Risk Factors for Abdominal Surgical Site Infection after Exploratory Laparotomy among Combat Casualties

Abstract

Background:

Surgical site infections (SSI) are well-recognized complications after exploratory laparotomy for abdominal trauma; however, little is known about SSI development after exploration for battlefield abdominal trauma. We examined SSI risk factors after exploratory laparotomy among combat casualties.

Methods:

Military personnel with combat injuries sustained in Iraq and Afghanistan (6/2009–5/2014) who underwent laparotomy and were evacuated to participating U.S. military hospitals were included. Log-binominal regression was used to identify SSI risk factors.

Results:

Of 4304 combat casualties, 341 patients underwent a total of 1053 laparotomies. Abdominal SSIs were diagnosed in 49 (14.4%) patients; 8% with organ space SSI, 4% with deep incisional SSI, and 4% with superficial SSIs (4 patients had multiple SSIs). Patients with SSIs had more colorectal (p<0.001), small bowel (p=0.010), duodenum (p=0.006), pancreas (p=0.032), and abdominal vascular injuries (p=0.040), as well as prolonged open abdomen (p=0.004) and more infections diagnosed before the SSI (or final ex-lap) versus non-SSI patients (p<0.001). Sustaining colorectal injuries (risk ratio [RR]: 3.20; 95% CI: 1.58–6.45), duodenum injuries (RR: 6.71; 95% CI: 1.73–25.58), and being diagnosed with prior infections (RR: 10.34; 95% CI: 5.05–21.10) were independently associated with any SSI development. For either organ space or deep incisional SSIs, non-intraabdominal infections, fecal diversion, and duodenum injuries were independently associated, while being injured via an improvised explosive device (IED) was associated with reduced likelihood compared to penetrating non-blast (e.g., gunshot wounds) injuries. Non-intraabdominal infections and hypotension were independently associated with organ space SSIs development alone, while sustaining blast injuries were associated with reduced likelihood.

Conclusions:

Despite severity of injuries and the battlefield environment, the combat casualty laparotomy SSI rate is relatively low at 14% and is comparable to risk factors and rates reported following severe civilian trauma.

Level of Evidence:

Epidemiological, Level III

Background:

Trauma-related laparotomies were performed in up to 13% of combat-injured service members in recent overseas conflicts.^1,2 Among civilian trauma patients who undergo abdominal surgery, up to 36% may develop a surgical site infection (SSI),^3–6 which is a significant source of morbidity and mortality, as well as added healthcare costs.⁷ Damage control laparotomy (DCL) is frequently performed for combat-injured patients with abdominal injuries.^8–10 Accordingly, patients are often left with open abdomens, temporizing hemostasis measures, or in bowel discontinuity until they are evacuated to higher echelons of care for definitive surgical treatment.^1,9,11,12 Characterization of SSI after exploratory laparotomy for battlefield trauma is poorly defined, as prior studies have usually reported short-term and in-theater outcomes, and either broadly grouped or only reported specific infectious categories.^{1,2,8,9,12–15} Furthermore, battlefield casualties are subject to devastating patterns of injury resulting from blast mechanisms and high-velocity rounds, often sustained in inherently dirty and unique geographic environments.^11,16,17 Their clinical course is frequently complicated by complex soft-tissue injuries, traumatic amputations, and invasive fungal and multidrug-resistant organisms that are endemic to the battlefield region.^1,11,16,18 Unfortunately, high-velocity gunshot wounds and blast injuries have been recently encountered in the civilian trauma setting, and the sequela of such mechanisms are now similar to battlefield casualties.^19–21

We have previously performed an initial assessment of the epidemiology of abdominal SSIs following combat-related laparotomy.¹⁰ The objective of this analysis was to identify risk factors for the development of SSI after exploratory laparotomy in battlefield trauma patients.

Methods:

Study Design

We have previously described the data repository from the Trauma Infectious Disease Outcomes Study (TIDOS), which is an observational study of infectious complications following combat trauma.^18,22 Criteria for this analysis included any military personnel that underwent exploratory laparotomy during combat operations in Iraq and Afghanistan from 1 June 2009 through 31 May 2014 and were medically evacuated from the combat zone to Landstuhl Regional Medical Center (LRMC, Germany) before being transferred to a participating military hospital in the United States. The Institutional Review Board of the Uniformed Services University of the Health Sciences (Bethesda, MD) approved the study protocol and all research was conducted in accordance with the Health Insurance Portability and Accountability Act requirements.

