The association between ischemic and jeopardized myocardia and all-cause mortality in patients with peripheral artery disease

Abstract

Peripheral artery disease (PAD) is associated with increased mortality and concomitant coronary artery disease (CAD). However, it is unclear whether uncovering the presence of functional coronary ischemia by single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) would further help stratifying that excess risk. From January 2000 to 2009, 4294 individuals underwent cardiac stress testing within 180 days of ankle–brachial index (ABI) measurements. Of these, 645 had PAD and SPECT MPI stress testing. Abnormal ABI was defined as ⩽0.9 or prior lower extremity arterial revascularization. Myocardial ischemic burden and total jeopardized myocardium were represented by the summed difference score (SDS) and summed stress score (SSS), respectively. Univariate and multivariable Cox proportional hazard analyses were used to study the impact of SDS and SSS on all-cause mortality. Additionally, using a hierarchical approach, we examined the step-wise addition of post-stress test coronary and lower extremity arterial revascularizations as time-varying covariates on outcomes. We found no significant difference in all-cause mortality between patients with ischemic myocardium (SDS>0) and those without (SDS=0) (adjusted HR: 0.94, 95% CI: 0.53–1.69; p=0.84). Similarly, the presence of jeopardized myocardium (SSS>0) did not have a significant impact on mortality (adjusted HR: 1.16, 95% CI: 0.67–2.00; p=0.59). Moreover, adjustment for post-testing coronary and lower extremity arterial revascularizations did not affect our results. In conclusion, ischemic and jeopardized myocardia are not predictors of all-cause mortality in PAD; thus, SPECT MPI does not appear to be a useful risk stratification tool in these patients.

Keywords

cardiac stress testing mortality peripheral artery disease imaging SPECT MPI risk stratification

Introduction

Peripheral artery disease (PAD) and coronary artery disease (CAD) coexist. The ankle–brachial index (ABI) is a commonly used test to diagnose PAD and is associated with all-cause mortality;¹ moreover, measuring ABI in CAD patients has an incremental prognostic value.^2–4 Single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) is also used to evaluate for CAD, and its scores are known to be associated with high risk of all-cause mortality in general.^5–7 Given the high prevalence of CAD and cardiovascular morbidity and mortality in patients with PAD,^8,9 it remains unclear whether the identification of functional coronary ischemia allows better risk stratification in these patients.¹⁰ We therefore examined the impact of myocardial ischemic burden and total jeopardized myocardium on all-cause mortality among patients with PAD.

Methods

Study population

During the period from January 2000 to January 2009, we identified all patients (n = 4294) who had a cardiac stress test 180 days before or after assessment for PAD with ABI measurement at our tertiary care institution. Of these, the 645 who had PAD (an abnormal ABI or a history of lower extremity arterial intervention) and underwent cardiac stress testing with SPECT MPI comprised the study population. Others who underwent imaging with echocardiography or positron emission tomography (PET) were excluded to avoid differences related to imaging modality. The ABI measurements were performed for evaluation of lower extremity symptoms, signs, or asymptomatic screening in a laboratory accredited by the Intersocietal Accreditation Commission (IAC) in vascular testing and by certified technologists. Cardiac stress testing was performed for cardiac symptom evaluation in 43%; the remaining underwent testing for either resting electrocardiogram abnormalities or pre-operative assessment. The baseline characteristics of the study population, cardiac medication use, and cardiac stress test data were obtained from the prospectively collected stress test database at our institution.¹¹ Electronic medical records were reviewed to capture coronary revascularizations (open heart coronary bypass or percutaneous coronary intervention) if there was any pre- and/or post-cardiac stress testing. Similarly, lower extremity arterial revascularization procedures (open or endovascular) were obtained pre- and post-ABI measurement. These data were specifically captured to determine whether any revascularization as the result of cardiac stress testing would impact long-term outcome. The left ventricular ejection fraction (LVEF) was obtained from transthoracic echocardiography or SPECT MPI data during the same period. The study was approved by the Institutional Review Board, data were de-identified and informed consent was waived.

