Surgical Site Infection after Multiple Groin Incisions in Peripheral Vascular Surgery

Abstract

Background:

Patients with peripheral arterial disease (PAD) are at risk for revision surgery in the groin and therefore at potential risk for surgical site infections (SSIs). In an observational study, a cohort of patients with peripheral arterial disease was followed to examine the effect of different incision intervals on SSI-free survival.

Methods:

Patients, needing peripheral vascular surgery because of PAD, were retrieved from a prospectively collected database on SSIs after vascular surgery between March 2009 and January 2012, the group consisting of 720 patients. Of these, 255 patients were selected (age 71.9±10.4 y). Cox proportional hazards models were used for event-history analyses. The effect of incision interval was estimated with adjustment for a number of potential confounders. Effects were quantified by means of hazard ratios (HRs) with 95% confidence intervals (CIs).

Results:

No significant effect on the incision interval on SSI-free survival was observed. After separating incisional SSIs into superficial- and deep-seated, a significant linear trend effect of the groin incision interval on deep-incisional SSI development was observed: the shorter the interval, the higher the event rate (HR 1.5 per category, 95% CI 1.1–2.1, p=0.22). Besides the incision interval, the Rutherford classification was a significant risk factor for SSI development (HR 3.0; 95% CI 2.1–4.2; p<0.0005).

Conclusion:

Revision surgery in the groin puts patients at risk for deep-incisional SSI. No effect on superficial incisional SSI development was observed. Besides the incision interval, the Rutherford classification was a significant risk factor for both superficial- and deep-incisional SSI. Quality improvement and better risk stratification schemes are suggested.

Severe peripheral arterial disease (PAD) is associated with major complications after vascular surgery, as well as a higher cardiovascular event rate, comorbidities, and death [1,2]. Given this information, better outcome predictions for this particular group are needed to improve treatment strategies.

A common complication after vascular surgery is a surgical site infection (SSI) [3], which is associated with increased hospital length of stay (LOS), morbidity, readmissions, re-interventions, and death [4]. Deep-incisional SSI could, with the use of prosthetic grafts, lead to high morbidity and mortality rates. Reduction of the incidence of these SSIs therefore is necessary.

A groin incision for femoral artery access is a risk factor for these SSIs [5,6]. Additionally, in an aging population wherein risk factors for the development of severe PAD increase, a higher demand for peripheral surgical revascularization and re-intervention is expected [4,7,8]. Multiple groin incisions could lead to a high incidence of complications; e.g., SSI.

To our knowledge, there is little information on the effect of different groin incision intervals on patient outcomes such as SSI. This lack of knowledge could have serious clinical consequences.

The present study was performed to evaluate the effect of different groin incision intervals after elective vascular procedures on patients with moderate to severe PAD with regard to SSI development. Previously, groin incisions, performed at different times, were analyzed for their influence on SSI-free survival [6].

Patients and Methods

Data collection and study population

A prospectively collected database on SSIs after 720 central and peripheral vascular procedures performed between March 2009 and January 2012 [3] was studied to retrieve all patients with PAD seen in the Amphia Hospital. For this study, the hospital's Infection Control Committee and the Board of Directors, as part of the patient safety program and the Medical Ethical Committee, approved this study and waived informed consent.

All patients were selected on the basis of operation codes corresponding to endarterectomy of the femoral artery and peripheral bypass surgery. Patients undergoing popliteal aneurysm repair, endovascular procedures of the lower and upper extremities, central vascular procedures, or endarterectomy of the carotid artery were excluded from this study.

Procedures were performed at an institution with five vascular surgeons. All patients were worked up via duplex ultrasound examination and magnetic resonance angiography. Endartectomies were performed according to standard vascular procedures. Whenever possible, autologous grafting was performed in peripheral bypass surgery. Alternatively, an expanded polytetrafluoroethylene (ePTFE) pre-cuffed Dynaflo^® graft (Bard Peripheral Vascular Inc., Tempe, AZ USA) was used for supragenicular femoropopliteal bypass reconstructions and the Distaflo^® (Bard Peripheral Vascular) for infragenicular femoropopliteal and femoralcrural bypass reconstructions [9].

