Clinical practice patterns and ascertainment bias for cardiovascular events in a randomized trial: A survey of investigators in the BEST-CLI trial

Abstract

Ascertainment bias is a well-recognized source of bias in research, but few studies have systematically analyzed sources of ascertainment bias in randomized trials in which blinding is not possible and endpoint assessment is not protocolized. In the current study, we sought to evaluate differences in the clinical practice patterns of trial investigators with respect to bias in the ascertainment of pre-revascularization patient risk and the incidence of secondary endpoints post-revascularization. We conducted a cross-sectional survey of active investigators (n = 936) from the Best Endovascular Versus Best Surgical Therapy for Patients with Critical Limb Ischemia (BEST-CLI) trial. The total survey response rate was 19.6% (183/936). Vascular surgeons were more likely than nonsurgical interventionalists to order tests for cardiac complications after both surgical bypass (p < 0.001) and endovascular revascularization (p = 0.038). Post-procedure, investigators were more likely to order additional testing for cardiac complications in open surgery versus endovascular cases (7% vs 16% never, 41% vs 65% rarely, 43% vs 17% sometimes, 9% vs 2% always, respectively; p < 0.0001). Significant variation in practice patterns exist in the pre- and post-procedure assessment of cardiac risk and events for patients with CLI undergoing revascularization. Variation in the ascertainment of risk and outcomes according to the type of revascularization procedure and physician specialty should be considered when interpreting the results of clinical studies, such as the BEST-CLI trial. ClinicalTrials.gov Identifier: NCT02060630

Keywords

bias critical limb ischemia (CLI)randomized controlled trial revascularization

Introduction

Critical limb ischemia (CLI) is the most advanced manifestation of lower extremity peripheral artery disease (PAD) and is characterized by pain at rest in the ankle or foot and/or the presence of ischemic ulcerations or necrotic tissue. Approximately 20–40% of patients with CLI will require amputation and greater than 20% of patients will die without successful lower extremity revascularization.^1,2 Equipoise exists in the optimal technique (open vs endovascular) for revascularization in patients with CLI.^3,4

The Best Endovascular Versus Best Surgical Therapy for Patients With Critical Limb Ischemia (BEST-CLI) trial is a prospective randomized study comparing the safety and efficacy of open surgical versus endovascular lower extremity revascularization in patients with CLI.⁵ Blinding of treatment strategy is often not feasible in interventional trials such as BEST-CLI where a surgical operation is compared to a minimally invasive catheter-based procedure.⁶ This lack of blinding introduces the possibility of ascertainment bias, a well-recognized source of bias even in randomized controlled trials.⁷ Ascertainment bias is the systematic distortion of the assessment of outcome measures by the investigators or trial participants because they are aware of treatment allocation. This is particularly problematic for trial endpoints that do not have protocol-mandated testing for their ascertainment, as may be the case for observational studies comparing minimally invasive to open surgical interventions.^8
–10 The detection of such endpoints is therefore subject to the practice patterns of the individual investigators or study sites.

In the current study, we sought to characterize the clinical practices of trial investigators with respect to pre- and post-procedure testing for cardiac events using a specially designed survey to assess ascertainment bias. The primary purpose of the survey was to determine whether the potential for ‘missed’ secondary endpoints in BEST-CLI, such as myocardial infarction (MI), and other cardiovascular events, might differ between patients randomized to open surgery versus endovascular revascularization due to dissimilarities in individual or institutional practice patterns. We also compared practice patterns between vascular surgeons and investigators from other specialties.

Methods

Survey and study design

We conducted a cross-sectional survey of BEST-CLI investigators. The survey was distributed via Survey Monkey to the 936 active investigators as of June 2016. We analyzed practice patterns according to (1) the type of revascularization (endovascular vs open surgical); and (2) medical subspecialty. All ethical and institutional review boards of the participating BEST-CLI centers approved this study.

