Preoperative Intraaortic Balloon Pump Improves Early Outcomes Following High-Risk Coronary Artery Bypass Graft Surgery: A Meta-Analysis of Randomized Trials and Prospective Study Design

Background. Despite widespread use, evidence to support preemptive intraaortic balloon pump (IABP) insertion for patients undergoing high-risk coronary artery bypass graft (CABG) surgery remains sparse, and in need of a well-defined clinical trial. To inform the design of a prospective trial, we sought to review outcomes in randomized controlled trials (RCTs) of anticipatory IABP use vs control in patients undergoing high-risk CABG through meta-analysis. The primary endpoint was all-cause mortality within 30 days of surgery. The secondary endpoint was major adverse cardiac and cerebrovascular event (MACCE), a composite of death, myocardial infarction, stroke, or repeat revascularization. Methods. Using Ovid MEDLINE, we systematically reviewed all RCTs comparing preoperative IABP with control in patients undergoing high-risk CABG, defined as: left ventricular ejection fraction (LVEF) ≤40%, left main stenosis ≥70%, unstable angina, recent myocardial infarction, or prior myocardial revascularization undergoing elective or emergent CABG on or off pump. Results. Of 950 articles assessed for eligibility, 10 RCTs of 1261 subjects (mean age, 65.0 years; 21.8% women; mean LVEF, 35%) were included. Mortality was significantly lower in patients receiving IABP compared with control (relative risk [RR], 0.48; 95% confidence interval [CI], 0.30-0.76; P<.01). The risk of MACCE was also lower with IABP (RR, 0.67; 95% CI, 0.54-0.84; P<.001). No significant differences in major bleeding events (RR, 1.27; 95% CI, 0.44-3.72) or vascular complications (RR, 1.13; 95% CI, 0.42-3.06) were detected. Conclusions. A strategy of routine prophylactic IABP use may reduce short-term mortality and MACCE in high-risk CABG patients. A definitive, adequately powered, prospective, randomized trial is warranted to confirm these results.

J INVASIVE CARDIOL 2018;30(1):2-9. Epub 2017 September 15.

Key words: intraaortic balloon pump, coronary heart disease, coronary artery bypass graft surgery, outcomes, mortality, major adverse cardiovascular and cerebrovascular effects

The utility of intraaortic balloon pump (IABP) counterpulsation in patients with acute coronary heart disease (CHD) has been a long-standing source of debate. Recent studies suggest that routine IABP use does not improve outcomes in patients with acute myocardial infarction (MI), even when complicated by cardiogenic shock.^1-3 There is controversy regarding the utility of IABPs in the setting of high-risk coronary artery bypass graft (CABG) surgery. A recent randomized controlled trial (RCT) showed no benefit of routine preemptive IABP deployment in high-risk patients with reduced left ventricular ejection fraction (LVEF) undergoing CABG, and the authors concluded that prophylactic IABP deployment in this patient population is not justified.⁴ However, subsequent meta-analyses that included this trial suggested a benefit of preoperative IABP deployment in elective patients undergoing high-risk CABG.^5-7 These meta-analyses, however, failed to include a number of eligible RCTs, limiting their utility, particularly with respect to their ability to inform about the effect of the IABP in important subgroups. Additionally, none have examined the clinically useful combined endpoint of major adverse cardiac and cerebrovascular event (MACCE) or fully extrapolated data on bleeding and vascular complications frequently associated with IABP use. Current CABG surgery guidelines reflect this uncertain evidence. Both American and European guidelines cite class IIa indications for IABP use in the preoperative setting; however, American guidelines recommend preoperative IABP use for high-risk patients,⁸ while European guidelines only recommend preoperative IABP use in the setting of a mechanical complication.⁹ 

Given that individual RCTs of preemptive IABP use in high-risk patients undergoing CABG were underpowered to detect differences in individual cardiovascular outcomes and adverse events,^4,10-21 and that prior meta-analyses omitted relevant trials,^5-7,22-26 we set out to systematically evaluate and meta-analyze cardiovascular outcomes and adverse events reported in all available RCTs of preoperative IABP vs control in high-risk patients undergoing CABG. Our aim was to determine which patient populations might specifically benefit from prophylactic IABP placement in a large RCT. 

