Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2024 Mar 1.
Published in final edited form as: J Vasc Surg. 2022 Oct 28;77(3):836–847.e3. doi: 10.1016/j.jvs.2022.10.025

Early Peripheral Vascular Interventions for Claudication are Associated with Higher Rates of Late Interventions and Progression to CLTI

Rebecca Sorber 1, Chen Dun 2, Qingwen Kawaji 3, Christopher J Abularrage 1, James H Black III 1, Martin A Makary 4, Caitlin W Hicks 1,2
PMCID: PMC10242207  NIHMSID: NIHMS1854056  PMID: 37276171

Abstract

Objectives

Despite societal guidelines that peripheral vascular interventions (PVI) should not be first line therapy for intermittent claudication, a significant number of patients undergo PVI for claudication within six months of diagnosis. The aim of this study was to investigate the association of early PVI for claudication with subsequent interventions.

Methods

100% Medicare fee-for-service claims were used to identify all beneficiaries newly diagnosed with claudication between 1/1/2015–12/31/2017. The primary outcome was late intervention, defined as any femoropopliteal PVI performed >6 months after claudication diagnosis (through 6/30/2021). Kaplan-Meier curves were used to compare the cumulative incidence of late PVI for claudication patients receiving early (≤6 months) PVI vs. those without early PVI. A hierarchical Cox proportional hazards model was used to evaluate patient and physician-level characteristics associated with late interventions.

Results

A total of 187,442 patients were newly diagnosed with claudication in the study period, of which 3.2% (6,069) underwent early PVI. After a median follow-up of 4.39 years (IQR 3.62, 5.17), 22.5% of early PVI patients underwent late PVI vs. 3.6% of patients who did not receive early PVI (P<0.001). Patients treated by high physician users of early PVI (≥2 SD, physician outliers) were more likely to receive late PVI than patients treated by standard physician users (9.8% vs. 3.9%, P<0.001). Patients undergoing early PVI (16.4% vs. 7.8%) and patients treated by outlier physicians (9.7% vs. 8.0%) were more likely to convert to CLTI (both P< 0.001). After adjustment, patient factors associated with late PVI included receipt of early PVI (aHR 6.89, 95%CI:6.42–7.40) and Black race (vs. White, aHR 1.19, 95%CI:1.10–1.30). The only physician factor associated with late PVI was majority practice in an ASC/OBL, where increasing proportion of ASC/OBL services was associated with significantly increased rates of late PVI (Q4 vs Q1, aHR 1.57, 95%CI:1.41–1.75).

Conclusions

Early PVI following diagnosis of claudication is associated with higher late PVI rates compared to patients treated with early nonoperative management. High physician users of early PVI for claudication perform more late PVI than their peers, particularly those primarily delivering care in high reimbursement settings. The appropriateness of early PVI for claudication needs critical evaluation, as do incentives surrounding delivery of these interventions in ambulatory intervention suites.

Keywords: Claudication, CLTI, progression, peripheral vascular interventions, ambulatory surgery center, office-based lab

Table of Contents Summary:

In this review of 187,442 Medicare beneficiaries newly diagnosed with claudication, patients who were managed with peripheral vascular intervention in the first six months following diagnosis were more likely to undergo additional PVI, convert to CLTI and undergo amputation compared to those treated with early medical management alone. The authors suggest that appropriateness of early PVI for claudication needs critical evaluation.

Introduction

Intermittent claudication is the most common clinical manifestation of peripheral artery disease (PAD)1,2. While it is estimated that less than 15% of patients with PAD-associated claudication progress to chronic limb threatening ischemia (CLTI) within 5 years3–5, claudication symptoms may be severe and lifestyle-limiting. Multiple societal guidelines clearly state that peripheral vascular interventions (PVI) should not be first line therapy for intermittent claudication5,6. However, a substantial number of patients undergo PVI for claudication within six months of their initial diagnosis7.

It has been previously demonstrated that the cumulative burden of interventions for claudication is associated with an increased rate of progression to CLTI3 and major amputation8. Despite this, there is a growing body of published data suggesting that early PVI for claudication is increasingly widespread7,9–11. The expansion of PVI may be driven by physician-level factors12–14 and has been connected to its application in high reimbursement settings such as ambulatory surgery centers (ASCs) and office-based laboratories (OBLs)15,16. Taken together, these data raise concerns that patients who receive early PVI for claudication may initiate a domino-like effect of additional open and endovascular interventions and, consequently, worse limb and functional outcomes8.

The aim of this study was to investigate the association of early PVI for claudication with cumulative subsequent interventions and progression to CLTI among Medicare beneficiaries. Additionally, physician-level data was queried to examine the relationship between individual physician utilization of early PVI and subsequent interventions and limb outcomes.

Methods

Study Design

This was a retrospective cross-sectional study utilizing 100% Medicare fee-for-service claims to identify all Medicare beneficiaries who were newly diagnosed with claudication between January 1, 2015 and December 31, 2017 using International Classification of Diseases (ICD) 9 and 10 codes (Supplementary Table 1). Patients who had a prior PVI (N=11,953), any diagnosis code corresponding to CLTI before or within 6 months after the first claudication diagnosis (N=1,402,572), without 12 months continuous enrollment before diagnosis of claudication (N=38,271), died or lost to follow-up within 6 months after the diagnosis date (N=7,540), or missing demographic information (N=90) were excluded. We selected 6 months as the time period in which one could be expected to see results from supervised exercise therapy17 or other forms of medical management. The Institutional Review Board at the Johns Hopkins University School of Medicine approved this study and waived informed consent requirements given that this was a retrospective analysis of claims data.

Main Exposures of Interest

Patients were stratified by whether or not they received a femoropopliteal PVI ≤6 months of receiving their new diagnosis of claudication (termed “early PVI”). PVI was defined as any peripheral vascular intervention, including angioplasty, stenting or atherectomy, as defined by Current Procedural Technology (CPT) codes (Supplementary Table 2).

As a secondary analysis, physician-level data was also utilized to stratify patients by whether or not they received care from a high vs. standard physician user of early PVI for claudication. High physician users of early PVI for claudication were defined as physicians whose rate of early PVI was ≥2 standard deviations above the mean, as previously described7. All other physicians were defined as standard physician utilizers of PVI for claudication.

