic renal failure, diabetes mellitus, cirrhosis, dementia, psychological illness, cancer, metastatic cancer, and active use of chemotherapy.

Injury variables collected included: Abbreviated Injury Score (AIS) abdomen, Injury Severity Score (ISS), American Society of Anesthesiologists score (ASA), trauma mechanism, perforated organ(s), and peri-operative presence of shock (systolic blood pressure <90 mm Hg). Treatment variables such as injury to operation time, length of operation, type of operation (TL vs. DCL), operative temperature nadir, pre-incision antibiotic agent data (usage, bacterial coverage, timing of delivery, and appropriateness of cessation), peri-operative blood product use, intra-operative culture, post-operative glucose control up to post-operative day (POD) 3, time from date of operation to normalization of lactate and base deficit (lactate <2.5 mmol/L and base deficit between −2–2 mmol/L), and closure method.

The operations in this study refer to the operations in which the fascia was closed. Thus, it was either the operation at the initial TL or in patients who had DCL, at the take-back operation in which the fascia was closed during their initial hospitalization. Patients who did not have primary fascial closure during their initial hospitalization were excluded from the study. Post-operative antibiotic use was deemed appropriate if stopped within 24 hours post-operatively, or if continued for documented diagnosis. Peri-operative blood product use was defined to include the use of any packed red blood cells, fresh frozen plasma, platelets, or cryoprecipitate intra-operatively or within 24 hours from the operation. Skin closure methods included primary closure using suture or staple, intermittently stapled with wicks in between staples, and open with only a few staples approximating the incision (typically around the umbilicus). Incisions closed by other means were not included in the comparison of outcomes by closure method. The DPC is used infrequently at our institution and therefore was not included in this analysis.

Primary outcome variables were any SSI, superficial/deep incisional SSI, and surgical incision complications. SSI included superficial incisional, deep incisional, and organ space infection as defined by the CDC [3] and was determined through documented occurrence in either inpatient and outpatient records. Surgical incision complications were defined as occurrence of any superficial or deep incisional SSI or other complication such as wound dehiscence, hernia, seroma, hematoma, and enterocutaneous fistula. Thus, isolated organ space infections were not considered a surgical incision complication. Secondary outcome variables included intensive care unit (ICU) LOS, hospital LOS, in-hospital death, and incision healing time. Demographics, injury, treatment, and outcome variables were compared based on skin closure methods. Subgroup analyses for TL patients, DCL patients, and high-risk patients with perforated stomach, colon, or intra-abdominal rectum were also performed.

Statistical analyses were performed using IBM SPSS^© version 23 (IBM, Armonk, NY). Continuous variables were compared using Student t-test or Kruskal-Wallis test with post hoc analysis as appropriate. Means in this study are reported as mean ± standard deviation, and medians are reported as median (25th–75th percentile). Categoric data were analyzed using either chi-square test or Fisher exact test. Bivariable analyses were performed for factors that may associate with superficial/deep incisional SSI or surgical incision complications. Appropriate variables identified with a p ≤ 0.05 after bivariable analyses were then analyzed using backward stepwise multivariable logistic regression (MLR) to determine independent predictors of superficial/deep incisional SSI and surgical incision complications. Receiver-operating curves were generated from the logistic regression analyses, and the areas under the curve were calculated and expressed as the C-statistics. The odds ratios are reported with 95% confidence intervals. A p value of 0.05 was used to determine statistical significance. This study was approved by the MetroHealth Medical Center Institutional Review Board.

Results

Evaluation by skin closure method

All patients

A total of 348 patients were included; 21% sustained blunt injuries. The majority of patients (76%) underwent TL and 24% underwent DCL. Three hundred and forty-one patients had incision closure by one of the three defined categories; 213 (63%) patients had primary closure, 28 (8%) patients had intermittent staple with wicks, and 100 (29%) patients had open skin incisions. The overall rate of SSI was 18.2%. Superficial/deep incisional SSI occurred in 12.6%, and surgical incision complications developed in 21.9% of patients. Overall in-hospital deaths were 4%. The three groups were similar in age, gender, trauma mechanism, BMI, and co-morbidities (Table 1). Patients with primary skin closure had slightly lower AIS abdomen, ISS, and ASA score compared with the other two groups (all p < 0.05). They also had a lower proportion of colon injury, small intestine injury, and peri-operative shock. In contrast, patients with open skin incisions had the highest proportion of colon and small intestinal injuries, as well as peri-operative shock.

Table 1.

