Surgical Site Complications After Complex Iliofemoral Reconstruction and the Role of Negative Pressure Wound Therapy: A Retrospective, Single-center Study

Surgical site infection (SSI) of femoral incisions after vascular surgery occurs in up to 44% of patients.¹ These infections are a significant source of morbidity and are associated with increased rates of readmission and reoperation, increased overall cost of care, and longer hospital length of stay.² In addition to clinical morbidity, the patient-reported care experience is negatively affected by this seemingly avoidable complication. Currently, limited guidance exists regarding best management practices to consistently mitigate wound complications associated with this common incision site for vascular and cardiovascular procedures.

The femoral incision is used in many different types of vascular intervention, including for thoracic and abdominal aortic aneurysm, aortoiliac occlusive disease, and infrainguinal revascularization. Increasingly, hybrid reconstructions that include common and deep femoral endarterectomy combined with aortoiliac endovascular intervention are supplanting historical approaches to Trans-Atlantic Inter-Society Consensus Document on Management of Peripheral Arterial Disease (TASC) C and D aortoiliac artery occlusive lesions. Durable outcomes with the hybrid approach are predicated on the complete removal of plaque burden, which frequently necessitates extensive vessel exposure and lymphatic disruption over a longer operation. Both are clinical predictors of increased risk of surgical site occurrence and SSI.

The role of closed incision negative pressure therapy (CiNPT) as an adjunct to the primarily closed femoral incision after vascular surgery is unclear, and the data are seemingly inconclusive. Positive results with use of CiNPT have been reported in several randomized controlled trials (RCTs).^3,4 The Inguinal Vascular Surgical Wound Protection by Incisional Negative Pressure Wound Therapy (INVIPS) trial demonstrated a significantly decreased SSI incidence with use of CiNPT.⁵ However, a Cochrane review published in 2019 concluded that use of CiNPT may reduce the rate of SSI, noting a “low certainty of evidence” and “high risk of bias.”⁶ These mixed results achieved with CiNPT, coupled with the additional procedural expense associated with the technology, may affect its adoption for widespread use in the clinical setting, because there are less expensive standard wound dressings available. Currently, no standardized protocol is available to prospectively enhance the identification of high-risk patients who would derive the most benefit from such therapy.

Study population

A retrospective chart review of all adult patients (age, ≥18 years) who underwent femoral endarterectomy at the Cleveland Clinic Foundation from 2009 through 2017 was performed after receiving approval from the institutional review board. Inclusion criteria were femoral incisions for iliofemoral endarterectomy with or without profundoplasty, as well as hybrid procedures, which were defined as open procedures that involved endovascular iliac stenting. Exclusion criteria consisted of femoral incision for bypass procedure, or repair of endovascular aneurysm.

Primary closure was performed on all patients followed by CiNPT or application of the standard dressing (Primapore; Smith+Nephew). The negative pressure wound therapy (NPWT) system used (Prevena; 3M; henceforth known as product A) was applied per the manufacturer’s instructions for use and left in place for 5 to 7 days during the hospital stay. The incisional closure technique and standard dressing choice were at the surgeon’s discretion. The hybrid procedure included performing the common femoral endarterectomy followed by direct puncture through the femoral patch or femoral artery to facilitate the iliac intervention. Bovine pericardial patches were used in this study.

End points

The primary end point was SSI within 30 days postoperatively or at the first postoperative visit; SSI was graded using the Szilagyi classification.⁷ Secondary end points included readmission, reoperation, and length of stay.

Statistical analysis

Continuous and categorical variables were compared with analysis of variance and chi-square analysis. Continuous variables are presented as mean ± standard deviation. Univariate and multivariable analysis using logistic regression were used to identify high-risk factors for SSI. Product A and standard dressing patients were matched based on body mass index (BMI), surgical duration, insulin use, chronic obstructive pulmonary disease (COPD), previous femoral cutdown, immunosuppression use, and prosthetic use to determine whether CiNPT conferred any benefit compared with standard dressings.