Methods

Search strategy. Using Ovid MEDLINE (1946-2016 July week 4), we performed a systematic review according to the recommendations of the Cochrane Collaboration²⁷ and the PRISMA guidelines, and identified all RCTs comparing preoperative IABP with controls in high-risk adult CABG patients (Figure 1). Keyword search terms included “intra-aortic balloon pump,” “cardiac surgical procedures,” “preoperative,” and “adverse effects” (see Supplemental Appendix S1 for details of search parameters). The search was not restricted to any language. Bibliographies of retrieved articles and pertinent systematic reviews were scrutinized for other relevant studies. 

Selection criteria. Inclusion criteria consisted of RCTs comparing preoperative IABP placement to control, and adult patients characterized as high risk, undergoing elective or emergent CABG with either cardiopulmonary bypass (CPB) or off-pump coronary artery bypass (OPCAB) procedures. High-risk patients were defined as having at least one of the following: LVEF ≤40%; left main stenosis ≥70%; unstable angina (current); recent MI (within 10 days); prior CABG; or an estimated perioperative risk estimated by an additive EuroScore of ≥6.^28,29 Studies of patients in cardiogenic shock were excluded. Two investigators (RZ and JU) independently determined study eligibility. Results were compared and any disagreements were ultimately resolved by consensus agreement. Data extracted from published articles were confirmed and validated by contacting the corresponding author of each study for insight and clarification. Data regarding outcomes by sex and surgical urgency were obtained directly from primary authors.^{4,17,18,20,21}   

Study variables. The primary endpoint of this study was all-cause mortality during the postoperative period, measured either in hospital or within 30 days. The secondary endpoint was MACCE, a composite of all-cause death, MI, cerebrovascular accident or transient ischemic attack, or need for repeat revascularization either in hospital or within 30 days. A composite of mortality and MI was reported when all MACCE variables were not available. Where we could not exclude double-counting of events and primary authors could not be reached for clarification, the more conservative number was utilized. An additional endpoint included a composite of low cardiac output, as determined by an estimated cardiac index ≤1.8 L/min/m² or postoperative congestive heart failure (CHF). Adverse events including vascular complications and major bleeding were examined, as defined in each trial (see Supplemental Appendix S2 for definitions).²⁷ 

Validity assessment. Utilizing the Cochrane Collaboration risk of bias assessment tool, the methodological quality of each RCT was evaluated on the following: (1) sequence generation; (2) allocation concealment; (3) participant, personnel, and outcome assessor blinding; (4) incomplete outcome data; (5) selective outcome reporting; and (6) other sources of bias.²⁷ Studies were characterized as low, unclear, or high risk.

Statistical analysis. All RCT data were evaluated on an intention-to-treat basis. Study population characteristics, preoperative and baseline demographic data, and timing of IABP deployment and retrieval were detailed. Analyses of published data for primary and secondary endpoints were stratified by preoperative IABP deployment vs control. Each dichotomous endpoint was analyzed by the Mantel-Haenszel method with a random-effects model. Results were expressed as relative risk (RR) with 95% confidence intervals (CIs). Significant heterogeneity of the data set was determined by the Chi-squared test <0.1, and I² ≥50%. Subgroup analyses to explore potential heterogeneity were conducted based on sex, timing of IABP placement, emergent vs elective procedures, and CABG with CPB vs OPCAB. The effect of prophylactic IABP insertion on postoperative heart failure or shock as estimated by low cardiac output (defined as ≤1.8 L/min/m²) was determined. To assess the potential for publication bias, funnel plots (precision [inverse standard error] vs log RR) were constructed. Further analysis of funnel plots by Begg’s and Egger’s analyses was completed with Comprehensive Meta-Analysis Software version 3. Data analysis was otherwise performed using the Cochrane Collaboration web-based software, Review Manager version 5.3.5. 

Results

Of 950 articles assessed for eligibility, 13 RCTs^4,10-21 were initially identified. Of these, 10 RCTs,^{4,10-13,15,17-20} containing 1261 subjects, compared preoperative IABP with no intervention. Two studies were excluded that compared a preoperative intervention vs a strategy of intraoperative or postoperative IABP placement.^14,16 A single study was also excluded that examined two different timings (very early [12 hours] vs early [2 hours] preoperative IABP placement).²¹ Baseline characteristics of the 1261 subjects (mean age, 65.0 years; 21.8% women; mean LVEF, 35%) are summarized in Table 1. Emergency surgeries comprised 507 (40.2%) of all procedures. Both on-pump (n = 999; 79.2%) and off-pump (n = 262; 20.8%) surgeries were performed. The timing of IABP placement relative to surgery varied between studies, although the most commonly tested strategy of preoperative IABP placement <2 hours prior to the first incision was observed in 520 patients (82.9%). 