The Department of Defense Trauma Registry (DoDTR) was used to collect patient demographics, injury characteristics, trauma history, and early trauma care.²³ Infection-related information was obtained through the DoDTR TIDOS Infectious Disease module.²² Retrospective abstraction of all individual medical records was further used to elicit information relating to abdominal surgery and additional information regarding their hospitalization and outcomes during the inpatient period.

Abdominal SSIs were identified based on a combination of clinical findings, medical record documentation, imaging, and laboratory results and were classified by data abstractors in accordance with the Centers for Disease Control and Prevention National Healthcare Safety Network and American College of Surgeons guidelines.^5,7,24 An organ space SSI involved any part of the abdomen deep to the fascia or muscle, while a deep incisional SSI involved the fascia or muscle layers, and a superficial incisional SSI involved the skin or subcutaneous tissue.⁷

Statistical Analysis

Chi-square or Fisher’s exact test were used to detect differences between patients with and without SSIs for categorical variables, and the Kruskal-Wallis test was used for continuous variables. Risk factors for the development of abdominal SSIs following exploratory laparotomy were assessed in univariate and multivariable log-binomal regression models. Variables significant in the univariate model were assessed for inclusion in the multivariable model using the backward selection method. Variables that had substantial missing observations (>35% of patients missing data) were excluded from the multivariable model. Significance was defined as a p-value of <0.05. Statistical analysis was completed using SAS version 9.4 (SAS, Cary, NC).

Results:

Study Population

A total of 4304 combat casualties were evaluated, of which 341 (7.9%) underwent trauma-related laparotomy and met inclusion criteria for analysis. The majority of patients were male (99%) with a median age of 24 years old. Blast-related mechanisms accounted for 72.4% of injuries (56.9% were caused by improvised explosive devices [IED], of which 75% occurred while on foot patrol). The median Injury Severity Score (ISS) was 33, and 53.4% of patients received a massive blood transfusion (≥10 units packed red blood cells or whole blood) within 24 hours of injury (Table 1). Extremity trauma was common in the study population with 35% sustaining amputations, 37% with open fractures, and 16% with other open extremity injuries. The median Sequential Organ Failure Assessment (SOFA) score was 8 upon admission to LRMC.

Table 1:

Characteristics of Combat Casualties who Underwent Exploratory Laparotomy

Characteristics, No. (%)	Total Patients (N=341)	Patients with Abdominal SSIs (N=49)	Patients without Abdominal SSIs (N=292)	P-value
Male	336 (98.5)	49 (100)	287 (98.3)	~1.00
Age at injury, median years (IQR)	24 (21, 27)	24 (21, 26)	24 (21, 28)	0.366
Injured in Afghanistan	327 (95.9)	46 (93.9)	281 (96.2)	0.434
Injury Mechanism				0.109
IED Blast	194 (56.9)	26 (53.1)	168 (57.5)
Non-IED Blast	53 (15.5)	4 (8.2)	49 (16.8)
Penetrating Non-Blast†	94 (27.6)	19 (38.8)	75 (25.7)
Injury Severity Score, median (IQR)	33 (26, 45)	38 (30, 50)	33 (26, 45)	0.070
0 – 9 (minor)	10 (2.9)	0	10 (3.4)	0.438
10 – 15 (moderate)	10 (2.9)	0	10 (3.4)
16 – 25 (severe)	51 (15.0)	8 (16.3)	43 (14.7)
≥26 (critical)	270 (79.2)	41 (83.7)	229 (78.4)
Abdominal Severity, median (IQR) ‡	3 (2, 4)	4 (3, 4)	3 (2, 4)	0.011*
PRBC units within 24 hrs post-injury, median (IQR)	15 (6, 30)	24 (6, 40)	14 (6, 28)	0.034*
None / Missing	63 (18.5)	7 (14.3)	56 (19.2)	0.212
1 – 9	96 (28.2)	12 (24.5)	84 (28.8)
10 – 20	73 (21.4)	8 (16.3)	65 (22.3)
≥20	109 (32.0)	22 (44.9)	87 (29.8)
Extremity injuries				0.426
Amputation	119 (34.9)	19 (38.9)	100 (34.3)
Open fracture	125 (36.7)	15 (30.6)	110 (37.7)
Other open injury	54 (15.8)	6 (12.2)	48 (16.4)
LRMC Admission SOFA Score, median (IQR)	8 (4, 11)	9 (6, 12)	7 (3, 10)	0.009*
Mechanical Ventilation	248 (72.7)	43 (87.8)	205 (70.2)	0.011*
Ventilation prior to SSI§	178 (52.2)	41 (83.7)	137 (46.9)	<0.001*
Intensive Care Unit (ICU) Admission	304 (89.1)	47 (95.9)	257 (88.0)	0.100
Number of Operating Room Visits, median (IQR)||	6 (2, 12)	9 (4, 18)	5 (2, 11)	0.001*
Maximum Clinical Characteristics, median (IQR)§
WBC count (cells/mm3)	9.9 (7.9, 13.2)	12.3 (9.3, 15.6)	9.5 (7.8, 12.2)	0.007*
BUN (mg/dL)	16 (12, 24)	17 (14, 39.3)	15.5 (11, 23)	0.024*
Creatinine (mg/dL)	1.1 (0.9, 1.6)	1.1 (0.9, 2.0)	1.1 (1.0, 1.6)	0.969
Any infection diagnosed§	85 (24.9)	34 (69.4)	51 (17.5)	<0.001*
Sepsis	14 (4.1)	6 (12.2)	8 (2.7)	0.002*
Skin and soft-tissue infection	55 (16.1)	26 (53.1)	29 (9.9)	<0.001*
Bloodstream infection	33 (9.7)	13 (26.5)	20 (6.9)	<0.001*
Pneumonia	32 (9.4)	11 (22.5)	21 (7.2)	0.001*
Osteomyelitis	8 (2.4)	2 (4.1)	6 (2.1)	0.323
Urinary tract infection	2 (0.6)	2 (4.1)	0	0.020*
Clostridioides difficile	3 (0.9)	2 (4.1)	1 (0.3)	0.055
Sinusitis	1 (0.3)	1 (2.0)	0	0.144
Empyema without Pneumonia	2 (0.6)	1 (2.0)	1 (0.3)	0.267
Received perioperative and postoperative antibiotics¶	282 (82.7)	44 (89.8)	238 (81.5)	0.158
Number of exploratory laparotomies, median (IQR)	2 (1, 3)	4 (2, 6)	2 (1, 3)	<0.001*
Prior to SSI§	2 (1, 3)	3 (2, 5)	2 (1, 3)	<0.001*
Primary Fascial Closure, median days (IQR)††	3 (3, 3)	3 (3, 3)	3 (3, 3)	0.564
Length of hospitalization, median days (IQR)	35 (19, 63)	59 (30, 96)	34 (17, 55)	<0.001*
Survival	329 (96.5)	48 (98.0)	281 (96.2)	~1.00