Questionnaire

The survey tool was developed by the principal investigators of the BEST-CLI trial. The survey evaluated operator background and demographics, self-reported primary specialty, practice location (zip code), and individual volume of CLI cases. The survey then explored the individual and institutional practice patterns with respect to pre- and/or post-procedure cardiac testing. Finally, survey participants were asked a series of questions to characterize which circumstances (patient subgroup or particular symptoms/signs) would trigger pre-procedure cardiac testing for risk assessment. Furthermore, post-procedure evaluation for MI for both endovascular and open procedures were used to determine potential differences in operator practice between modes of intervention (see the supplementary online material for the complete survey).

Statistical analysis

Fisher’s exact test was used to compare response rates between different investigator specialties. Descriptive statistics for continuous outcomes were presented using medians and IQR range. Dot plots were used to present data rather than box-and-whisker plots because the latter can be uninformative for extremely skewed data.

Comparisons of practice patterns for patients undergoing open surgery versus endovascular revascularization took into account the paired nature of these data (i.e., one response for open surgery patients and one response for endovascular revascularization patients). For yes/no responses, McNemar’s test was used. For continuous outcomes, we used the Wilcoxon signed rank test. For ordered categorical outcomes, we assigned consecutive scores for the possible responses (e.g., if the possible responses were never, rarely, sometimes, and always, the scores were 0, 1, 2, and 3, respectively), and then used the Wilcoxon signed rank test to compare differences between the investigator’s response for open surgery patients and endovascular revascularization patients. Missing data were rare. Investigators who were missing one or both of the paired responses were dropped from that particular analysis.

Comparisons of responses from vascular surgeons versus other specialties used Fisher’s exact test for binary outcomes, the Wilcoxon test for continuous outcomes, and an exact version of the Mantel–Haenszel chi-squared test for ordered categorical outcomes. All analyses were done with SAS software, version 9.4 (SAS Institute Inc., Cary, NC, USA).

Results

Provider characteristics and response rate

Overall demographic characteristics are shown in Table 1. A majority of respondents were high volume operators including 76% of respondents who reported caring for over 25 patients with CLI per year. The vast majority were vascular surgeons (77%); interventional cardiologists (13%) were the next largest group of respondents (online Supplemental Figure 1). Most respondents had approximately 20 years of clinical service. The total response rate for the survey was 19.6% (183/936), with some variation according to specialty: 19.9% (140/702) of vascular surgeons, 21.4% (24/112) of interventional cardiologists, 13.9% (16/115) of interventional radiologists, 100% (3/3) of vascular medicine physicians, and 0% (0/4) of other specialties.

Table 1.

Demographics of respondents.

	Overalln = 183 surveys		Vascular surgeonn = 140		IC/IR/VMn = 43		p-value
	Missing data	n (%)	Missing data	n (%)	Missing data	n (%)
Sex, male	0	166 (91)	0	124 (89)	0	42 (98)	0.08
Race	1		1		0		0.18
Asian		22 (12)		17 (12)		5 (12)
Black		5 (3)		2 (1)		3 (7)
Caucasian		145 (80)		110 (79)		35 (81)
Hispanic/Latino		7 (4)		7 (5)		0 (0)
Other		3 (2)		3 (2)		0 (0)
Specialty
Interventional cardiologist		24 (13)
Interventional radiologist		16 (9)
Vascular medicine		3 (2)
Vascular surgery		140 (77)
Years in service							0.37
0–10		21 (11)		17 (12)		4 (9)
11–20		69 (38)		50 (36)		19 (44)
21–30		53 (29)		38 (27)		15 (35)
31–40		34 (19)		30 (21)		4 (9)
>40		6 (3)		5 (4)		1 (3)
Years in service, median (Q1, Q3)	0	21 (14, 30)	0	21 (14, 30.5)	0	19 (12, 26)	0.13

IC/IR/VM, interventional cardiologists/interventional radiologists/vascular medicine physicians.

