Association between guideline-directed medical therapy and reintervention risk following peripheral vascular interventions in patients with peripheral artery disease

Abstract

Introduction:

Reintervention following peripheral vascular intervention (PVI) for peripheral artery disease (PAD) is common. Guideline-directed medical therapy (GDMT) is recommended post-PVI, yet its association with reintervention outcomes remains unclear.

Methods:

We analyzed Vascular Quality Initiative registry data linked with Medicare outcome for patients undergoing PVI for PAD (2017–2018). GDMT was defined as the receipt of statin, antiplatelet, and angiotensin-converting enzyme or angiotensin receptor blocker (ACE/ARB) therapy if hypertensive at discharge. Competing risk analyses and conditional risk models assessed the reintervention outcome, and the recurrent reintervention outcomes within 2 years, by GDMT receipt, compliance with each GDMT element, the number of elements received, and GDMT rate across sites and operators in a 1:1 propensity score-matched cohort.

Results:

We included 13,244 patients (mean age 72.0 ± 9.9, women 41.0%). The reintervention outcome did not differ by GDMT receipt (cumulative incidence: 43.0% [95% CI 41.0–44.9%] in no GDMT vs 41.2% [95% CI 39.4– 43.0%] in GDMT; subhazard ratio (sHR): 1.03 [95% CI 0.97–1.10]), compliance with GDMT elements, the number of elements received, or site and operator GDMT rates (sHR per 10% increase: 1.00 [95% CI 0.98–1.03] and 1.00 [95% CI 0.98–1.02]) (all p > 0.05). However, a higher operator GDMT rate reduced the recurrent reintervention risk (HR: 0.98 [95% CI 0.97–1.00], p = 0.026).

Conclusion:

Around 40% of patients undergoing a PVI experience reintervention within 2 years, but the outcome was not reduced with GDMT receipt, and higher GDMT rates by site and operators were not associated with reintervention risk. Future studies should focus on medication adherence, refills, and more granular GDMT data for PAD care surveillance postrevascularization.

Keywords

guideline-directed medical therapy peripheral artery disease (PAD)peripheral vascular intervention reintervention

Introduction

Peripheral artery disease (PAD) is an increasingly prevalent disease, currently affecting over 230 million people worldwide and representing over $4.37 billion in costs for the United States healthcare system, with procedures bearing a significant portion of the costs.^1,2 Administrating optimal guideline-directed medical therapy (GDMT) – antiplatelet therapy, statin medication, and antihypertensive agents – is strongly recommended per American Heart Association/American College of Cardiology (AHA/ACC) guidelines to reduce cardiovascular risk in patients with PAD.^3,4 Despite these recommendations, only about half of the patients undergoing peripheral vascular interventions (PVI) received optimal GDMT with a high variability across the health system and operators.^4,5 Older patients presenting multiple comorbidities and those with advanced PAD, termed chronic limb-threatening ischemia (CLTI), face a higher risk of reintervention following initial PVI and are less likely to be discharged on GDMT.^6
–8

Reinterventions after a PVI are common. In the BASIL-2 and BEST-CLI trials, the reintervention rate at 30 days for CLTI was 19.0% and 23.5%, respectively.^9,10 This rate increased to nearly 40% at 3 years across the PAD spectrum.¹¹ Despite the high prevalence of repeat lower-limb revascularization and its associated costs, there is a lack of studies evaluating effective clinical strategies to reduce the need for reintervention after initial revascularization for PAD. GDMT has been associated with a lower risk of major amputation and mortality at 2 years, but its association with reintervention remains unclear and mostly extrapolated from the evidence in the coronary field.⁴ We therefore aim to assess: (1) the association between the receipt of GDMT, the compliance with each GDMT element, and the total number of GDMT elements received, and 2-year reintervention risk following the index PVI; and (2) the GDMT rate at discharge following the index PVI across sites and operators and subsequent reintervention risk. These aims were addressed using the national Medicare-linked Vascular Quality Initiative (VQI) registry data of patients undergoing PVI for PAD.

Methods

Data source and studied population

Our study population, derived from a claims-linked registry dataset, has been previously described by Smolderen et al.⁴ Briefly, we derived a dataset that combines the VQI registry data – capturing the demographic, clinical, and procedural data of patients undergoing endovascular procedures in the lower extremities performed at over 800 centers nationwide – with Medicare claims outcomes data.^12
–14 Consistent with our prior work, we included adults with PAD who underwent PVI between January 1, 2017 and December 31, 2018.⁴ We excluded patients with acute limb ischemia, those without Medicare fee-for-service coverage, individuals treated at centers with a high missing rate (defined as the third quartile + 1.5 × IQR), and patients with incomplete information on GDMT.⁴ Approval for the study was granted by the VQI Research Advisory Committee and the Yale Institutional Review Board.

