Surgical Site and Graft Infections in Endovascular and Open Abdominal Aortic Aneurysm Surgery

Abstract

Background:

Surgical site infections (SSIs) constitute one of many major complications after aortic aneurysm surgery and its details and outcome have not been evaluated extensively. This research evaluates the incidence and outcome of SSI and graft infection in open and endovascular abdominal aortic aneurysm surgery.

Patients and Methods:

A retrospective study was performed, including all patients who underwent surgery for aortoiliac aneurysmatic disease between January 2000 and December 2016 in the Amphia Hospital, Breda, The Netherlands. Surgical site infections were defined in accordance with criteria of the U.S. Centers for Disease Control and Prevention (CDC). Aortic graft infection was diagnosed by a combination of computed tomography imaging and microbiologic results.

Results:

Between January 2000 and December 2016, 845 patients (mean age, 72.80 ± 7.15 y, 86.3% male) underwent abdominal aneurysm surgery (477 endovascular, 368 open). The total SSI rate was 3.1% (12 endovascular [2.5%], 14 open [3.8%], p = 0.318) with 0.8% deep SSI (26.9% of SSIs). No difference in SSIs was found between acute and elective surgery (2.9% vs. 3.1%, p = 1.000). Patients with SSI had a longer duration of stay (mean, 27.65 ± 30.74 d vs. 9.79 ± 12.50 d, p < 0.0001) but no difference in 30-day or 1-year mortality. Twenty-four graft infections occurred (4.3% of open vs. 1.7% of endovascular surgery, p = 0.023) of which 20 (83.3%) required surgery. Two-year mortality was higher when graft infections occurred (33.3% vs. 16.3%, p = 0.046). Surgical site infections (odds ratio [OR] 7.10, 95% [CI] 2.10–23.98) and mycotic aneurysms (OR 9.38, 95% CI 1.78–54.19) are independent determinants for graft infections.

Conclusion:

This study shows that SSIs after endovascular aortic surgery are as common as after open aortic surgery. Furthermore, this study supports the concept that aortic graft infection is highly associated with SSIs and mycotic aneurysms. Studies evaluating the prevention or early diagnosis of graft infection after SSIs through intensified imaging follow-up or even long-term antibiotic treatment should be considered.

Infections are major complications after aortic aneurysm surgery [1]. The incidence of surgical site infections (SSIs) is reported to be between 0.5% and 1.5% for endovascular surgery and between 2% and 14% for open surgery [2 –5]. Most studies that describe SSIs after abdominal aortic aneurysm surgery only report incidence numbers, without providing details of cause and outcome after SSIs. We hypothesize that the incidence of SSIs in endovascular aortic aneurysm surgery (EVAR) is higher than reported in the literature.

Deep SSIs are associated with aortic graft infection, sepsis, aortic rupture, and early mortality. Prosthetic aortic graft infection is a severe complication after open and endovascular abdominal aneurysm surgery with a mortality rate of 20%–40%, often requiring surgical intervention, with high amputation rates of 11% [6 –8]. Trends have been reported suggesting a link between peri-operative infections and the development of aortic graft infection [9], although no conclusive evidence can be found.

This research evaluates the incidence of SSIs and graft infection in open and endovascular abdominal aortic aneurysm surgery. It also reports the outcome of patients who develop an SSI or graft infection after abdominal aneurysm surgery.

Patients and Methods

Design

A retrospective study was conducted, including all patients who underwent surgery for aortoiliac aneurysmatic disease between January 2000 and December 2016 in the Amphia Hospital, Breda, The Netherlands.

Patients and data collection

Patient data were gathered from a database of all patients undergoing aortoiliac surgery in the aforementioned timeframe, selecting the patients with aneurysmatic disease, rather than occlusive aortoiliac disease. Both endovascular and open surgical procedures were included.