Overall, a low risk of bias was noted among the 10 RCTs, with the exception that all trials were not blinded to the intervention (Figure 2 and Supplemental Figure S1). All 10 RCTs demonstrated rigorous random-sequence generation.^{4,10,12,13,16-21} Of these, 8 studies also provided sufficient evidence of allocation concealment.^{4,10,12,13,16,17,20,21} Given the intervention, all trials were biased due to lack of blinding of participants and study personnel. Outcomes were defined in all studies by intention to treat analysis. Neither attrition nor selection bias were evident in any of the studies. Data published by Christenson et al were queried for duplication of subjects across studies given the relatively small sample sizes; however, previous documentation by the Cochrane Heart Group confirmed the independence of these data.²⁵ 

Preoperative IABP and mortality. Among the 10 RCTs^{4,10-13,15,17-20} with data reported for mortality, controls only included patients allocated to no intervention. Thirty patients were excluded from a single trial with a third intervention arm that included IABP plus intervention.²⁰ Additional data from an ongoing RCT that randomized patients to either preoperative IABP (n = 20) or no intervention (n = 19) were also included in the mortality analysis. In total, 1270 patients were included in the mortality analysis, with 627 patients (49.4%) assigned preoperative IABP and 643 patients (50.6%) assigned as controls. The primary outcome measure reported in the trials was mortality in hospital (n = 455) or within 30 days (n = 815). There were 24 deaths (3.8%) in the IABP group vs 55 deaths (8.6%) in the control group of standard care (RR, 0.48; 95% CI, 0.30-0.76; P<.01; I² = 0%; Figure 3). This difference represented a 4.8% absolute risk reduction (ARR) and a number needed to treat (NNT) of 21 to prevent one early postoperative death. Exclusion of the unpublished data did not materially change the results (3.8% IABP vs 8.8% control; RR, 0.46; 95% CI, 0.29-0.73; P<.01; I² = 0%), nor did exclusion of data published by Christenson et al (4.1% IABP vs 6.9% control; RR, 0.59; 95% CI, 0.35-1.00; P=.05; I² = 0%). Further sensitivity analyses excluding studies from the 1990s and the Christenson data still produced an I² value of 0%. The results remained statistically significant, with a P=.05, although the Z value fell from 3.13 to 1.84.

Subgroup analyses of the risk of mortality according to differences in baseline characteristics are shown in Table 2. The benefit of IABP on mortality did not differ with the timing of IABP placement (>2 hours vs <2 hours prior to surgery) in the 8 RCTs^{4,10,11,15,17-20} with sufficient data (Pinteraction=.93), although the benefit in each group was directionally consistent. Mortality also did not appear to be significantly different between on-pump vs off-pump CABG patients in 10 RCTs with available data^{4,10-13,15,17-20} (Pinteraction=.93), although again the trend toward benefit with IABP was consistent. Sex (Pinteraction=.83) and the urgency of surgery (Pinteraction=.68) had no significant effect on the benefit of preemptive IABP (Table 2). 

Preoperative IABP and MACCE. In total, 835 patients from 7 RCTs were included in the MACCE analysis; data from an ongoing RCT that randomized patients to either preoperative IABP (n = 20) or no intervention (n = 19) were also included in the MACCE analysis.^{10,12,15,17,18,20} The risk of MACCE at 30 days was lower in patients preemptively treated with an IABP compared with standard care (22.4% IABP vs 34.1% control; RR, 0.67; 95% CI, 0.54-0.84; P<.001; I² = 0%; Figure 4). This difference represented an 11.7% ARR and an NNT of 9 to prevent one postoperative MACCE event with preemptive IABP. Again, exclusion of data published by Christenson et al did not significantly change the results (25.8% IABP vs 36.4% control; RR, 0.70; 95% CI, 0.56-0.88; P<.01; I² = 0%). The only available variable to explore for a differential effect according to baseline characteristics was sex, for which there was no significant interaction (Pinteraction=.96) (Table 2). Preemptive treatment with IABP was also associated with a lower composite endpoint of heart failure or shock events (22.5% vs 31.0%; RR, 0.70; 95% CI, 0.50-0.97; P=.03; I² = 46%) (Supplemental Figure S2).^{10-13,15,18-20} 