BUN – blood urea nitrogen; IQR – interquartile range; LRMC – Landstuhl Regional Medical Center; PRBC – packed red blood cells or whole blood; SOFA – Sequential Organ Failure Assessment; WBC – white blood cell

^*Denotes statistical significance, p<0.05

^†Non-blast mechanism are primarily gunshot wounds (Total N=83; SSI patients N=17; non-SSI patients N=66).

^‡Abdominal severity defined using Abbreviated Injury Scale codes

^§Values are prior to diagnosis of abdominal SSI or prior to the final exploratory laparotomy for patients who did not develop an abdominal SSI. For clinical characteristics, they are the maximum values over the entire hospitalization period for the non-SSI patients.

^||Includes visits for laparotomies as well as other surgical procedures (e.g. for extremity injuries).

^¶Includes antibiotic post-trauma prophylaxis and perioperative antibiotics for concomitant injuries

^††Data missing for 5 patients (3 with SSI and 2 without SSI)

The median number of operating room visits (not including only laparotomies, such as for extremity procedures) for the study population was 6 (9 for patients with SSIs and 5 for those without SSIs, p=0.001). Approximately 83% of patients received both perioperative and postoperative antibiotics. The majority of patients (89%) were admitted to an intensive care unit (ICU) and 73% of patients required post-operative mechanical ventilation (Table 1). Overall, patient survival was 96.5%, with no significant differences observed between patients with and without SSI (98% vs. 96.2%, p=1.00). The median length of hospitalization was 35 days in all patients, including 59 days (interquartile range [IQR]: 30–96) for patients with SSI and 34 days (IQR: 17–55) for patients without SSI (p<0.001).

Laparotomy Indications and Findings

A total of 1053 laparotomies (index and subsequent operations) were performed among the 341 patients. The median number of laparotomies per patient was 2 (IQR: 1–3; range: 1–28; Table 1). The most common trauma indication for index laparotomy was penetrating injury (57% of 341) with abdominal explorations primarily being performed for open abdomen at subsequent operation (73%) (Table 2). Approximately 49% of patients sustained a hollow viscus injury, 29% of patients suffered a solid organ injury, and 37% had a colorectal (large bowel) injury. Damage control laparotomy (defined as a positive traumatic finding with temporary abdominal closure at index operation) was performed in 47% of patients. The overall rate of negative laparotomy (i.e., no findings) was 19.9%, while 23% of patients underwent laparotomy for proximal control of the iliac vessels and/or fecal diversion. Negative pressure wound therapy (NPWT) was utilized for vacuum-assisted temporary abdominal closure in 72% of all laparotomies. Twenty-seven patients underwent repeat laparotomy after prior abdominal closure (e.g., small bowel obstruction, missed injury, and SSI management) during their admission.