Data were gathered retrospectively by reviewing medical files in the hospital's electronic medical record. Patient characteristics included were age at the time of surgery, gender, aneurysm location, aneurysm size at the time of surgery, aneurysm morphology (mycotic, saccular, or para-anastomotic), American Society of Anesthesiologists Physical Status classification (ASA score), and specific risk factors such as smoking, diabetes mellitus, hypertension, chronic obstructive pulmonary disease, chronic renal disease, history of congestive heart failure, and pre-operative statin use. The post-operative information included duration of stay, time and number of re-interventions, pulmonary infections, urinary tract infections, bowel ischemia, and expansive information regarding SSIs, and graft infections.

Surgical site infections were defined, in accordance with the criteria of the U.S. Centers for Disease Control and Prevention (CDC) [10], as the presence of heat, swelling, or pain at or around the surgical site within 90 days after the initial procedure. Alternatively, a positive incision culture and/or drainage of pus from the incision or aspiration of pus in a diagnostic puncture were also criteria. Aortic graft infection was diagnosed by a combination of positron emission tomography or computed tomography imaging and positive blood culture or a culture from the graft during surgery.

Surgical procedures

All procedures were performed at the Amphia Hospital, Breda, The Netherlands. A regular operating room was used for both endovascular and open surgery, using a mobile C-arm for imaging support for endovascular operations. All EVAR procedures were performed using open surgical access to both femoral arteries. All patients received two grams of cefazolin as standard pre-operative antibiotic prophylaxis.

The choice for endovascular or open treatment and choice of (endo)graft type were made by the treating physician based on patient age, comorbidities, and aneurysm morphology. Acute surgery was scored when patients had a ruptured aneurysm at the time of surgery. Symptomatic aneurysm was scored when patients received aneurysm surgery within 24 hours of emergency admittance as a result of suspicion of symptomatic aneurysm through abdominal or flank pain, thromboembolism, and some cases of mycotic aneurysm. Open acute surgery constituted of either aortic tube or aortic bifurcation implantation. Endovascular acute surgery was performed by unilateral iliacal stenting with surgically implanted femorofemoral crossover. Among endovascular grafts used were Cook Zenith^® (Cook Medical LLC, Bloomington, IN), Gore Excluder^® (W. L. Gore & Associates, Inc., Flagstaff, AZ), Endologix AFX^® and Nellix^® (Endologix Inc., Irvine, CA), and Medtronic Endurant^® (Medtronic Minneapolis, MN).

Follow-up

Surgical sites were treated with a dry adhesive bandage for the first 48 hours after surgery. Transcutaneous stitches or staples, if used, were removed 14 days after surgery.

All patients were followed up by clinical examination by a vascular surgeon on the outpatient clinic for a period of three months. Additional imaging was only performed upon suspicion of complications. Patients were informed to contact a doctor immediately when signs of infection were present.

Loss to follow-up was scored after a patient was not seen in our hospital for a period longer than 24 months, except in the case of patient death. The last patient visit was extracted from our hospital's electronic patient records.

Statistical analysis

A Pearson χ² test was used to compare categorical variables; continuous variables were analyzed using independent t-tests. Logistic regression was used to evaluate risk factors. Statistical analysis was performed using SPSS software v24.0 (SPSS Inc., Chicago, IL). P values used were two-sided, unless explicitly mentioned as one-sided. A p value below 0.05 (or a one-sided p value below 0.025) indicated statistical significance.

Results

Between January 2000 and December 2016, 845 patients underwent aortoiliac aneurysm surgery. Endovascular procedures accounted for 477 cases, of which 412 (86.4%) were elective procedures. A total of 368 patients underwent open surgery, of which 224 (60.9%) procedures were elective. The baseline characteristics are shown in Table 1. The ratio of male patients was significantly higher in the endovascular group, whereas substantially more iliac aneurysms were treated with open procedures. The mean aneurysm diameter at surgery was higher in the group of patients treated with open surgery. Endovascular procedures were performed using Medtronic Endurant (n = 332), Gore Excluder (n = 20), Cook Zenith (n = 121), Endologix AFX (n = 1), and Endologix Nellix (n = 1). Two device brands could not be recovered from medical records. A mean follow-up duration of 49 months was observed, with 20 (2.37%) patients lost to follow-up.