Preoperative IABP and safety endpoints. No significant differences in major bleeding events among 579 patients with available data^{4,12,13,15,19,20} (2.84% vs 2.01%; RR, 1.27; 95% CI, 0.44-3.72) (Figure 5) or vascular complications among 732 patients with available data^{4,12,13,15,18,20} (4.42% vs 3.11%; RR, 1.13; 95% CI, 0.42-3.06) (Figure 6) were detected.

Publication bias. Visual inspection of the funnel plot for mortality did not suggest reporting bias (Supplemental Figure S3); as confirmed by both Begg’s and Mazumdar’s rank correlation test with continuity correction (Tau = -0.18; one-tailed P=.22) and Egger’s regression intercept (B₀ = -0.62; df = 9; one-tailed P=.14). Likewise, funnel-plot analysis of MACCE studies was not suggestive of reporting bias by both Begg’s analysis (Tau = -0.095; one-tailed P=.38) and Egger’s analysis (B₀ = -0.25; df = 5; one-tailed P=.33) (Supplemental Figure S4). 

Discussion

The major finding of our review of 1261 patients from RCTs is that high-risk patients randomized to receive preoperative IABP placement prior to CABG derived a significant mortality benefit, as well as a decreased risk of MACCE compared with patients randomized to standard care. The benefits of IABP were consistent irrespective of sex, timing of preemptive IABP deployment, urgency of the surgery, or use of off-pump CABG. Adverse effects, such as vascular injury and major bleeding, were not significantly different between groups, although the absolute number of events reported in the trials was low. 

To our knowledge, this is the most comprehensive meta-analysis of prophylactic IABP deployment trials in high-risk patients undergoing CABG to date. The study has a number of strengths compared to recent meta-analyses of this subject.^6,7 The literature review was not limited to English language studies and individual trial authors were contacted to obtain primary data where possible. Our eligibility criteria were limited to RCTs and did not include observational studies. Finally, our study also focused on clinically relevant secondary outcomes of MACCE, bleeding events, and vascular complications that have not been previously assessed. 

Study limitations. There are several inherent limitations to trial-level meta-analyses. All trials lacked a sham intervention as a control. Lack of blinding could have biased the reporting of postoperative outcomes aside from death. Second, the quality of some of the included trials was unclear due to insufficient reporting of randomization and allocation concealment or inadequate blinding of outcome assessment.^4,17,21 Criticisms of earlier meta-analyses have questioned the significance of studies by Christenson et al stemming from a single center; however, the Cochrane meta-analysis cited no overlap in these populations, and our data remain significant even with their exclusion.²⁵ Also, the study population examined may not be representative of contemporary high-risk CABG patients as trials were conducted over the course of the past 40 years, prior to minimally invasive CABG techniques and hybrid approaches, with limited granularity on the acuity of the population or the urgency for surgery. Likewise, the relative spectrum of acute coronary syndromes varied across studies, including the window between the index MI and time of surgery as well as the degree of ongoing ischemia. Last, while our MACCE analysis was consistent with a previously reported RCT,¹⁷ both MACCE and safety analyses are likely underreported, as data were not available across all studies. 

Appropriate use of prophylactic IABP prior to CABG. The clinical context for prophylactic IABP use prior to CABG has evolved considerably over the last 40 years. In part, this may be due to complex patient populations, improved surgical techniques, and hemodynamic monitoring, as well as new pharmacologic and mechanical modalities offering additional hemodynamic support. Given these developments, the efficacy of preoperative IABP may vary over time as a result of overall advances in cardiac care. An increasingly comorbid surgical population may benefit more from additional hemodynamic support. Alternatively, IABP effects may be outweighed by improvement in other areas such as surgical technique.³⁰ 

