Outcomes Following Peripheral Vascular Interventions in Patients with End-Stage Renal Disease

Abstract

Patients with End-Stage Renal Disease (ESRD) represent a very fragile population, presenting with higher rates of complications and morbidity following vascular interventions. This study aims to analyze post-operative outcomes, such as mortality, readmissions, and amputations, in patients with Peripheral Arterial Disease (PAD) and ESRD on dialysis. Methods: A retrospective cohort study was conducted of patients with PAD and ESRD on dialysis who underwent vascular interventions between 2015 and 2017. This study focused on post-operative outcomes, including mortality, readmissions, and amputations. The data were analyzed to identify patterns and correlations. Results: This study found that patients with PAD and ESRD have long hospital stays, high amputation rates, high readmission rates, and treatment failure. Above-knee amputation (AKA) and female gender were associated with higher mortality rates, while prior stroke was associated with higher odds of readmissions. Conclusions: This study highlights the need for further studies with larger patient populations to identify independent predictors of negative outcomes. The findings suggest that specific factors, such as AKA, female gender, and prior stroke, significantly impact post-operative outcomes in this patient population.

1. Introduction

End-Stage Renal Disease (ESRD) patients on hemodialysis are increasing worldwide and have exorbitant healthcare costs compared to the general population, roughly USD 40 billion in the United States alone [1,2,3]. The United States Renal Data System (USRDS) has reported a steady rise in dialysis dependence, fueled in part by growing rates of diabetes, hypertension, and an aging population. Notably, between 2000 and 2020, the global dialysis population more than doubled, and projections suggest that this number will continue to rise, especially in low- and middle-income countries [3].

In the last decade, improved dialysis techniques and infection control strategies have extended survival by more than 25%, creating a new cohort of older and more medically complex surgical candidates [3]. However, longer survival comes with trade-offs. Patients are now living long enough to face the downstream vascular complications of uremia and atherosclerosis. Surgical teams are increasingly confronted with decisions about complex revascularization or limb salvage in patients whose baseline physiology often confers poor healing potential and high procedural risk.

Perioperative complications remain disproportionately common among dialysis patients. These complications contribute to longer hospitalizations, increased readmission rates, and cumulative healthcare expenditures [4]. Although the reasons for poor surgical outcomes in ESRD remain only partially understood, studies suggest that age over 80, low serum albumin, and comorbid conditions like diabetes and CHF are among the strongest predictors of post-operative morbidity and mortality [5,6].

Peripheral Arterial Disease (PAD) is widely prevalent, affecting an estimated 5 to 12 million Americans, and is strongly associated with both age and diabetes mellitus, which alone confers over a three-fold increase in risk [7,8]. Among patients with ESRD, the burden of PAD is substantially higher, with reported prevalence approaching 40% in some cohorts [9]. More concerning is the rate at which PAD progresses to chronic limb-threatening ischemia (CLTI) in this population, a condition marked by rest pain, non-healing ulcers, or gangrene. CLTI in ESRD patients represents not just a localized disease process, but a systemic failure of vascular health and wound healing capacity.

ESRD patients with CLTI often face poor limb salvage rates and dismal survival. Even with attempted revascularization, three-year survival rarely exceeds 35%, and in some studies is as low as 25% [10,11,12]. These statistics reflect the cumulative effect of systemic inflammation, impaired immunity, vascular calcification, and microvascular dysfunction unique to advanced kidney disease. Moreover, the success of revascularization is frequently undermined by poor outflow vessels, non-compressible arteries, and limited physical reserves to tolerate repeated interventions.

Patients with ESRD are also less likely to receive guideline-directed PAD therapies such as statins or antiplatelets, either due to therapeutic nihilism or concern over bleeding risk and polypharmacy [13,14]. Adherence to lifestyle interventions such as supervised exercise is also reduced, further accelerating functional decline. The heightened risk of infection, sepsis, and delayed wound healing post-intervention places these patients at an elevated risk of treatment failure and amputation [9,14,15,16,17,18].