Comparison of practice patterns for patients treated with open surgical vs endovascular revascularization

Standardized pre-procedural practice patterns

In their personal practice, 65% of investigators reported following a standardized algorithm for the cardiac evaluation of patients prior to open surgical bypass, compared to 32% who followed a standardized algorithm prior to endovascular revascularization (p < 0.0001) (online Supplemental Figure 2A). Institutional algorithms for pre-revascularization cardiac evaluation were reported by 34% of investigators for open surgery, and by 12% for endovascular procedures (p < 0.0001). Pre-procedure, 90% of investigators ordered electrocardiograms (ECGs) for all open surgery patients, and 63% ordered them for all endovascular revascularization patients (Figure 1). Cardiac stress tests, cardiac catheterizations, and cardiac biomarker tests were less commonly ordered (medians of 50%, 10%, and 0% of open surgery patients and 5%, 0%, and 0% of endovascular patients, respectively) (Figure 1). All four types of pre-procedure cardiac evaluation were ordered more often for patients undergoing open surgery (p < 0.0001 for each comparison) (Figure 1).

Figure 1.

Reported percentage of patients with CLI undergoing pre-procedure cardiac evaluations before open surgical and endovascular revascularization procedures.

Post-procedural practice patterns

Following revascularization, investigators were more likely to order additional testing for cardiac complications in open surgery cases (7% never, 41% rarely, 43% sometimes, 9% always) than endovascular cases (16%, 65%, 17%, and 2%, respectively; p < 0.0001) (Figure 2B). In their personal practice, 29% of investigators claimed to follow a standardized algorithm for cardiac evaluation of open surgery patients, compared to 21% for endovascular patients (p = 0.0004). Institutional algorithms for cardiac evaluation were reported by 15% of investigators for open surgery and 9% for endovascular procedures folllowing revascularization (p = 0.0008) (online Supplemental Figure 2B). Post-procedure ECG, creatinine phosphokinase (CPK), creatinine phosphokinase-myocardial band (CPK-MB), and troponin tests were all ordered more often for open surgery patients than for endovascular patients (p < 0.0001 for each comparison) (Figure 3).

Figure 2.

Frequency in which additional testing is ordered to assess for cardiac complications following revascularization in patients with CLI (A) by procedure type and by investigator specialty for (B) open surgery and (C) endovascular revascularization subjects.

Figure 3.

Reported percentage of patients with CLI in which the specified test is collected following revascularization, by procedure type.

Comparison of practice patterns of vascular surgeons vs nonsurgical investigators

Revascularization techniques

Vascular surgeons reported the following breakdown of revascularization techniques for patients under their care with CLI: a median 50% (IQR: 40–62.5%) receive endovascular revascularization, 30% (5–40%) receive open surgical revascularization, and 10% (7.5–20%) undergo hybrid revascularization (endovascular + open surgical) (online Supplemental Figure 3). Nonsurgical interventionalists reported a median of 85% (75–95%), 10% (0–20%), and 5% (0–10%) of their patients with CLI receive endovascular, open surgical, and hybrid revascularization, respectively (online Supplemental Figure 3). Vascular surgeons were more likely to report performance of open surgical and hybrid revascularization, whereas nonsurgical interventionalists more commonly reported performing endovascular revascularization (p < 0.001 for all comparisons) (online Supplemental Figure 3).

Pre-procedural cardiac assessment of patients undergoing open surgical revascularization

For patients undergoing open surgical revascularization, 67% of vascular surgeons and 56% of other investigators reported a personal algorithm for pre-procedure cardiac evaluation (p = 0.20); 33% of vascular surgeons and 37% of other investigators reported institutional algorithms (p = 0.71) (online Supplemental Figure 2A).

Vascular surgeons were more likely than other investigators to order ECGs before open surgical revascularization (p = 0.02) (Figure 2). Cardiac stress tests and cardiac biomarker testing before open surgical revascularization did not differ by investigator type. Regarding invasive evaluation, there was a trend towards fewer cardiac catheterizations among vascular surgery versus nonvascular surgeon investigators (10% (IQR: 5–20%) vs 15% (IQR: 5–25%), p = 0.06) (Figure 1).