Prediction model

A scoring system was created using a logistic regression model to predict SSI based on high-risk factors identified in this cohort: BMI, insulin-dependent diabetes mellitus (DM), COPD, immunosuppression, hybrid surgery, presence of prosthetic material, previous femoral cutdown, and surgical duration. Factors were removed based on their ability to predict the full model predictions using a step-down approach. The parameter estimates were reviewed, and, given their similarity, a sum of risk factors was used (weighted operation duration as 1 point per 5 hours). The scoring system was validated using bias-adjusted discrimination and calibration measures calculated using bootstrap resampling.

Baseline demographics

A total of 422 femoral incisions for femoral endarterectomy were performed on 337 patients (85 procedures were bilateral). Most of these surgical procedures were hybrid (77%), 89% of all surgeries used longitudinal incisions, and 85.5% involved prosthetic with bovine pericardial product or polyester graft for the patch angioplasty. A comparison of patient demographics and procedural characteristics is shown in Table 1. The SSI group was significantly more likely than the group without SSI to have a BMI greater than 30 (P <.001) and to have insulin-dependent DM (P =.008), COPD (P =.031), and a higher hemoglobin A_1c level (P =.045).

Postoperative outcomes

The incidence of SSI during the follow-up period was 16.1% (68/422). Patients with SSI had a higher rate of readmission (44.1% vs 2.3%; P <.001), reoperation (38.2% vs 0.28%; P <.001), and longer hospital length of stay (8.5 vs 5.1 days; P =.016) than those without SSI. Most of these infections (58/68; 85.3%) were Szilagyi grade I or II (Table 2). The multivariable analysis demonstrated that obesity, insulin-dependent DM, and COPD were associated with SSI (Table 3). Unplanned return to the operating room was required to manage groin wound complications associated with 27 incisions (6.4%). Most of these surgical procedures (15/ 27 [55.5%]) required debridement with or without NPWT. Vessel reconstruction with greater saphenous or femoral vein was required in only 5 of 27 patients (18.5%), and 11 of 27 patients (40.7%) required muscle coverage with a sartorius or rectus femoris muscle flap.

The product A dressing was used in 47 of 337 patients (13.9%) at the surgeon’s discretion, with a standard dressing used for the remainder. There was no statistical difference in wound complications in either the unmatched (Table 4) or the matched analysis (Table 5). However, there was a trend towards decreased SSI severity with use of the product A dressing. All 10 of the Szilagyi grade III complications occurred in patients treated with the standard dressing. Although patients treated with product A required reoperation, they did not require vessel reconstruction or muscle flap coverage.

Prediction model

The final prediction model used 5 variables, with 1 point assigned to each: obesity (BMI > 30), insulin use, COPD, immunosuppression, and surgical duration. The median score was 2; 44% of patients had a score of 1, and 37% had a score of 2. Figure 1 shows the discrimination and calibration with a concordance index of 0.67. Figure 2 demonstrates the SSI rate increases as the sum score increases greater than or equal to 3.

This study was undertaken to describe SSI after complex iliofemoral reconstruction, describe the associated risk factors, and better define the effect of CiNPT in this clinical setting. The femoral incision is the foundation of many types of vascular reconstruction. Although this exposure affords the surgeon the desired access, the devastating consequences of femoral artery infection, vessel blowout, and consequent hemorrhage may occur. In addition to patient morbidity, wound complications dramatically alter the patient’s overall care experience. Finally, public reporting of process, outcomes, and surgical quality measures demands that health care providers identify and implement measures to reduce the incidence of avoidable wound complications.