Study Outcomes

The primary outcome of this study was late PVI intervention, defined as any PVI performed >6 months from the patient’s diagnosis of claudication (in patients with no early PVI), or >6 months after the initial PVI for claudication (in patients with early PVI). Late PVI interventions included both femoropopliteal and tibial PVI, and were identified using Current Procedural Terminology (CPT) codes (Supplementary Table 2). Late interventions were accrued in a cumulative fashion until 6/30/21, but only the first late PVI intervention was counted for the purposes of the analysis. Secondary outcomes included receipt of an open lower extremity bypass, lower extremity major amputation, and subsequent diagnosis of CLTI (Supplementary Table 3). CLTI was defined as atherosclerosis with rest pain, ulcer or gangrene based on ICD-10 codes; diagnoses defining acute limb ischemia or aneurysm were excluded (Supplementary Table 1).

Patient and Physician-level Covariates

Patient demographic data was derived from the Medicare Master Beneficiary Summary File18. State, census region, and population density of the patient residence were defined using Federal Information Processing Standard (FIPS) Crosswalk. Patient comorbidities were defined from claims data with ICD-9 and 10 codes within the 12 months preceding their claudication diagnosis (Supplementary Table 3). Comorbidities were assigned based one diagnosis from inpatient claims or at least two diagnoses recorded >30 days apart from outpatient or carrier claims. Physician level data was assigned to individual providers based on their National Provider Identification (NPI) number. Only physicians who treated more than 10 new claudication patients during the baseline study period (2015–2017) were included in the analysis, in accordance with our data use agreement. Physician-level characteristics including sex, years since medical school graduation, specialty, region of practice, population density of practice, and number of patients diagnosed with claudication during the study period were obtained from the Medicare Data on Provider Practice and Specialty and the Medicare Physician National Downloadable File. Each physician’s percentage of services delivered in an ASC or OBL were estimated from summary statistics reported in the Medicare Data on Provider Practice and Specialty19.

The Medicare Data on Provider Practice and Specialty (MD-PPAS) file included percentage of line items delivered in different place of service (including office, retail clinics, urgent care centers, inpatient hospital, hospital outpatient department, emergency room, nursing facility or skilled nursing facility, ambulatory surgery center (ASC), patient’s residence, and other places of service) and count of line items billed by the provider. We calculated the proportion of services delivered in an ASC or OBL using providers’ percentage of line items delivered in office and ambulatory surgery center multiple by the number of line items then divided by the number of line items to get each physician’s percentage of services delivered in an ASC or OBL.

Statistical Analysis

Descriptive statistics were used to describe patient demographics and comorbidities as well as physician practice characteristics with χ2 and Student’s t tests used for discrete and continuous variables, respectively. Kaplan-Meier curves and log-rank tests were used to compare the cumulative incidence of late PVI interventions for claudication patients 1) receiving early (≤6 months) PVI vs. those without early PVI (patient-level analysis); and 2) receiving care from high physician users of early PVI for claudication vs. standard physician users of early PVI for claudication (physician-level analysis). A hierarchical Cox proportional hazards model was used to evaluate patient (first-level covariates) and physician (second level-covariates)-level characteristics associated with late PVI interventions. The model was adjusted for patient factors in level 1 (age, sex, race, comorbidities, population density and census region of residence, receipt of early PVI) and physician factors in level 2 (physician sex, years since medical school graduation, population density and census region of practice, primary specialty, volume of patients diagnosed with claudication and percentage of services delivered in an ASC or OBL).. We also assessed the association of early PVI for claudication with subsequent major amputation and conversion to CLTI.

Notably, intervention laterality is coded in the Medicare database, but there is a large amount of missing and inaccurate data around this. In addition, laterality is not commonly coded for claudication diagnoses. We therefore performed a sensitivity analysis limited to patients who had early or late intervention laterality coded at the time of their index intervention for claudication, as well as for their follow-up reinterventions. Cumulative ipsilateral reintervention rates were compared for patients who underwent early versus late PVI for claudication, and for high versus standard physician users of early PVI using Kaplan-Meier curves and log-rank tests.

All statistical analyses were performed using SAS Enterprise version 7.1 (SAS Institute, Cary, North Carolina). All values were two-tailed and reported as statistically significant at a level of α≤0.05.

Results

Study Cohort

Overall, 187,442 Medicare patients received a new diagnosis of claudication during the baseline study period (Table 1). Of these, 6,069 (3.2%) underwent early PVI and 181,373 (96.8%) did not undergo early PVI. Median age for the cohort was 71.5 years, with the early PVI group slightly younger than the no early PVI group (70.4 years vs 71.5 years, P<0.001). A smaller proportion of female patients (38.3% early PVI vs. 43.6% no early PVI; P<0.001) and non-White patients (16.4% vs. 18.4%; P<0.001) underwent early PVI. More patients undergoing early PVI were ever smokers (61.9% vs. 43.0%; P<0.001) and had hypertension (89.7% vs. 87.5%; P<0.001). Early PVI was also more prevalent among patients with a rural domicile (25.4% vs. 20.3%; P<0.001) and among those who lived in the Midwest (22.6% vs 20.7%) and South (50.9% vs 48.3%) (both, P<0.001).

Table 1.