Surgical site infections, defined according to the criteria of the U.S. Centers for Disease Control and Prevention [10], were monitored prospectively. A distinction was made between superficial and deep-incisional SSIs. Registration of SSIs was performed by dedicated and trained infection control personnel. The presence of an SSI was ascertained by a microbiologist and a vascular surgeon [3].

Previous procedures were recorded (before and at inclusion) and procedures during the follow-up period in which the groin was involved. A previous groin procedure was defined as a procedure at the groin to improve vascular circulation (patch, bypass, or thrombectomy) or to stop major bleeding, there being no indication of an infection. Drainage from abscesses, seromas, hematomas, or wound debridement was not included in the analyses as a previous groin incision. Also, percutaneous transluminal angioplasty (PTA) and thrombolysis were not included in the groin procedures.

Other data, aside from SSIs and procedures, collected from patient records were age, gender, the American Society of Anesthesiologists (ASA) score; height, weight, and body mass index (BMI); presence of diabetes mellitus, chronic obstructive pulmonary disease (COPD), or chronic kidney disease; Rutherford classification; the use of synthetic grafts (bovine patches or ePTFE bypasses); amputation; and death.

Outcome measures

The outcome variable to be analyzed was SSI-free survival time. The endpoint was either an SSI or death while SSI-free. Also, a distinction was made between superficial and deep SSI. Censoring of the follow-up was at day 90 after the inclusion operation and could therefore properly be assumed to be non-informative for the incidence time of the endpoint as defined.

Follow-up

Surveillance was performed during a time period in which the patient was considered most prone to infections. All patients were followed until 90 d after the date of the initial procedure (during admission and in the outpatient clinic) or until the day of death when earlier than 90 d.

Statistical analysis

The purpose of the statistical analyses was to estimate the effects of a number of independent variables on SSI-free survival time using event-history analysis. Our primary interest was in the effect of previous groin procedures on SSI-free survival time. Also of interest was the interval between two consecutive incisions. This interval was defined as the time between the two last incisions prior to the occurrence of an event and so was accounted for in the analysis as a time-dependent covariable. The following categorization was used for the groin incision interval between the last-but-one incision and the last incision: less than two weeks, two weeks to six months, and more than six months vs. no previous groin operation.

The Cox proportional hazards model was used for the event-history analyses. First, an unadjusted effect was estimated of the incision interval, as defined above. In more elaborate analyses, the effect of the incision interval was estimated with adjustment for a number of potentially confounding variables. Along with incision interval, age (in years) was entered in the model, being an important risk factor in elderly patients. Occurrence of an amputation during follow-up was considered a confounding covariable to be adjusted for when considering the effect of incision interval on SSI-free survival time. Therefore, amputation (Y/N) was entered in the model as a time-dependent covariable. Model entry of other explanatory variables was based on a backward elimination procedure so as to delete variables from the model having a p value >0.20 in a stepwise manner. Those other candidate variables were mentioned above. The synthetic graft variable was based on the last operation before the occurrence of an event and thus was entered in the model as a time-dependent covariable. The effects of the independent variables on the various survival times were quantified by means of hazard ratios (HRs) with 95% confidence intervals (CIs) and p values. Logistic regression analysis was performed among the patients with an SSI in order to investigate which variables were able to discriminate between deep-seated and superficial SSIs. Those discriminating variables should have an effect on the occurrence rate of deep SSIs that differs from that of the occurrence rate of superficial SSIs.

Statistical analyses were done using the SPSS^® software program version 20 (IBM, Armonk, New York, USA). P values<0.05 were considered to denote statistical significance.

Results

This study included 255 consecutive patients who had a surgical revascularization because of PAD (Rutherford class 3–6) in our hospital. Their mean age was 71.9±10.4 y, and 173 (68%) were male. At the time of inclusion, 184 (72.2%) underwent a primary procedure and did not have a previous groin incision. None had a groin incision less than two weeks ago, 8 (3.1%) had an incision between two weeks and six months earlier, and 63 (24.7%) had an incision more than six months earlier. These numbers were updated during follow-up after each operation. After 90 d of follow-up, 236 patients were still alive, 160 (67.8%) did not undergo another operation, 15 (6.4%) had a groin incision less than two weeks previously, 12 (5.1%) had an incision between two weeks and six months earlier, and 49 (20.8%) had a groin incision more than six months earlier. The 255 patients contributed 292 operations, in which in 173 (59.2%) synthetic grafts were used. Other basic patient characteristics are listed in Table 1.