A recent multicenter trial (BEST-CLI) confirmed that renal dysfunction remains among the most potent predictors of both limb loss and death following lower extremity revascularization, regardless of whether patients receive open or endovascular procedures [19].

Despite this, there is a lack of consensus on optimal treatment strategies for PAD in dialysis-dependent patients. Current guidelines offer little granularity for tailoring care to this subgroup, and existing scoring systems inadequately capture the compounding risk of renal failure. While the WIfI (Wound, Ischemia, and foot Infection) classification is often used, its predictive accuracy in ESRD patients is limited due to the unique pathophysiology involved.

In this context, a more detailed understanding of how ESRD modifies risk in patients undergoing vascular intervention is essential. Although prior studies have explored limb salvage and short-term outcomes in this group, few have provided comprehensive evaluations of predictors of mortality, readmissions, or long-term failure. To address this knowledge gap, we conducted a retrospective cohort study to evaluate post-operative outcomes in patients with PAD and ESRD on dialysis who underwent vascular interventions. We specifically aimed to identify factors associated with mortality, readmissions, and limb loss, with the goal of informing future patient selection and optimizing resource allocation in this vulnerable population.

2. Materials and Methods

2.1. Design

This is a retrospective analysis of all patients above 18 years old with a history of PAD and ESRD on dialysis, who had vascular intervention from 2015 to 2017 at a single institution. Indications for surgical intervention were chronic limb-threatening ischemia (CLTI), disabling claudication, rest pain, and non-healing ulcer(s) or gangrene. Patients were included if they met the following inclusion criteria: adults aged over 18 who have PAD and ESRD on dialysis [19].

2.2. Variables and Outcomes

Patient demographic information, including age and gender and medical comorbidities, such as ESRD, congestive heart failure (CHF), diabetes mellitus (DM), hypertension, myocardial infarction (MI), PAD, stroke/cerebrovascular accident (CVA), and transient ischemic attack (TIA) were collected. Prior interventions, such as percutaneous coronary intervention (PCI), coronary artery bypass graft (CABG), prior lower extremity bypass grafts, and medications, such as antiplatelet, statins, anticoagulants, or antibiotics for a prior infection were recorded for the aforementioned comorbidities. Details on vascular surgery interventions including open procedures, such as endarterectomy, lower extremity bypass, debridement performed in the operating room for the index, non-healing or infected wounds, amputation types such as toes, trans-metatarsal amputation (TMA), below-knee amputation (BKA), above-knee amputation (AKA), and endovascular interventions, such as balloon angioplasty, vessels treated, and type of stents such as drug-eluting stents or bare-metal stents were also noted. Patient outcomes, including hospital length of stay, mortality, amputations, bypass failure, and readmissions at 30, 90, and 120 days were analyzed. The study cohort was also analyzed for independent predictors of mortality and readmissions.

2.3. Definitions

In this study, a priori definitions for ESRD, PAD, and bypass failure were used. ESRD was defined as stage 5 of the National Kidney Foundation Kidney Disease Outcomes Quality Initiative classification of chronic kidney disease (CKD), indicating individuals requiring dialysis irrespective of glomerular filtration rate [20]. PAD was defined as a chronic arterial occlusive disease of the lower extremities caused by atherosclerosis [21]. Bypass failure was defined as stenosis or occlusion of bypass requiring angioplasty or redo bypass. Amputation of the limb after bypass was also considered as bypass failure.

2.4. Statistics

This study used summary statistics to present frequencies and percentages for categorical variables and mean and standard deviation for continuous data. Multivariate logistic regression was used to determine the independent predictors of readmissions and mortality. Univariable predictors for readmissions and mortality with p < 0.20 were entered in multivariable binary logistic regression and eliminated by backward regression until the final model had all predictors with p < 0.05. The receiver operating characteristic curve (ROC curve) was constructed for a graphical illustration of the area under the curve (AUC) to ascertain the discrimination power of our predictive models for readmissions and mortality. Data were analyzed using SPSS 26 software (IBM, Inc., Armonk, NY, USA). The statistical significance was defined as p < 0.05. This study was reported according to strengthening the reporting of observational studies in epidemiology (STROBE) criteria [22].