Patient demographics and comorbidities stratified by receipt of early PVI

Patient Characteristics All Patients (N=187,442) Early PVI (N=6,069) No Early PVI (N=181,373) P value
Age (years) Median (IQR) 71.47 (67.23, 77.79) 70.41 (66.86, 76.13) 71.51 (67.25, 77.84) <0.001
 ≤64 24586 (13.12) 819 (13.49) 23767 (13.10)
 65–74 97473 (52.00) 3505 (57.75) 93968 (51.81)
 75–84 51246 (27.34) 1411 (23.25) 49835 (27.48)
 85–94 13643 (7.28) 329 (5.42) 13314 (7.34)
 ≥95 494 (0.26) 5 (0.08) 489 (0.27)
Sex <0.001
 Female 81313 (43.38) 2322 (38.26) 78991 (43.55)
 Male 106129 (56.62) 3747 (61.74) 102382 (56.45)
Race <0.001
 White 153086 (81.67) 5064 (83.44) 148022 (81.61)
 Black 19891 (10.61) 656 (10.81) 19235 (10.61)
 Asian 3508 (1.87) 63 (1.04) 3445 (1.90)
 Hispanic 4449 (2.37) 109 (1.80) 4340 (2.39)
 North America Native 939 (0.50) 32 (0.53) 907 (0.50)
 Other/unknown 5569 (2.97) 145 (2.30) 5424 (2.99)
Comorbidities
 ESRD 5644 (3.01) 163 (2.69) 5481 (3.02) 0.13
 Diabetes 75781 (40.43) 2454 (40.43) 73327 (40.43) 0.99
 Hypertension 164093 (87.54) 5442 (89.67) 158651 (87.47) <0.001
 Smoking 81755 (43.62) 3758 (61.92) 77997 (43.00) <0.001
Population density of residence <0.001
 Urban 149075 (79.53) 4526 (74.58) 144549 (79.70)
 Rural 38367 (20.47) 1543 (25.42) 36824 (20.30)
Census region of residence <0.001
 Midwest 38976 (20.79) 1371 (22.59) 37605 (20.73)
 North East 30640 (16.35) 658 (10.84) 29982 (16.53)
 South 90640 (43.36) 3089 (50.90) 87551 (48.27)
 West 26791 (14.29) 944 (15.55) 25847 (14.25)
 Other 395 (0.21) 7 (0.12) 388 (0.21)
ABI within 3 months of claudication diagnosis <0.001
 No 99791 (53.24) 2935 (48.36) 96856 (53.40)
 Yes 87651 (46.76) 3134 (51.64) 84517 (46.60)
Open in a new tab

IQR, interquartile range; ESRD, end stage renal disease

Physician practice patterns

When the cohort was stratified by physician practice patterns (Table 2), 6,812 patients with newly diagnosed claudication (4.3%) were treated by physicians who performed high rates of early PVI for claudication. There were 5,459 physicians who diagnosed >10 patients with claudication and were included in this analysis. The mean (SD) early PVI rate was 3.5% (5.4%) for the entire physician group, 19.5% (5.9%) for high use group (n=318), and 2.5% (3.5%) for standard use group (n=5141). High use physicians treated fewer women (42.1% vs. 43.9%; P=0.004) and non-White patients (17.2% vs. 18.3%; P<0.001) with claudication compared to standard physician users. High use physicians treated significantly more patients with a smoking history (49.0% vs. 43.0%; P<0.001) and were more likely to practice in a rural area (25.3% vs. 20.5%; P<0.001). High physician users of early PVI for claudication were less likely to have a high-volume claudication practice (P<0.001) but were more likely to deliver their services in an ASC or OBL setting (P<0.001; Table 2).

Table 2.

Patient demographics and comorbidities stratified by treatment by standard and high-use physicians

Patient Characteristics Treated by Standard-Use Physician (N=153,339) Treated by High-Use Physician (N=6,812) P value
Age (years)
 median, IQR 71.43 (67.23, 77.70) 71.09 (67.08, 77.45) 0.18
 ≤64 20041 (13.07) 944 (13.86)
 65–74 80203 (52.30) 3542 (52.00)
 75–84 41831 (27.28) 1856 (27.25)
 85–94 10862 (7.08) 458 (6.72)
 ≥95 402 (0.26) 12 (0.18)
Sex 0.004
 Female 67343 (43.92) 2871 (42.15)
 Male 85996 (56.08) 3941 (57.85)
Race <0.001
 White 125225 (81.67) 5639 (82.78)
 Black 16230 (10.58) 773 (11.35)
 Asian 2908 (1.90) 76 (1.12)
 Hispanic 3633 (2.37) 145 (2.13)
 North American Native 758 (0.49) 31 (0.46)
 Other/unknown 4585 (2.99) 148 (2.17)
Comorbidities
 ESRD 4151 (2.71) 234 (3.44) <0.001
 Diabetes 62139 (40.52) 2723 (39.97) 0.37
 Hypertension 134411 (87.66) 5967 (87.60) 0.88
 Smoking 65901 (42.98) 3340 (49.03) <0.001
Population density of residence <0.001
 Urban 121849 (79.46) 5089 (74.71)
 Rural 31490 (20.54) 1723 (25.29)
Census region of residence <0.001
 Midwest 30850 (20.12) 1779 (26.12)
 North East 25345 (16.53) 529 (7.77)
 South 76805 (50.09) 3219 (47.25)
 West 20092 (13.10) 1283 (18.83)
 Other 247 (0.16) 2 (0.03)
No. of patients diagnosed with claudication for the first time during study period <0.001
 <=17 24883 (16.23) 2101 (30.84)
 18–28 35239 (22.98) 2480 (36.41)
 >=29 93217 (60.79) 2231 (32.75)
Percentage of services delivered in ASC or OBL <0.001
 0%–24% 34480 (22.49) 1047 (15.37)
 25%–51% 36798 (24.00) 1846 (27.10)
 52%–74% 41284 (26.92) 2401 (35.25)
 75%–100% 40777 (26.59) 1518 (22.28)
Open in a new tab

IQR, interquartile range; ESRD, end stage renal disease; ASC, ambulatory surgery center; OBL, office based laboratory

Outcomes by early PVI status

Median follow up time for the cohort was 4.4 years (95% CI 3.7, 5.2 years) (Table 3). For those patients who received an early PVI for claudication, the time to late PVI intervention was significantly shorter (346 days vs 554 days) and the mean number of late PVI interventions was significantly higher (2.9 vs 2.5) than for patients who did not receive an early PVI (both, P<0.001). Kaplan Meier analysis (Figure 1) demonstrated a significant difference in cumulative late PVI interventions over time (P<0.001). The 5-year cumulative late PVI intervention rate was 23.5% (95% CI 21.9%, 25.0%) for patients who received early PVI, compared to 3.8% (95% CI 3.7%, 4.0%) for patients who did not receive an early PVI. Patients undergoing early PVI had higher rates of subsequent open surgical bypass (2.8% vs 0.5%: P<0.001) and major amputation (1.0% vs. 0.4%; P<0.001). The estimated 5-year major amputation rate was 1.01% (95% CI 0.75%, 1.37%) for patients who received early PVI, compared to 0.40% (95% CI 0.035%, 0.49%) for patients who did not receive an early PVI (P<0.001). Patients undergoing early PVI also had higher rates of conversion to CLTI compared to patients not receiving early PVI (16.4% vs. 7.8%; P<0.001).