Exposure: Guideline-directed medical therapy (GDMT)

GDMT was defined according to the 2016 AHA/ACC guidelines, which were the most current recommendations available for our study population.³ This guideline recommended prescribing the three following medication elements upon discharge: statin, antiplatelet, and angiotensin-converting enzyme inhibitor or angiotensin receptor blocker (ACEi/ARB) if the patient had hypertension. A detailed definition of the optimal GDMT receipt has been previously published.⁴ Briefly, to determine receipt of optimal GDMT, we first defined compliance for each GDMT medication element. For both statins and antiplatelet therapies, compliance was defined as either the patient receiving the medication or being deemed ineligible due to documented medical contraindications. For ACEi/ARB therapy, compliance was defined similarly for patients with hypertension. In patients without hypertension, ACEi/ARB compliance was considered achieved irrespective of whether they received the medication.

Patients who met guideline compliance for all three medications were then classified into the GDMT group; otherwise, they were classified into the no GDMT group. Similarly, we also examined the compliance for each GDMT element. Additionally, we examined the number of GDMT elements received (0, 1, 2, or 3) among patients with hypertension and without contraindication for statin, antiplatelet, and ACEi/ARB medications (i.e., eligible patients receiving full optimal GDMT).

Finally, we calculated the GDMT rate by sites and operators (the number of patients receiving GDMT divided by the total patients by sites and operators, respectively). The GDMT rates by sites and operators were examined as quartiles (Q) and as continuous linear per 10% increase.

Reintervention outcome

Our primary outcome was first reintervention following index PVI derived from the corresponding Current Procedural Terminology (CPT) and International Classification of Diseases (ICD) codes queried in Medicare claims data, and included any endovascular (defined as any catheter-based therapy) or surgical intervention (including endarterectomy, bypass, or other surgical treatment) performed after an initial endovascular procedure. The CPT and ICD codes can be found in the Supplementary Table 1. We considered both ipsilateral and contralateral reintervention, as the Medicare claims data received within the VISION platform does not distinguish for laterality. To account for the competing risk of death, the patients’ vital status was derived from the Centers for Medicare and Medicaid Services’ vital status files. Prior work has demonstrated that the linked VQI–Medicare database achieves a sensitivity of 92% and a specificity of 96% in identifying reintervention events and an accuracy of > 99% for identifying a death in the registry.^15,16 Patients were followed until the first reintervention, up to the 2-year follow up, until December 31, 2018, or until death, whichever occurred first. As an exploratory analysis, we also examine the risk of undergoing multiple reinterventions, including all reinterventions following the index PVI.

Statistical analysis

To reduce selection bias between patients who did not receive optimal GDMT versus those who did, the analyses were performed in a 1:1 propensity score (PS)-matched cohort as described previously.^4,17 The overlap of the Kernel density plot of PSs and the negligeable effect size between the two groups (absolute standardized differences, d < 0.2) demonstrated the effectiveness of the 1:1 PS matching method employed.¹⁸

Patients and procedural characteristics, along with the granular information on receipt of GDMT at discharge, were described for the overall 1:1 PS matched cohort, by GDMT groups, and by PAD symptoms. Continuous data were summarized as mean ± SD and were compared using Student’s t-test. Categorical variables were summarized as count (percent) and were compared using the chi-squared test and Fisher’s exact test, as appropriate. Additionally, both the GDMT and the PAD symptoms groups were compared based on standardized differences with an absolute value < 0.20, 0.20–0.49, 0.50–0.79, and ≥ 0.80 indicating negligeable, small, medium, and large effect size, respectively.^19
–21 Similarly, available data on reintervention outcome at 2 years following the index PVI were described and compared by GDMT groups.

We conducted time-to-event competing risk analyses to assess the 2-year major amputation outcome while accounting for competing risk of death.^22,23 We used the Aalen–Johansen estimator to calculate the 2-year cumulative incidence of reintervention by GDMT groups (no GDMT vs GDMT) and using Gray’s test for comparison. The 2-year risk of reintervention in no GDMT versus GDMT groups was estimated using the subhazard ratio (sHR) with a 95% CI derived from a Fine–Gray competing risk regression model.²⁴

Additionally, we examined the association between compliance for each GDMT element (statin, antiplatelet, and ACEi/ARB) and the 2-year reintervention risk. We also assessed the reintervention risk associated with the number of GDMT elements received (0, 1, or 2 vs 3 GDMT element[s]) in patients with hypertension and no contraindication of receiving statin, antiplatelet, and ACEi/ARB therapy.

Owing to high variability in GDMT rates by sites and operators, we examined the 2-year cumulative incidence of reintervention by site and operator quartiles of GDMT rates, and the 2-year reintervention risk associated with a 10% increase in GDMT rate across sites and operators, as described above.