Table 1.

Patient Characteristics

Characteristics	Endovascular (n = 477)	Open (n = 368)	p
Mean (SD) age	73.82 (7.06)	71.47 (7.06)	<0.0001
Gender, no. male (% male)	423 (88.7)	306 (83.2)	0.026
ASA score ≥3 (%)	286 (61.5)	222 (64.7)	0.377
Smoking (%)	153 (32.1)	157 (42.9)	0.002
Diabetes mellitus (%)	81 (17.0)	59 (16.0)	0.780
Hypertension (%)	269 (56.8)	212 (59.6)	0.435
Chronic renal disease (%)	19 (4.1)	21 (5.8)	0.327
Cardiac history (%)	218 (45.9)	139 (39.3)	0.065
Ruptured aneurysm (%)	45 (9.4)	125 (34.0)	<0.0001
Aneurysm characteristics
Mean (SD) aneurysm diameter in mm	62.99 (11.98)	70.31 (17.31)	<0.0001
Mycotic (%)	3 (0.6)	9 (2.5)	0.037
Anastomotic (%)	6 (1.3)	3 (0.8)	0.432
Saccular (%)	6 (1.3)	1 (0.3)	0.081
Iliac aneurysm (%)	34 (7.1)	55 (14.9)	<0.0001

SD = standard deviation; ASA = American Society of Anesthesiologists.

Surgical site infections

A total of 26 (3.1%) SSIs occurred, of which eight were after acute surgery (Table 2). After elective surgery, 11 endovascular surgery patients (2.7%) and nine open surgery patients (4.0%) developed an SSI (p = 0.352), as shown in Table 3. Of these, three (0.5%) deep SSIs occurred after endovascular treatment and three (1.3%) deep SSIs occurred after open abdominal access (p = 0.351). Table 4 presents the outcome after acute surgery. After acute surgery for ruptured or symptomatic abdominal aneurysm, one (1.5%) SSI occurred after endovascular treatment and five (3.5%) after open treatment (p = 0.668). All infections after acute endovascular procedures were deep infections, compared with one-fourth (1; 0.7%) of the infections after acute open surgery. All but one patient were treated with antibiotic agents after diagnosis of the SSI. This patient underwent incision and drainage of the surgical site abscess without further antibiotic treatment. All seven (0.8% of all) patients with deep SSIs required short-term surgery for infection control. Four patients required surgical evacuation of infected hematoma, one patient required drainage of abdominal abscesses that had contributed to abdominal evisceration. Removal of the prosthesis was necessary in two patients, in whom the synthetic femorofemoral crossovers were replaced with venous crossovers. One patient died as a direct consequence of deep SSI.

Table 2.

Overall Infection Rates and Secondary Outcomes

	Endovascular (n = 477)	Open (n = 368)	p
Surgical site infection rate (%)	12 (2.5)	14 (3.8)	0.318
Superficial infection (%)	9 (1.9)	10 (2.7)	0.477
Deep infection (%)	3 (0.6)	4 (1.1)	0.477
Re-admission for infection (%)	9 (1.9)	5 (1.4)	0.599
Re-intervention for infection (%)	4 (0.8)	4 (1.1)	0.734
Duration of hospital stay (mean, SD)^a	6.55 (7.56)	15.26 (17.83)	<0.0001
30-d mortality (%)	15 (3.1)	47 (12.8)	<0.0001
6-mo mortality (%)	31 (6.5)	65 (17.7)	<0.0001
1-y mortality (%)	40 (8.4)	69 (18.8)	<0.0001

including duration of readmissions.

SD = standard deviation.

Table 3.