Table 3.

Unadjusted limb outcomes stratified by receipt of early PVI

All Patients (N=187,442) Early PVI (N=6,069) No Early PVI (N=181,373) P-value
Mean follow-up time (days±SD) 1621.54±336.81 1598.29±327.86 1622.31±337.08 <0.001
Median follow-up time (days, IQR) 1624 (1351, 1900) 1589 (1343, 1855) 1625 (1352, 1902) <0.001
Time to first PVI reintervention (days), median (IQR) 510 (197, 969) 346 (130, 743) 554 (221, 1013) <0.001
Mean number of PVI reinterventions (±SD) 2.55±2.79 2.85±3.81 2.49±2.51 0.001
Receipt of open bypass (N, %) 1065 (0.57) 168 (2.77) 897 (0.49) 0.001
Major amputation (N, %) 728 (8.42) 59 (0.97) 669 (0.37) <0.001
Conversion to CLTI (N, %) 15195 (8.11) 992 (16.35) 14203 (7.83) <0.001
Open in a new tab

IQR, interquartile range; SD, standard deviation; CLTI, chronic limb threatening ischemia

Figure 1.

Figure 1.

Open in a new tab

Kaplan-Meier curve depicting cumulative incidence of late PVI stratified by whether or not patients received early PVI for claudication (defined as within first 6 months of diagnosis)

Outcomes by physician practice pattern status

Significantly more patients receiving treatment from high physician users of early PVI for claudication underwent late PVI intervention compared to patients receiving treatment from standard physician users (Figure 2). The 5-year cumulative late PVI intervention rate was 10.4% (95% CI 9.4%, 11.5%) for patients who were treated by high physician users, compared to 4.2% (95% CI 4.0%, 4.4%) for patients treated by standard physician users. The median time to first late PVI intervention was significantly shorter (378 days vs. 523 days; P<0.001) and the mean number of late PVI interventions was significantly higher (2.9±2.8 vs 2.5±2.9; P<0.001) for those patients treated by high physician users of early PVI). Patients undergoing treatment by high physician users of early PVI for claudication also had higher rates of conversion to CLTI compared to patients receiving treatment from standard user physicians (9.7% vs. 8.0%; P<0.001).

Figure 2.

Figure 2.

Open in a new tab

Kaplan-Meier curve depicting cumulative incidence of late PVI stratified by physician utilization of early PVI for claudication, where high denotes 4th quartile physician utilizers and standard denotes 1st-3rd quartile physician users

Sensitivity analysis

A sensitivity analysis limited to patients who had intervention laterality coded at the time of their index early or late PVI and any reintervention. The overall findings were similar to the main analysis. Patients who underwent early PVI for claudication had a higher incidence of ipsilateral reintervention compared with the late PVI group (Supplementary Figure 1A). The incidence of ipsilateral reintervention was also persistently higher among high vs. standard physician users of early PVI for claudication (Supplementary Figure 1B).

Patient and physician factors associated with late PVI interventions

A hierarchical Cox proportional hazards model was used to investigate patient- and physician-level factors associated with receiving a late PVI reintervention (Table 4). Receipt of early PVI for claudication was associated with a nearly 7-fold increased odds of receiving a late PVI intervention (aHR 6.89; 95% CI: 6.42,7.40). Female sex was associated with a lower odds of receiving a late PVI intervention (aHR 0.80, 95% CI: 0.76, 0.84), while Black and Hispanic patients had a higher odds of receiving late PVI compared with White patients (aHR 1.19 for both, P<0.05). All comorbidities were associated with a significantly increased odds of receiving a late PVI intervention, with ESRD having the largest magnitude of effect (aHR 2.52; 95% CI: 2.25,2.82). Patients living in the West also had a higher odds of late PVI intervention (aHR 1.47, 95% CI: 1.02, 2.13).

Table 4.