The majority of patients (90%) underwent pharmacologic vasodilator stress testing at the discretion of the clinician. On SPECT MPI, using the 17-segment model and 5-point scale (0 = normal, 1 = mild, 2 = moderate, 3 = severe, 4 = absence of uptake), the summed rest score (SRS) and summed stress score (SSS) were obtained by addition of the scores of the 17 segments at rest and peak stress, respectively. There was a maximum possible overall score of 68.¹² The summed difference score (SDS) was defined as the difference between the SSS and SRS. An SDS > 0 indicated the presence of stress-induced reversible myocardial ischemia, which was used to define a positive stress test. The SDS represented total ischemic myocardium (myocardial ischemic burden), while the SSS signified total myocardium that was ischemic and infarcted (total jeopardized myocardium).

For ABI, systolic arm and leg blood pressures were first obtained from all extremities using a Doppler-based technique with the patient at rest and in the supine position. For each lower extremity, the ABI was calculated by dividing the higher of two ankle systolic blood pressures (measured at dorsalis pedis or posterior tibial artery) by the higher of the two brachial systolic blood pressures. The lower of the two ABIs, right or left leg, was the ABI used for analysis. All ABIs reported in this study were performed at rest. Patients were classified as having PAD if they had an ABI ⩽ 0.9¹³ or a prior history of lower extremity arterial intervention regardless of their ABI values (n = 293). Death was defined as all-cause and obtained using the Social Security Death Index (SSDI) with a closing date of 25 February 2011.

Statistical analysis

Simple descriptive statistics were used to summarize the data. Continuous variables are presented as mean ± standard deviation. Comparisons were made using the Wilcoxon rank-sum non-parametric test. Categorical data are described using frequencies and percentages. Comparisons were made using the chi-squared test or Fisher’s exact test if the frequency was less than 5. Mean imputation was used to impute the missing values (only race and body mass index variables had missing values) with < 1% of the data missing. The distribution of SDS and SSS among the study population was left skewed with a large proportion having SDS = 0 (454 (70%) patients) and SSS = 0 (357 (55%) patients) (Figure 1). Plots of SDS and SSS against survival were conducted to examine the optimal cutoff for these variables; however, linear relationships were observed. Therefore, we considered both elements of SDS > 0 and the continuous SDS (likewise with the SSS elements) in the analyses because the data were zero inflated.

Figure 1.

Histogram of SDS and SSS on SPECT MPI in peripheral artery disease patients. Histograms of the summed difference score (SDS) and summed stress score (SSS) for the entire study population are depicted in panels A and B, respectively. The dataset is ‘zero-inflated’ with the majority of SDS and SSS values equal to zero. (MPI, myocardial perfusion imaging; SPECT, single-photon emission computed tomography.)

We employed a multivariable Cox proportional hazard regression model using reliable predictors obtained by bootstrapping methods to examine whether ischemic or jeopardized myocardium had any impact on the occurrence of subsequent (post-stress test) coronary revascularization in PAD patients. We adjusted for history (pre-stress test) of coronary and lower extremity arterial interventions.

All-cause mortality was studied non-parametrically by the Kaplan–Meier method with the date of cardiac stress test as time zero. Univariate and multivariable Cox proportional hazard models using reliable predictors obtained by bootstrapping methods were also employed to examine the association of all-cause mortality in PAD patients with SDS and SSS. Again, we adjusted for a pre-test history of coronary and lower extremity arterial interventions. In addition, we repeated all-cause mortality analyses after excluding ejection fraction from the models. Furthermore, we re-examined the impact of SDS and SSS on all-cause mortality only in those whose cardiac stress testing was performed for evaluation of cardiac symptoms. In the univariate models, each SDS, SDS > 0, SSS, and SSS > 0 was modeled individually, while all of them were included together in the multivariable models. Importantly, a hierarchical approach examined the step-wise addition of post-testing coronary and lower extremity arterial interventions to the full mortality models to understand their impact on the above associations. Bootstrap bagging methods^14,15 were used to determine reliable predictors for the above models with random re-sampling and automated step-wise selection (entry criteria of p ⩽ 0.1 and retention criteria of p ⩽ 0.05). Variables or clusters of variables which entered more than 50% of 1000 models were chosen for the final model. The results were presented as unadjusted and adjusted hazard ratios (HRs) with 95% confidence intervals (CIs). The proportional hazards assumption was tested. We used SAS (v9.2; SAS Inc., Cary, NC, USA) for statistical analyses with a p-value < 0.05 for statistical significance.