All Patients: Evaluation by Skin Closure Method

Patient factors	Primary closure (n = 213)	Intermittent staple w/wicks (n = 28)	Open management (n = 100)	p
Median age (y)	30 (22–41)	35 (26 − 56)	32 (22–41)	0.160
Female	24 (11.3%)	7 (25.0%)	16 (16.0%)	0.105
Blunt injury	38 (17.8%)	8 (28.6%)	26 (26.0%)	0.154
Median BMI	26.2 (23.0–31.1)	27.2 (23.5–32.4)	27.5 (24.1–31.3)	0.126
Co-morbidity present	61 (28.9%)	9 (33.3%)	29 (29.3%)	0.893
Injury variables
Mean AIS abdomen	2.6 ± 0.9^a	3.1 ± 1.0	3.0 ± 0.8	<0.001
Mean ISS	14.7 ± 10.4^b	17.3 ± 10.8	18.3 ± 10.4	0.003
Mean ASA	2.9 ± 0.9^b	3.1 ± 0.8	3.2 ± 0.9	0.029
Colon injury	39 (18.3%)	12 (42.9%)	54 (54.0%)	<0.001
Stomach injury	23 (10.8%)	5 (17.9%)	16 (16.0%)	0.316
Small intestine injury	68 (31.9%)	14 (50.0%)	55 (55.0%)	<0.001
Rectal injury	5 (2.3%)	2 (7.1%)	3 (3.0%)	0.368
Bladder injury	16 (7.5%)	2 (7.1%)	6 (6.0%)	0.88
Ureter injury	3 (1.4%)	1 (3.6%)	2 (2.0%)	0.699
Peri-operative shock	55 (26.1%)	9 (32.1%)	48 (48.0%)	0.001
Treatment variables
DCL	30 (14.1%)	9 (32.1%)	39 (39.0%)	<0.001
Median injury to operative time (min)	150 (95–294)^b	359 (100–2427)	383 (95–3309)	0.006
Median operative duration (min)	110 (78–145)^b	130 (79–167)	139 (101–202)	<0.001
Median operative temp nadir (^oC)	35.6 (35.0–36.2)^b	36.1 (35.2–36.9)	36.1 (35.0–36.9)	0.013
Pre-incision abx given	198 (93.4%)	24 (85.7%)	90 (90.0%)	0.284
Pre-incision abx given within 1 h	168 (84.8%)	21 (87.5%)	77 (85.6%)	0.938
Peri-operative abx stopped appropriately	183 (97.9%)	28 (100%)	82 (93.2%)	0.076
Outcomes
Surgical incision complication	37 (18.6%)	4 (14.8%)	29 (31.2%)	0.034
Hernia	11 (5.5%)	1 (3.7%)	10 (10.8%)	0.206
Wound dehiscence	4 (2.0%)	0	11 (11.8%)	0.001
Necrotizing fasciitis	1 (0.5%)	0	2 (2.2%)	0.345
SSI	26 (13.1%)	6 (22.2%)	25 (26.9%)	0.013
Superficial/deep SSI	21 (10.6%)	3 (11.1%)	15 (16.1%)	0.392
Organ space SSI	9 (4.5%)	3 (11.1%)	21 (22.6%)	<0.001
Median ICU LOS (d)	0 (0–3.0)^b	2.5 (0–11.5)	3.5 (0–8.0)	<0.001
Median LOS (d)	6.0 (4.0–9.0)^a	13.0 (6.3–23.5)	13.0 (7.0–22.0)	<0.001
Median time to heal (d)	20 (14–31)^a	68 (25–125)	71 (45–108)	<0.001
Deaths	6 (2.8%)	0	9 (9.0%)	0.022

Normally distributed values are presented as mean ± standard deviation, all other values are presented as median (interquartile range). Superscripts denote results of post –hoc analysis: ^aSignificantly different compared with other two groups; ^bsignificantly different between primary and open.

BMI = body mass index; AIS = Abbreviated Injury Score; ISS = Injury Severity Score; ASA = American Society of Anesthesiologists score; DCL = damage control laparotomy; abx = antibiotic agents; SSI = surgical site infection; ICU = intensive care unit; LOS = length of stay.

Patients with primary closures had the shortest length of operation and had the lowest intra-operative temperature. They also had a lower proportion of DCL and significantly shorter time from injury to operation, where the fascia was closed (Table 1). All three groups were similar in terms of transfusion needs, post-operative glucose management, and time to normal lactate and base deficit.

Overall SSI rates were lowest in the primary closure group (p = 0.013); however, no differences in rates of superficial/deep incisional SSI were detected. Similarly, surgical incision complications were lowest for the primarily closed and the intermittently stapled groups. The primary closure group had significantly shorter ICU LOS, hospital LOS, and time for incision to heal (Table 1).

TL patients

A total of 263 patients had TL with fascial closure during the same operation; 183 (70%) patients had primary skin closure, 19 (7%) patients had intermittently stapled skin, and 61 (23%) patients had open skin incisions. The three groups were similar in age, gender distribution, trauma mechanism, BMI, co-morbidities, ISS, and ASA. Patients with primary closures, however, had slightly lower AIS abdomen (2.6 vs. approximately 3.0 in the other two groups) and significantly fewer colon injuries. Patients with intermittently stapled skin had the lowest proportion of peri-operative shock (Table 2). Primary skin closures had the shortest median operative time, 110 minutes (80–144), compared with 141 (83–177) and 150 (126–209) minutes for the intermittently stapled and open groups, respectively (p < 0.001). Post-operative antibiotic agents were continued inappropriately more often in patients with open skin incisions (11% vs. 2% in the primary closure group, and 0% in the intermittently stapled group, p = 0.02). It also took a slightly longer time to achieve normal base deficit in patients with open skin incisions (0.9 days vs. 0.1 to 0.2 days in the other groups, p = 0.003).