Patient and physican-level predictors of receipt of late PVI

Covariate N=156,372 Unadjusted Hazard Ratio (95% CI) Adjusted Hazard Ratio (95% CI)
Patient-level characteristics
Early PVI for Claudication
 Yes 5,155 (3.30) 7.66 (7.20, 8.15) 6.89 (6.42, 7.40)
 No 151,217 (96.70) Ref Ref
Age (years)
 ≤64 20,528 (13.13) 0.97 (0.90, 1.04) 0.81 (0.75, 0.87)
 65–74 81,838 (52.34) Ref Ref
 75–84 42,568 (27.22) 0.77 (0.73, 0.82) 0.82 (0.77, 0.87)
 85–94 11,033 (7.06) 0.51 (0.45, 0.58) 0.59 (0.52, 0.67)
 ≥95 405 (0.26) 0.38 (0.18, 0.79) 0.54 (0.26, 1.12)
Sex
 Female 68,504 (43.81) 0.73 (0.69, 0.77) 0.80 (0.76, 0.84)
 Male 87,868 (56.19) Ref Ref
Race
 White 127,738 (81.69) Ref Ref
 Black 16,556 (10.59) 1.32 (1.23, 1.42) 1.19 (1.10, 1.30)
 Asian 2,942 (1.88) 0.54 (0.42, 0.69) 0.50 (0.37, 0.67)
 Hispanic 3,709 (2.37) 1.30 (1.12, 1.50) 1.19 (1.01, 1.40)
 North America Native 771 (0.49) 1.40 (1.04, 1.89) 1.30 (0.95, 1.76)
 Other/unknown 4,656 (2.98) 0.97 (0.84, 1.13) 0.89 (0.77, 1.04)
Comorbidities
 ESRD 4,295 (2.75) 2.75 (2.50, 3.03) 2.52 (2.25, 2.82)
 Diabetes 63,355 (40.52) 1.59 (1.51, 1.67) 1.49 (1.42, 1.57)
 Hypertension 137,055 (87.65) 1.45 (1.33, 1.58) 1.17 (1.07, 1.28)
 Smoking 67,599 (43.23) 1.46 (1.39, 1.53) 1.27 (1.20, 1.34)
Population density of residence
 Urban 123,959 (79.27) Ref Ref
 Rural 32,413 (20.73) 1.04 (0.98, 1.11) 1.02 (0.94, 1.11)
Census region of residence
 Midwest 31,977 (20.45) 1.05 (0.99, 1.12) 1.21 (0.98, 1.49)
 North East 25,553 (16.34) 0.81 (0.75, 0.87) 0.87 (0.65, 1.15)
 South 77,769 (49.73) Ref Ref
 West 20,824 (13.32) 1.16 (1.08, 1.24) 1.47 (1.02, 2.13)
 Other 249 (0.16) 0.28 (0.09, 0.87) 0.50 (0.13, 1.96)
Physician-level characteristics
Sex
 Male 150,744 (96.40) Ref Ref
 Female 5,628 (3.60) 0.91 (0.79, 1.04) 1.00 (0.84, 1.21)
Years since medical school graduation
 <10 8,061 (5.16) 0.95 (0.84, 1.07) 1.04 (0.84, 1.21)
 11–20 51,307 (32.81) 1.07 (1.00, 1.14) 1.06 (0.97, 1.16)
 21–30 55,736 (35.64) 0.99 (0.93, 1.05) 0.98, 0.90, 1.08)
 ≥31 41,268 (26.39) Ref Ref
Census region of practice location
 Midwest 31,350 (20.05) 1.04 (0.98, 1.11) 0.96 (0.77, 1.19)
 Northeast 25,380 (16.23) 0.81 (0.75, 0.87) 1.06 (0.80, 1.42)
 South 78,481 (50.19) Ref Ref
 West 20,990 (13.42) 1.13 (1.05, 1.21) 0.80 (0.55, 1.18)
 Other 171 (0.11) 0.14 (0.02, 0.96) 0.30 (0.03, 2.71)
Population density of practice location
 Urban 144,699 (92.54) Ref Ref
 Rural 11,673 (7.46) 0.97 (0.89, 1.07) 0.97 (0.86, 1.10)
Primary specialty
 Vascular surgery 50,066 (32.02) Ref Ref
 Cardiology 81,870 (52.36) 1.04 (0.99, 1.10) 1.01 (0.94, 1.09)
 General surgery 6,586 (4.21) 0.79 (0.65, 0.95) 0.82 (0.70, 0.97)
 Radiology 11,631 (7.44) 0.76 (0.68, 0.85) 0.99 (0.87, 1.13)
 Cardiothoracic surgery 3,464 (2.22) 0.79 (0.65, 0.95) 0.81 (0.67, 0.98)
 Other 2,755 (1.76) 1.04 (0.68, 0.85) 0.95 (0.75, 1.20)
No. of patients diagnosed with new claudication during study period
 ≤17 26,366 (16.86) Ref Ref
 18–28 36,864 (23.57) 1.02 (0.95, 1.10) 1.00 (0.92, 1.08)
 ≥29 93,142 (59.56) 0.98 (0.91, 1.05) 0.96 (0.89, 1.04)
Percentage of services delivered in ASC or OBL
 0%–24% 34,601 (22.13) Ref Ref
 25%–51% 38,127 (24.38) 1.32 (1.22, 1.42) 1.27 (1.16, 1.39)
 52%–74% 42,660 (27.28) 1.43 (1.32, 1.54) 1.38 (1.25, 1.52)
 75%–100% 40,984 (26.21) 1.55 (1.44, 1.67) 1.57 (1.41, 1.75)
Open in a new tab

ESRD, end stage renal disease; ASC, ambulatory surgery center; OBL, office based laboratory

When examining physician-level factors, the only significant covariate associated with late PVI interventions was the proportion of a physician’s PVI practice delivered in an ASC or OBL; there was stepwise increase in aHR with increasing percentage of delivery of services in an ambulatory setting (Q4 vs Q1, aHR 1.57, 95% CI: 1.41–1.75). Physician specialty, sex, years since medical school graduation, practice location, urban density, and volume of patients with claudication diagnosed during the study period were all examined as potential factors, but none of them were significantly associated with late PVI interventions after adjustment.

Discussion

We assessed the association of early (≤6 months) PVI with subsequent late interventions among Medicare beneficiaries treated for a new diagnosis of claudication and followed for nearly 5 years. We found that patients who undergo PVI within the first six months following their diagnosis of claudication have increased subsequent late PVI interventions and, overall, inferior limb outcomes when compared to patients undergoing initial nonoperative management. Additionally, patients receiving care from high physician users of early PVI for claudication had similar patterns of increased cumulative late interventions and inferior limb outcomes. Notably, the only physician characteristic associated with late PVI interventions was percentage of practice in a high-reimbursement setting (ASC/OBL). Taken together, these data raise concern that early application of PVI for claudication not only triggers a cascade of increased subsequent interventions, but also that its application may be driven by incentive structures rather than patient factors.

We found a nearly a seven-fold increase in the odds of receiving a late PVI intervention for patients who received an early PVI for claudication compared to patients who received early conservative management. These data are enlightening, as the comparative efficacy of an intervention vs. no intervention strategy is historically difficult to study using traditional outcomes databases, and data from randomized controlled trials assessing the benefits of endovascular intervention compared to conservative management are largely limited to 1–2 years of follow-up20,21. The EUCLID trial (Examining Use of Ticagrelor In PAD) reported that the risk of acute limb ischemia was 4-fold higher among patients with peripheral artery disease and a history of lower extremity revascularization22 over a median follow-up time of 30 months. A prospective cohort study of 456 patients treated with early PVI versus initial conservative management of intermittent claudication demonstrated a 5-fold increase in major amputation among patients receiving early PVI23. Based on a secondary data analysis, the estimated 5-year incidence of subsequent revascularization was 49% and 32% in patients undergoing early revascularization and initial conservative treatment respectively. The 5-year reintervention rate that we report is quite a bit lower than that prior study, but persistently higher among patients receiving early PVI compared to those who did not.