Infection Rates and Secondary Outcomes after Elective Surgery

	Endovascular (n = 412)	Open (n = 224)	p
Surgical site infection rate (%)	11 (2.7)	9 (4.0)	0.352
Superficial infection (%)	9 (2.2)	6 (2.7)	0.351
Deep infection (%)	2 (0.5)	3 (1.3)	0.351
Re-admission for infection (%)	8 (1.9)	4 (1.8)	1.000
Re-intervention for infection (%)	3 (0.7)	4 (1.8)	0.357
Duration of hospital stay (mean, SD)^a	5.96 (7.19)	13.86 (14.76)	<0.0001
Re-intervention within 30 days	22 (5.3)	32 (14.3)	<0.0001
Re-intervention within 1 year	46 (11.2)	43 (19.2)	0.008
30-d mortality (%)	4 (1.0)	11 (4.9)	0.004
6-mo mortality (%)	18 (4.4)	16 (7.1)	0.144
1-y mortality (%)	24 (5.8)	19 (8.5)	0.247

Including duration of re-admissions.

SD = standard deviation.

Table 4.

Infection Rates and Secondary Outcomes in Acute Surgery

	Endovascular (n = 65)		Open (n = 144)
	Rupture (n = 45)	Symptomatic (n = 20)	Rupture (n = 125)	Symptomatic (n = 19)	p
Surgical site infection rate (%)	1 (1.5)	0	5 (3.5)	0	0.668
Superficial infection (%)	0	0	3 (2.1)	0	0.526
Deep infection (%)	1 (1.5)	0	2 (1.4)	0	0.526
Re-admission for infection (%)	1 (1.5)	0	1 (0.7)	0	0.526
Re-intervention for infection (%)	1 (1.5)	0	1 (1.5)	0	0.526
Duration of hospital stay (mean, SD)^a	10.78 (8.14)	6.60 (3.45)	18.49 (22.81)	10.37 (7.46)	0.011
30-d mortality (%)	10 (15.4)	1 (1.5)	33 (22.9)	3 (2.1)	0.215
6-mo mortality (%)	12 (18.5)	1 (1.5)	46 (31.9)	3 (2.1)	0.049
1-y mortality (%)	14 (21.5)	2 (3.0)	47 (32.6)	3 (2.1)	0.153

Including duration of re-admissions.

SD = standard deviation.

The mean hospital duration of stay was longer for patients with SSIs (9.79 ± 12.50 d without SSI vs. 27.65 ± 30.74 d with SSI; p < 0.0001). Table 5 presents the outcomes grouped for patients with SSI.

Table 5.

Characteristics and Outcome by Surgical Site Infection

	With SSI (n = 26)	Without SSI (n = 819)	p
Mean duration of hospital stay (SD)	27.65 (30.74)	9.79 (12.50)	<0.0001
Re-intervention <30 y (%)	9 (34.6)	91 (11.1)	0.002
Re-intervention <1 y (%)	11 (42.3)	135 (16.5)	0.002
30-d mortality	1 (3.8)	61 (7.4)	0.714
6-month mortality	3 (11.5)	93 (11.4)	1.000
Graft infection	4 (15.4)	20 (2.4)	0.005

SSI = surgical site infection; SD = standard deviation.

In 10 patients with SSIs, wound cultures yielded Staphylococcus aureus as the infectious pathogen. Three surgical site swabs tested positive for Streptococcus epidermidis, two for Bacteroides fragilis, and two for Escherichia coli. Pseudomonas aeruginosa, Morganella morganii, undetermined coagulase-negative staphylococci, and otherwise unspecified gram-positive rods were all isolated once, from different sites. Two cultures yielded mixed flora and in three patients a positive culture was not obtained.

The mean time to infection was 22 days (range, 5–65). This duration did not differ between elective and acute surgery (mean, 21.8 ± 14.8 vs. mean, 27.5 ± 17.9; p = 0.434) or endovascular and open surgery (mean, 21.1 ± 12.6 vs. mean, 24.8 ± 17.8; p = 0.552).

A logistic regression revealed no independent risk factor for developing surgical site infections. A Mantel-Cox log rank test showed a one-year mortality of 12.8% without SSI versus 15.4% with SSI with a p-value of 0.657.