Table 2:

Indications for Exploratory Laparotomy, Findings, and Outcomes at Exploration, No. (%)

Exploratory Laparotomy Characteristic†	Total Patients (N=341)	Patients with Abdominal SSIs (N=49)	Patients without Abdominal SSIs (N=292)	P-value
Damage Control Laparotomy	161 (47.2)	28 (57.1)	133 (45.6)	0.132
Exploratory Laparotomy Indication
Open Abdomen	248 (72.7)	44 (89.8)	204 (69.9)	0.004*
Discontinuity	102 (29.9)	25 (51.0)	77 (26.4)	<0.001*
Penetrating injury	194 (56.9)	34 (69.4)	160 (54.8)	0.056
Hypotension	116 (34.0)	27 (55.1)	89 (30.5)	<0.001*
Positive FAST exam	35 (10.3)	4 (8.2)	31 (10.6)	0.800
Computed tomography findings of intra-abdominal injury	101 (29.6)	12 (24.5)	89 (30.5)	0.395
Fecal diversion	62 (18.2)	17 (34.7)	45 (15.4)	0.001*
Proximal control of iliac vessels	75 (22.0)	12 (24.5)	63 (21.6)	0.649
Exploratory Laparotomy Type
Fecal Diversion and/or Proximal Control	78 (22.9)	12 (24.5)	66 (22.6)	0.771
Negative finding (no injuries)	68 (19.9)	5 (10.2)	63 (21.6)	0.065
Positive findings	225 (66.0)	45 (91.8)	180 (61.6)	<0.001*
Planned Reoperation	244 (71.6)	42 (85.7)	202 (69.2)	0.018*
Exploratory Laparotomy Injury Findings
Hollow Viscus Injury	167 (49.0)	36 (73.5)	131 (44.9)	<0.001*
Solid Organ Injury	100 (29.3)	17 (34.7)	83 (28.4)	0.372
Colorectal Injury	125 (36.7)	32 (65.3)	93 (31.9)	<0.001*
Individual Organ Injury
Spleen	61 (17.9)	10 (20.4)	51 (17.5)	0.619
Small Bowel Injury	77 (22.6)	18 (36.7)	59 (20.2)	0.010*
Colon Injury	82 (24.1)	22 (44.9)	60 (20.6)	<0.001*
Rectum Injury	57 (16.7)	14 (28.6)	43 (14.7)	0.016*
Duodenum Injury	12 (3.5)	5 (10.2)	7 (2.4)	0.006*
Liver	48 (14.1)	6 (12.2)	42 (14.4)	0.690
Kidney	27 (7.9)	5 (10.2)	22 (7.5)	0.522
Gastric	15 (4.4)	2 (4.1)	13 (4.5)	~1.00
Biliary tree	5 (1.5)	0	5 (1.7)	~1.00
Pancreas	15 (4.4)	5 (10.2)	10 (3.4)	0.032*
Bladder	25 (7.3)	6 (12.2)	19 (6.5)	0.154
Diaphragm	25 (7.3)	5 (10.2)	20 (6.9)	0.404
Retroperitoneal Hematoma Injury	69 (20.2)	15 (30.6)	54 (18.5)	0.051
Abdominal Vascular Injury	20 (5.9)	6 (12.2)	14 (4.8)	0.040*
Iliac vascular injury	16 (4.7)	3 (6.1)	13 (4.5)	0.712

FAST – focused assessment with sonography in trauma

^*Denotes statistical significance, p<0.05

^†Many of the patients had multiple indications, findings, or outcomes

Abdominal Surgical Site Infections

Forty-nine (14.4%; 95% Confidence Interval [CI]: 10.6–18.1) patients who underwent combat-associated exploratory laparotomy developed an abdominal SSI. Superficial SSIs were diagnosed in 12 (3.5%; 95% CI: 1.6–5.5), deep incisional SSIs in 14 (4.1%; 95% CI: 2.0–6.2), and organ space SSIs in 27 (7.9%; 95% CI: 5.1–10.8). Two patients had both an organ space SSI and a deep incisional SSI and two patients had both an organ space SSI and a superficial SSI. Supplemental Tables 1–4 provide characteristics for patients with organ space or deep incisional SSIs, as well as for patients with only organ space SSIs.