Table 2.

Trauma Laparotomy Patients: Evaluation by Skin Closure Method

Patient factors	Primary closure (n = 183)	Intermittent staple w/wicks (n = 19)	Open management (n = 61)	p
Median age (y)	28 (22–40)	28 (24–39)	30 (21–39)	0.851
Blunt injury	16 (8.7%)	2 (10.5%)	8 (13.1%)	0.609
Co-morbidity present	131 (71.6%)	13 (72.2%)	43 (71.7%)	0.998
Injury variables
Mean AIS abdomen	2.6 ± 0.8^a	3.2 ± 1.0	3.0 ± 0.6	<0.001
Mean ISS	12.8 ± 8.8	15.8 ± 11.0	14.0 ± 6.9	0.102
Mean ASA	2.8 ± 0.8	2.9 ± 0.9	2.9 ± 0.9	0.763
Colon injury	34 (18.6%)	9 (47.4%)	40 (65.6%)	<0.001
Stomach injury	18 (9.8%)	3 (15.8%)	9 (14.8%)	0.476
Small intestine injury	63 (34.4%)	12 (63.2%)	39 (63.9%)	0.476
Rectal injury	5 (2.7%)	2 (10.5%)	3 (4.9%)	0.209
Bladder injury	16 (8.7%)	1 (5.3%)	5 (8.2%)	0.872
Ureter injury	1 (0.5%)	0	1 (1.6%)	0.644
Peri-operative shock	35 (19.1%)	2 (10.5%)	20 (32.8%)	0.038
Outcome variables
Surgical incision complication	28 (16.0%)	3 (15.8%)	18 (31.6%)	0.033
Hernia	8 (4.6%)	1 (5.3%)	8 (14.0%)	0.046
Wound dehiscence	4 (2.3%)	0	7 (12.3%)	0.004
Necrotizing fasciitis	1 (0.6%)	0	2 (3.5%)	0.184
SSI	21 (12.0%)	4 (21.1%)	18 (31.6%)	0.003
Superficial/deep SSI	16 (9.1%)	2 (10.5%)	9 (15.8%)	0.372
Organ space SSI	8 (4.6%)	2 (10.5%)	16 (28.1%)	<0.001
Median ICU LOS (d)	0 (0 − 2)	0 (0–3)	0 (0–4)	0.188
Median LOS (d)	5 (4 − 8)^b	9 (4–18)	10 (6–19)	<0.001
Deaths	1 (0.5%)	0	3 (4.9%)	0.046
Median time to heal (d)	19 (13–25)^a	61 (21–104)	70 (41–113)	<0.001

Normally distributed values are presented as mean ± standard deviation; all other values are presented as median (interquartile range). Superscripts denote results of post hoc analysis: ^aSignificantly different compared with other two groups; ^bsignificantly different between primary and open.

AIS = abbreviated injury score; ISS = injury severity score; ASA = American Society of Anesthesiologists score; SSI = surgical site infection; ICU = intensive care unit; LOS = length of stay.

Patients with open incisions had the highest proportion of SSI (p = 0.003) and surgical incision complications (p = 0.033), but there was no significant difference in the rate of superficial/incisional SSI. Patients with primary closures had similar ICU LOS as the remaining groups; however, they had significantly shorter overall hospital LOS and median time to heal (Table 2).

DCL patients

A total of 78 patients underwent DCL and had fascial closure during their initial hospitalization; 30 (38%) patients had primary skin closures, nine (12%) patients had skin intermittently stapled with wicks, and 39 (50%) patients had their skin incisions left open. The three groups had similar demographics, except that the intermittently stapled group was older (p = 0.05). Injury variables, including AIS abdomen and injured organs, among others were also similar, and no differences in treatment variables were seen (Table 3).

Table 3.

Damage Control Laparotomy Patients: Evaluation by Skin Closure Method

Patient factors	Primary closure (n = 30)	Intermittent staple w/wicks (n = 9)	Open management (n = 39)	p
Median age (y)	38 (24–52)	60 (37–74)^a	32 (24–48)	0.050
Blunt injury	22 (73.3%)	6 (66.7%)	18 (46.2%)	0.066
Co-morbidity present	19 (67.9%)	5 (55.6%)	27 (69.2%)	0.729
Injury variables
Mean AIS abdomen	2.9 ± 1.0	2.9 ± 1.0	3.3 ± 1.0	0.207
Mean ISS	26 ± 12	20 ± 10	25 ± 12	0.555
Mean ASA	3.8 ± 0.7	3.6 ± 0.5	3.7 ± 0.7	0.505
Colon injury	5 (16.7%)	3 (33.3%)	14 (35.9%)	0.199
Stomach injury	5 (16.7%)	2 (22.2%)	7 (17.9%)	0.930
Intestine injury	5 (16.7%)	2 (22.2%)	16 (41.0%)	0.078
Bladder injury	0	1 (11.1%)	1 (2.6%)	0.181
Ureter injury	2 (6.7%)	1 (11.1%)	1 (2.6%)	0.513
Peri-operative shock	20 (71.4%)	7 (77.8%)	28 (71.8%)	0.928
Outcome variables
Surgical incision complication	9 (37.5%)	1 (12.5%)	11 (30.6%)	0.415
Hernia	3 (12.5%)	0	2 (5.5%)	0.419
Wound dehiscence	0	0	4 (11.1%)	0.151
Necrotizing fasciitis	0	0	0	–
SSI	5 (20.8%)	2 (25.0%)	7 (19.4%)	0.939
Superficial/deep SSI	5 (20.8%)	1 (12.5%)	6 (16.7%)	0.845
Organ space SSI	1 (4.2%)	1 (12.5%)	5 (13.9%)	0.467
Median ICU LOS (days)	9 (4–17)	12 (8–16)	8 (5–19)	0.651
Median LOS (d)	16 (9–25)	22 (14–29)	19 (12–28)	0.446
Deaths	5 (16.7%)	0	6 (15.4%)	0.429
Median time to heal (d)	57 (35–105)	111 (42–143)	72 (59–110)	0.301