Graft infections

A total of 24 (2.8%) graft infections occurred, 16 open reconstructive grafts (4.3% of all open grafts) and eight endografts (1.7% of endovascular grafts, p = 0.023). All cases were diagnosed by a combination of positive blood culture and radiographic imaging. The mean time to graft infection was 392 days (±696; range, 12–3,289) with a median of 118 days.

Surgery was required in 20 (83.3%) patients with graft infection, although five of them died before surgery or were deemed inoperable because of rapid decline. Of 15 operations, 11 consisted of open graft replacement, with nine rifampicin-soaked graft replacements, one autologous aortibifurcation replacement using the superficial femoral vein, and one autologous venous replacement of a femorofemoral crossover. Three of these removed grafts were endografts after EVAR, constituting 0.6% of all endovascular treatments, compared with eight (2.2%) graft removals after open surgery (p = 0.066). The remaining four procedures were because of bleeding from infectious pseudo-aneurysm (n = 3) and drainage of abdominal abscesses and evisceration treatment (n = 1).

Patients with graft infections were treated with antibiotic agents for an average of 325 days, ranging from 0 to 2,447 days. When discarding the extremes, such as three patients who were put on life-long antibiotic treatment and three patients who died within a week of being diagnosed with graft infection, patients were treated for a mean of 138 days, ranging from 14 to 440 days. The choice for the length of antibiotic treatment was made by the treating physician and clinical microbiologist based on the patient's symptoms, infectious pathogen, radiographic imaging, and laboratory tests.

The 30-day mortality for patients with graft infections was not significantly different from patients without graft infection (3.7% vs. 7.5%, p = 0.714). Over time, however, a difference is revealed with 25.0% one-year mortality for infected grafts versus 12.5% for grafts free of infection (p = 0.111). It becomes even more distinct two years after surgery, seen in a two-year mortality of 33.3% after graft infection, compared with 16.3% without infection (p = 0.046). Ten patients (41.7% of patients with graft infection) died as a direct consequence of the graft infection.

A logistic regression revealed that patients who develop an SSI (both superficial and deep) are 7.1 (95% CI 2.1–24.0) times more likely to develop a graft infection (p < 0.0001). The complete results from the logistic regression are shown in Table 6.

Table 6.

Determinants for Graft Infection with Corresponding Odds Ratios (OR)

Covariate	Adjusted OR	95% Confidence Interval
Surgical Site Infection	7.10	2.10 – 23.98
Endovascular Surgery	0.62	0.24 – 1.65
Mycotic Aneurysm	9.83	1.78 – 54.19
Diabetes Mellitus	1.18	0.38 – 3.70
Smoking	1.54	0.65 – 3.64
Ruptured Aneurysm	0.97	0.36 – 2.59
Pulmonary infection <30 days of surgery	2.91	1.01 – 8.39

A striking, although perhaps not surprising find was the high percentage of graft infections after surgery for suspected mycotic aneurysms (OR 7.71, p = 0.011). Two of 12 patients with mycotic aneurysm surgery developed a graft infection despite rigorous post-operative antibiotic treatment.

Secondary outcomes

Limiting the analysis to elective surgery, there were more overall re-interventions within 30 days of surgery after open surgery than after endovascular surgery (14.3% vs. 5.3%, p ≤ 0.0001). This difference in re-interventions was also present after one year (19.2% vs. 11.2%, p = 0.008). The 30-day mortality after elective surgery was 1.0% for endovascular surgery versus 4.9% for open surgery (p = 0.004). This difference dissipated over the months, with 4.4% versus 7.1% (p = 0.144) after six months and 5.8% versus 8.5% (p = 0.247) after one year.

After acute surgery the rate of re-intervention within 30 days was 20.0% after endovascular surgery and 27.2% after open surgery (p = 0.425). After one year, this difference disappeared with 31.1% after endovascular treatment versus 30.4% after open surgery (p = 1.000). The 30-day mortality after endovascular intervention for ruptured aneurysm was 22.2%, opposed to 26.4% after open aneurysm repair (p = 0.691). One-year mortality in this group was 31.1% for endovascular surgery versus 37.6% after open treatment (p = 0.474).