As an exploratory analysis, we also examined the risk of undergoing multiple reinterventions (i.e., risk of recurrent reintervention), including all reinterventions following the index PVI. The recurrent reintervention outcome includes all reinterventions occurring during the patient’s follow up after the index PVI. Recurrent reinterventions were identified using corresponding CPT and ICD codes from Medicare claims data, as described above and listed in Supplementary Table 1. To analyze the recurrent reintervention risk, we used a conditional risk set model (Prentice–Williams–Peterson regression model).²⁵ This model enables the risk of recurrent reintervention (expressed as hazard ratio (HR) with 95% CI) to be assessed while accounting for competing risk of death, by GDMT groups, by compliance for each GDMT element received, by number of GDMT elements received, and per 10% increase in the GDMT rate across sites and operators.

Missing data

As outlined in prior work, multiple imputation by chained equation and the Rubin’s rule were applied to account for missing data in the PS calculation.^26,27

All statistical tests were conducted as two-sided, with a significance level set at p < 0.05. To handle multiple testing, p-values from pairwise comparison were corrected using the Simes method.²⁸ All analyses were performed using Stata/MP 18 (StataCorp LLC, College Station, TX, USA).

Results

Study population

The 1:1 PS-matched cohort, as previously described by Smolderen et al.,⁴ included 13,244 patients (6622 in each GDMT group). The PS balance between GDMT groups was met after 1:1 matching, as shown in Supplementary Figure 1. Overall, the mean age was 71.8 ± 9.9 years, 41.0% were women, 15.8% were Black/African American, and 4.1% were Hispanic/Latino. Patient and procedural characteristics, as well as granular information on GDMT receipt at discharge, are summarized for the overall cohort and by GDMT groups in Supplementary Table 2. Of patients eligible for receiving the three elements of optimal GDMT (n = 11,771), 0.9% received none of the GDMT elements, 8.0% received one element, 35.9% received two elements, and 42.8% received all three elements of optimal GDMT.

The 1:1 PS-matched cohort included 58.7% of patients presenting CLTI symptoms. The GDMT rates and patient characteristics did not differ between the PAD symptoms groups, as shown in Table 1.

Table 1.

Patient characteristics, procedural characteristics, and granular information on GDMT receipt at discharge stratified by peripheral artery disease symptoms.^a

	No CLTI(n = 5445)	CLTI(n = 7746)	p-value	d
Demographics
Age, years, mean (SD)	71.8 (8.5)	71.8 (10.8)	0.695	–0.007
Female sex	2244 (41.2)	3162 (40.8)	0.653	0.008
Race			< 0.001	0.008
White	4605 (84.6)	5808 (75.0)
Black or African American	624 (11.5)	1454 (18.8)
Asian	47 (0.9)	80 (1.0)
American Indian or Alaskan Native	17 (0.3)	60 (0.8)
Other or unknown	152 (2.8)	344 (4.4)
Hispanic ethnicity	130 (2.4)	409 (5.3)	< 0.001	0.151
Primary insurer			0.010	0.061
Medicare	5202 (95.7)	7315 (94.6)
Medicaid	23 (0.4)	61 (0.8)
Commercial	191 (3.5)	319 (4.1)
Other	21 (0.4)	40 (0.5)
Preoperative living location at home	5389 (99.1)	7186 (92.9)	< 0.001	0.320
Medical history and risk factors
Smoking status			< 0.001	0.396
Never	773 (14.2)	2341 (30.3)
Prior	2931 (53.8)	3525 (45.6)
Current	1739 (31.9)	1872 (24.2)
Hypertension	5173 (95.0)	7189 (92.8)	< 0.001	0.092
Congestive heart failure	883 (16.2)	2354 (30.4)	< 0.001	0.340
Coronary artery disease	2003 (36.8)	3034 (39.2)	0.006	0.049
Coronary artery bypass grafting	1214 (22.3)	1868 (24.2)	0.014	0.043
Percutaneous coronary intervention	1456 (26.8)	1993 (25.8)	0.199	0.023
Diabetes	2061 (37.9)	5239 (67.7)	< 0.001	0.625
Chronic kidney disease	1764 (34.6)	3345 (46.2)	< 0.001	0.237
Cerebrovascular disease	742 (13.6)	1519 (19.6)	< 0.001	0.161
Chronic lung disease	1699 (31.2)	2184 (28.2)	< 0.001	0.066
Carotid artery disease	665 (12.2)	649 (8.4)	< 0.001	0.126
Prior amputation	267 (4.9)	2005 (25.9)	< 0.001	0.608
Prior peripheral procedure	2490 (45.7)	3079 (39.8)	< 0.001	0.121
CLTI laterality				0.776
Asymptotic	307 (5.6)	0 (0.0)	< 0.001
Left	1061 (19.5)	2824 (36.5)
Right	1090 (20.0)	2893 (37.3)
Both	2987 (54.9)	2029 (26.2)
Peripheral vascular intervention
Aorta/iliac artery treated by endovascular intervention	2419 (44.4)	1797 (23.2)	< 0.001	0.461
Common onto external iliac artery treated by endovascular intervention	326 (6.0)	319 (4.1)	< 0.001	0.085
Profunda artery treated by endovascular intervention	72 (1.3)	101 (1.3)	0.925	0.002
SFA/popliteal artery treated by endovascular intervention	3176 (58.4)	4796 (61.9)	< 0.001	0.073
Below-knee artery treated by endovascular intervention	678 (12.5)	3399 (43.9)	< 0.001	0.746
Plain balloon	3732 (68.5)	6135 (79.2)	< 0.001	0.244
Atherectomy	1189 (21.8)	1534 (19.8)	0.005	0.050
Special balloon	1988 (36.5)	2538 (32.8)	< 0.001	0.079
Stent	2907 (53.4)	3079 (39.7)	< 0.001	0.276
Bailout stent	105 (1.9)	148 (1.9)	0.942	0.001
Bailout stent graft	15 (0.3)	15 (0.2)	0.331	0.017
Preprocedural granular information on receipt of GDMT
Receipt of GDMT at prior procedure^b	2439 (44.8)	3236 (41.8)	< 0.001	0.061
Compliance with aspirin/antiplatelet therapy^c	4693 (86.2)	6249 (80.7)	< 0.001	0.148
Compliance with statin therapy^c	4345 (79.8)	5913 (76.4)	< 0.001	0.084
Compliance with ACEi/ARB therapy^d	3162 (58.1)	4295 (55.5)	0.003	0.053
Granular information on receipt of GDMT at discharge
Receipt of GDMT at discharge^b	2683 (49.3)	3914 (50.5)	0.156	0.025
Compliance with aspirin/antiplatelet therapy^c	5207 (95.6)	7298 (94.2)	< 0.001	0.064
Compliance with statin therapy^c	4531 (83.2)	6443 (83.2)	0.957	0.001
Compliance with ACEi/ARB therapy^d	3135 (57.6)	4345 (56.1)	0.091	0.030
Number of GDMT elements received			< 0.001	0.092
0	35 (0.6)	82 (1.1)
1	388 (7.1)	658 (8.5)
2	2060 (37.8)	2680 (34.6)
3	2411 (44.3)	3231 (41.7)
< 3 because patient did not have hypertension	272 (5.0)	557 (7.2)
< 3 because patient not eligible^e	279 (5.1)	538 (6.9)