AIS = Abbreviated Injury Score; ISS = Injury Severity Score; ASA = American Society of Anesthesiologists score; SSI = surgical site infection; ICU = intensive care unit; LOS = length of stay.

There were no differences in SSI or surgical incision complications based on closure method. Primarily closed incisions healed in a median of 57 days, compared with 111 and 72 days for the intermittently stapled and open groups, respectively; however, this was not statistically significant (Table 3). There were no differences in secondary outcomes such as ICU LOS and overall LOS.

High-risk patients

A total of 144 high-risk patients had a perforation of the stomach, colon, or intra-abdominal rectum; 61 (42%) patients had primary skin closure, 17 (12%) patients had skin intermittently stapled with wicks, and 66 (46%) patients had their skin incision left open. All three groups were similar in patient factors, except that patients with primary closures had slightly lower median BMI (24 vs. 27 in the remaining groups, p = 0.004) and fewer were in peri-operative shock. Also, patients with open skin incisions had an AIS abdomen of 3.3 compared with 2.9 for the other two groups (Table 4). There were no significant differences in peri-operative factors among the three groups.

Table 4.

High-Risk Patients: Evaluation by Skin Closure Method

Patient factors	Primary closure (n = 61)	Intermittent staple w/wicks (n = 17)	Open management (n = 66)	p
Median age (y)	25 (21–38)	29 (24–42)	31 (24–39)	0.138
Blunt injury	7 (11.5%)	3 (17.6%)	8 (12.1%)	0.787
Co-morbidity present	49 (80.3%)	12 (75.0%)	50 (76.9%)	0.852
Injury variables
Mean AIS abdomen	2.9 ± 1.0	2.9 ± 1.0	3.3 ± 1.0^a	0.004
Mean ISS	13 (9–19)	17 (12–24)	14 (9–22)	0.314
Mean ASA	3 (2–3)	3 (2–5)	3 (2–4)	0.198
Stomach injury	23 (37.7%)	5 (29.4%)	16 (24.2%)	0.257
Small intestine injury	21 (34.4%)	9 (52.9%)	35 (53.0%)	0.086
Colon injury	39 (63.9%)	12 (70.6%)	54 (81.8%)	0.075
Rectal injury	5 (8.2%)	2 (11.8%)	3 (4.5%)	0.510
Bladder injury	2 (3.3%)	2 (11.8%)	5 (7.6%)	0.368
Ureter injury	3 (4.9%)	1 (5.9%)	1 (1.5%)	0.489
Peri-operative shock	11 (18.0%)	5 (29.4%)	28 (42.4%)	0.014
Outcome variables
Surgical incision complication	11 (18.6%)	3 (18.8%)	25 (39.1%)	0.028
Hernia	2 (3.4%)	1 (6.3%)	8 (12.5%)	0.168
Wound dehiscence	1 (1.7%)	0	9 (14.1%)	0.015
Necrotizing fasciitis	0	0	2 (3.1%)	0.305
SSI	10 (16.9%)	5 (31.3%)	24 (37.5%)	0.038
Superficial/deep SSI	8 (13.6%)	2 (12.5%)	14 (21.9%)	0.412
Organ space SSI	5 (8.5%)	3 (18.8%)	20 (31.3%)	0.007
Median ICU LOS (d)	0 (0–3)^b	2 (0–15)	3 (0−6)	0.039
Median LOS (d)	7 (5–15)^b	14 (7–25)	14 (8–23)	<0.001
Deaths	0	0	4 (6.1%)	0.088
Median time to heal (d)	20 (13–47)^a	85 (44–123)	74 (46–124)	<0.001

AIS = Abbreviated Injury Score; ISS = Injury Severity Score; ASA = American Society of Anesthesiologists score; SSI = surgical site infection; ICU = intensive care unit; LOS = length of stay.

High-risk patient subgroup (perforated stomach, colon, or intra-abdominal rectum).

