Abstract
Background:
Peripheral artery disease (PAD) guidelines recommend revascularization only for patients with lifestyle-limiting claudication that is refractory to goal-directed medical therapy (class IIA, level of evidence A). However, real-world invasive treatment patterns and predictors of revascularization in patients with symptomatic lower-extremity PAD are still largely unknown.
Aim:
We aimed to examine rates, patient-level predictors, and site variability of early revascularization in patients with new or worsening PAD symptoms.
Methods:
Among patients with new-onset or recent exacerbation of PAD in the 10-center Patient-centered Outcomes Related to TReatment practices in peripheral Arterial disease: Investigating Trajectories (PORTRAIT) study enrolled between June 2011 and September 2015, we classified early revascularization (endovascular or surgical) as procedures being performed within 3 months of presentation. Hierarchical logistic regression was used to identify patient characteristics associated with early revascularization. Variability across sites was estimated using the median odds ratio (OR).
Results:
Among 797 participants, early revascularization procedures were performed in 224 (28.1%). Rutherford class 3 (vs Rutherford class 1; OR=1.86, 95% confidence interval [CI] 1.04–3.33) and having lesions in both iliofemoral and below-the-knee arterial segments (vs below the knee only; OR=1.75, 95% CI: 1.15–2.67) were associated with a higher odds of revascularization. Longer PAD duration >12 months (vs 1–6 months; OR=0.50, 95% CI: 0.32–0.77), higher ankle-brachial index scores (per 0.1 unit increase; OR=0.86, 95% CI: 0.78–0.96), and higher Peripheral Artery Questionnaire Summary scores (per 10 unit increase; OR=0.89, 95% CI: 0.80–0.99) were associated with a lower odds of revascularization. The raw rates for revascularization in different sites ranged from 6.25% to 66.28%, and the median OR was 1.88, 95% CI: 1.38–3.57.
Conclusions:
About 1 in 3 patients with symptomatic PAD received early revascularization. A more extensive disease and symptom burden were the main predictors of receiving early revascularization in PAD. There was significant site variability in revascularization patterns, and further studies will better understand the source of this variability and optimal selection criteria for early revascularization.
Clinical Impact
Real world patterns and predictors of early revascularization in peripheral artery disease are not well understood. In this retrospective analysis of the POTRAIT study, about 1 out of 3 patients with PAD symptoms received early revascularization, with significant site variability. A more extensive disease and symptom burden were the main predictors of receiving early revascularization in PAD.
Keywords
Introduction
Lower-extremity peripheral artery disease (PAD) affects approximately 8.5 million Americans and 202 million people worldwide.1,2 Lower-extremity pain, limited mobility, and overall diminished quality of life (QoL) result from PAD.3,4 Treatment goals in patients with symptomatic PAD are directed toward minimizing patients’ cardiovascular risk and improving their functional status and QoL. 5 To improve survival and prevent disease progression, guidelines recommend antiplatelet and high-intensity statin therapy, smoking cessation, and treatment of cardiovascular risk factors in all patients with PAD.5–9
For PAD symptom relief, supervised exercise therapy (SET) is recommended to improve patients’ functional status and, to a lesser extent, to reduce the need for claudication medications. 10 According to the American College of Cardiology/American Heart Association guidelines, invasive revascularization should be considered when PAD symptoms are lifestyle-limiting and refractory to medical management and SET.5,11 Despite the guidelines, revascularization patterns appear to differ markedly across centers, 12 mandating a better understanding of which patient or center characteristics are most associated with revascularization treatment. 13
To better understand real-world treatment patterns, we studied the use of early (within 3 months) revascularization patterns in a multicenter cohort study of patients presenting with new or worsening PAD symptoms to vascular specialty care centers in the United States. Specifically, we sought to illuminate patient-level predictors of early revascularization and to adjust for these to better examine site-level variability. Addressing these gaps in knowledge can support both medical decision-making and efforts to improve the quality and consistency of care.