Results

Patient characteristics

Of the 645 patients with PAD, 288 (45%) had an abnormal SPECT MPI stress test (SSS > 0). The baseline characteristics of patients stratified by SSS are presented in Table 1. Patients with an abnormal stress test (SSS > 0) were more likely to be male with a significantly higher prevalence of myocardial infarction, heart failure, coronary bypass grafting, prior coronary intervention, and cerebrovascular accident. Patients with abnormal stress tests were also more likely to be on baseline aspirin, statin, or angiotensin-converting enzyme inhibitors, and tended to have lower ejection fractions (Table 1). Pharmacologic stress was performed in the majority of patients without significant difference between the two groups (Table 1). We also presented baseline characteristics of the study population based on the indication of cardiac stress testing. Those with cardiac symptoms were more likely to have cardiac morbidities, abnormal or positive stress tests, or higher stress testing scores, while those with no cardiac symptoms were more likely to have end-stage renal disease, aortoiliac aneurysm, or cancer, which possibly reflects the indication for pre-operative stress testing in these patients (Appendix, Table 1). The mean ± standard deviation follow-up among the study population was 3.2 ± 1.8 years (median = 3.0 years). A total of 2036 patient years were available for analysis. Ten percent had follow-up of nearly 6 years (5.7 years) or more.

Table 1.

Selected baseline characteristics of the study peripheral artery disease population according to SSS^a.

Characteristics	SSS = 0 (n = 357)	SSS > 0 (n = 288)	p-value
Age, years	67 ± 12.3	69 ± 10.4	0.06
Male	196 (55%)	198 (69%)	< 0.001
White/Caucasian^b	245 (70%)	223 (78%)	0.02
Black race^b	93 (26%)	51 (18%)	0.009
Other race^b	13 (3.7%)	13 (4.5%)	0.6
Body mass index^c, kg/m²	28 ± 5.8	29 ± 5.8	0.02
Hypertension	311 (87%)	264 (92%)	0.06
Diabetes: on oral agents only	74 (21%)	78 (27%)	0.06
Diabetes: requiring insulin	74 (21%)	60 (21%)	0.97
Current or ex-smoker (⩽ 1 year)	125 (35%)	84 (29%)	0.11
History of end-stage renal disease	31 (8.7%)	13 (4.5%)	0.04
History of myocardial infarction	59 (17%)	119 (41%)	< 0.001
History of coronary artery disease	147 (41%)	186 (65%)	< 0.001
History of congestive heart failure	41 (11%)	64 (22%)	< 0.001
History of cerebrovascular accident	35 (9.8%)	45 (16%)	0.03
History of claudication	205 (57%)	166 (58%)	0.96
History of CABG	72 (20%)	123 (43%)	< 0.001
History of PCI	93 (26%)	105 (36%)	0.004
History of aortoiliac aneurysm	60 (17%)	65 (23%)	0.07
History of deep vein thrombosis	27 (7.6%)	13 (4.5%)	0.11
History of cancer	35 (9.8%)	23 (8%)	0.4
Baseline medications
Aspirin	219 (61%)	206 (72%)	0.007
ACE inhibitors or ARB	140 (39%)	140 (49%)	0.02
Beta receptor antagonists	223 (62%)	195 (68%)	0.17
Non-dihydropyridine calcium channel blockers	40 (11%)	19 (6.6%)	0.04
Treatment for dyslipidemia	212 (59%)	205 (71%)	0.002
Pharmacologic stress test	326 (91%)	256 (89%)	0.3
Summed rest score	0 ± 0.00	7.0 ± 8.4	< 0.001
Summed stress score	0 ± 0.00	11 ± 8.6	< 0.001
Summed difference score	0 ± 0.00	4.1 ± 5.0	< 0.001
Left ventricular ejection fraction, %	56 ± 7.8	50 ± 11.5	< 0.001

Data presented as number (percentage) for categorical variables and mean ± standard deviation for continuous variables.