Use of mechanical ventilation, maximum SOFA score, maximum blood urea nitrogen, occurrence of any infection prior to the diagnosis of the SSI (or prior to final ex-lap) were significantly associated (p<0.05) with any SSI risk on univariate analysis (Table 3). In addition, maximum abdominal severity defined by Abbreviated Injury Scale codes, blood transfusion within 24 hours of injury, number of exploratory laparotomies prior to SSI, exploratory laparotomy indications (i.e., hypotension, fecal diversion, open abdomen, and discontinuity), and exploratory laparotomy findings (i.e., hollow viscus injury, abdominal vascular injury, colorectal injuries, and retroperitoneal hematoma, as well as injuries to the small bowel, colon, rectum, pancreas, and duodenum) were also associated with the development of SSIs.

Table 3.

Log-binomial Regression Analysis of Univariate Risk Factors for the Development of Abdominal Surgical Site Infections Following Exploratory Laparotomy

Factor	Univariate Risk Ratio (95% CI)	P-value
Injury severity score, continuous	1.02 (1.00, 1.03)	0.137
Injury mechanism		0.087
Penetrating non-blast†	Reference
IED Blast	0.66 (0.39, 1.14)
Non-IED Blast	0.37 (0.13, 1.04)
Max Abdominal Severity	1.34 (1.08, 1.67)	0.007*
Blood transfusion within 24 hours post-injury, continuous‡	1.02 (1.00, 1.03)	0.010*
Mechanical ventilation§	4.69 (2.27, 9.71)	<0.001*
Maximum SOFA score, continuous§,||	1.06 (1.01, 1.11)	0.029*
Maximum WBC count, continuous§,¶	1.00 (0.98, 1.03)	0.883
Maximum BUN, continuous§,††	1.01 (1.00, 1.02)	0.008*
Any non-Intraabdominal Infection§	13.27 (6.46, 27.25)	<0.001*
Pneumonia	2.80 (1.59, 4.91)	<0.001*
Skin and soft-tissue infection	5.88 (3.63, 9.51)	<0.001*
Bloodstream infection	3.37 (2.00, 5.69)	<0.001*
Damage Control Laparotomy	1.49 (0.88, 2.52)	0.136
Number of exploratory laparotomies prior to SSI§	1.28 (1.11, 1.48)	0.001*
Exploratory Laparotomy Indication
Penetrating Injury	1.72 (0.97, 3.03)	0.062
Hypotension	2.38 (1.42, 3.99)	0.001*
Fecal diversion	2.39 (1.42, 4.02)	0.001*
Open abdomen	3.30 (1.35, 8.07)	0.009*
Discontinuity	2.44 (1.47, 4.06)	0.001*
Exploratory Laparotomy Injury Findings
Small Bowel	1.99 (1.18, 3.36)	0.010*
Colon	2.57 (1.55, 4.27)	<0.001*
Rectum	1.99 (1.15, 3.46)	0.014*
Pancreas	2.47 (1.15, 5.32)	0.021*
Duodenum	3.12 (1.51, 6.42)	0.002*
Retroperitoneal Hematoma	1.74 (1.01, 3.01)	0.047*
Abdomen Vascular	2.24 (1.08, 4.62)	0.029*
Hollow Viscus	2.89 (1.59, 5.24)	0.001*
Colorectal	3.25 (1.89, 5.61)	<0.001*

BUN – blood urea nitrogen; CI – confidence interval; IED – improvised explosive device; SOFA – sequential organ failure assessment; WBC – white blood cell

^*Denotes statistical significance, p<0.05

^†Non-blast mechanism was primarily gunshot wounds

^‡Blood transfusion information is missing for 60 patients.

^§Period for variable was until the diagnosis of the abdominal SSI or the time of the final exploratory laparotomy for patients who did not develop an abdominal SSI

^||SOFA score is missing for 122 patients. Due to the missing numbers, SOFA score was not included in the multivariable analysis.

^¶WBC count is missing for 143 patients. Due to the missing numbers, WBC count was not included in the multivariable analysis.

^††BUN missing for 144 patients. Due to the missing numbers, WBC count was not included in the multivariable analysis.

Following assessment in the multivariable model, sustaining a colorectal injury (risk ratio [RR]: 3.20; 95% CI: 1.58–6.45, p=0.001), duodenum injury (RR: 6.71; 96% CI: 1.73–25.58, p=0.006), and being diagnosed with an infection prior to the SSI diagnosis (RR: 10.34; 95% CI: 5.05–21.10, p<0.001) remained independently associated with any SSI development (Table 4).