3. Results

There was a total of 74 patients who met the inclusion criteria.

3.1. Patient-Related Factors

The mean age of the study cohort was 66 ± 12.3 years with a range of 31–90 years. The elderly (aged ≥ 65 years) constituted about 55% of the cohort, and males (73%) were the predominant gender. The most common comorbidities were hypertension (85%), diabetes mellitus (63.5%), congestive heart failure (CHF) (37%), myocardial infarction (MI) (24%), and prior stroke (16.2%). The pertinent medications, i.e., antiplatelets, statins, anticoagulants, insulin/oral hypoglycemics, and antibiotics for prior infection, were taken by 62%, 57%, 50%, 63%, and 22% of patients, respectively (

Table 1. Patient-related factors.

Variables	N (%)
Age (mean ± SD years)	66 ± 12.3 (31–90)
Elderly (aged ≥ 65 years)	41 (55%)
Gender [n (%)]
Males	54 (73%)
Comorbidities
Prior stroke	12 (16.2%)
Prior lower extremity bypass graft surgery	15 (21%)
Percutaneous coronary intervention (PCI)	18 (24.3%)
Coronary artery bypass (CABG)	16 (21.6%)
Myocardial infarction (MI)	18 (24%)
Congestive heart failure (CHF)	27 (37%)
Angina	4 (3.1%)
Transient ischemic attacks	4 (5%)
Hypertension	63 (85%)
Diabetes Miletus	44 (63.5%)
Hyperlipidemia	42 (57%)
Medications
Insulin/oral hypoglycemic	47 (63.5%)
Antiplatelet	46 (62%)
Statins	42 (57%)
Anticoagulant	37 (50%)
Antibiotics for prior infection	16 (22%)

).

3.2. Description of Surgical Interventions

Amongst open procedures, surgical debridement on both index and surgical wounds was performed in 54%, endarterectomy in 23%, and lower extremity bypass in 23% of the patients. Amongst type of bypass performed; femoro-popliteal bypass (11%), popliteo-tibial bypass (11%), femoral-tibial bypass (8%), fem-fem bypass (4%), and tibio-peroneal bypass (1.3%) were seen in decreasing order of frequency. PTFE graft (12%) was the most common conduit used for bypass, followed by cadaveric vein (10%) and autologous vein (7%) in the order of frequency(

Table 2. Description of surgical interventions.

Variables	N (%)
Open surgical intervention
Endarterectomy	17 (23%)
Lower extremity bypass	17 (23%)
Debridement	40 (54%)
Amputation type
Toe amputation	35 (47%)
Trans-metatarsal amputation (TMA)	23 (31%)
Below-knee amputation (BKA)	14(19%)
Above-knee amputation (AKA)	9 (8%)
Endovascular surgical intervention
Angioplasty	68 (91%)
-  Iliac angioplasty	15 (21%)
-  Femoral angioplasty	17 (23%)
-  Popliteal angioplasty	6 (8%)
-  Tibial angioplasty -  Peroneal angioplasty	22 (30%) 8 (11%)
Drug-eluting stents	15 (20%)
Bare-metal stents	14 (19%)
Type of Bypass
Aorto-femoral	2 (3%)
Femoro-popliteal	8 (11%)
Popliteo-tibial	8 (11%)
Femoral-tibial	6 (8%)
Fem-fem bypass	3 (4%)
Tibio-peroneal	1 (1.3%)
Type of Conduit used
Cadaveric vein	7 (10%)
Autologous vein	5 (7%)
PTFE graft	9 (12%)

).