Normally distributed values are presented as mean ± standard deviation; all other values are presented as median (interquartile range). Superscripts denote results of post hoc analysis: ^aSignificantly different compared to other two groups; ^bsignificantly different between primary and open.

Patients with primary closures had the lowest rates of SSI, but there were no significant differences in rates of superficial/deep incisional SSI. They also had the shortest ICU and hospital LOS, and their incisions healed in significantly less time (Table 4). Surgical incision complications and wound dehiscence in particular were higher in the open group compared with the other closure methods.

Factors associated with superficial/ deep SSI and surgical incision complications

One of the specific purposes of this study was to evaluate the factors related to the development of surgical incision complications and infections—in particular, superficial and deep incisional SSI. We evaluated TL and DCL patient subgroups separately given the distinct characteristics and peri-operative management of these populations.

TL: Superficial or deep incisional SSI

Superficial/deep incisional SSI developed in 11.1% of the TL population. Factors associated with superficial/deep incisional SSI in this group included older age, diabetes mellitus, colon injuries, and higher median net fluid on POD 0. (Table 5) There were no significant differences in other associated injuries or peri-operative factors such as shock, blood transfusion, and operative duration. Primary closure rates were not statistically different. The TL patients with superficial/deep incisional SSI had significantly longer hospitalization, ICU days, and time to heal, as well as higher rates of wound dehiscence and development of surgical incision complications other than SSI. The MLR demonstrated that older age, diabetes mellitus, and colonic injury were independently associated with superficial/deep incisional SSI in TL patients (Table 5).

Table 5.

Trauma Laparotomy Patients: Factors Associated with Superficial or Deep Incisional Surgical Site Infection (n = 253)

Patient/ injury variables	No SSI (n = 225)	Superficial/ deep SSI (n = 28)	p
Median age (y)	28.0 (21.5–39.0)	38.0 (29.3–51.8)	0.002
Female	29 (12.9%)	0	0.053
Blunt injury	24 (10.7%)	2 (7.1%)	0.749
Median BMI	26.5 (22.8–29.9)	27.3 (24.0–30.1)	0.282
Diabetes mellitus	8 (3.6%)	4 (14.3%)	0.032
Obesity	57 (25.3%)	8 (28.6%)	0.819
Any comorbidity	160 (71.4%)	23 (85.2%)	0.170
Mean AIS abdomen	2.7 ± 0.81	3.0 ± 0.9	0.144
Mean ISS	13.7 ± 8.9	12.4 ± 8.6	0.461
Mean ASA	2.8 ± 0.9	2.9 ± 0.8	0.858
Colon injury	67 (29.8%)	16 (57.1%)	0.005
Stomach injury	27 (12.0%)	2 (7.1%)	0.752
Small intestine injury	100 (44.4%)	12 (42.9%)	1.00
Rectal injury	10 (4.4%)	0	0.608
Bladder injury	20 (8.9%)	1 (3.6%)	0.485
Ureter injury	2 (0.9%)	0	1.00
Peri-operative shock	50 (22.2%)	5 (17.9%)	0.808
Treatment variables
Median injury to operative time (min)	137.0 (91.0–222.8)	100.0 (67.5–144.0)	0.106
Median operative duration	121.5 (90.0–159.8)	143.0 (102.5–199.3)	0.081
Median operative temp nadir (°C)	35.6 (34.9–36.2)	35.9 (35.3–36.7)	0.083
Pre-incision abx given	210 (93.3%)	26 (92.9%)	1.00
Pre-incision abx given within 1 h	185 (88.1%)	23 (88.5%)	1.00
Post-op abx stopped appropriately	191 (92.3%)	20 (83.3%)	0.138
Blood transfusion	71 (31.6%)	11 (39.3%)	0.401
Median total blood product units	0 (0 − 2.0)	0 (0 − 2.0)	0.674
Primary closure	159 (70.7%)	16 (57.1%)	0.192
Median net fluid on POD 0 (L)	1.9 (0.94 − 3.4)	2.7 (1.4–5.4)	0.038
POD 0 glucose <180	161 (89.9%)	26 (96.3%)	0.479
Outcomes
Median ICU LOS (d)	0 (0–2.0)	1.5 (0–4.0)	0.030
Median LOS (d)	6.0 (4.0–9.0)	11.0 (5.25–22.0)	0.009
Median time to heal (d)	20.0 (15.0–41.0)	123.0 (65.5–211)	<0.001
Surgical incision complication other than SSI	21 (9.3%)	10 (35.7%)	0.001
Hernia	14 (6.2%)	3 (10.7%)	0.414
Wound dehiscence	5 (2.2%)	7 (25.0%)	<0.001
Necrotizing Fasciitis	2 (0.9%)	1 (3.6%)	0.298
Death	2 (0.9%)	1 (3.6%)	0.298

Multivariable Logistic Regression for Risk of Superficial/Deep SSI in TL Patients
C-statistic = 0.775	Odds ratio	95% CI	p
Age	1.03	1.0–1.0	0.035
Diabetes mellitus	5.9	1.1–31.1	0.036
Colon injury	3.1	1.2–7.9	0.018

Normally distributed values are presented as mean ± standard deviation, all other values are presented as median (interquartile range).