An updated RCT examining the use of prophylactic IABPs in high-risk patients undergoing CABG should be performed to answer outstanding questions of both efficacy and safety. Specifically, a new study must delineate the appropriate context for IABP use. Whether preemptive IABP therapy is warranted in high-risk patients scheduled for both elective and emergent CABG is not known. The high-risk population would be defined as having at least one of the following: LVEF ≤40%; left main stenosis ≥70%; unstable angina (current); recent MI (within 10 days); prior CABG; or a high perioperative risk estimated by either an additive EuroScore or Society of Thoracic Surgeons (STS) score. The randomized cohort should be well described in regard to acuity by the Simplified Acute Physiological Score 2,³¹ as well as severity of coronary artery disease by Canadian Cardiovascular Society class and Syntax score.³² Subjects should be stratified according to elective CABG (operating room booked >24 hours) or emergent CABG (operating room  booked <24 hours). A well-designed trial must also explore utility based on timing of IABP deployment. Most contemporary studies have assumed this to be <2 hours prior to first incision based on an RCT of 60 patients by Christenson et al.¹³ An adaptive sequential design based on two different time points for IABP placement (preoperative insertion vs >2 hours) could be refined to a single time point based on interim analysis and guidance by an independent party. 

A new RCT should also consider the use of IABPs in the context of refined surgical techniques and contemporary ventricular assist modalities. Many high-risk CABG patients now undergo OPCAB as the preferred surgical approach.^33,34 Existing data on IABPs in the setting of OPCAB are limited, but suggest a mortality benefit in these patients.^15,16,19,21 Despite more contemporary hemodynamic support devices, the IABP remains an accessible intervention in non-tertiary centers where access to mechanical support devices may be limited by lack of expertise, prohibitive cost, or uncertain safety profile. Documentation of clinical failure of IABP or sham arms necessitating alternative hemodynamic supports such as left ventricular assist device or extracorporeal membrane oxygenation may help garner insights toward IABP utility in the modern era.  

Finally, the therapeutic index of IABPs needs to be stringently evaluated. It is unclear whether previous studies truly yielded a low adverse event rate or were simply underpowered to detect harm. In addition to safety endpoints, clinically relevant surrogate outcome variables should be examined. Potential increases in the length of stay in the intensive care unit secondary to IABP use, interval changes in sequential organ failure assessment (SOFA), life-support free days, and other intermediate endpoints may garner both physiological and clinical insights.^35,36 Reporting events by means of blinded central adjudication according to standardized definitions of MACCE, bleeding severity, and vascular complications is required, in addition to adjudicating cause of death.³⁷ Based on the event rate observed in this analysis, an RCT adequately powered for mortality would require a total of 1090 patients. Similarly, to detect a magnitude of reduction in MACCE may require a total of 458 patients. 

Conclusion

In this meta-analysis of over 1200 patients, a strategy of routine preemptive IABP insertion reduced the risk of 30-day mortality and MACCE in patients undergoing CABG surgery considered at high risk for adverse perioperative events. The benefit of this strategy extended across a broad range of patient characteristics and surgical revascularization strategies; however, further data on individual cardiovascular and safety endpoints are needed. A definitive, adequately powered outcomes trial appears warranted to confirm these findings prior to routine adoption of this strategy as the standard of care. 

References

1.    Thiele H, Zeymer U, Neumann FJ, et al. Intraaortic balloon support for myocardial infarction with cardiogenic shock. N Engl J Med. 2012;367:1287-1296.

2.    Thiele H, Zeymer U, Neumann FJ, et al. Intra-aortic balloon counterpulsation in acute myocardial infarction complicated by cardiogenic shock (IABP-SHOCK II): final 12 month results of a randomised, open-label trial. Lancet. 2013;382:1638-1645.

3.    Ahmad Y, Sen S, Shun-Shin MJ, et al. Intra-aortic balloon pump therapy for acute myocardial infarction: a meta-analysis. JAMA Intern Med. 2015;175:931-939.

4.    Ranucci M, Castelvecchio S, Biondi A, et al. A randomized controlled trial of preoperative intra-aortic balloon pump in coronary patients with poor left ventricular function undergoing coronary artery bypass surgery^*. Crit Care Med 2013;41:2476-2483.

5.    Zangrillo A, Pappalardo F, Dossi R, et al. Preoperative intra aortic balloon pump to reduce mortality in coronary artery bypass graft: a meta-analysis of randomized controlled trials. Criti Care. 2015;19:10.

6.    Pilarczyk K, Boening A, Jakob H, et al. Preoperative intra-aortic counterpulsation in high-risk patients undergoing cardiac surgery: a meta-analysis of randomized controlled trialsdagger. Eur J Cardiothorac Surg. 2016;49:5-17.