Methods
Study Design and Population
Details about the Patient-centered Outcomes Related to TReatment practices in peripheral Arterial disease: Investigating Trajectories (PORTRAIT) study have been described previously. 14 Briefly, PORTRAIT enrolled 1275 patients with an abnormal ankle-brachial index (ABI) score of <0.90 with new or worsened claudication symptoms from 16 PAD-specialty clinics in the Unites States, the Netherlands, and Australia between June 2, 2011, and December 3, 2015. Patients with noncompressible ABI scores (≥1.3) and critical limb ischemia were excluded. Patients were excluded if they had a prior lower-limb revascularization procedure (atherectomy, endarterectomy, bypass surgery, angioplasty) in the same leg where they were currently having symptoms. For the present study, we focused on the US sites (n=797) to illuminate care within a single health care system. From them, eight were academic centers, one was academic affiliated, and another one did not have an official affiliation with an academic institution. All study participants provided either written or telephonic informed consent. Consistent with the design of the study, patients’ primary treatment regimens up to 3 months were abstracted from the medical records. The study protocol was approved by the institutional review boards at all participating sites.
Data Collection and Definitions
Early revascularization was defined as endovascular or surgical treatment within 3 months of presentation. Endovascular treatment was defined as percutaneous angioplasty with or without stent placement or atherectomy. Surgical treatment was defined as bypass surgery or endarterectomy. All revascularization procedures were adjudicated. 15 Information about demographics, cardiovascular risk factors, psychosocial and socioeconomic characteristics, preferences for shared decision-making, PAD symptoms, and health status were obtained before treatment. Data on demographic, socioeconomic (insurance, marital status, finances, employment, cost of care), symptom status (new onset vs exacerbation, severity of claudication, duration of PAD), and lifestyle (activity level, smoking status) factors were collected through patient interviews at the initial visit. Other medical comorbidities and medications were obtained from the medical record by trained study personnel.
Health Status Assessment
Disease-specific health status was assessed using the Peripheral Artery Questionnaire (PAQ).16–18 The PAQ is a 20-item validated, multidimensional, PAD-specific health status instrument that measures the following health domains: physical limitation, symptoms, symptom stability, social limitation, treatment satisfaction, and QoL. Scores range from 0 to 100, with higher scores correlating with a better health status. A summary score, which also ranges from 0 to 100, is calculated by averaging PAQ physical limitation, symptom, social limitation, and QoL subscales. Trained study personnel administered the PAQ in person at the initial visit and by telephone at 3, 6, and 12 months of follow-up. Depressive symptoms at baseline were assessed with the 8-item Patient Health Questionnaire (PHQ-8) depression scale, a screening instrument for major depression. The PHQ-8 scores range from 0 to 24, with higher scores indicating greater severity of depressive symptoms. A score of 10 or higher implies clinically relevant symptoms that warrant further assessment. Patients with depression were not excluded. Psychosocial factors were also assessed through the ENRICHD Social Support Inventory, which is a 7-item measure, used in clinical trials, that assesses the 4 defining attributes of social support: emotional, instrumental, informational, and appraisal. 19
Statistical Analysis
Baseline patient characteristics were compared by revascularization status. Continuous and categorical variables were compared using Students t tests and χ2 tests, respectively. To determine predictors of revascularization, we performed hierarchical logistic regression, with a random intercept for site. Predictors were chosen based on prior literature review and clinical importance. They included demographic and socioeconomic characteristics, cardiovascular risk factors, lifestyle, medical history, PAD severity, history of diagnostic test for PAD, prior medical and invasive treatment for PAD, health status, and psychosocial factors. We also examined variation in the rate of revascularization therapy across study sites in the United States with the use of the median odds ratio (MORs). The MOR estimates the median difference in the odds of a patient undergoing endovascular or surgical treatment at 1 random hospital as compared with another, after adjusting for site and patient factors.
Baseline covariate data were nearly complete, with 98.3% of patients missing 1 or fewer covariates. The covariate with the most missing data was “duration of pain,” which was missing in 7.8% of patients, followed by PHQ-8 Depression score missing in 2.8% of patients. Data were assumed to be missing at random and were imputed using IVEWARE, which computed a single imputation model that contained all variables from the multivariable model. 20 All statistical analyses were performed using SAS 9.4 (SAS Institute, Inc, Cary, NC). A 2-tailed alpha level of 0.05 was used to determine statistical significance.
Results
Baseline Characteristics
Our analytic cohort consisted of 797 patients who were treated in the United States (Figure 1). Early revascularization procedures were performed in 224 patients (28.1%) versus 573 (71.9%) patients who were managed medically (Table 1). The median age in the entire cohort was 69 (interquartile range: 62–75 years) and did not differ by treatment strategy. African American patients were less often treated with early revascularization (16.5% vs 25.5%, p=0.025). Marriage status, insurance status, and sex were similar between the two groups (p value nonsignificant). Patients with Rutherford 1 symptoms (14% vs 22.3%, p<0.001) and Rutherford 2 symptoms (37.8% vs 56.4%, p<0.001) were less likely to receive revascularization than patients who had Rutherford 3 symptoms (48.2% vs 21.3%, p<0.001). Patients in the revascularization group were more likely to have a history of prior endovascular or surgical revascularization (42.4% vs 32.1%, p=0.006).