Table 4.

Log-binomial Regression Multivariable Analysis of Risk Factors for the Development of Surgical Site Infections (SSI) Following Exploratory Laparotomy

Factor	Any SSI RR (95% CI)	P-value	Organ Space or Deep Incisional SSI RR (95% CI)	P-value	Organ Space SSI only RR (95% CI)	P-value
Injury mechanism				0.001*		<0.001*
Penetrating non-blast†	–	–	Reference		Reference
IED Blast	–	–	0.16 (0.06, 0.43)		0.14 (0.05, 0.39)
Non-IED Blast	–	–	0.28 (0.06, 1.12)		0.13 (0.02, 0.90)
Any Infection‡	10.34 (5.05, 21.10)	<0.001*	8.11 (4.29, 12.53)	<0.001*	5.95 (2.65, 10.93)	<0.001*
Exploratory Laparotomy Indication
Hypotension	–	–	–	–	2.83 (1.22, 5.89)	0.017*
Fecal diversion	–	–	3.21 (1.53, 5.71)	0.003*	–
Exploratory Laparotomy Injury Findings
Colorectal	3.20 (1.58, 6.45)	0.001*	–	–	–	–
Duodenum	6.71 (1.73, 25.58)	0.006*	4.27 (1.47, 7.32)	0.011*	–	–

CI – confidence interval; IED – improvised explosive device; RR – risk ratio

^*Denotes statistical significance, p<0.05

^†Non-blast mechanism was primarily gunshot wounds

^‡Period for variable was until the diagnosis of the abdominal surgical site infection or the time of the final exploratory laparotomy for patients who did not develop an abdominal surgical site infection

Factors associated with the development of either organ space or deep incisional SSIs were also assessed (Table 4). Sustaining a non-intraabdominal infection (RR: 8.11; 95% CI: 4.29–12.53; p<0.001), fecal diversion as the indication for the laparotomy (RR: 3.21; 95% CI: 1.53–5.71; p=0.003); and a duodenum injury (RR: 4.27; 95% CI: 1.47–7.32; p=0.011) were independently associated with the development of an organ space or deep incisional SSI. Furthermore, sustaining an injury from an IED (RR: 0.16; 95% CI: 0.06–0.43; p=0.001) was associated with a reduced likelihood infection risk compared to penetrating non-blast mechanisms (88% had a gunshot wound).

After restricting to only patients with organ space SSIs, non-intraabdominal infections (RR: 5.95; 95% CI: 2.65–10.93; p<0.001) and hypotension as the indication for the laparotomy (RR: 2.83; 95% CI: 1.22–5.89; p=0.017) were independently associated with organ space SSI development (Table 4). Sustaining a blast injury (p<0.001), either from an IED (RR: 0.14; 95% CI: 0.05–0.39) or a non-IED, such as a grenade, (RR: 0.13; 95% CI: 0.02–0.90) were associated with a reduced infection likelihood compared to penetrating non-blast injury mechanisms (87% had a gunshot wound).

Discussion:

Surgical site infections after combat laparotomy developed in 14% of patients in our study, which analyzed wounded military personnel evacuated from the combat theater back to the United States. These casualties were severely injured, suffered a high rate of concurrent soft tissue injuries, extremity fractures, and amputations, and frequently received massive blood transfusions. The prolonged hospitalization in our military healthcare system allowed us to follow these patients for at least a median of 30 days after index operation. Furthermore, utilization of multiple databases and individual chart review provided the framework for robust retrospective data analysis.

In our population, colorectal and duodenum injuries were independently associated with the development of abdominal SSIs. Duodenum injuries were also significant risk factors for organ space or deep incisional SSI development. Hollow viscus injuries, especially large bowel, pose well-recognized risks for SSIs after trauma laparotomy.^3–5 Other reported SSI risk factors after abdominal trauma include wound classification, pancreatic injury, transfusion requirements, hypothermia, and higher abdominal injury severity.^4,25,26 Our study population consisted of young, previously healthy, and primarily male patients with severe injuries (median ISS of 33) and high blood transfusion requirements (median 15 units packed red blood cells or whole blood within 24 hours of injury). The high injury severity observed in both groups (median ISS of 38 in SSI patients and 33 in non-SSI patients) did not bear out to statistical significance; however, no ISS of an SSI patient was less than 16 (minor to moderate range). While the high overall transfusion requirement (median 24 units in SSI patients and 14 in non-SSI patients) was significantly associated with the development of SSI in the univariate model, data missingness (60 patients, 17.6%), unfortunately, precluded inclusion for multivariable analysis. Although we were unable to ascertain degree of abdominal contamination in our chart review, it is possible that sustainment of colorectal (large bowel) injuries serves as a surrogate for intra-operative contamination.