Variable has seven (1%) missing values.

Variable has two (< 1%) missing values.

ACE, angiotensin-converting enzyme; ARB, angiotensin receptor blocker; CABG, coronary artery bypass grafting; PCI, percutaneous coronary intervention; SSS, summed stress score.

Association of SDS > 0 with subsequent coronary revascularization

Fifteen percent of our study cohort underwent coronary revascularization after undergoing stress testing; 30% of those with an SDS > 0. A multivariable Cox proportional hazard regression analysis showed that a positive stress test (SDS > 0) was associated with a significant increase in the occurrence of future coronary revascularization (adjusted HR: 5.46, 95% CI: 1.99–15.00; p = 0.001). This was despite adjustments for pre-stress test coronary intervention in addition to the reliable baseline predictors of post-stress test coronary revascularization as identified by the bootstrapping methods: history of myocardial infarction, family history of CAD, history of pulmonary hypertension, abdominal aortic aneurysm, and shortness of breath.

All-cause mortality in PAD – the impact of SDS and SSS

There were 192 deaths (30%) among our study patients during follow-up. Mortality in the positive stress test group (SDS > 0) was 35% versus 27% in the negative stress test group (SDS = 0), and 35% in the SSS > 0 group versus 25% in the SSS = 0 group. The Kaplan–Meier curves showed no significant mortality difference between positive and negative cardiac stress test cohorts (SDS = 0 vs SDS > 0, p = 0.6) (Figure 2, Panel A) or when comparing SSS > 0 vs SSS = 0 groups, p = 0.1 (Figure 2, Panel B). In unadjusted univariate analysis, only the magnitude of jeopardized myocardium (SSS) was associated with a slight increase in all-cause mortality (HR: 1.02, 95% CI: 1.003–1.04; p = 0.02) (Table 2).

Figure 2.

Unadjusted comparison of all-cause mortality among peripheral artery disease (PAD) patients according to SDS (ischemic myocardium) (Panel A) and SSS (jeopardized myocardium) (Panel B). Kaplan–Meier curves for all-cause mortality in PAD patients show no significant differences between SDS = 0 and SDS > 0 groups (Panel A), or between SSS = 0 and SSS > 0 groups (Panel B). (SDS, summed difference score; SSS, summed stress score.)

Table 2.

Univariate and multivariable analyses for all-cause mortality in peripheral artery disease patients.

Stress test variable	Hazard ratio	95% CI	p-value
SSS > 0 compared to SSS = 0
Unadjusted	1.24	0.93–1.65	0.14
Adjusted for baseline characteristics^{a, b}	1.16	0.67–2.00	0.59
Adjusted for baseline characteristics + post-stress test coronary^c and peripheral revascularization^{b, d}	1.14	0.66–1.98	0.64
SSS
Unadjusted	1.02	1.003–1.04	0.02
Adjusted for baseline characteristics^{a, b}	1.01	0.99–1.04	0.36
Adjusted for baseline characteristics + post-stress test coronary^c and peripheral revascularization^{b, d}	1.01	0.99–1.04	0.34
SDS > 0 compared to SDS = 0
Unadjusted	1.09	0.81–1.46	0.59
Adjusted for baseline characteristics^{a, b}	0.94	0.53–1.69	0.84
Adjusted for baseline characteristics + post-stress test coronary^c and peripheral revascularization^{b, d}	0.96	0.53–1.72	0.88
SDS
Unadjusted	1.00	0.97–1.04	0.96
Adjusted for baseline characteristics^{a, b}	1.00	0.94–1.06	0.93
Adjusted for baseline characteristics + post-stress test coronary^c and peripheral revascularization^{b, d}	0.99	0.93–1.05	0.79

Baseline characteristics include age, body mass index, diabetes requiring insulin, history of end-stage renal disease, history of myocardial infarction, history of congestive heart failure, history of deep vein thrombosis, history of significant carotid stenosis (> 60%), history of aortoiliac aneurysm, left ventricular ejection fraction, resting heart rate, type of stress test, and pre-stress test coronary and peripheral interventions.