Flow diagram showing the selection of the included patients.
Baseline Characteristics of the Study Population Presented by the Receipt of Invasive Treatment at 3 Months vs. Medical Management Only.
Abbreviations: ACE, angiotensin-converting enzyme; BL, Baseline; ESSI, ENRICHD social support inventory; IQR, interquartile range; PAQ, Peripheral Artery Questionnaire; PHQ-8, Patient Health Questionnaire-8.
The median ABI score (0.63 vs 0.70, p<0.001) and the PAD-specific health status (a lower PAQ Summary score) (43.1 vs 49.3, p<0.001) were lower in the revascularization group. In total, 24.1% patients had iliofemoral disease only (21.8% in the revascularization group vs 24.9% in the medical management group); 40.1% had below-the-knee (infrapopliteal) lesions, with lower rates in the revascularization group (32.3% vs 43.2%, p<0.001); and 35.8% of the patients had both iliofemoral and infrapopliteal lesions, with higher rates in the revascularization group (45.9% vs 31.9%, p<0.001). Similarly, early revascularization was more common in patients with bilateral lower-extremity claudication symptoms versus patients with unilateral-only symptoms, while chronic kidney disease was more common in the medical management group.
Predictors of revascularization
Independent predictors for revascularization in the adjusted hierarchical model were severe claudication (Rutherford class 3) versus mild claudication (Rutherford class 1; OR=1.86, [95% CI: 1.04–3.33]) and lesion location in both iliofemoral and infrapopliteal (vs infrapopliteal only) with an OR of 1.75 (95% CI: 1.15–2.67). On the other hand, PAD symptoms duration >12 months (vs 1–6 months; OR=0.5 [95% CI: 0.32–0.77]), ABI (per 0.1 unit increase; OR=0.86 [95% CI: 0.78–0.96]), and higher PAQ Summary scores (per 10 unit increase; OR=0.89 [95% CI: 0.80–0.99]) were less often treated with early revascularization (Figure 2).
Forest plot for the fully adjusted hierarchical logistic regression model with parameter estimates presented as odds ratios and 95% confidence intervals. PAD, peripheral artery disease; CAD, coronary artery disease; CVA, cerebrovascular accident; ABI, ankle-brachial index; PAQ, Peripheral Artery Questionnaire; PHQ, Patient Health Questionnaire; ACE, angiotensin-converting enzyme; ESSI, ENRICHD social support inventory.
The raw rates of revascularization in different sites ranged from 6.25% to 66.28%. The adjusted rates (model adjusted for sites with small sample sizes) range from 11.0% to 63.7%. The adjusted MOR for site variability was 1.88 (95% CI: 1.38–3.57), indicating a large amount of variability in treatment decisions across sites in the United States.
Discussion
Understanding patient-level factors, and the variability across sites, with early use of revascularization in patients with PAD is critical to improve care for this population. Using rigorously collected prospective data reflecting a real-world practice pattern, our study provides significant insights into revascularization patterns for PAD. Patients with more advanced Rutherford class and those with diffuse disease in both proximal and distal vessels were more likely to undergo revascularization in our adjusted analysis, whereas patients with a longer history of PAD, higher ABIs, and better baseline health status (higher PAQ Summary scores) were less likely to undergo revascularization. There was also marked variation in the use of revascularization therapy across sites, with an average 88% likelihood that statistically identical patients would undergo revascularization at one site as compared with another in the PORTRAIT study.
In patients with lifestyle-limiting PAD, revascularization therapy may be recommended based on factors such as inadequate response to medical treatment or exercise therapy, anatomical considerations, and importantly patient preferences. 11 The primary goal of revascularization in PAD is to improve symptoms and QoL. 5 Using the PAD questionnaire in our analysis aimed—among others—to capture frequently unmeasured variables such as PAD-related functioning and QoL. Revascularization treatment entails higher upfront risk and is expensive but can potentially offer faster symptom relief. After accounting for patient preferences, other important factors to consider include limited response to medical and exercise therapy, status of comorbid conditions, favorable risk-benefit profile such as vascular anatomy, and likelihood of achieving durable symptom relief. 5 Despite adjusting for these factors, however, we still found marked variability in treatment patterns across sites.