The observed overall SSI rate of 16.1% in complex femoral reconstruction in the current study is lower than the previously reported rate (44%) of SSI after groin incisions in vascular surgery.¹ Multivariable analysis in the current study showed obesity, insulin-dependent DM, and COPD are associated risk factors for SSI. Closed incision negative pressure therapy was not associated with a statistically significant decrease in SSI, though its use appeared to decrease the severity of infections in all 10 Szilagyi grade III infections occurring in patients treated with the standard dressing. Moreover, no infection in the CiNPT group required arterial reconstruction with autologous vein or myofascial flap coverage, both of which are indications of advanced, deep space infection. A predictor model based on the 5 variables of obesity, insulin use, COPD, immunosuppression, and surgical duration was presented to identify patients at high risk for developing SSI after femoral incisions.

Closed incision negative pressure therapy was initially presented as a novel approach to prevent SSI and its complications, and multiple trials have been conducted to prove its efficacy. The CiNPT device used in the current study was the first product approved by the US Food and Drug Administration for such indications. This battery-driven unit delivers continuous negative pressure through a 2-layer dressing. The mechanism is believed to decrease edema, reduce distracting skin forces, reduce external wound contamination, and decrease bacterial colony counts.^8,9 The early results of this product have been mixed. To date, 7 RCTs have reported encouraging but inconclusive data.^3,4,10-15 The cost of this product should factor into the decision to use it rather than the standard dressing ($495 vs $18, respectively).¹⁶ However, SSI is among the most common causes of readmission, and weighted analysis of previous studies demonstrated that the cost of each additional hospital day was $4140 and that the readmission cost was approximately $28 178.^2,17–19 Furthermore, Kwon et al³ demonstrated in an RCT that the use of product A in high-risk patients was associated with an average cost savings of $6045 per patient.

Groin incisions for vascular reconstructions in vascular surgery are a known risk factor for SSI.²⁰ The authors of the current study believe there is a role for CiNPT to mitigate groin incision SSI, but CiNPT should be used selectively. This study used data from a single institution to successfully develop a prediction model to determine which patients were at highest risk for SSI and thereby identify those who might benefit from the application of CiNPT. Wiseman et al²¹ used the National Surgical Quality Improvement Program database to develop prediction models for SSI after major vascular surgery (eg, open abdominal aortic aneurysm, open aortoiliac, and lower extremity revascularization). They identified 24 variables that influenced SSI, including 4 of the 5 variables (excluding immunosuppression) that were included in the current study. The Wiseman et al²¹ study was limited by the use of a national database, with its inherent limitations, and by the inclusion of multiple procedures. Furthermore, based on a point scale in the prediction model used in that study, lower limb revascularization had the highest association with SSI; this finding further indicates the need for a prediction model solely for femoral incisions.

It is important to emphasize that myriad additional factors, including perioperative glucose optimization, femoral skin colonization, intraoperative handling of skin and soft tissue, incision orientation, wound closure technique, and postoperative wound surveillance, may affect the development of wound complications, including surgical site occurrence and SSI. Accordingly, a program to reduce wound occurrences should optimize care at each of these clinical junctures. Efforts to completely adjust for these confounding variables are fraught with statistical shortcomings.

The limitations of the current study are inherent in its retrospective nature. Surgeon selection bias may have influenced the use of the product A device in certain incisional environments judged to be at greater risk for infection complications. If true, this may suggest that product A enhances healing in a way that protects against deep space infection. However, the cohorts in this study seemed to be matched evenly across a comorbidity profile. It is possible that Szilagyi grade I and II wound infections may have been underreported. Additionally, this study, like other RCTs, was not adequately powered to detect a statistically significant benefit with product A, and the possibility of type II error was increased by the relatively lower rate of SSI relative to historically published rates.

Patients with obesity, COPD, and insulin-dependent DM are at increased risk for SSI after femoral incisions for peripheral revascularization. A prediction model may assist in identifying patients at high risk for SSI so that targeted risk reduction strategies can be implemented to decrease both patient morbidity and economic burden to the health care system. Targeted use of CiNPT may help reduce the severity of SSI in these at-risk patients.