Toe amputation (47%) was the most frequent, followed by TMA (31%), BKA (19%), and AKA (12%). Among the endovascular procedures, angioplasty was performed in 93% of the patients. Tibial artery angioplasty (30%) was the most common type of angioplasty, followed by angioplasty of femoral artery (23%), iliac artery (21%), peroneal artery (11%), and popliteal artery (8%). The drug-eluting stents and bare-metal stents were used with similar frequency (20% vs. 19%, respectively).

3.3. Patient-Related Outcomes

The mean hospital stay of the cohort was 38 ± 47 days with a mortality rate of 10%. Angioplasty failure leading to bypass or amputation was seen in 32% of the patients, and bypass failure was seen in 12% of the patients. The overall readmission rate in our study cohort was 36%. The 30-day, 90-day, and 120-day readmission rates were 15%, 20%, and 27%, respectively. The most common reason for readmission was wound-related problems in 23% (n = 17), followed by bypass or stent failures in 12.2% (n = 9) of patients (Figure 1). The mean number of days to readmission was 12 ± 34 (

Table 3. Patient-related outcomes.

Variables	Mean ± SD
Hospital length of stay (HLOS) (mean ± SD) (days)	38 ± 47 (1–285)
Toes amputated	1.2 ± 1.8 (0–10)
Variables	N (%)
Angioplasty failure leading to bypass/amputation	24 (32%)
Bypass failure	9 (12.1%)
Readmissions
30-day readmission	11 (15%)
90-day readmission	15 (20%)
120-day readmission	20 (27%)
Days to readmission (mean ± SD)	12 ± 34 (0–159)
Mortality	7 (10%)

).

3.4. Predictive Modeling for Readmissions and Mortality

There were three univariable predictors for readmissions, namely, prior stroke, age, and females. However, on multivariable binary logistic regression, stroke was independently associated with high odds of readmission (

Table 4. Multivariable binary logistic regression model predicting readmissions after surgery for PAD in HD-dependent patients with ESRD.

Variables	Univariable Predictors			Multivariable Predictors
	OR	95% CI	p-Value	OR	95% CI	p-Value
Stroke	5.27	1.43–19.3	0.01	6.37	1.62–35	<0.01
Age	0.96	0.92–0.99	0.10	0.94	0.91–0.99	0.05
Females	0.59	0.17–2.00	0.41	Eliminated by backward regression

). The discriminatory power of predictive models for readmission as obtained on the ROC curve with the AUC was 0.714 (Figure 2).

For mortality, there were five univariable predictors, i.e., female gender, AKA, lower extremity bypass failure, TMA, and age. However, on multivariable binary logistic regression, AKA and female gender were found to be independently associated with higher odds of mortality (

Table 5. Multivariable binary logistic regression model predicting mortality after surgery for PAD in HD-dependent patients with ESRD.

Variables	Univariable Predictors			Multivariable Predictors
	OR	95% CI	p-Value	OR	95% CI	p-Value
Female	6.00	1.44–24.00	0.01	8.01	1.21–53.3	0.03
Above-knee amputation	16.00	2.41–106.0	<0.01	14.45	1.73–120.75	0.01
Lower extremity bypass failure	7.62	1.37–42.4	0.02	Eliminated by backward regression
Trans-metatarsal amputation	3.30	0.68–16.4	0.13
Age	1.05	0.97–1.13	0.16

). The discriminatory power of predictive models for mortality as obtained on the ROC curve with the AUC was 0.840 (Figure 3).

4. Discussion

This retrospective study analyzed 74 patients with PAD and ESRD on dialysis who underwent vascular intervention. Our results highlight the vulnerability of this patient population, which is associated with a higher risk of poor outcomes, including higher rates of readmissions, post-operative complications, and mortality. Notably, our study illustrated a high readmission rate of 36%, with rates of 15%, 20%, and 27% at 30, 90, and 120 days, respectively. Wound-related issues were the most common cause for readmission, occurring in 23% of our cohort. This finding is consistent with previous research indicating that patients on dialysis are more likely to present with limb-threatening wound infections and experience higher rates of limb loss following revascularization compared to patients not on dialysis [14].