Table 1.

Clinical Characteristics of Study Population (n=255 [%])

Age (y), mean±SD	71.9±10.4
Male/female (%)	173 (68)/82 (32)
American Society of Anesthesiologists class
I or II	54 (21.2)
III, IV, or V	200 (78.4)
Body mass index (mean±SD)	25.3±3.6
Co-morbidities
Chronic obstructive pulmonary disease	50 (19.6)
Chronic kidney disease	30 (11.8)
Diabetes	86 (33.7)
Rutherford class
3	97 (38)
4	70 (27.5)
5/6	88 (34.5)
Total procedures performed	292
Primary procedure at inclusion (%)	184 (72.2)
Synthetic grafts	143 (55.2)
Procedures at inclusion
Endarterectomy, common femoral artery	97 (38)
Supragenicular bypass	61 (23.9)
Infragenicular bypass	48 (18.8)
Femorocrural bypass	49 (19.2)

The 255 patients having an operation contributed a total of 21,863 person-days alive after operation. During follow-up, 19 patients (7.5%) died, and 38 patients (14.9%) developed a SSI (Staphylococcus aureus was cultured from 45%). Of these 38 patients, 15 (5.9%) turned out to have a deep-incisional SSI.

SSI-free survival

During follow-up, the following events were registered in 51 patients: 38 developed an SSI and 13 died while SSI-free. The unadjusted overall effect of the incision interval on event-free survival was not significant (p=0.38). In the subgroup with an incision interval of less than two weeks, a non-significant higher event rate was observed than in the subgroup with no former incision (HR=2.09; 95% CI 0.87–5.02; p=0.10).

After adjusting for potential confounders, no significant effect of the groin incision interval on SSI-free survival was observed (p=0.97) (Table 2). An incision interval of less than two weeks resulted in a 1.2-fold times higher event rate than in the case of no previous groin incision, although this difference was not statistically significant. The Rutherford class (HR=3.0/point) had a significant effect on SSI-free survival. During follow-up, 14 patients underwent a major amputation while SSI free. After an amputation, the event rate was not significantly higher (HR=1.4; p=0.56) than in patients who had not had an amputation. Also, the effect of age (a 2% event increase per year) was not significant (p=0.36). Each of the variables that were backwards-eliminated from the model (gender, diabetes mellitus, synthetic graft, COPD, renal impairment, ASA score, or BMI) did not have a significant effect on SSI-free survival when added to the final model presented in Table 2 (p values ranging from 0.5 to 0.9; data not shown).

Table 2.

Results of Cox Proportional Hazards Regression Analysis on Surgical Site Infection-Free Survival Based on 51 Events in 255 Patients

Explanatory variable	Coefficient	Hazard ratio (95 % CI)	p
Interval between two last groin incisions			0.97
None	0	1
<2 wks	0.192	1.21 (0.42–3.48)
2 wks–6 mo	−0.255	0.78 (0.18–3.29)
>6 mos	−0.015	0.99 (0.51–1.92)
Rutherford class (+1 point)	1.101	3.01 (2.15–4.20)	<0.0005
Age (+1 y)	0.015	1.02 (0.98–1.05)	0.36
Amputation (time dependent)	0.307	1.36 (0.48–3.84)	0.56
Model test: χ²=68.860 (df=6); p<0.0005

Deep-incisional SSI

In 15 of the 38 patients with an SSI, the SSI turned out to be deep seated. The variables that differed significantly between patients with a deep-incisional SSI and those with a superficial-incisional SSI were last incision interval before the SSI (the shorter the interval, the higher the probability that the SSI was deep), Rutherford class (effect in favor of a deep SSI), and BMI (effect in favor of a superficial-incisional SSI). In these analyses, incision interval as a linear trend variable (with no previous incision defined as the highest category) gave a better fit than incision interval as a categorical nominal variable. After having entered these three variables simultaneously in a logistic regression model, the Rutherford class lost its significance (p=0.41). The variables incision interval and BMI kept their effects in the directions just mentioned (both p values 0.003).