7.    Poirier Y, Voisine P, Plourde G, et al. Efficacy and safety of preoperative intra-aortic balloon pump use in patients undergoing cardiac surgery: a systematic review and meta-analysis. Int J Cardiol. 2016;207:67-79.

8.    Hillis LD, Smith PK, Anderson JL, et al. 2011 ACCF/AHA guideline for coronary artery bypass graft surgery. A report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Developed in collaboration with the American Association for Thoracic Surgery, Society of Cardiovascular Anesthesiologists, and Society of Thoracic Surgeons. J Am Coll Cardiol. 2011;58:e123-e210.

9.    Windecker S, Kolh P, Alfonso F, et al. 2014 ESC/EACTS Guidelines on myocardial revascularization: the task force on myocardial revascularization of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS). Developed with the special contribution of the European Association of Percutaneous Cardiovascular Interventions (EAPCI). Eur Heart J. 2014;35:2541-2619. Epub 2014 Aug 29. 

10.    Christenson JT, Simonet F, Badel P, Schmuziger M. Evaluation of preoperative intra-aortic balloon pump support in high risk coronary patients. Eur J Cardiothorac Surg. 1997;11:1097-103; discussion 104.

11.    Christenson JT, Simonet F, Badel P, Schmuziger M. The effect of preoperative intra-aortic balloon pump support in patients with coronary artery disease, poor left-ventricular function (LVEF <40%), and hypertensive LV hypertrophy. Thorac Cardiovasc Surg. 1997;45:60-64.

12.    Christenson JT, Badel P, Simonet F, Schmuziger M. Preoperative intraaortic balloon pump enhances cardiac performance and improves the outcome of redo CABG. Ann Thorac Surg. 1997;64:1237-1244.

13.    Christenson JT, Simonet F, Badel P, Schmuziger M. Optimal timing of preoperative intraaortic balloon pump support in high-risk coronary patients. Ann Thorac Surg. 1999;68:934-939.

14.    Marra C, De Santo LS, Amarelli C, et al. Coronary artery bypass grafting in patients with severe left ventricular dysfunction: a prospective randomized study on the timing of perioperative intraaortic balloon pump support. Int J Artif Organs. 2002;25:141-146.

15.    Christenson JT, Licker M, Kalangos A. The role of intra-aortic counterpulsation in high-risk OPCAB surgery: a prospective randomized study. J Cardiac Surg. 2003;18:286-294.

16.    Qiu Z, Chen X, Xu M, et al. Evaluation of preoperative intra-aortic balloon pump in coronary patients with severe left ventricular dysfunction undergoing OPCAB surgery: early and mid-term outcomes. J Cardiovasc Surg. 2009;4:39.

17.    Wilczynski M, Krzych LJ, Bis J, Szmagala P, Ulczok R, Bochenek A. Effect of gender on efficacy of preoperative intra-aortic balloon pump in high risk patients undergoing surgical coronary revascularisation. Kardiologia Polska. 2010;68:1361-1368.

18.    Metz D, Stiller M, Silber RE, Kroll H, Hofmann HS, Diez C. [Prophylactic intraaortic balloon pumping in high-risk cardiac surgery patients]. Med Klin Intensivmed Notfmed. 2011;106:125-131.

19.    Shi M, Huang J, Pang L, Wang Y. Preoperative insertion of an intra-aortic balloon pump improved the prognosis of high-risk patients undergoing off-pump coronary artery bypass grafting. J Int Med Res. 2011;39:1163-1168.

20.    Lomivorotov VV, Boboshko VA, Efremov SM, et al. Levosimendan versus an intra-aortic balloon pump in high-risk cardiac patients. J Cardiothorac Vasc Anesth. 2012;26:596-603.

21.    Mannacio V, Di Tommaso L, De Amicis V, Stassano P, Musumeci F, Vosa C. Preoperative intraaortic balloon pump for off-pump coronary arterial revascularization. Ann Thorac Surg. 2012;93:804-809.

22.    Hosmane SR, Dawson AG. In patients coming to theatre with an intra aortic balloon pump, is it better to turn it off or keep it on while on bypass? Interact Cardiovasc Thorac Surg. 2010;11:314-321.