SSI = surgical site infection; BMI = body mass index; AIS = Abbreviated Injury Score; ISS: Injury Severity Score; ASA = American Society of Anesthesiologists score; abx = antibiotic agents; post-op = post-operative; POD = post-operative day; ICU = intensive care unit; LOS = length of stay; TL = trauma laparotomy; CI = confidence interval.

TL: Surgical incision complications

Surgical incision complications occurred in 50 (19.8%) TL patients, 62% of whom had a complication other than a SSI. Factors associated with surgical incision complications in the TL group are shown in Table 6. The MLR demonstrated that diabetes mellitus and colon injury were independently associated with surgical incision complications for patients undergoing TL, but skin closure method was not (Table 6).

Table 6.

Trauma Laparotomy Patients: Factors Associated with Surgical Incision Complications

Patient/ injury variables	No surgical incision complication (n = 203)	Surgical incision complication (n = 50)	p
Median age (y)	27.0 (22.0–39.0)	33.5 (25.3–46.0)	0.025
Female	26 (12.8%)	3 (6.0%)	0.221
Blunt injury	21 (10.3%)	5 (10.0%)	1.00
Median BMI	26.2 (22.9–29.5)	27.3 (23.5–32.5)	0.192
Diabetes mellitus	6 (3.0%)	6 (12.0%)	0.016
Obesity	48 (23.6%)	17 (34.0%)	0.150
Cirrhosis	0 (%)	2 (4.0%)	0.038
Any comorbidity	142 (70.3%)	41 (83.7%)	0.073
Mean AIS abdomen	2.7 ± 0.8	3.0 ± 0.9	0.025
Mean ISS	13.6 ± 8.5	13.3 ± 10.1	0.847
Mean ASA	2.8 ± 0.9	2.8 ± 0.8	0.790
Colon injury	56 (27.6%)	27 (54.0%)	0.001
Stomach injury	26 (12.8%)	3 (6.0%)	0.221
Small intestine injury	87 (42.9%)	25 (50.0%)	0.427
Rectal injury	9 (4.4%)	1 (2.0%)	0.692
Bladder injury	16 (7.9%)	5 (10.0%)	0.576
Ureter injury	1 (0.5%)	1 (2.0%)	0.357
Peri-operative shock	46 (22.7%)	9 (18.0%)	0.568
Treatment variables
Median injury to operative time (min)	139 (92.0–221.3)	117 (71.5–233.8)	0.167
Median operative duration	118 (88.5–157.3)	135.5 (97.0–204.0)	0.034
Median operative temp nadir (°C)	35.6 (34.9–36.2)	35.8 (35.1–36.4)	0.080
Pre-incision abx given	190 (93.6%)	46 (92.0%)	0.752
Pre-incision abx given within 1 h	166 (87.4%)	42 (91.3%)	0.614
Post-op abx stopped appropriately	172 (92.5%)	39 (86.7%)	0.238
Blood transfusion	65 (32.0%)	17 (34.0%)	0.866
Median total blood product units	0 (0 − 2)	0 (0 − 2)	0.971
Primary closure	147 (72.4%)	28 (56.0%)	0.027
Outcomes
Median ICU LOS (d)	0 (0–2)	0.5 (0–4.5)	0.023
Median LOS (d)	6 (4–9)	10 (5 − 20)	<0.001
Median time to heal (d)	20.0 (15.0–31.0)	79.0 (32.5–132.5)	<0.001
Death	2 (1.0%)	1 (2.0%)	0.485

Multivariable Logistic Regression for Risk of Surgical Incision Complications in TL Patients
C-statistic = 0.698	Odds ratio	95% CI	p
Diabetes mellitus	7.4	2.0–26.5	0.002
Colon injury	3.1	1.5–6.1	0.001
AIS abdomen	1.5	0.994–2.2	0.054

Normally distributed values are presented as mean ± standard deviation, all other values are presented as median (interquartile range).

BMI = body mass index; AIS = Abbreviated Injury Score; ISS: Injury Severity Score; ASA = American Society of Anesthesiologists score; abx = antibiotic agents; post-op = post-operative; ICU = intensive care unit; LOS = length of stay; TL = trauma laparotomy; CI = confidence interval.

DCL: Superficial or deep incisional SSI

The rate of superficial/deep incisional SSI was 18.1% in the DCL subgroup. There were no significant differences in demographics, enteric injuries, and peri-operative factors associated with the development of superficial/deep incisional SSI. Primary closure rates were similar between the two groups: 38.5% vs. 32.2%, p = 0.749. The DCL patients with superficial/deep SSI had longer hospital LOS (27 [19–37] vs. 15 [10–22] days, p = 0.002), and time to heal (150 [67–218] vs. 65 [43–31] days, p = 0.018), but no significant difference in ICU days or surgical incision complications other than SSI.