We further elicited that concurrent infections diagnosed prior to SSI were independently associated with any type of SSI development. The risk of concurrent infections is well-described in the orthopaedic surgery literature, and other studies have reported a similar association with SSI.²⁷ Perhaps the susceptibility to multiple infections is due to an underlying immunocompromised state or global organ dysregulation in response to insults, such as injury or massive transfusion. In the univariate model, SOFA scores were associated with risk for developing SSIs. The utility of SOFA score is characterized in sepsis literature, as it describes a continuum of organ dysfunction and estimates the risk of morbidity and mortality in sepsis.²⁸ It is also a reliable indicator of prolonged length of stay and death in trauma patients.^29,30 Nevertheless, the SOFA score does not include an abdominal component, so it may underestimate risk for laparotomy-related complications. Unfortunately, due to data missingness, we were unable to incorporate SOFA score into our multivariable analysis. It is also possible that the environmental contamination from battlefield injury may lead to organism seeding in the abdomen (such as through injury tracts), as well as close proximity soft tissue colonization. Over half (53%) of patients who developed an SSI also suffered from a skin or soft-tissue infection. This was likely instrumental in the elicited relationship between SSI development and concurrent infections.

Blast injuries (IED or non-IED) were independently associated with a reduced likelihood organ space SSI, compared to penetrating non-blast injury mechanism (87% of organ space SSI patients with gunshot wounds). Similarly, IED injury mechanism was associated with a reduced likelihood of organ space or deep incisional SSI. It is possible that penetrating mechanisms, such as gunshot wounds, tended to produce greater intra-abdominal contamination, thus predisposing combat casualties to a greater risk of organ space and/or deep incisional SSI. Unfortunately, record limitations restricted our ability to ascertain the degree of contamination at index exploration, as well as performance of lavage and control of contamination.

Abdominal wounds represent approximately 10% of the battlefield injury burden, and exploratory abdominal surgery contributes to approximately 13% of all modern combat-associated surgical procedures.^1,2,31 Although there have been studies in civilian trauma populations, combat trauma care is inherently different from civilian trauma care (e.g., combat environment, transport logistics and duration, and mechanism of injury). The far-forward surgeon may have reduced capability in an isolated “primitive” tent or structure-based operating room with limited personnel assets (Role II) compared to the surgeon at a robust field hospital with tertiary support (Role III). Battlefield wounds are usually inflicted by high-energy blast and gunshot mechanisms, which cause a high frequency of concurrent injuries, such as traumatic amputations and soft-tissue defects.^16,17,31,32 Non-abdominal infections are common and complicate the care and recovery of these combat-injured patients.^22,32 A typical patient in this population may have suffered an abdominal injury, traumatic amputation(s), extensive soft-tissue destruction, received a massive transfusion, and undergone initial surgical care in a small military base with a tent-based, or equivalent, operating room.¹⁶ Despite these and other variables that would presumably lead to higher SSI rates in combat casualties, we found rates and associations comparable to civilian trauma populations.^3–6

Trauma remains a leading preventable cause of death and disability in the civilian world, and SSI after laparotomy is a substantial source of morbidity.^3,12,33 While risk factors for abdominal SSIs in civilian trauma have been well-described, combat trauma literature is sparse.^3–5,10 Bograd and colleagues described a complication rate as high as 40% after DCL from 2003–2009, but grouped non-infectious and infectious complications together.¹² Examination of 135 combat casualties who had battlefield exploratory laparotomies found that 10% developed a SSI with a higher proportion among those who had proximal vascular control vs non-proximal control (23% vs 4%).³⁴ In another study, Smith and colleagues reported an SSI rate of 18% for a cohort of 82 combat theater patients definitely treated for abdominal trauma without long-term follow up.⁹ Two other studies described abdominal infectious complications in patients managed for combat-associated hollow viscus injuries. One study found a 21.5% rate of intra-abdominal abscess in 65 patients with colonic injuries, while the other reported a 14.6% rate of “wound infections” and 4.1% for abscess formation in 171 patients that underwent small bowel resection.^15,35 Short-term outcomes, usually at Role II or III facilities in the area of combat operations, have been reported in additional series, but these studies lacked the longitudinal follow-up to adequately capture infectious outcomes.^8,13,14,36