Pre-procedural cardiac assessment of patients undergoing endovascular revascularization

There was no statistical difference between the percentage of vascular surgeons versus other investigators reporting a personal algorithm for pre-procedure cardiac evaluation for patients scheduled for endovascular procedures (34% vs 28%, p = 0.58); similarly, there was no difference between these groups reporting institutional algorithms (12% vs 12%, p = 1.00) (online Supplemental Figure 2A).

ECGs were commonly performed by both vascular surgeons and nonsurgical interventionalists (100% (IQR: 65%–100%) vs 100% (40%–100%), p = 0.08) before endovascular revascularization. Cardiac stress tests, cardiac catheterization, and cardiac biomarkers were rarely ordered for these patients, and vascular surgeons were less likely to order cardiac catheterization compared to their nonsurgical counterparts (0% (0–5%) vs 5% (0–10%), p = 0.02) (Figure 2).

Pre-procedural triggers for cardiovascular assessment

A minority of both vascular surgeons and other investigators reported performing pre-procedure cardiac evaluation in every patient before revascularization (15% and 19%, p = 0.63). Specific triggers for cardiac evaluation for vascular surgeons compared to nonsurgical interventionalists included history of congestive heart failure (CHF) (69% vs 53%), history of coronary artery disease (CAD) (56% vs 67%), presence of cardiac risk factors (52% vs 53%), history of arrhythmia (34% vs 26%), and older age (28% vs 19%). Pre-procedural triggers did not significantly differ by investigator type.

Post-procedure cardiovascular assessment

A personal standardized algorithm for cardiac evaluation following open surgical bypass was reported by 31% of vascular surgeons and 25% of other investigators (p = 0.56); 14% of vascular surgeons and 20% of other investigators reported the existence of an institutional algorithm (p = 0.46). A personal standardized algorithm for cardiac evaluation after endovascular intervention was reported by 20% of vascular surgeons and 23% of other investigators (p = 0.67); 7% of vascular surgeons and 16% of other investigators reported the existence of an institutional algorithm (p = 0.13) (Figure 2B).

Figures 2B and 2C show how often vascular surgeons and nonsurgical interventionalists reported performing post-procedure testing for cardiac complications in patients who have undergone open surgical and endovascular revascularization. Vascular surgeons were more likely than nonsurgical interventionalists to order tests for cardiac complications after both surgical bypass (p < 0.001) and endovascular revascularization (p = 0.0381).

Figure 4A shows the frequency by which vascular surgeons and nonsurgical interventionalists order various types of cardiac evaluations after open surgical bypass. There was no difference in frequency of ECGs ordered by specialty type (p = 0.71); however, following open surgical revascularization, vascular surgeons were more likely than nonsurgical interventionalists to order CPK, CPK-MB, or troponin (p = 0.05, 0.004, and 0.004, respectively).

Figure 4B depicts the practice patterns of vascular surgeons and nonsurgical interventionalists regarding ordering cardiac tests after endovascular revascularization. The specialty groups did not differ in how often CPK and CPK-MB were ordered (p = 0.23 and 0.16, respectively), but vascular surgeons were less likely to order ECGs (p = 0.02) and more likely than nonsurgical interventionalists to order troponin (p = 0.02).

Figure 4.

Reported percentage of patients with CLI in which specified test is collected following revascularization, by investigator specialty for (A) surgical bypass and (B) endovascular revascularization subjects.