Data presented as n (%) unless otherwise specified.

n = 53 not presented here due to missing data on PAD symptoms.

Patients who met guideline compliance for aspirin/antiplatelet, statin, and ACEi/ARB therapies.

Patients receiving the medication or not eligible for compliance with guideline due to documented medical contraindication(s).

Patients with hypertension receiving the medication or not eligible for compliance with guideline due to documented medical contraindication(s); or patients without hypertension as not eligible for compliance.

Patient not eligible for compliance with guideline due to documented medical contraindication(s).

ACEi/ARB, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker; CLTI, chronic limb-threatening ischemia; GDMT, guideline-directed medical therapy; PAD, peripheral artery disease; SFA, superficial femoral artery; d, standardized differences, with an absolute value < 0.20, 0.20–0.49, 0.50–0.79, and ≥ 0.80 which indicated negligeable, small, medium, and large effect size, respectively.

The mean time to follow-up was 0.9 ± 0.6 years, 3.2% of patients were lost to follow up, and 11.3% died before undergoing reintervention (i.e., competing event of death). A third of patients underwent reintervention. On average, patients received 0.6 ± 0.18 reinterventions per year, and mostly for the femoro- and infrapopliteal artery (Supplementary Table 3).

Reintervention outcome by receipt of GDMT

Receipt of GDMT was not associated with reintervention at 2 years (cumulative incidence in no GDMT: 43.0% [95% CI 41.0–44.9%] vs GDMT: 41.2% [95% CI 39.4– 43.0%], p = 0.293; sHR: 1.03 [95% CI 0.97–1.10], p = 0.293) (Figure 1A and Table 2). Similarly, guideline compliance for each GDMT element and the number of GDMT elements were not associated with the 2-year reintervention risk (Table 2).

Figure 1.

The 2-year cumulative incidence of reintervention following PVI in patients with PAD by receipt of GDMT at discharge (A), and by quartile of GDMT rate across site (B) and operators (C).

Table 2.

The 2-year reintervention risk^a following index peripheral vascular intervention in patients with PAD.

	sHR^a	95% CI	p-value
Receipt of optimal GDMT
No GDMT vs GDMT	1.03	0.97–1.10	0.293
Compliance for each GDMT element
No compliance for statins	1.00	0.92–1.08	0.986
No compliance for antiplatelets	0.96	0.83–1.11	0.577
No compliance for ACEi/ARB	0.96	0.90–1.02	0.204
Number of GDMT elements received ^b
0	0.83	0.59–1.16	0.273
1	0.99	0.88–1.12	0.899
2	0.97	0.90–1.04	0.338
3 (full optimal GDMT)	Reference
Per 10% increase in GDMT rate
Across sites	1.00	0.98–1.03	0.760
Across operators	1.00	0.98–1.02	0.774

Risk of reintervention (i.e., considering the first occurrence of reintervention only), derived from Fine–Gray competing risk regression model accounting for competing risk of death, using a 1:1 PS-matched cohort of patients with PAD.