In addition to having a higher cumulative incidence of late PVI interventions, patients receiving an early PVI had inferior outcomes in nearly every other metric we studied, including a higher mean number of overall interventions, shorter time to late interventions, increased conversion to CLTI, and increased rates of major amputation. These findings are consistent with published data that implicates the cumulative burden of intervention among patients with claudication as a risk factor for progression to CLTI3 and increased major amputation7. Multiple prior research studies have suggested that repeated interventions are associated with a 2 to 3-fold increase in risk of progression to CLTI3,9. Consistent with this notion, early PVI patients had a 2-fold increase in the odds of conversion to CLTI in conjunction with markedly increased rates of late PVI interventions in our study. Overall, ours and prior studies23,24 suggest that the decision to intervene starts a “ticking clock” that ultimately increases a patient’s long-term risk of reinterventions, open surgical revascularization, and major amputation.

The only physician factor associated with late PVI intervention in our study was an increasing proportion of their practice delivered in an ASC or OBL. The clinical significance of this observation is strengthened by the progressive increase in the magnitude of the effect size with increasing proportions of services delivered in an ASC/OBL. Even after risk adjustment, patients treated by physicians who deliver more than 75% of their services in an ASC/OBL have a 157% increased chance of requiring a late PVI intervention compared to a patient treated by a physician who delivers the majority of their care in a hospital-based setting. It should be noted that both standard and high-use physician groups delivered a large proportion of their services in an ASC/OBL, with more that 50% of physicians in both groups delivering more than half of their services in such a setting. It is possible that patients being treated in ambulatory settings have more severe disease or more severe symptoms, which we cannot quantify using claims data. However, the concept that physician factors play a role in the data we report is supported by our finding that high physician users of early PVI for claudication have higher rates of subsequent late PVI interventions as well.

The findings we report are similar to those reported in our index report on overapplication of PVI in claudication7. They also support published data suggesting that transitioning practice to OBL is associated with higher PVI and atherectomy rates among individual physicians9,10. Although cost data is not part of this analysis, prior research has shown that increased utilization of OBLs increase Medicare payments to physicians on a per-patient basis11. While it remains unclear the exact mechanisms by which practice in an ambulatory setting seems to drive increased utilization of certain types of PVI and PVI overall, establishing practice guidelines and regulation of incentives for ambulatory intervention suites has received increasing attention7,15,16,25,26. As a result, the Society for Vascular Surgery has recently partnered with the American College of Surgeons to initiate a Vascular Center Verification and Quality Improvement Program27 that may be critical for maintaining value-based care in high-reimbursement settings. There are several accreditation programs for ambulatory and office-based procedural sites that exist28,29, which will likely be integral in the establishment of regulations and ongoing incorporation of these tools into the future of vascular care.

It should be noted that out of more than 185,000 patients with a new diagnosis of claudication in our study, only 3.2% underwent early PVI. This indicates that the vast majority of physicians captured in our analysis were compliant with societal guidelines with respect to best practices for the management of claudication. While we have no reference point to suggest what an optimal threshold is for early PVI and the lack of anatomic and disease severity information in this dataset limits the conclusions gathered here, the early PVI rate of >14% by the high physician user group is concerning and this and other studies3,7,8,23,24 suggest early invasive treatment of claudication is harmful. Both the Society for Vascular Surgery and the American Heart Association / American College of Cardiology guidelines explicitly recommend a trial of conservative management with best medical therapy, smoking cessation, and supervised exercise therapy prior to consideration of revascularization4,6. Prior data has shown that a supervised exercise therapy program for improving claudication symptoms requires a program duration of 6 months17. The incidence of minor or major amputation among patients with claudication is estimated to be less than 10% over 10 years of follow-up30, and the average rate of ankle brachial-index decline is 0.014 per year31. Although recent studies have shown improved quality of life for patients with claudication who are offered early revascularization32, these benefits are not sustained long-term33,34. Considering the lack of a sustained improvement in quality of life along with the increased risk of poor limb outcomes observed here, it is our belief that the use of early PVI for claudication should be reserved to select patients, reflecting the Society for Vascular Surgery appropriate use criteria for claudication35.

Limitations of this study include those inherent to a retrospective, large-scale claims analysis. Use of claims data necessitates use of ICD codes to assign diagnoses, which raises the possibility of miscoding bias both of claudication and CLTI. We elected not to evaluate rates of supervised exercise therapy and smoking cessation counseling due to high rates of missing data and known coding revisions that occurred during the study period. Only about 30% of this study population had Medicare Part D, which limited our ability to examine medication use. Likewise, Medicare does not have any anatomic data regarding level or severity of disease or disease symptoms; there is also a large amount of missing data with regards to laterality, which we attempted to address with a sensitivity analysis. Because of this we elected only to included femoropopliteal interventions to limit increasing cohort heterogeneity further. We are also unable to stratify this dataset by intervention type as the majority of the patients included did not undergo any intervention. Medicare also, by definition, excludes patients with private insurance. Strengths of our study include the large Medicare dataset (>180,000 patients), inclusion of non-operatively as well as operatively managed patients, and length of median follow up (nearly 5 years), which allows a robust analysis of cumulative peripheral vascular interventions over time. Use of Medicare claims data also allows analysis of physician-level factors, including specialty and practice patterns, that are not available in other datasets.

Conclusions

Early PVI following a new diagnosis of claudication is associated with higher rates of late PVI interventions compared to patients treated with early nonoperative management. High physician users of early PVI for claudication perform more late PVI than their peers, particularly those primarily delivering care in high reimbursement settings. These findings raise concern that early application of PVI for claudication breeds increased subsequent interventions and worse limb outcomes, but also that its application may be driven by incentive structures inherent to providing care in the ambulatory setting. The appropriateness of early PVI for claudication needs critical evaluation, as do incentives surrounding delivery of these interventions in ambulatory intervention suites.