Post-procedural triggers for cardiovascular assessment

Indications that trigger post-procedure cardiac evaluation, testing, or referral included clinical signs of ischemia (98% of vascular surgeons and 93% of other investigators (p = 0.14)), ECG changes during or after the procedure (94% vs 84%, p = 0.0499), clinical signs of decompensated heart failure (96% vs 81%, p = 0.01), hemodynamic instability during or after the procedure (89% vs 79%, p = 0.13), and arrhythmia during or after the procedure (89% vs 72%, p = 0.01). Only 2% of vascular surgeons and 5% of other investigators routinely evaluate all patients for cardiac complications (p = 0.34).

Discussion

The validity and relevance of clinical trial results may be critically affected by bias. Ascertainment bias is an important type of bias that may be difficult to detect and appraise, particularly in interventional trials comparing invasive and minimally invasive therapies.^11,12 To our knowledge, this survey is the first study to systematically examine perioperative cardiac and neurologic assessments of patients by investigators in a clinical intervention research study. In the current study, we sought to explore how nonblinded detection of secondary outcomes may be affected by ascertainment bias in the BEST-CLI trial. The principal findings of our study are that, both pre- and post-procedure, assessment of cardiac risk and events is significantly more intense for CLI patients scheduled for open surgery compared to endovascular revascularization. Furthermore, vascular surgeons are more likely to perform post-procedure cardiac biomarker assessment than their nonsurgical interventional colleagues.

Pre-procedure practice patterns

We observed that a minority of investigators reported the existence of an institutional algorithm for cardiac risk assessment prior to either surgical or endovascular revascularization. However, both surgical and nonsurgical investigators reported following a personal protocol for preoperative cardiac risk assessment (e.g., cardiac stress tests, cardiac catheterizations, and cardiac biomarker tests) more commonly before open surgical compared to endovascular revascularization. Although vascular surgeons reported obtaining an ECG more frequently before surgical revascularization, pre-surgical cardiac stress testing did not differ by investigator type.

The American Heart Association guidelines recommend additional cardiovascular testing in patients at elevated risk of a perioperative cardiovascular event with poor or unknown functional capacity.¹³ One major predictor of such perioperative events is suprainguinal vascular surgery.¹⁴ Although most patients with CLI tend to have infrainguinal and infrapopliteal PAD,¹⁵ many of these patients have poor functional capacity¹⁶ that may prompt clinicians to pursue additional testing, as we observed among BEST-CLI investigators. Various forms of stress testing have been shown to improve risk stratification in patients at elevated risk for a perioperative event; however, no study has shown that interventions (i.e., coronary revascularization)¹⁷ subsequent to such testing improves outcomes. Although uncommon for all specialties, pre-revascularization (both open and endovascular) cardiac catheterization was more frequently reported to be performed on patients cared for by nonsurgical investigators than their surgical colleagues. Current guidelines do not recommend routine coronary angiography and only recommend such testing in the presence of high-risk noninvasive findings that would otherwise prompt angiography irrespective of surgery.¹³

Post-procedure practice patterns

No differences were reported in the presence of an institutional or personal standardized algorithm for post-procedure cardiac assessment by investigator-type. However, post-procedure assessment for cardiac complications was more common after open surgery than endovascular revascularization. The most common post-procedure test was the ECG. According to existing guidelines, the utility of postoperative ECG or cardiac biomarkers in patients at high risk for perioperative MI is ‘uncertain’.¹³ Although no data exist describing the frequency of ECG testing following open bypass versus endovascular revascularization, our survey findings seem consistent with the common clinical practice of obtaining an ECG following surgery as opposed to minimally invasive therapeutic counterparts.

Vascular surgeons were more likely than nonsurgical interventionalists to order CPK, CPK-MB, or troponin after bypass surgery, and less likely to order ECG but more likely to order troponin after endovascular procedures. The pattern of increased testing following open surgery by vascular surgeons may reflect the possibility that surgeons are more frequently the responding clinicians for patients in the immediate postoperative period following surgical bypass. Nonsurgical investigators may not routinely be involved in the immediate postoperative care of their patients referred for surgical bypass and their responses may be more theoretical than reflective of their clinical practice. The reasons underlying increased utilization of cardiac biomarkers by surgeons following endovascular revascularization are unclear and require further study. Although it is well established that postoperative elevations in cardiac biomarkers is associated with a worse prognosis,¹⁸ in the absence of any demonstrated therapeutic approach to elevated biomarkers in this setting, the benefit of routine biomarker screening remains circumspect.