DCL: Surgical incision complications

The overall surgical incision complication rate in the DCL group was 31.9%. There were no statistically significant differences in patient and injury variables with relation to surgical incision complications. Median days from injury to operation, where fascia was closed (2.7 [2.3–6.8] vs. 2.0 [1.7–2.5], p = 0.001), and inappropriate continuation of post-operative antibiotic agents without documented indication (25.0% vs. 4.4%, p = 0.025) were higher in the DCL patients with surgical incision complications. There were no differences detected in other peri-operative factors, including rates of primary closure, 39.1% in those with surgical incision complications compared with 30.6% without a surgical incision complication (p = 0.593). The DCL patients with surgical incision complications had significantly longer ICU days (12 [8–26] vs. 8 [5–14], p = 0.023), LOS (24 [15–39] vs. 15 [10–22] days, p = 0.003), and time to heal (115 [57–183 vs. 65 [39–80] days, p = 0.008) than those without.

Discussion

This study sought to evaluate the impact of skin closure method on SSI and surgical incision complications in trauma patients with contaminated abdominal wounds. Our overall rate of SSI was 18.2%, and 12.6% had a superficial/deep incisional SSI. Rates of SSI in high-risk abdominal wounds vary widely in the trauma literature, between 3% and 45% [20 –24], with surgical incision infection rates reported up to 56% for primarily closed surgical incisions [25].

Much to our surprise, patients with primary skin closure had the lowest rates of SSI in our study at 13%; however, there were no significant differences in superficial/deep incisional SSI based on skin closure method. In this study, injuries and patient factors had a greater effect on the development of superficial/deep incisional SSI rather than the method of closure. Evaluation of our TL group demonstrated diabetes mellitus and colon injuries as independent predictors of superficial/deep SSI and surgical incision complications. Interestingly, TL patients in whom surgical incision complications developed were more likely to have had open skin incisions (36.7%) versus any other closure (19.3%, p = 0.013), but this was not independently associated once controlling for other significant variables in our MLR. Despite several subgroup analyses, we were unable to identify a particular population who benefitted from open skin incision management. Even in our high-risk cohort, incision closure method did not significantly change the rates of superficial/deep SSI, which ranged from 12.5% to 21.9%.

Our results are in contrast to other studies that have found open skin incision management to be associated with decreased incidence of SSI in contaminated or dirty abdominal wounds [20 –25]. Seamon et al. [20] found that leaving skin incisions open decreased post-operative SSI, and specifically superficial SSI, after high-risk TL when compared with any skin closure (9.8% vs. 31.1%). They also identified DCL as an additional risk factor for superficial SSI. Their overall SSI rate of 44.5% was markedly higher than the 18.2% in this study. This may in part be because of a higher proportion of DCL patients (35.6%) in their study compared with 24% in ours. The DCL patients in our study had a 17.6% rate of superficial/deep incisional SSI, which was slightly higher in the primary closure group than the other two closure methods, but this was not statistically significant.

Our rates are higher than the 7.3% superficial/deep SSI Pommerening et al. found in their cohort of DCL patients [23]. Their primary skin closure group had 13.5% superficial/deep SSI compared with 3.7% in the open group. Our DCL population had greater rates of organ space infections at 10.3% compared with the 2.0% in theirs, which may have contributed to some of the differences seen. Despite their findings, the authors suggested that “primary skin closure may be an acceptable strategy after DCL depending on the level of risk of superficial or deep SSI that patients and surgeons are willing to accept” given that superficial/deep SSI developed in fewer than 85% of their DCL patients with primary closure [23].

While closure method may have played less of a role in the prevention of superficial/deep incisional SSIs in our study, it did have a substantial effect on patient outcomes. Patients with open skin incisions took the longest time to heal, a median of 10 weeks, and had the highest rates of surgical incision complications in all of our subgroups except for the DCL patients. Wound dehiscence in particular remained consistently higher for those with open skin incision management. Shorter incision healing times may translate to improved patient satisfaction and healthcare costs.

Because of the retrospective nature of this study and the complexity of these patients, there are several limitations. First, the number of patients was not distributed evenly across the three skin closure groups, with only 28 patients in the intermittently stapled group. In addition, the DCL group had the least amount of patients of our subgroups analysis. Both of these may have introduced a type II error. Second, the groups were different in the distribution of gastrointestinal tract injuries and other peri-operative parameters. These may reflect some degree of selection bias by the operating surgeons. We attempted to adjust for this, however, via subgroup analyses and MLR, and our results still demonstrated that open management did not decrease SSI and surgical incision complications.

Third, intra-operative antibiotic re-dosing was not captured in this study. However, surgeons in this study were highly compliant with pre-incisional antibiotic agents. Lastly, this study attempted to capture post-operative SSIs and surgical incision complications through review of clinical assessments documented in the electronic medical records. We recognize that tracking post-discharge events retrospectively is very difficult, and it is likely that not all post-discharge surgical incision complications were captured in this study. A randomize controlled trial is likely needed to further elucidate the optimal surgical site closure method for this high-risk population.

Conclusion

The results of this study contrast our own hypothesis and traditional thinking with regard to the benefit of open skin incision management. While there is no question that we should make every effort to prevent SSIs, we should begin to question whether open skin incision management is truly worth the associated morbidity. Are we actually benefiting patients by leaving the skin open or are we just making ourselves feel better?