Supplementary Material

1

Supplementary Figure 1. Kaplan-Meier curves showing cumulative ipsilateral late PVI (>6 months after diagnosis) for A) patients who did vs. did not undergo early intervention (within 6 months) for new diagnosis of claudication and B) patients who were treated for a new diagnosis of claudication stratified by initial treating physician outlier status

Supplementary Table 1. International Classification of Disease (ICD) 9 and 10 codes used to define study population

Supplementary Table 2. Current Procedural Technology (CPT) codes used to classify and define both early and late peripheral vascular interventions

Supplementary Table 3. Current Procedural Technology (CPT) codes used to classify and define open surgical interventions, both revascularizations and major amputations

Article Highlights:

Type of Research:

Retrospective cross-sectional study utilizing Medicare claims data

Key Findings:

A review of 187,442 Medicare beneficiaries newly diagnosed with claudication demonstrated a nearly 7-fold increase in rates of late peripheral vascular intervention (PVI) for those patients treated with early peripheral vascular intervention (≤6 months) versus those with no early PVI. Patients treated with early PVI for claudication were more likely to convert to CLTI. The only physician-level factor associated with increased rates of late PVI was majority practice in a high reimbursement setting, which was associated with a stepwise increase in late PVI with increasing percentage of practice in an ambulatory surgery center or office-based laboratory.

Take Home Message:

Medicare beneficiaries who underwent PVI in the first 6 months following diagnosis of claudication were at risk for higher late PVI rates and impaired limb outcomes compared to patients treated with early medical management alone.

Footnotes

Author conflicts of interest:

This work was supported by a grant from the American College of Surgeons. Dr. Hicks is also supported by a grant from the NIH/NIDDK K23DK124515. Dr. Hicks also declares relationships with Cook Medical, LLC and W.L. Gore that are unrelated to this work.

Presented at the 2022 Society for Vascular Surgery Vascular Annual Meeting, Boston, MA, June 15–18, 2022