Study limitations

We acknowledge our study has several limitations. Only 20% of BEST-CLI investigators responded to the survey. However, the relative proportion of investigator specialists who did respond was relatively balanced. Given that survey response rates in the 20% range have been considered adequate in providing a representative sample, particularly when total responses exceed 100,¹⁹ our results are likely to reflect the general practice patterns of the investigators in the BEST-CLI trial. Indeed, the approximate 20% response rate was observed among vascular surgeons and interventional cardiologists who represent the majority of investigators in the BEST-CLI trial. Regarding bias in the ascertainment of outcomes between revascularization strategies, the survey did not collect data on triggers for cardiac evaluation separately for open surgery and endovascular revascularization. Finally, our survey was administered exclusively to investigators in the BEST-CLI trial, which may not reflect the clinical practice patterns of the broader vascular community.

Conclusions

Our findings have important implications for the clinical contextualization of the main results of the BEST-CLI trial, including the possibility that: (a) more intense pre- and post-procedure evaluation for patients with CLI undergoing open revascularization may create the ‘appearance’ of greater baseline procedural risk, in spite of the prospective randomized trial design; and (b) the detection of post-procedure myocardial injury or infarction may be increased among patients cared for by vascular surgeons due to a greater tendency to perform post-procedure troponin testing among these providers. Findings from this study indicate the potential for ascertainment bias in the important secondary endpoint of MI, wherein patients undergoing endovascular revascularization could be diagnosed less frequently with an MI, based on variations in testing that reflect biases and differences in routine clinical practice patterns.

Supplemental Material

sj-docx-2-vmj-10.1177_1358863X21995897 – Supplemental material for Clinical practice patterns and ascertainment bias for cardiovascular events in a randomized trial: A survey of investigators in the BEST-CLI trial

Supplemental material, sj-docx-2-vmj-10.1177_1358863X21995897 for Clinical practice patterns and ascertainment bias for cardiovascular events in a randomized trial: A survey of investigators in the BEST-CLI trial by Mazen S Albaghdadi, Michael N Young, Mohammed M. Chowdhury, Susan Assmann, Taye Hamza, Sandra Siami, Maria Villarreal, Michael Strong, Matthew Menard, Alik Farber and Kenneth Rosenfield in Vascular Medicine

Supplemental Material

sj-pdf-1-vmj-10.1177_1358863X21995897 – Supplemental material for Clinical practice patterns and ascertainment bias for cardiovascular events in a randomized trial: A survey of investigators in the BEST-CLI trial

Supplemental material, sj-pdf-1-vmj-10.1177_1358863X21995897 for Clinical practice patterns and ascertainment bias for cardiovascular events in a randomized trial: A survey of investigators in the BEST-CLI trial by Mazen S Albaghdadi, Michael N Young, Mohammed M. Chowdhury, Susan Assmann, Taye Hamza, Sandra Siami, Maria Villarreal, Michael Strong, Matthew Menard, Alik Farber and Kenneth Rosenfield in Vascular Medicine

Footnotes

Acknowledgements

We would like to thank Dr Eric G Campbell PhD, formerly of the Massachusetts General Hospital Institute for Health Policy and now Director of Research at the University of Colorado’s Center for Bioethics and Humanities, for his assistance in developing the survey tool used in this study.

Declaration of conflicting interests

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

Funding

This work was supported by the National Institutes of Health (1U01HL107407).

Supplementary material

The supplementary material is available online with the article.

ORCID iD

Mazen S Albaghdadi

References

Benoit

O’Donnell

Jr Kitsios

, et al. Improved amputation-free survival in unreconstructable critical limb ischemia and its implications for clinical trial design and quality measurement. J Vasc Surg 2012; 55: 781–789.