To investigate further the effects of these variables on SSI-free survival while distinguishing between deep and superficial-incisional SSIs, we entered these variables simultaneously in two Cox proportional hazards regression analyses, one for deep-incisional SSI, with 29 events (15 SSIs and 14 deaths), and the other for superficial-incisional SSI, with 41 events (23 SSIs and 18 deaths). Incision interval appeared to have a significant linear trend effect on deep-SSI-free survival: One category, shorter incision interval, was associated with an HR of 1.5 (95% CI 1.1–2.1; p=0.022). A non-significant reverse linear trend effect of incision interval was seen on superficial SSI-free survival: HR=0.75 per category of shorter incision interval (p=0.19). The Rutherford class had a large and significant positive significant effect on the event rates concerning deep as well as superficial SSI-free survival, with respective HRs of 6.7 and 3.0 per point higher on the scale (p<0.0005). The BMI appeared to have a significant reverse effect on the event rate concerning deep SSI-free survival: HR=0.89/unit kg/m² (95% CI 0.81–0.99; p=0.025). A non-significant positive effect of BMI was found on the event rate concerning superficial SSI-free survival: HR=1.06/unit kg/m² (95% CI 0.98–1.14; p=0.15). The extent to which these results differed between deep- and superficial-incisional SSI-free survival was in line with the above-mentioned results of the logistic regression analysis on the 36 SSIs, as it should be.

Discussion

The results from the present study demonstrate that the frequency of repeated vascular revision surgery at the groin is not a risk factor for SSI development. Only after separating SSIs into superficial and deep-seated, a significant linear trend effect of the groin incision interval on deep-SSI development was observed: the shorter the interval, the higher the event rate. This suggests early revision surgery in the groin to be a risk factor for deep-SSI development. The incision interval had no significant effect on superficial-incisional SSI development. Besides the incision interval, the Rutherford classification is a significant risk factor for superficial as well as deep-seated incisional SSI development. This effect did not differ significantly between superficial and deep-seated-incisional SSIs.

Each shorter incision interval was associated with an increased risk of deep-incisional SSI development (1.5 per category). Kolakowski et al. reported early revision of lower extremity bypass grafts (within one month), compared with late revision (after one month) to be a risk factor for graft infection [11] and stated reintervention should be delayed until one month after initial surgery. Unfortunately, in most cases, revision surgery cannot be delayed because of severe, even critical, limb ischemia [12]. Alternative procedures for vascular revision could therefore be considered. For example, extra-anatomic surgical approaches and percutaneous transluminal interventions might be performed to resolve stenosis or occlusions [13], but current literature on these interventions performed in this situation is lacking.

The Rutherford classification turned out to be a strong risk factor for SSI development. We could not demonstrate that the effect of the Rutherford classification on deep-incisional SSI development differed from that on superficial-incisional SSI development (p=0.41 adjusted for incision interval). The poor peripheral and subcutaneous vascularization status in which wound healing might be impaired and micro-organisms have a better chance of growth could be an explanation.

The BMI was a non-significant risk factor for superficial-incisional SSI development in this study and had a significant protective effect on deep-incisional SSI development. This effect of BMI differed significantly between superficial and deep-seated incisional SSIs. In the literature, BMI is associated with SSI development [14,15], but no distinction was made between superficial and deep-incisional SSIs in these studies. A potential explanation is that in the case of a high BMI, prosthetic grafts are better embedded in subcutaneous tissue.

A synthetic graft infection is a well-known risk factor for amputation [15]. A logical explanation for this could be the fact that, at the time of presentation, no reconstructive options are available if a limb is not salvageable because of critical limb ischemia. Besides this, SSIs are also is associated with higher morbidity and mortality rates [4]. The reduction of SSI is therefore highly desirable. In our hospital, several interventions have been undertaken to reduce SSIs, namely the introduction of the Surgical Patient Safety System (SURPASS) checklist [16] and of a bundle of care consisting of ensuring compliance with four evidence-based measures: perioperative normothermia, hair removal, perioperative antibiotic prophylaxis, and discipline in the operating room. [3]. In this study, we found a significant association between early revision surgery (within two weeks) and deep-seated SSI development. Future trials should be performed on alternative interventions or strategies to reduce SSI after peripheral vascular surgery.