Discussion

To our knowledge, this is the first study giving an in-depth report of SSIs after abdominal aortic aneurysm surgery. Most studies report incidence numbers (between 0.5% and 14%) but do not describe the effects and outcome of SSIs [3 –5,11,12]. Our study shows that with an incidence of 3.4%, SSIs are a relatively frequent complication with serious consequences. In our population, contrary to findings in the literature [4,5], SSI after endovascular aortic aneurysm repair is as prevalent as SSI after open aortic aneurysm surgery. Deep SSIs warrant major surgery in more than 90% of cases and are associated with a mortality of 28%. This corresponds with findings in the literature [8]. As other studies have shown [1], the duration of hospital stay is longer for patients developing SSIs. Although we did not conduct a cost-benefit analysis, it is clear that the longer hospital stay (on average more than 18 days longer) leads to increased healthcare costs.

Potential bias of this study lies in the retrospective nature of the research. Some data accessed were more than 10 years old and had to be gathered from minimal or brief patient information. This might be cause for an underestimation of the found number of SSIs. Also, the comparison between open and endovascular procedure may be biased by the indication for one or the other procedure. For instance, while emergency endovascular aorta repair is increasing, in this cohort the majority of ruptured aneurysms were treated with open abdominal surgery. One-third of all open surgeries were because of ruptured aneurysm, compared with only 10% acute surgery in the endovascular group. Furthermore, there are certain patient characteristics that may sway the choice toward open or endovascular treatment.

Endovascular aorta repair is characterized by both frequent and long-term imaging follow-up for early endoleak diagnosis. Therefore, an otherwise healthy 50-year-old patient could be treated better with open surgery to prevent a high follow-up burden for the patient. Conversely, an older patient with considerable comorbidities might be better off undergoing a less invasive endovascular repair, which is known to have lower short-term mortality [13]. Last, regardless of patient age or comorbidity, some aneurysms have a morphology poorly suited for endoluminal stenting, ruling out these patients for endovascular surgery altogether. Because the indications for these two types of surgery may vary, some outcome measures are influenced by the choice of treatment. Ruptured aneurysms are associated with high 30-day (18%–35% for endovascular and 24%–37% for open treatment) and one-year mortality (37% for endovascular and 45% open treatment) [14]. This might explain the difference in mortality between the two groups in this study.

The incidence of graft infections corresponds with the range of 0.6%–3.0% reported in the literature [15 –17]. In accordance with Vogel et al. [9], we find that SSI is a risk factor for development of aortic graft infection. In addition to SSI, mycotic aneurysm was an independent risk factor for developing a graft infection, and a trend was seen for pulmonary infection within 30 days of surgery, although this was not significant. Although it seems intuitive that patients with peri-operative infectious disease have a higher risk for graft infections, these are also the patients who are rigorously treated with antibiotic agents. This would mean that the antibiotic treatment that suffices for treating SSI or pulmonary infection is insufficiently adequate in preventing bacteremia or the bacterial colonization of the graft. It might be worth investigating whether routine long-term antibiotic prophylaxis in patients with a mycotic aneurysm or an SSI after aortic graft surgery reduces the risk of aortic graft infection. Another, less invasive option would be to subject patients who develop an SSI after aortic graft implantation to intensified follow-up with more frequent radiographic imaging. We suggest that this follow-up period should constitute at least the first year after surgery, because 70% of graft infections occurred within one year.

As mentioned above, most SSIs were caused by Staphylococcus aureus. Previous studies show a reduction in one-year mortality after decreased amounts of Staphylococcus aureus SSI [18]. Our recent research has shown that SSIs with Staphylococcus aureus can be prevented in nasal carriers undergoing aortoiliac surgery by means of eradication therapy [2]. This could mean that by preventing Staphylococcus aureus SSIs, considerable amounts of subsequent graft infections could be avoided.