23.    Chen S, Yin Y, Ling Z, Krucoff MW. Short and long term effect of adjunctive intra-aortic balloon pump use for patients undergoing high risk reperfusion therapy: a meta-analysis of 10 international randomised trials. Heart. 2014;100:303-310.

24.    Dyub AM, Whitlock RP, Abouzahr LL, Cina CS. Preoperative intra-aortic balloon pump in patients undergoing coronary bypass surgery: a systematic review and meta-analysis. J Cardiac Surg. 2008;23:79-86.

25.    Theologou T, Bashir M, Rengarajan A, et al. Preoperative intra aortic balloon pumps in patients undergoing coronary artery bypass grafting. Cochrane Database of Systematic Reviews. 2011:CD004472.

26.    Sa MP, Ferraz PE, Escobar RR, et al. Prophylactic intra-aortic balloon pump in high-risk patients undergoing coronary artery bypass surgery: a meta-analysis of randomized controlled trials. Coron Artery Dis. 2012;23:480-486.

27.    Higgins JP, Altman DG, Gotzsche PC, et al. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. Br Med J. 2011;343:d5928.

28.    Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothor Surg. 1999;16:9-13.

29.    Michel P, Roques F, Nashef SA, Euro SPG. Logistic or additive EuroSCORE for high-risk patients? Eur J Cardiothorac Surg. 2003;23:684-687; discussion 7.

30.    Diodato M, Chedrawy EG. Coronary artery bypass graft surgery: the past, present, and future of myocardial revascularisation. Surg Res Practice. 2014;2014:726158.

31.    Le Gall JR, Lemeshow S, Saulnier F. A new simplified acute physiology score (SAPS II) based on a European/North American multicenter study. JAMA. 1993;270:2957-2963.

32.    Sianos G, Morel MA, Kappetein AP, et al. The SYNTAX score: an angiographic tool grading the complexity of coronary artery disease. EuroIntervention. 2005;1:219-227.

33.    Sepehripour AH, Harling L, Ashrafian H, Casula R, Athanasiou T. Does off-pump coronary revascularization confer superior organ protection in re-operative coronary artery surgery? A meta-analysis of observational studies. J Cardiothorac Surg. 2014;9:115.

34.    Dhurandhar V, Saxena A, Parikh R, et al. Outcomes of on-pump versus off-pump coronary artery bypass graft surgery in the high risk (AusSCORE >5). Heart Lung Circ. 2015;24:1216-1224. Epub 2015 Feb 23.

35.    Rubenfeld GD, Angus DC, Pinsky MR, Curtis JR, Connors AF Jr, Bernard GR. Outcomes research in critical care: results of the American Thoracic Society Critical Care Assembly Workshop on Outcomes Research. The Members of the Outcomes Research Workshop. Am J Resp Crit Care Med. 1999;160:358-367.

36.    de Grooth HJ, Geenen IL, Girbes AR, Vincent JL, Parienti JJ, Oudemans-van Straaten HM. SOFA and mortality endpoints in randomized controlled trials: a systematic review and meta-regression analysis. Crit Care. 2017;21:38.

37.    Mehran R, Rao SV, Bhatt DL, et al. Standardized bleeding definitions for cardiovascular clinical trials: a consensus report from the Bleeding Academic Research Consortium. Circulation. 2011;123:2736-2747.

From the ¹Division of Cardiology, Department of Medicine, Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada; ²Women’s College Research Institution and Cardiovascular Division, Department of Medicine, Women’s College Hospital, Toronto, Ontario, Canada; ³Division of Cardiac Surgery, Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada; ⁴Heart and Stroke Richard Lewar Centre of Excellence in Cardiovascular Research, Toronto, Ontario, Canada; and ⁵University of Toronto, Toronto, Ontario, Canada. 

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.

Manuscript submitted April 3, 2017, provisional acceptance given April 17, 2017, final version accepted May 5, 2017.

Address for correspondence: Vladimír Džavík, MD, Peter Munk Cardiac Centre, 6-246 EN Toronto General Hospital, 200 Elizabeth Street, Toronto, ON M5G 2C4. Email: vlad.dzavik@uhn.ca; or Jacob A. Udell, MD, MPH, FRCPC, Cardiovascular Division, Peter Munk Cardiac Centre, Toronto General Hospital and Women’s College Hospital, University of Toronto, 76 Grenville Street, Toronto, ON M5S 1B1, Canada. Email: jay.udell@utoronto.ca