Each multivariable analysis included all SSS > 0, SSS, SDS > 0, and SDS variables together in the same model along with other variables as indicated above.

Open heart coronary bypass or percutaneous coronary intervention.

Open or endovascular intervention.

CI, confidence interval; SDS, summed difference score; SSS, summed stress score.

The reliable baseline predictors of mortality identified by bootstrapping methods were age, body mass index, diabetes requiring insulin, end-stage renal disease, history of congestive heart failure, myocardial infarction, abdominal aortic aneurysm, significant carotid stenosis (> 60%), lower extremity arterial intervention, or deep vein thrombosis (marginally significant), LVEF, and resting heart rate. In a multivariable Cox proportional hazard model that adjusted for reliable predictors and pre-stress test coronary and lower extremity arterial revascularizations, positive stress test (SDS > 0) (adjusted HR: 0.94, 95% CI: 0.53–1.69; p = 0.84) or an abnormal stress test (SSS > 0) (adjusted HR: 1.16, 95% CI: 0.67–2.00; p = 0.59) did not have any significant effect on survival. More importantly, in a hierarchical model, further adjustments for post-stress test coronary and lower extremity arterial interventions had no significant impact on all-cause mortality or on the association between the SDS or SSS and all-cause mortality (Table 2).

After excluding ejection fraction from all-cause mortality models, and adjusting for reliable predictors and pre-stress test coronary and lower extremity arterial revascularization, only continuous SSS became statistically significant (adjusted HR: 1.03, 95% CI: 1.002–1.06; p = 0.03). Adjustments for post-stress test coronary and lower extremity arterial interventions in a hierarchical approach did not impact these results (Appendix, Table 2).

In separate analyses including only those with cardiac symptoms (n = 278), SDS and SSS variables still had no significant association with all-cause mortality, similar to our results in the whole study population (Appendix, Table 3).

Discussion

This study found no association between all-cause mortality and functional markers for CAD represented by SDS and SSS in PAD patients, even among those with cardiac symptoms. Therefore, flow-limiting CAD and infarcted scar tissue (SDS and SSS), important precursors for ventricular arrhythmia, do not seem to predict the increased all-cause mortality observed in the PAD population. Our findings are based on adjusted analyses that included multiple baseline characteristics, LVEF, medication use, type of stress test, and any coronary or lower extremity revascularization after obtaining the result of the stress test and ABI. LVEF might be a better marker of myocardial response to ischemic burden; therefore, analyses were repeated without ejection fraction in the models. Of all 12 models tested in Table 2, only continuous SSS became statistically significant (Appendix, Table 2). Ejection fraction is an important prognostic variable; however, from a clinical standpoint, a low LVEF does not preclude ischemic work-up when indicated.

Data have repeatedly shown that patients with PAD have an increased risk of all-cause mortality,¹ with the majority dying from cardiovascular causes.¹⁶ In addition, observational data show a high prevalence of CAD among individuals with PAD,^8,9 mostly asymptomatic due to the claudication-limited physical activity in these patients. As a result, there has been an emerging clinical practice trend towards recommending testing for CAD in PAD patients;⁸ however, it remains unknown whether uncovering the presence of coexisting CAD would help further risk stratification of these patients beyond what has been recommend by the guidelines.¹³ In this study, we used hierarchical multivariable models to understand the role of SDS and SSS (markers of myocardial ischemia and scar) as possible predictors of all-cause mortality in this population. Available evidence shows that SPECT MPI perfusion scores (SDS and SSS), in addition to providing a quantitative assessment of total ischemic burden and total jeopardized myocardium, are independent predictors of all-cause mortality.^5–7 Therefore, we believe, SDS and SSS are well suited to address the study objective.