Although our study documents the overall variation in revascularization across sites, further studies should examine how much of this variation is due to reasonable patients’ preferences and expectations (i.e., appropriate variation in care) versus variations that might be related to physicians’ preferences, compensation, and patient demographics or other characteristics.21–23 The fact that we found a relatively higher African American representation in the medically managed group versus the early revascularization might explain in part the variability that is noticed across the United States in treatment patterns in PAD. However, we were not able to find a difference in the adjusted analysis. Larger studies should confirm this hypothesis. Furthermore, lifestyle-limiting PAD is defined purely by patients and not by any testing or laboratory work. It is thus expected to have some variation in revascularization procedures across sites because of patient preferences.
Future studies are needed to understand if the threshold—in terms of patient symptoms—leading to revascularization is different across treatment sites and what is the main driver of the different threshold. 24 In our analysis, 37.8% of patients with Rutherford 2 underwent revascularization, which could be considered premature. Potential explanations include patient preferences, practice variations which may or may not be linked to financial incentives, PAD provider variability, selection biases related with sociodemographic factors, and barriers in health care, as well as variability in symptom tolerance across patient communities in different parts of the United States.22,25 Current US guidelines recommend considering invasive treatment after inadequate response to exercise therapy if anatomy allows and after taking into account patient’s treatment preference. 5 Barriers to the use of SET in the United States remain, and it is unclear whether the further uptake of SET will change the revascularization rates for PAD. 26 Future studies should examine whether the predictors that we found can be generalized outside of the United States and in larger samples. Developing data-driven risk-benefit scenarios tailored to patients’ characteristics and the outcomes that they value, including their health status and functioning, as well as long-term durability of patency results are needed to understand which patients would benefit the most from invasive treatment and can help patients and clinicians personalize treatment approaches in PAD.
Limitations
Our findings should be interpreted in the context of several potential limitations. First, the sample size was modest, and only 10 US centers were included. A larger sample size may have identified additional factors associated with early revascularization, and more sites might provide a better description of the variability in care across the United States. Including sites outside of the United States would probably enhance the generalizability of our findings and offer opportunities for comparisons between different health care systems, but we decided to focus only in the US sites to have a more homogeneous patient population and health care network. Besides, we had adjudicated data only for the US centers for invasive treatment. Sites participating in the current registry could be highly motivated and may have higher rates of guideline-concordant practice than sites not participating in such a registry. Given the fact that PAD is a chronic disease not defined by a single episode of symptoms, our cohort was mixed with patients having “de novo” PAD, and others an exacerbation of prior PAD, and even though we included prior medical therapy and history of prior revascularization in our models, it is possible that different selection mechanisms may have played a role in those presenting with newly detected PAD versus those with exacerbations. Finally, we stratified the participants based on whether they received early revascularization, and alternative definitions may have impacted our findings. Nevertheless, we would argue that any revascularization within 3 months is consistent with the primary planned treatment strategy for patients with stable PAD.
Conclusions
Among patients presenting for care with new onset or a recent exacerbation of PAD, almost 28% received early revascularization. Severe claudication and its more extensive forms (concomitant multilevel with iliofemoral and infrapopliteal disease) were associated with higher rates of revascularization, whereas higher ABI, longer symptom duration, and better initial health status were associated with lower rates of revascularization. There was significant site variability in revascularization patterns. Future studies should investigate the source of this variability and help us better understand which are the drivers of different revascularization patterns among sites both in the US population and also in other populations.
Footnotes
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: The views in this work are solely the responsibility of the authors and do not necessarily represent the views of the Patient-Centered Outcomes Research Institute, its Board of Governors or Methodology Committee. J.A.S. co-owns the copyright to the Peripheral Artery Questionnaire. He provides consultative services to Janssen, Bristol Meyers Squibb, Novartis, Myokardia, Bayer, Merck, and Pfizer and has research grants from Myokardia, Janssen, and Abbott Vascular. C. M.-H. is a consultant for Abbott, COOK, Cardinal Health, and Optum Labs. K.G.S. receives unrestricted grants from Cardiva, Johnson&Johnson, Abbott, Merck, and Cardinal Health and is also a consultant for OptumLabs and Abbott. The other authors have no conflicts to disclose.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