Agarwal

Sud

Shishehbor

MH.

Nationwide trends of hospital admission and outcomes among critical limb ischemia patients: From 2003–2011. J Am Coll Cardiol 2016; 67: 1901–1913.

Adam

Beard

Cleveland

, et al.; BASIL trial participants. Bypass versus angioplasty in severe ischaemia of the leg (BASIL): Multicentre, randomised controlled trial. Lancet 2005; 366: 1925–1934.

Gerhard-Herman

Gornik

Barrett

, et al. 2016 AHA/ACC guideline on the management of patients with lower extremity peripheral artery disease: Executive Summary: A report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol 2017; 69: 1465–1508.

Menard

Farber

Assmann

, et al. Design and rationale of the Best Endovascular Versus Best Surgical Therapy for Patients With Critical Limb Ischemia (BEST-CLI) Trial. J Am Heart Assoc 2016; 5: e003219.

Probst

Grummich

Heger

, et al. Blinding in randomized controlled trials in general and abdominal surgery: Protocol for a systematic review and empirical study. Syst Rev 2016; 5: 48.

Piaggio

Elbourne

Schulz

, et al.; WHO Research Group to evaluate Misoprostol in the Management of the Third Stage of Labour. The reporting of methods for reducing and detecting bias: An example from the WHO Misoprostol Third Stage of Labour equivalence randomised controlled trial. BMC Med Res Methodol 2003; 3: 19.

McNulty

Spertus

, et al. Surgical candidacy and selection biases in nonemergent left main stenting: Implications for observational studies. JACC Cardiovasc Interv 2011; 4: 1020–1027.

Bhatti

Grayson

Grotte

, et al.; North West Quality Improvement Programme in Cardiac Interventions. The logistic EuroSCORE in cardiac surgery: How well does it predict operative risk? Heart 2006; 92: 1817–1820.

10.

Flaherty

Davidson

CJ.

Coronary-artery bypass grafting versus stent implantation. N Engl J Med 2005; 353: 735–737; author reply 735–737.

11.

Farooq

Serruys

PW.

‘Cherry-picking’ patients for randomized, controlled trials—Reliving the past. J Am Coll Cardiol 2013; 61: 2492.

12.

Testa

Latib

Bollati

, et al. Unprotected left main revascularization: Percutaneous coronary intervention versus coronary artery bypass. An updated systematic review and meta-analysis of randomised controlled trials. PloS One 2017; 12: e0179060.

13.

Fleisher

Fleischmann

Auerbach

, et al. 2014 ACC/AHA guideline on perioperative cardiovascular evaluation and management of patients undergoing noncardiac surgery: Executive Summary: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014; 130: 2215–2245.

14.

Ford

Beattie

Wijeysundera

DN.

Systematic review: Prediction of perioperative cardiac complications and mortality by the revised cardiac risk index. Ann Intern Med 2010; 152: 26–35.

15.

Shishehbor

White

Gray

, et al. Critical limb ischemia: An Expert Statement. J Am Coll Cardiol 2016; 68: 2002–2015.

16.

Landry

GJ.

Functional outcome of critical limb ischemia. J Vasc Surg 2007; 45(Suppl A): A141–148.

17.

McFalls

Ward

Moritz

, et al. Coronary-artery revascularization before elective major vascular surgery. N Engl J Med 2004; 351: 2795–2804.

18.

Writing Committee for the VISION Study Investigators; Devereaux

Biccard

Sigamani

, et al. Association of postoperative high-sensitivity troponin levels with myocardial injury and 30-day mortality among patients undergoing noncardiac surgery. JAMA 2017; 317: 1642–1651.

19.

Nulty

The adequacy of response rates to online and paper surveys: What can be done?

Assess Eval High Educ 2008; 33: 301–314.

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

0.24 MB

0.21 MB