In the subcohort of patients eligible for receiving full optimal GDMT (i.e., with hypertension and no medical contraindication to statins, antiplatelets, or ACEi/ARB GDMT elements).

ACEi/ARB, angiotensin-converting enzyme inhibitor or angiotensin receptor blocker; GDMT, guideline-directed medical therapy; PAD, peripheral artery disease; PS, propensity score; PVI, peripheral vascular interventions; sHR, subhazard ratio.

Reintervention outcomes by GDMT rates across sites and operators

The 2-year cumulative incidence of reintervention differed between the third and fourth quartiles of GDMT rates across sites only (cumulative incidence in Q3: 41.9% [95% CI 39.3–44.5%] vs Q4: 42.2% [95% CI 39.8–44.6%], p = 0.016; Figure 1B). However, the 2-year reintervention risk per 10% increase in GDMT rates by sites was not significant (sHR: 1.00 [95% CI 0.98–1.03], p = 0.760) (Table 2).

The variability in the GDMT rates by operator was not associated with the 2-year reintervention outcome, with the cumulative incidence ranging from 41.2 (95% CI 38.2–44.0%) in Q1 to 42.3 (95% CI 39.1–45.6%) in Q4 (p = 0.700; Figure 1C); the sHR per 10% increase in GDMT rate by operator was 1.00 (95% CI 0.98–1.02, p = 0.774) (Table 2).

Exploratory analysis: Multiple reintervention outcome by receipt of GDMT and by GDMT rates across sites and operators

The receipt of GDMT was not significantly associated with the 2-year risk of recurrent reintervention, nor the compliance for each GDMT element or the number of GDMT elements received. Additionally, though the risk of recurrent reintervention was not significantly associated with GDMT rate by sites, this risk decreased significantly by 2% per 10% increase in GDMT rate by operator (HR: 0.98 [95% CI 0.97–1.00], p = 0.026) (Supplementary Table 4).

Discussion

Using a nationwide vascular registry for patients undergoing PVI, we found that more than 40% of patients will undergo a reintervention within 2 years. Around 43% of those patients who underwent a PVI complied with the three elements of GDMT at discharge. Using our propensity-matched cohort, we were unable to document a significant association between GDMT discharge rates, either in full, or its partial components, and subsequent reintervention risk for all disease stages (claudication, CLTI), or when considering recurrent reinterventions. Our findings also did not show an association between variability in GDMT discharge rates as assessed using VQI-VISION data by site and operators and the subsequent 2-year reintervention risk.

Previous work has shown an association between GDMT prescription at discharge and all-cause mortality and major amputation at 2 years, as well as 90-day readmission rates, for patients with PAD undergoing an index PVI.^4,29,30 Variability in GDMT prescription rates at sites and operators was also associated with the risk for mortality, major amputation, and 90-day readmission.^4,29,30 Despite using a similar methodology and cohorts, our results did not support an association of partial or full GDMT nor GDMT variability with reintervention rates at 2 years. These results, however, are aligned with previously published work exploring factors associated with reintervention in single-center studies, which reported a nonsignificant association of aspirin and statins with reintervention rates after open or endovascular revascularization among patients with PAD and CLTI.^31,32

The nonsignificant association of GDMT use and reintervention rates can potentially be explained by multiple factors. Up to two-thirds of patients with PAD have been shown to have suboptimal medication adherence,³³ and receipt of GDMT at discharge may not be correlated with long-term medication adherence. Moreover, there is no information about the doses or regimens of GDMT prescribed in current datasets. Another potential explanation is the barriers in medication access or suboptimal refills. In chronic cardiovascular conditions, up to 40% of patients are reported to have suboptimal refill rates.³⁴ Another potential explanation could be the follow-up time needed for GDMT to significantly reduce the incidence and risk of reintervention, and longer follow up might be needed to demonstrate a difference. Lastly, guideline recommendations such as lifestyle changes, cardiovascular risk reduction, diabetes management, and referral to exercise therapy are not available in current national datasets and might also be associated with reintervention rates.

Patients eligible to undergo revascularization are at a crucial time in the disease and treatment process, and a higher emphasis on life-long therapy is warranted and may improve vascular outcomes. GDMT after the index revascularization has been shown to reduce mortality, major amputation, and 90-day readmissions.^4,29,30 Other interventions such as exercise therapy, lifestyle changes, and cardiovascular risk reduction have also been shown to improve major adverse limb and cardiovascular events.³⁵ Addressing concomitant risk factors and adhering to the most updated guideline recommendations might also improve the patency of procedures in real-world scenarios, supporting lower reintervention rates. For example, in recent trials, low-dose rivaroxaban and aspirin was shown to decrease reintervention rates as a secondary outcome, and was also added as a new class IA recommendation in the 2024 PAD guidelines to reduce major cardiovascular and limb events.^36
–38 As patients undergoing a revascularization are likely more symptomatic, this should be an important timepoint for the clinicians to emphasize the importance of GDMT, lifestyle changes, cardiovascular risk reduction, exercise therapy, and addressing mental health comorbidities to complement the revascularization, and this may aid in improving the patency of procedures and reduce reinterventions.³⁹