In trauma patients with class III and IV wounds, primary skin closure was associated with the lowest SSI in overall patients, TL patients, and high-risk patients. Primary closure was also associated with shortest LOS and healing time. Further, open management did not decrease the rate of superficial/deep SSI in any specific groups. Risk factors such as diabetes mellitus and colon injuries were found to be more significant in the development of superficial/deep incisional SSI and surgical incision complications. These findings contrast the dogmatic approach of keeping high-risk abdominal incisions open and suggest that primary closures can be performed more safely than commonly perceived.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

References

Magill

, Edwards

, Bamberg

, et al. Multistate point-prevalence survey of health care-associated infections. N Engl J Med, 2014; 370:1198–1208.

National and State Healthcare Associated Infections Progress Report, based on 2014 data, published in 2016. 2016. www.cdc.gov/HAI/pdfs/progress-report/hai-progress-report.pdf. Last accessed March 3, 2017.

Surgical Site Infection (SSI) Event Procedure-associated Module. www.cdc.gov/nhsn/pdfs/pscmanual/9pscssicurrent.pdf. 2017. Last accessed March 3, 2017.

Mangram

, Horan

, Pearson

, et al. Guideline for Prevention of Surgical Site Infection, 1999. Centers for Disease Control and Prevention (CDC) Hospital Infection Control Practices Advisory Committee. Am J Infect Control, 1999; 27:97–132.

Anderson

, Podgorny

, Berrios-Torres

, et al. Strategies to prevent surgical site infections in acute care hospitals: 2014 update. Infect Control Hosp Epidemiol, 2014; 35(Suppl 2):S66–S88.

Grace

, Cox

. Incidence of incisional hernia following dehiscence of the abdominal wound. Proc R Soc Med, 1973; 66:1091–1092.

Bucknall

, Cox

, Ellis

. Burst abdomen and incisional hernia: A prospective study of 1129 major laparotomies. Br Med J, 1982; 284:931–933.

Irvin

, Stoddard

, Greaney

, Duthie

. Abdominal wound healing: A prospective clinical study. Br Med J, 1977; 2:351–352.

Cruse

, Foord

. The epidemiology of wound infection. A 10-year prospective study of 62,939 wounds. Surg Clin North Am, 1980; 60:27–40.

10.

Cruse

Wound infection surveillance. Rev Infect Dis, 1981; 3:734–737.

11.

Zimlichman

, Henderson

, Tamir

, et al. Health care-associated infections: A meta-analysis of costs and financial impact on the US health care system. JAMA Intern Med, 2013; 173:2039–2046.

12.

Hunter

A Treatise on the Blood, Inflammation, and Gunshot Wounds. London: Nicol, 1794: 216–220.

13.

Wangensteen

, Wangensteen

, Klinger

. Wound management of Ambroise Pare and Dominique Larrey, great French military surgeons of the 16th and 19th centuries. Bull Hist Med, 1972; 46:207–234.

14.

Bender

JS.

Factors influencing outcome in delayed primary closure of contaminated abdominal wounds: A prospective analysis of 181 consecutive patients. Am Surg, 2003; 69:252–255.

15.

Cohn

, Giannotti

, Ong

, et al. Prospective randomized trial of two wound management strategies for dirty abdominal wounds. Ann Surg, 2001; 233:409–413.

16.

Duttaroy

, Jitendra

, Duttaroy

, et al. Management strategy for dirty abdominal incisions: Primary or delayed primary closure? A randomized trial. Surg Infect (Larchmt), 2009; 10:129–136.

17.

Smilanich

, Bonnet

, Kirkpatrick

. Contaminated wounds: The effect of initial management on outcome. Am Surg, 1995; 61:427–430.

18.

Weiss

Delayed closure in the management of the contaminated wound. Int Surg, 1982; 67(4 Suppl):403–404.

19.

Bhangu

, Singh

, Lundy

, Bowley

. Systemic review and meta-analysis of randomized clinical trials comparing primary vs delayed primary skin closure in contaminated and dirty abdominal incisions. JAMA Surg, 2013; 148:779–786.

20.

Seamon

, Smith

, Capano-Wehrle

, et al. Skin closure after trauma laparotomy in high-risk patients: Opening opportunities for improvement. J Trauma Acute Care Surg, 2013; 74:433–439.

21.

Adkins

Jr ., Zirkle

, Waterhouse

Penetrating colon trauma. J Trauma, 1984; 24:491–499.

22.

O'Neill

, Kirton

, Dresner

, et al. Analysis of 162 colon injuries in patients with penetrating abdominal trauma: Concomitant stomach injury results in a higher rate of infection. J Trauma, 2004; 56:304–312.

23.

Pommerening

, Kao

, Sowards

, et al. Primary skin closure after damage control laparotomy. Br J Surg, 2015; 102:67–75.

24.

Velmahos

, Vassiliu

, Demetriades

, et al. Wound management after colon injury: Open or closed? A prospective randomized trial. Am Surg, 2002; 68:795–801.

25.

Voyles

, Flint

Jr . Wound management after trauma to the colon. South Med J, 1977; 70:1067–1069.