In one study (1992–1995) by Thatipelli et al. (n = 395), no correlation was found between the wall motion score index on dobutamine stress echocardiography and ABI. Moreover, in multivariable analysis, inducible and fixed wall motion abnormalities on dobutamine stress echocardiography did not predict all-cause mortality.¹⁰ This study, however, was limited by small sample size and the heterogeneity of its population (14% had no PAD), though ABI was adjusted for in the multivariable analysis. In another study, Ishihara et al.¹⁷ found that the rate of all-cause mortality in PAD patients was lowest in the no-CAD group and highest in the triple-vessel CAD group. The limitations included small sample size, lack of functional assessment of CAD (coronary angiography was used), and lack of adjusted analysis. Our study has a larger sample size, applies multivariable adjustment using reliable predictors from bootstrapping methods, and also involves a contemporary patient population on appropriate guideline-recommended medications. Moreover, we present for the first time a step-wise hierarchical approach to test the impact of post-stress test coronary and peripheral interventions on all-cause mortality and its association with SDS and SSS (Table 2).

There are several studies that have shown that the SPECT MPI cardiac stress test can predict early^18,19 and late^20,21 coronary revascularization. Similarly in our study, patients with a positive stress test (SDS > 0) were more likely to have subsequent (post-stress test) coronary revascularization despite adjustment for all reliable predictors identified by bootstrapping methods in addition to history of (pre-stress test) coronary and lower extremity arterial intervention. However, there was still no association between SDS > 0 and all-cause mortality in PAD patients, even after adjusting for post-stress test coronary revascularization. Therefore, it appears that patients with PAD and a positive stress test are more likely to undergo coronary revascularization; however, in this analysis, revascularization did not impact clinical outcomes.

Limitations

While this study proposes a new perspective to the well-known association between PAD and mortality, a few limitations need to be highlighted. First, ABI measurements and cardiac stress testing were, for the most part, symptom-driven in a population with multiple co-morbidities typical for a tertiary care center. Hence, the results may not be easily generalized to the asymptomatic ‘silent’ PAD population. Second, patients with an abnormal stress test (SSS > 0) as a group had a significantly higher prevalence of cardio-protective medication use that included aspirin, statin, angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, and beta-blockers that may have modified the association between SSS and mortality. Nevertheless, our results are based on adjustments for cardio-protective medication use. Lastly, despite multiple adjustments and prospectively collected data, this is a retrospective analysis and the results must be verified by well-designed randomized clinical trials.

Conclusions

In this study, there was no association between the presence or extent of functionally significant CAD and all-cause mortality in patients with PAD. The myocardial ischemic burden and total jeopardized myocardium do not appear to predict the high mortality seen in the PAD population. Furthermore, neither coronary nor lower extremity arterial revascularization after cardiac stress and ABI testing had any significant impact on all-cause mortality. Hence, cardiac stress testing is likely of limited additional value for risk stratification in the PAD population beyond the aggressive risk stratification recommended by the guidelines. However, it is often unavoidable in the setting of questionable cardiac symptoms.

Footnotes

Acknowledgements

Suzanne Turner, Department of Graphics, Heart and Vascular Institute, Cleveland Clinic and Kathryn Brock, BA, Editorial Services Manager, Heart and Vascular Institute, Cleveland Clinic.

Dr. Joshua Beckman served as guest editor for this manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Supplementary material

The supplementary material is available at

References

Resnick

Lindsay

McDermott

. Relationship of high and low ankle brachial index to all-cause and cardiovascular disease mortality: The Strong Heart Study. Circulation 2004; 109: 733–739.

Banerjee

Vinas

Mohammad

. Significance of an abnormal ankle-brachial index in patients with established coronary artery disease with and without associated diabetes mellitus. Am J Cardiol 2014; 113: 1280–1284.