In our cohort, certain risk factors were found to be associated with poor outcomes following vascular intervention. Prior stroke was found to be independently associated with higher odds of readmission, while AKA and female gender were independently associated with higher odds of mortality on multivariable binary logistic regression. These findings are consistent with prior data suggesting that neurologic impairment may lead to decreased mobility, impaired wound care, and difficulty adhering to post-discharge instructions, thereby increasing the risk of readmission [14].

Mao et al. reported that patients who underwent AKA after angioplasty demonstrated the highest rate of major adverse cardiovascular events and overall mortality [23]. The study highlighted that patients undergoing AKA often represent significant and advanced atherosclerotic disease and have other factors leading to poor outcomes. Although the underlying mechanisms are unclear, several studies have observed sex differences in the pathophysiology of diabetic vascular complications. Hirsch et al. reported that women had a significant increased risk of future cardiovascular events compared to men with an ankle-brachial index (ABI) < 0.7 [24]. Similarly, women had higher rates of cardiovascular death compared to men for an ABI < 0.7 as well as an ABI > 1.4, indicating non-compressible vessels and a sign of PAD [25]. Furthermore, Wisman et al. reported that women, regardless of age, had a five-fold risk of vascular death, major amputation, myocardial infarction, or stroke, while men younger than 60 years old had the best 5-year event-free survival following lower extremity bypass surgery [25].

The recent literature has reinforced these concerns. In the BEST-CLI trial, patients with chronic limb-threatening ischemia (CLTI) who had ESRD exhibited worse survival and limb salvage regardless of whether open surgery or endovascular treatment was selected [19].

Additionally, our findings of 32% angioplasty failure and 12% bypass failure reflect known limitations of revascularization durability in the ESRD population. The cause is likely multifactorial: a prothrombotic milieu, medial calcification, poor distal runoff, and impaired endothelial repair mechanisms all contribute [11,17]. This raises questions about whether traditional patency targets are even relevant in this group or whether limb preservation and quality of life should take precedence as primary endpoints.

Another notable finding from our study is the mean hospital stay of 38 days, which is considerably longer than PAD patients without ESRD, whose stays often range from 7 to 10 days for similar interventions [4]. This extended length of stay could be partially explained by wound complications, delayed graft healing, and the higher likelihood of in-hospital complications such as infection or decompensated heart failure. Optimizing discharge planning and early outpatient follow-up, particularly through coordinated multidisciplinary care (including nephrology, wound care, and vascular surgery), may help reduce these readmissions.

There is growing interest in preoperative frailty screening and individualized risk prediction models for vascular patients, especially those with ESRD. Emerging tools incorporating frailty indices, albumin levels, and walking speed may help identify patients who are unlikely to benefit from revascularization and who may be better served by palliative limb care or early amputation [26,27].

Despite our extensive institutional resources to enable advanced limb preservation capabilities, we found that limb preservation failed in nearly half (43%) of our patients with PAD and ESRD. This highlights the ongoing challenge of balancing procedural aggressiveness with realistic expectations of functional recovery in this frail group.

This study has potential limitations. Firstly, the small sample size could have limited the generalizability of our findings. Additionally, since this was a retrospective study, there could have been potential for selection bias. Furthermore, this study involved patients operated on by different vascular surgeons, which could influence intra- and post-operative decisions. Laboratory markers such as albumin, hemoglobin A1c, or CRP—which are known to correlate with outcomes—were not included in our model, but may offer further predictive value in future studies.

In conclusion, our study highlights the vulnerability of dialysis patients with PAD who are at a higher risk of mortality, complications, readmissions, and prolonged hospital stay. Stroke was associated with an increased likelihood of readmission, while AKA and the female gender were shown to be associated with a higher risk of mortality. Our findings underscore the importance of tailored surgical decision-making, better risk stratification tools, and interdisciplinary post-operative management in this challenging patient population.