Evaluating reintervention rates after the index revascularization have gained increasing relevance.³¹ In the BEST-CLI trial, the difference between surgical versus endovascular revascularization was mainly driven by the increased risk of reintervention in the endovascular group.¹⁰ Endovascular revascularization is burdensome for patients and systems alike. The mean total costs of the initial treatment and hospitalization ranges from USD10,000 to almost USD20,000 depending on the PAD staging, and the 2-year follow-up costs range from USD11,416 to USD25,720.⁴⁰ Thus, strategies for improving the patency of procedures could also be associated with improving vascular outcomes, quality of life, and reducing healthcare costs.

Assessment and evaluation of GDMT and reinterventions in a real-world cohort requires expanding current datasets to align with the latest recommendations. There are multiple ways that current datasets may expand to provide a more comprehensive overview of vascular care, including: data about surrogates and regimens of GDMT (creatinine, low-density lipoprotein levels, follow-up blood pressure measurements and body mass index, HbA1c, and medication doses and frequency), the use of cardiovascular risk-reducing medications (glucagon-like peptide-1 [GLP-1] analogs, sodium-glucose transport protein 2 [SGLT2] inhibitors, different antihypertensives classes, protein convertase subtilsin/kexin type 9 [PCSK9] inhibitors, anticoagulation), patient-reported outcomes (patient-reported adherence, Physical Activity Questionnaire [PAQ] score, EuroQol [EQ]-5, walking distance), use of smoking cessation strategies, billing codes for exercise referral, and diet and exercise changes.^41
–43 These variables are all frequent follow-up elements used daily in clinical practice by vascular specialists. To easily include in national datasets, machine-learning natural-language processing models may also aid in the review of electronic health records and could support the integration of these variables into large databases.^44,45

A complete cardiovascular risk evaluation and treatment of risk factors associated with vascular outcomes requires a team-based and multidisciplinary approach as advocated for in recent statements.^38,46–48 In addition, a granular collection of data regarding these multidisciplinary care pathways are needed to assess its real-world impact on outcomes. All these efforts would greatly support the goal of the AHA to reduce nontraumatic major amputation by 20% in 2030 and support the foundation of value-based care for patients with PAD.⁴⁹

Limitations

In addition to the previously mentioned limitations of the current datasets and the need to expand the data elements captured for pharmacological and nonpharmacologic interventions, our results should be interpreted with the following limitations in mind. Despite using a propensity-matched and well-balanced cohort, unmeasured confounding due to the observational nature of the study may have biased the results. Confounding might also stem from other unmeasured variables such as various social determinants of health not included in the dataset. Another potential confounding factor is the lack of data on the duration patients were on GDMT; longer durations could be associated with better outcomes. Additionally, we cannot rule out selection bias from using the VQI dataset, which might not represent all centers performing PVI in the United States. Finally, the current platform does not record the laterality of the procedure, preventing us from associating GDMT with ipsilateral reintervention, which may lead to an overestimation of reintervention rates.

Conclusion

Reintervention rates are frequent among patients with PAD, with more than 40% of patients experiencing a reintervention after a PVI. GDMT prescription and prescription rate variability is not associated with reintervention risk at 2 years. Updated registries and a holistic approach to GDMT and clinical care after revascularization are needed to potentially reduce reinterventions among patients with PAD undergoing revascularization.

Supplemental Material

sj-pdf-1-vmj-10.1177_1358863X251320347 – Supplemental material for Association between guideline-directed medical therapy and reintervention risk following peripheral vascular interventions in patients with peripheral artery disease

Supplemental material, sj-pdf-1-vmj-10.1177_1358863X251320347 for Association between guideline-directed medical therapy and reintervention risk following peripheral vascular interventions in patients with peripheral artery disease by Santiago Callegari, Gaëlle Romain, Isabella Capuano, Jacob Cleman, Lindsey Scierka, Kim G Smolderen and Carlos Mena-Hurtado in Vascular Medicine

Footnotes

Declaration of conflicting interests

Dr Mena-Hurtado reports unrestricted research grants from Philips and Shockwave and is a consultant for Abbott Vascular, Cook, Medtronic, and Terumo. Dr Smolderen reports unrestricted research grants from Philips, Merck, Shockwave, and Johnson & Johnson; she is a consultant for Terumo, Cook, and Twill, Inc. The other authors report no competing interests.

Funding

Research reported in this publication was supported by the National Heart, Lung, and Blood Institute of the National Institutes of Health under Award Number 1R01HL163640-01. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

ORCID iDs

Santiago Callegari

Gaëlle Romain

Jacob Cleman

Kim G Smolderen

Carlos Mena-Hurtado

Supplemental material

Supplemental material for this article is available online.