There were some limitations to this retrospective study that should be considered when interpreting the results. First, data were based on patients with peripheral arterial obstructive disease having peripheral reconstructive surgery. Patients with central reconstructive surgery were excluded from the analyses. Results therefore could not be applied to these patients. Second, the data were collected from one single center with five vascular surgeons; the SSI prevalence could differ from that at other centers. Finally, it should be appreciated that this single-center study nonetheless is based on a relatively large sample. There was a good collaboration between the medical microbiology department and the surgery department, so the presence of a surgical site infection was checked from both sides, limiting the possibility of detection bias.

The presence of an SSI is a risk factor for morbidity and death after peripheral vascular surgery. Reduction of SSI therefore is an important area for improvement. Early revision surgery in the groin, especially that performed within two weeks, is a risk factor for deep-incisional SSI. Besides this, the Rutherford class is a risk factor for both superficial and incisional-seated SSI. Further research has to be performed on risk stratification models to improve the outcomes of patients with peripheral arterial disease.

Footnotes

Acknowledgments and Author Disclosure Statement

The authors are grateful to Y. Hendriks and M.M.A. Romme-Verschuren for their help in obtaining the data.

The authors declare no conflicts of interest.

References

Criqui

, Langer

, Fronek

, et al. Mortality over a period of 10 years in patients with peripheral arterial disease. N Engl J Med, 1992; 326:381–386.

Murabito

, Evans

, Nieto

, et al. Prevalence and clinical correlates of peripheral arterial disease in the Framingham Offspring Study. Am Heart J, 2002; 143:961–965.

van der Slegt

, van der Laan

, Veen

, et al. Implementation of a bundle of care to reduce surgical site infections in patients undergoing vascular surgery. PloS One, 2013; 8:e71566.

Perencevich

, Sands

, Cosgrove

, et al. Health and economic impact of surgical site infections diagnosed after hospital discharge. Emerging Infect Dis, 2003; 9:196–203.

Siracuse

, Nandivada

, Giles

, et al. Prosthetic graft infections involving the femoral artery. J Vasc Surg, 2013; 57:700–705.

Derksen

, Verhoeven

, van de Mortel

, et al. Risk factors for surgical-site infection following common femoral artery endarterectomy. Vasc Endovasc Surg, 2009; 43:69–75.

Hirsch

, Hartman

, Town

, Virnig

. National health care costs of peripheral arterial disease in the Medicare population. Vasc Med, 2008; 13:209–215.

Burns

, Gough

, Bradbury

. Management of peripheral arterial disease in primary care. BMJ, 2003; 326:584–588.

Donker

, Ho

, Te Slaa

, et al. Midterm results of autologous saphenous vein and ePTFE pre-cuffed bypass surgery in peripheral arterial occlusive disease. Vasc Endovasc Surg, 2011; 45:598–603.

10.

Horan

, Gaynes

, Martone

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

11.

Kolakowski

Jr , Dougherty

, Calligaro

. Does the timing of reoperation influence the risk of graft infection?. J Vasc Surg, 2007; 45:60–64.

12.

Lepantalo

, Matzke

. Outcome of unreconstructed chronic critical leg ischaemia. Eur J Vasc Endovasc Surg, 1996; 11:153–157.

13.

Adam

, Beard

, Cleveland

, et al. Bypass versus angioplasty in severe ischaemia of the leg (BASIL): Multicentre, randomised controlled trial. Lancet, 2005; 366:1925–1934.

14.

Giles

, Hamdan

, Pomposelli

, et al. Body mass index: surgical site infections and mortality after lower extremity bypass from the National Surgical Quality Improvement Program 2005–2007. Ann Vasc Surg, 2010; 24:48–56.

15.

Siracuse

, Nandivada

, Giles

, et al. Prosthetic graft infections involving the femoral artery. J Vasc Surg, 2013; 57:700–705.

16.

de Vries

, Prins

, Crolla

, et al. Effect of a comprehensive surgical safety system on patient outcomes. N Engl J Med, 2010; 363:1928–1937.