Surgical care during the wars in Iraq and Afghanistan underwent maturation and standardization. Comprehensive Clinical Practice Guidelines were instituted through the Joint Trauma System and refined for various aspects of trauma care for combat casualties.³⁷ Point of injury care, rapid transport to surgical capability, ratio-driven resuscitation, minimization of crystalloids, and careful placement of surgical teams were instrumental in this development.^{11,17,23,38–41} Damage control laparotomy in the far-forward environment with surgical stabilization, use of NPWT for vacuum-assisted abdominal closure, aggressive debridement of infected tissue, evacuation to higher echelons of care for definitive surgery, and timely abdominal closure were integral aspects of combat casualty surgical care.^9,17,37,40

Damage control laparotomy was utilized in 42.7% of our study population and was not independently associated with SSI development. Although DCL has been associated with anastomotic-related complications, risk for enteric fistula development, and abdominal infection, it does appear to confer a survival advantage, particularly in the care of combat casualties.^9,40–45 Timely closure of the open abdomen after DCL has been found to reduce the risk for SSI and abdominal-related complications, which occurred at a median of three days in our population.^42,43 Furthermore, over 70% of patients were treated with NPWT for vacuum-assisted abdominal closure, which may be associated with higher fascial closure rates.^42,43,46 This finding may suggest that current combat surgical clinical practice measures (such as DCL with expeditious subsequent abdominal closure) are not placing patients at higher risk for SSI development. Furthermore, nontherapeutic laparotomy and indications for proximal vascular control or fecal diversion were not independently associated with the risk of all SSI subtypes, although fecal diversion was associated with increased likelihood of organ space or deep incisional SSI development. This is intuitive, as performing end or looped diversion with an unprepped colon, especially in a trauma patient may predispose to infection.⁴⁷

Despite this severely injured population, the SSI rate is comparable that reported in civilian trauma literature.^3–6 Moreover, we found that the mechanism of injury and affected organs largely accounted for SSI risk factors, rather than the characteristics or techniques of the military trauma system, such as initial location and capability of surgical care as well as the utilization of damage control laparotomy. The maturation and refinement of combat casualty surgical care has contributed to the highest survival rates in modern warfare.^11,17,38 It is imperative to maintain this skillset during peacetime, as well as to pass on this knowledge to civilian trauma practice that has recently encountered more severe mechanisms of injuries.^19–21,48

The lessons learned from wartime must not be forgotten, and military surgeons need to maintain trauma readiness during peacetime in preparation for the next conflict. Unfortunately, civilian trauma centers have recently been exposed to the care of high velocity gunshot wound victims and blast victims from domestic terrorist events.^19–21 This presents an opportunity for military and civilian partnerships to negate the “peacetime effect,” as well as to integrate successful military principles into civilian trauma care.^38,48

Limitations of this study include those innate to retrospective analysis. Documentation is at the mercy of combat zone logistics, and accurate information extraction by the DoDTR, TIDOS Infectious Disease Module, and retrospective examination of the medical records is inherently limited. It should also be noted that resuscitation and antibiotic stewardship practices have evolved in recent years, yet ratio-driven resuscitation guidelines and antibiotic protocols existed during this study period.^26,49,50 Patients were included for analysis if they were transferred back to the United States, so a survival bias was likely elicited as it does not include patients who underwent laparotomy and died before being evacuation from the combat zone or died at LRMC. This study population consisted of predominately severely injured young males with uncommon devastating mechanisms of injury, and 83% of patients received preoperative and postoperative antibiotics. This may be due to the high frequency of concurrent fractures, traumatic amputations, and other extremity and soft-tissue injuries that likely necessitated extended antibiotic therapy not otherwise indicated after laparotomy alone. As such, this severely injured patient population is unique from typical civilian trauma cohorts. Furthermore, the relatively small sample size of patients with SSIs and severity of injury (high overall ISS and blood product transfusion) limited our ability to observe statistically significant differences with the multivariable analysis. Despite these limitations, the continuity of care in patients in the military healthcare system allowed us to extract information from the individual records and trauma registries pertaining to outcomes of greater than a median of 30 days as casualties were transferred through various military treatment facilities.

Conclusions:

This work is the first to report risk factors for abdominal surgical site infections in modern combat injured patients. Despite the high injury severity sustained by the battlefield casualties, the rate of laparotomy-related SSI development is comparable to civilian populations, with similar risk factors. Current military trauma surgical practices do not place casualties at higher SSI risk, but mechanism of injury and affected organs exert the strongest SSI influence.