References

Hirsch

Hartman

Town

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

McDermott

MM.

Peripheral artery disease: Past and future. Circulation 2024; 149: 1151–1153.

Gerhard-Herman

Gornik

Barrett

, et al 2016 AHA/ACC guideline on the management of patients with lower extremity peripheral artery disease: A report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2017; 69: e71–e126.

Smolderen

Romain

Provance

, et al Guideline-directed medical therapy and long-term mortality and amputation outcomes in patients undergoing peripheral vascular interventions. JACC Cardiovasc Interv 2023; 16: 332–343.

Llanos-Chea

Jelani

Q-U-A

Trejo-Paredes

, et al Lack of guideline-directed medical therapy in patients undergoing endovascular procedures for critical limb ischemia. J Am Coll Cardiol 2021; 77: 1374–1375.

Altin

Castro-Dominguez

Kennedy

, et al Predictors of underutilization of medical therapy in patients undergoing endovascular revascularization for peripheral artery disease. JACC Cardiovasc Interv 2020; 13: 2911–2918.

Evans

Sobieszczyk

Eisenhauer

, et al Risk of mortality related to recurrent limb events after endovascular revascularization of the superficial femoral artery for peripheral artery disease: The Boston Femoral Artery Endovascular Revascularization Outcomes (Boston FAROUT) study. Am J Cardiol 2023; 205: 425–430.

Weissler

Wang

Gales

, et al Cardiovascular and limb events following endovascular revascularization among patients ≥65 years old: An American College of Cardiology PVI Registry analysis. J Am Heart Assoc 2022; 11: e024279.

Bradbury

Moakes

Popplewell

, et al A vein bypass first versus a best endovascular treatment first revascularisation strategy for patients with chronic limb threatening ischaemia who required an infra-popliteal, with or without an additional more proximal infra-inguinal revascularisation procedure to restore limb perfusion (BASIL-2): An open-label, randomised, multicentre, phase 3 trial. Lancet 2023; 401: 1798–1809.

10.

Farber

Menard

Conte

, et al Surgery or endovascular therapy for chronic limb-threatening ischemia. N Engl J Med 2022; 387: 2305–2316.

11.

Mohamedali

Kiwan

Kim

, et al Reinterventions in patients with claudication and chronic limb threatening ischemia. Ann Vasc Surg 2022; 79: 56–64.

12.

Columbo

Kang

Trooboff

, et al Validating publicly available crosswalks for translating ICD-9 to ICD-10 diagnosis codes for cardiovascular outcomes research. Circ Cardiovasc Qual Outcomes 2018; 11: e004782.

13.

Mao

Etkin

Lewallen

, et al Creation and validation of linkage between orthopedic registry and administrative data using indirect identifiers. J Arthroplasty 2019; 34: 1076–1081.e0.

14.

Mao

Moore

Columbo

, et al Validation of an indirect linkage algorithm to combine registry data with Medicare claims. J Vasc Surg 2022; 76: 266–271.e2.

15.

Butala

Strom

Faridi

, et al Validation of administrative claims to ascertain outcomes in pivotal trials of transcatheter aortic valve replacement. JACC Cardiovasc Interv 2020; 13: 1777–1785.

16.

Columbo

Kang

Hoel

, et al A comparison of reintervention rates after endovascular aneurysm repair between the Vascular Quality Initiative registry, Medicare claims, and chart review. J Vasc Surg 2019; 69: 74–79.e6.

17.

Austin

PC.

An introduction to propensity score methods for reducing the effects of confounding in observational studies. Multivariate Behav Res 2011; 46: 399–424.

18.

Austin

PC.

Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med 2009; 28: 3083–3107.

19.

Austin

PC.

Using the standardized difference to compare the prevalence of a binary variable between two groups in observational research. Commun Stat Simul Comput 2009; 38: 1228–1234.

20.

Cohen

Statistical power analysis for the behavioral sciences. L. Erlbaum Associates, 1988.

21.

Yang

Dalton

. A unified approach to measuring the effect size between two groups using SAS. Paper 335-2012. In: SAS Global Forum 2012, Orlando, FL, USA, April 22–25, 2012. https://support.sas.com/resources/papers/proceedings12/335-2012.pdf

22.

Austin

Lee

Fine

JP.

Introduction to the analysis of survival data in the presence of competing risks. Circulation 2016; 133: 601–609.

23.

Callegari

Smolderen

Cleman

, et al Competing risk analysis to estimate amputation incidence and risk in lower-extremity peripheral artery disease. Vasc Med 2024; 29: 496–506.

24.

Fine

Gray

RJ.

A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94: 496–509.

25.

Prentice

Williams

Peterson

AV.

On the regression analysis of multivariate failure time data. Biometrika 1981; 68: 373–379.

26.

Azur

Stuart

Frangakis

, et al

Multiple imputation by chained equations: What is it and how does it work?