Lee

. Prevalence and clinical implications of newly revealed, asymptomatic abnormal ankle-brachial index in patients with significant coronary artery disease. JACC Cardiovasc Interv 2013; 6: 1303–1313.

Mostaza

Manzano

Suarez

. Different prognostic value of silent peripheral artery disease in type 2 diabetic and non-diabetic subjects with stable cardiovascular disease. Atherosclerosis 2011; 214: 191–195.

Schelde

Schmidt

Madsen

. Impact of co-morbidity on the risk of first-time myocardial infarction, stroke, or death after single-photon emission computed tomography myocardial perfusion imaging. Am J Cardiol 2014; 114: 510–515.

Nakazato

Berman

Gransar

. Prognostic value of quantitative high-speed myocardial perfusion imaging. J Nucl Cardiol 2012; 19: 1113–1123.

Piccini

Horton

Shaw

. Single-photon emission computed tomography myocardial perfusion defects are associated with an increased risk of all-cause death, cardiovascular death, and sudden cardiac death. Circ Cardiovasc Imaging 2008; 1: 180–188.

Hirose

Chikamori

Hida

. Prevalence of coronary heart disease in patients with aortic aneurysm and/or peripheral artery disease. Am J Cardiol 2009; 103: 1215–1220.

Bhatt

Steg

Ohman

. International prevalence, recognition, and treatment of cardiovascular risk factors in outpatients with atherothrombosis. JAMA 2006; 295: 180–189.

10.

Thatipelli

Pellikka

McBane

. Prognostic value of ankle-brachial index and dobutamine stress echocardiography for cardiovascular morbidity and all-cause mortality in patients with peripheral arterial disease. J Vasc Surg 2007; 46: 62–70; discussion 70.

11.

Shishehbor

Litaker

Pothier

. Association of socioeconomic status with functional capacity, heart rate recovery, and all-cause mortality. JAMA 2006; 295: 784–792.

12.

Cerqueira

Weissman

Dilsizian

. Standardized myocardial segmentation and nomenclature for tomographic imaging of the heart. A statement for healthcare professionals from the Cardiac Imaging Committee of the Council on Clinical Cardiology of the American Heart Association. Circulation 2002; 105: 539–542.

13.

Aboyans

Criqui

Abraham

. Measurement and interpretation of the ankle-brachial index: A scientific statement from the American Heart Association. Circulation 2012; 126: 2890–2909.

14.

Breiman

. Bagging predictors. Mach Learn 1996; 24: 123–140.

15.

Efron

Tibshirani

. An Introduction to the Bootstrap. Monographs on Statistics and Applied Probability: 57. Boca Raton, FL: Chapman & Hall/CRC, 1998.

16.

Norgren

Hiatt

Dormandy

. Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). J Vasc Surg 2007; 45 Suppl S: S5–67.

17.

Ishihara

Iida

Tosaka

. Severity of coronary artery disease affects prognosis of patients with peripheral artery disease. Angiology 2013; 64: 417–422.

18.

Hachamovitch

Hayes

Friedman

. Comparison of the short-term survival benefit associated with revascularization compared with medical therapy in patients with no prior coronary artery disease undergoing stress myocardial perfusion single photon emission computed tomography. Circulation 2003; 107: 2900–2907.

19.

Miller

Roger

Hodge

. Gender differences and temporal trends in clinical characteristics, stress test results and use of invasive procedures in patients undergoing evaluation for coronary artery disease. J Am Coll Cardiol 2001; 38: 690–697.

20.

Dos Santos

Cwajg

Pantoja Mda

. Prognostic value of technetium-99m-labeled single-photon emission computerized tomography in the follow-up of patients after their first myocardial revascularization surgery. Arq Bras Cardiol 2003; 80: 25–30.

21.

Solodky

Assali

Mats

. Prognostic value of myocardial perfusion imaging in symptomatic and asymptomatic patients after percutaneous coronary intervention. Cardiology 2007; 107: 38–43.

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

0.20 MB