Conclusion

This study shows that SSIs after endovascular aortic surgery are as common as after open aortic surgery and are highly associated with graft infection. Also, patients with a primary mycotic aneurysm were at risk for developing graft infection. Studies evaluating the prevention or early diagnosis of graft infection after SSIs through intensified imaging follow-up or even long-term antibiotic treatment should be considered.

Footnotes

Acknowledgments

We would like to thank Jelle Raats, Kevin de Leur, and Ge Geenen for supplying part of the data used for analysis.

Author Disclosure Statement

No competing financial interests exist.

References

Faizer

, Dombrovskiy

, Vogel

. Impact of hospital-acquired infection on long-term outcomes after endovascular and open abdominal aortic aneurysm repair. Ann Vasc Surg, 2014; 28:823–830.

Langenberg

, Thomas

, Donker

, et al. Evaluation of Staphylococcus aureus eradication therapy in vascular surgery. PLoS One, 2016; 11:e0161058.

Giles

, Wyers

, Pomposelli

, et al. The impact of body mass index on perioperative outcomes of open and endovascular abdominal aortic aneurysm repair from the National Surgical Quality Improvement Program, 2005–2007. J Vasc Surg, 2010; 52:1471–1477.

De Rango

, Cao

. Long-term results of OVER: The dream of EVAR is not over. Eur J Vasc Endovasc Surg, 2013; 45:313–314.

Prinssen

, Buskens

, Blankensteijn

; The Dutch Randomised Endovascular Aneurysm Management (DREAM) trial. Background, design and methods. J Cardiovasc Surg (Torino), 2002; 43:379–384.

Mussa

, Hedayati

, Zhou

, et al. Prevention and treatment of aortic graft infection. Expert Rev Anti Infect Ther, 2007; 5:305–315.

Young

, Cherry

Jr , Davis

, et al. The results of in situ prosthetic replacement for infected aortic grafts. Am J Surg, 1999; 178:136–140.

Dirven

, van der Jagt

, Barendregt

, et al. The efficacy of autologous femoropopliteal vein reconstruction for primary aortic and aortic graft infection. Ann Vasc Surg, 2015; 29:1188–1195.

Vogel

, Symons

, Flum

. The incidence and factors associated with graft infection after aortic aneurysm repair. J Vasc Surg, 2008; 47:264–269.

10.

Horan

, Andrus

, Dudeck

. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control, 2008; 36:309–332.

11.

Peng

, Yuan

, Cui

, Fang

. Clinical outcomes of endovascular aneurysm repair of abdominal aortic aneurysm complicated with hypertension: A 5-year experience. Pak J Med Sci, 2016; 32:13–17.

12.

Inui

, Bandyk

. Vascular surgical site infection: Risk factors and preventive measures. Semin Vasc Surg, 2015; 28:201–207.

13.

United Kingdom EVAR Trial

Investigators

, Greenhalgh

, Brown

, et al. Endovascular versus open repair of abdominal aortic aneurysm. N Engl J Med, 2010; 362:1863–1871.

14.

Sweeting

, Balm

, Desgranges

, et al. Ruptured aneurysm, individual-patient meta-analysis of three randomized trials comparing endovascular versus open repair for ruptured abdominal aortic aneurysm. Br J Surg, 2015; 102:1229–1239.

15.

Kieffer

, Sabatier

, Plissonnier

, Knosalla

, Prosthetic graft infection after descending thoracic/thoracoabdominal aortic aneurysmectomy: Management with in situ arterial allografts. J Vasc Surg, 2001; 33:671–678.

16.

Swain

3rd, Calligaro

, Dougherty

. Management of infected aortic prosthetic grafts. Vasc Endovascular Surg, 2004; 38:75–82.

17.

O'Connor

, Andrew

, Batt

, Becquemin

. A systematic review and meta-analysis of treatments for aortic graft infection. J Vasc Surg, 2006; 44:38–45.

18.

Bode

, Rijen

, Wertheim

, et al. Long-term mortality after rapid screening and decolonization of Staphylococcus aureus carriers: Observational follow-up study of a randomized, placebo-controlled trial. Ann Surg, 2016; 263:511–515.