Int J Methods Psychiatr Res 2011; 20: 40–49.

27.

Rubin

Multiple imputation for nonresponse in surveys. John Wiley & Sons, 1987.

28.

Simes

RJ.

An improved Bonferroni procedure for multiple tests of significance. Biometrika 1986; 73: 751–754.

29.

Smolderen

Romain

Cleman

, et al Variability in guideline-directed medical therapy across sites and operators and long-term mortality and amputation outcomes risk in patients undergoing peripheral vascular interventions. Am Heart J 2024; 270: 75–85.

30.

Wittig

Pflug

Schmidt

, et al Impact of optimal medical therapy on reintervention and survival rates after endovascular infrapopliteal revascularization. J Clin Med 2023; 12: 5146.

31.

Ochoa Chaar

Gholitabar

DeTrani

, et al The Reintervention Index: A new outcome measure for comparative effectiveness of lower extremity revascularization. Ann Vasc Surg 2020; 69: 52–61.

32.

Mohamedali

Kiwan

Kim

, et al Reinterventions in patients with claudication and chronic limb threatening ischemia. Ann Vasc Surg 2022; 79: 56–64.

33.

Lampridou

Rawasdheh

Saghdaoui

, et al Factors affecting adherence to medication, smoking cessation, and exercise in patients with peripheral artery disease. JVS-Vascular Insights 2024; 2: 100074.

34.

Gazmararian

Kripalani

Miller

, et al Factors associated with medication refill adherence in cardiovascular-related diseases: A focus on health literacy. J Gen Intern Med 2006; 21: 1215–1221.

35.

Hamburg

Balady

GJ.

Exercise rehabilitation in peripheral artery disease: Functional impact and mechanisms of benefits. Circulation 2011; 123: 87–97.

36.

Hiatt

Bonaca

Patel

, et al Rivaroxaban and aspirin in peripheral artery disease lower extremity revascularization: Impact of concomitant clopidogrel on efficacy and safety. Circulation 2020; 142: 2219–2230.

37.

Eikelboom

Connolly

Bosch

, et al Rivaroxaban with or without aspirin in stable cardiovascular disease. N Engl J Med 2017; 377: 1319–1330.

38.

Gornik

Aronow

Goodney

, et al 2024 ACC/AHA/AACVPR/APMA/ABC/SCAI/SVM/SVN/SVS/SIR/VESS guideline for the management of lower extremity peripheral artery disease: A report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Circulation 2024; 149: e1313–e1410.

39.

Virani

Newby

Arnold

, et al 2023 AHA/ACC/ACCP/ASPC/NLA/PCNA guideline for the management of patients with chronic coronary disease: A report of the American Heart Association/American College of Cardiology Joint Committee on Clinical Practice Guidelines. Circulation 2023; 148: e9–e119.

40.

Magnuson

Vilain

, et al Two-year PAD-related health care costs in patients undergoing lower extremity endovascular revascularization: Results from the LIBERTY 360° trial. J Med Econ 2021; 24: 570–580.

41.

Belch

JJF

Brodmann

Baumgartner

, et al Lipid-lowering and anti-thrombotic therapy in patients with peripheral arterial disease: European Atherosclerosis Society/European Society of Vascular Medicine Joint Statement. Atherosclerosis 2021; 338: 55–63.

42.

Verma

Al-Omran

Leiter

, et al Cardiovascular efficacy of liraglutide and semaglutide in individuals with diabetes and peripheral artery disease. Diabetes Obes Metab 2022; 24: 1288–1299.

43.

Mlekusch

Exner

Sabeti

, et al Serum creatinine predicts mortality in patients with peripheral artery disease: Influence of diabetes and hypertension. Atherosclerosis 2004; 175: 361–367.

44.

Weissler

Zhang

Lippmann

, et al Use of natural language processing to improve identification of patients with peripheral artery disease. Circ Cardiovasc Interv 2020; 13: e009447.

45.

Lareyre

Nasr

Chaudhuri

, et al Comprehensive review of natural language processing (NLP) in vascular surgery. EJVES Vasc Forum 2023; 60: 57–63.

46.

Smolderen

Samaan

Decker

, et al Association between mental health burden, clinical presentation, and outcomes in individuals with symptomatic peripheral artery disease: A scientific statement from the American Heart Association. Circulation 2023; 148: 1511–1528.

47.

Allison

Armstrong

Goodney

, et al Health disparities in peripheral artery disease: A scientific statement from the American Heart Association. Circulation 2023; 148: 286–296.

48.

Smolderen

Alabi

Collins

, et al Advancing peripheral artery disease quality of care and outcomes through patient-reported health status assessment: A scientific statement from the American Heart Association. Circulation 2022; 146: e286–e297.

49.

Creager

Matsushita

Arya

, et al Reducing nontraumatic lower-extremity amputations by 20% by 2030: Time to get to our feet: A policy statement from the American Heart Association. Circulation 2021; 143: e875–e891.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.56 MB