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Cassar K “Intermittent claudication.” BMJ. 2006. Nov 11;333(7576):1002–5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Rantner B, Kollerits B, Pohlhammer J, Stadler M, C Lamina, Peric S, et al. “The fate of patients with intermittent claudication in the 21st century revisited—results from the CAVASIC study.” Sci Rep. 2017. Apr 3;8:45833. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Kim TI, Kiwan G, Mohamedali A, Zhang Y, Dardick A, Guzman RJ, et al. “Multiple reinterventions for claudication are associatd with progression to limb-threatening ischemia.” Ann Vasc Surg. 2021. Apr;72:166–74. [DOI] [PubMed] [Google Scholar]
  • 4.Gerhard-Herman MD, Gornik HL, Barrett C, Barshes NR, Corriere MA, Drachman DE, et al. “2016 AHA/ACC guideline on the management of patients with lower extremity peripheral artery disease: executive summary.” Circulation. 2017. Mar 21;135(12):e686–e725. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Dormandy J, Heeck L & Vig S. “The natural history of claudication: risk to life and limb.” Semin Vasc Surg. 1999. Jun;12(2):123–37. [PubMed] [Google Scholar]
  • 6.Conte MS, Pomposelli FB, Clair DG, Geraghty PJ, McKinsey JF, Mills JL, et al. “Society for Vascular Surgery practice guidelines for atherosclerotic occlusive disease of the lower extremities: management of asymptomatic disease and claudication.” J Vasc Surg. 2015. Mar 1;61(3):2S–41S.E1. [DOI] [PubMed] [Google Scholar]
  • 7.Hicks CW, Holscher CM, Wang P, Black JH III, Abularrage CJ & Makary MA. “Overuse of early peripheral vascular interventions for claudication.” J Vasc Surg. 2020. Jan;71(1):121–30.e1. [DOI] [PubMed] [Google Scholar]
  • 8.Levin SR, Farber A, Cheng TW, Arinze N, Jones DW, Rybin D, et al. “Patients undergoing interventions for claudication experience low perioperative morbidity but are at risk for worsening functional status and limb loss.” J Vasc Surg. 2020. Jul;72(1):241–9. [DOI] [PubMed] [Google Scholar]
  • 9.Itoga NK, Baker LC & Mell MW. “Impact of office-based laboratories on physician practice patterns and outcomes after percutaneous vascular interventions for peripheral artery disease.” J Vasc Surg. 2019. Nov;70(5):1524–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Smith ME, Sutzko DC, Beck AW & Osborne NH. “Provider trends in atherectomy volume between office-based laboratories and traditional facilities.” Ann Vasc Surg. 2019. Jul;58:83–90. [DOI] [PubMed] [Google Scholar]
  • 11.Itoga NK, Baker LC & Mell MW. “Initial financial impact of office-based laboratories on Medicare payments for percutaneous interventions for peripheral artery disease.” J Vasc Surg. 2020. Aug;72(2):686–691:e1. [DOI] [PubMed] [Google Scholar]
  • 12.Kawaji Q, Dun C, Walsh C, Sorber RA, Stonko DP, Abularrage CJ, et al. “Index atherectomy peripheral vascular interventions performed for claudication are associated with more reinterventions than nonatherectomy interventions.” J Vasc Surg. 2022. Mar 8;S0741–5214(22)00401–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Hicks CW, Wang P, Bruhn WE, Abularrage CJ, Perler BA, Black JH III, et al. “Race and socioeconomic differences associated with endovascular peripheral vascular interventions for newly diagnosed claudication.” J Vasc Surg. 2020. Aug;72(2):611–21.e5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Hicks CW, Holscher CM, Wang P, Dun C, Abularrage CJ, Black JH III, et al. “Use of atherectomy during index peripheral vascular interventions.” JACC Cardiovasc Interv. 2021. Mar 22;14(6):678–88. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Siracuse JJ, Woodson J, Ellis RP, Farer A, Roddy SP, Kalesan B et al. “Intermittent claudication treatment patterns in the commercially insured non-Medicare population.” J Vasc Surg. 2021. Aug;74(2):499–504. [DOI] [PubMed] [Google Scholar]
  • 16.Brown CS, Smith ME, Kim GY, Sutzko DC, Henke PK, Corriere MA, et al. “Exploring the rapid expansion of office-based laboratories and peripheral vascular interventions across the United States.” J Vasc Surg. 2021. Sep;74(3):997–1005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Gardner AW & Poehlman ET. “Exercise rehabilitation programs for the treatment of claudication pain. A meta-analysis.” JAMA. 1995. Sep;274(12):975–80. [PubMed] [Google Scholar]
  • 18.CMS. Master beneficiary summary file base. https://resdac.org/cms-data/files/mbsf-base. Accessed 25th May 2022.
  • 19.CMS. Medicare Data on Provider Practice and Specialty. https://resdac.org/sites/datadocumentation.resdac.org/files/MD-PPAS%20Record%20Layout%20-%20Version%202.4.pdf. Accessed 29th September 2022.
  • 20.Fakhry F, Spronk S, van der Laan L, Wever JJ, Tejink JAW, Hoffmann WH, et al. “Endovascular revascularization and supervised exercise for peripheral artery diesease and intermittent claudication: a randomized clinical trial.” JAMA. 2015. Nov 10;314(18):1936–44. [DOI] [PubMed] [Google Scholar]
  • 21.Nordanstig J, Taft C, Hensater M, Perlander A, Osterberg K & Jivegard L. “Two-year results from a randomized clincial trial of revascularization in patients with intermittent claudication.” Br J Surg. 2016. Sep;103(10):1290–9. [DOI] [PubMed] [Google Scholar]
  • 22.Jones WS, Baumgartner I, Hiatt WR, Heizer G, Conte MS, White CJ, et al. “Ticagrelor compared with clopidogrel in patients with prior lower extremity revascularizaiton for peripheral artery disease.” Circulation. 2017. Jan;135(3):241–50. [DOI] [PubMed] [Google Scholar]
  • 23.Golledge J, Moxon JV, Rowbotham S, Pinchbeck J, Yip L, Velu R, et al. “Risk of major amputation in patients with intermittent claudication undergoing early revascularization.” Br J Surg. 2018. May;105(6):699–708. [DOI] [PubMed] [Google Scholar]
  • 24.Madabhushi V, Davenport D, Jones S, Khoudoud SA, Orr N, Minion D, et al. “Revascularization of intermittet claudicants leads to more chronic limb threatening ischemia and higher amputation rates.” J Vasc Surg. 2021. Sep;74(3):771–9. [DOI] [PubMed] [Google Scholar]
  • 25.Lin PH, Chandra FA, Shapiro FE, Osman BM, Urman RD & Ahn SS. “The need for accrediatation of office-based interventional vascular centers.” Ann Vasc Surg. 2017. Jan;38:332–338. [DOI] [PubMed] [Google Scholar]
  • 26.Hodgson K “Get us back our turkey! 2020 Society for Vascular Surgery Presidential Address.” 2021. Delivered at the 2021 Vascular Annual Meeting, San Diego, CA. [Google Scholar]
  • 27.“SVS launches vascular verification program to enhance quality improvement.” Endovascular Today. 2021. Aug 20. Available online https://evtoday.com/news/svs-launches-vascular-verification-program-to-enhance-quality-improvement. Accessed 4 Jul 2022. [Google Scholar]
  • 28.IAC. “Vascular interventional accrediation.” https://intersocietal.org/programs/vascular-interventional/ Accessed 29 September 2022.
  • 29.The Joint Commission. “Accreditation for ambulatory surgery centers that delivers.” https://www.jointcommission.org/accreditation-and-certification/health-care-settings/ambulatory-health-care/learn/accreditation-options-certifications/ambulatorysurgery-center/ Accessed 29 September 2022.
  • 30.Muluk SC, Muluk VS, Kelley ME, Whittle JC, Tierney JA, Webster MW, et al. “Outcome events in patients with claudication: a 15 year study in 2777 patients.” 2001. Feb;33(2):251–7. [DOI] [PubMed] [Google Scholar]
  • 31.Aquino R, Johnnides C, Makaroun M, Whittle JC, Muluk VS, Kelley ME, et al. “Natural history of claudication: long-term serial follow-up study of 1244 claudicants.” J Vasc Surg. 2001. Dec;34(6):962–70. [DOI] [PubMed] [Google Scholar]
  • 32.Angraal S, Hejjaji V, Tang Y, Gosch KL, Patel MR, Heyligers J, et al. “One-year health status outcomes following early invasive and noninvasive treatment in symptomatic peripheral artery disease.” Circ Cardiovasc Interv. 2022. Jun;15(6):e011506. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Koelemay MJW, van Reijen NS, van Dieren S, Frans FA, Vermeulen EJG, Buscher HCJL, et al. “Editor’s choice—Randomised clinical trial of supervised exercise therapy vs. endovascular revascularisation for intermittent claudication caused by iliac artery obstruction: the SUPER study.” Eur J Vasc Endovasc Surg. 2022. Mar;63(3):421–9. [DOI] [PubMed] [Google Scholar]
  • 34.Fakhry F, Fokkenrood HJP, Spronk S, Teijink JAW, Rouwet EV, & Hunink MGM. “Endovascular revascularisation versus conservative management for itnermittent claudication.” Cochrane Database Syst Rev. 2018. Mar;3(3):CD010512. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Woo K, Siracuse JJ, Klingbeil K, Kraiss LW, Osborne NH, Singh N, et al. “Society for Vascular Surgery appropriate use criteria for management of intermittent claudication.” J Vasc Surg. 2022. Jul;76(1):3–22.e1. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Supplementary Figure 1. Kaplan-Meier curves showing cumulative ipsilateral late PVI (>6 months after diagnosis) for A) patients who did vs. did not undergo early intervention (within 6 months) for new diagnosis of claudication and B) patients who were treated for a new diagnosis of claudication stratified by initial treating physician outlier status

Supplementary Table 1. International Classification of Disease (ICD) 9 and 10 codes used to define study population

Supplementary Table 2. Current Procedural Technology (CPT) codes used to classify and define both early and late peripheral vascular interventions

Supplementary Table 3. Current Procedural Technology (CPT) codes used to classify and define open surgical interventions, both revascularizations and major amputations

NIHMS1854056-supplement-1.pdf (228.7KB, pdf)

RESOURCES