Body Surface Area-Weighted Left Ventricular Ejection Fraction Enhances Prediction Accuracy of OPCABG Outcomes: A Large Multi-Center Cohort Study

Abstract

Background:

We hypothesized that body surface area (BSA)-weighted left ventricular ejection fraction (LVEF) (bLVEF) would represent a superior predictor of mortality in off-pump coronary artery bypass grafting (OPCABG) patients than standard predictors. LVEF is associated with worse outcomes upon OPCABG, while referring left ventricular measurements to BSA should improve predictability.

Methods:

The bLVEF was calculated by multiplying the LVEF by the BSA. The primary endpoint was all-cause mortality within 30 days of hospitalization, while secondary endpoints included major postoperative complications.

Results:

A total of 7927 patients from five leading cardiac centers participating in the Chinese Cardiac Surgery Registry were included in the final analysis, of which 7093 (89.48%) had normal LVEF, 639 (8.06%) presented heart failure with mid-range ejection fraction (HFmrEF), and 195 (2.46%) exhibited heart failure with reduced ejection fraction (HFrEF). The average bLVEF in the cohort was 109.63 ± 18.16. Both the mortality (odds ratio (OR) 0.97) and secondary endpoints (OR 0.97) followed a similar trend with increasing bLVEF, indicating that bLVEF is a more reliable predictor of adverse outcomes. The individual components of bLVEF, including BSA (area under the curve (AUC) 0.63) and LVEF (AUC 0.64), made minor contributions to mortality risk with relatively low AUC values. However, these components were less impactful than bLVEF (AUC 0.70). Notably, patients with a bLVEF less than 85 had an increased mortality risk relative to those whose bLVEF was 85 or higher (adjusted OR 4.65 (95% confidence interval (CI): 3.81–5.83; p < 0.01)).

Conclusion:

The bLVEF serves as a key predictor of mortality in OPCABG patients, effectively eliminating BSA-related bias and demonstrating a strong capacity to predict mortality.

Clinical Trial Registration:

NCT02400125, https://www.clinicaltrials.gov/study/NCT02400125.

Graphical abstract

Keywords

BSA / LVEF / OPCABG / outcomes

1. Introduction

In the 1990s, interest in performing coronary artery bypass grafting (CABG) on a beating heart, avoiding cardiopulmonary bypass, experienced a resurgence, leading to renewed focus on off-pump surgery [1, 2]. Off-pump coronary artery bypass grafting (OPCABG) has been widely utilized in patients with heart failure (HF) due to its association with reduced perioperative complications and improved long-term outcomes compared to on-pump procedures [3, 4]. The left ventricular ejection fraction (LVEF) is an indicator to evaluate HF, which associates with the outcomes in cardiac surgery [5]. Patients with low LVEF face a high risk of perioperative complications, adversely affecting both immediate and long-term survival. This risk is particularly pronounced in those with ventricular contractility impairment following cardiopulmonary bypass and cardioplegia, regardless of other comorbidities [6, 7]. In OPCABG, reduced LVEF, often associated with ischemic cardiomyopathy, is linked to an increased risk of long-term complications [8, 9, 10]. Further investigation is needed to elucidate the relationship between LVEF and perioperative outcomes in OPCABG, offering greater insight into its prognostic significance.

LVEF refers to the percentage of blood volume in the left ventricle at the end of diastole that is pumped out with each contraction, which can be expressed mathematically as LVEF(\%)=\frac{\text{Stroke volume}}{\text{End-diastolic volume}}. End-diastolic volume is influenced by several factors, including a patient’s weight, body mass index (BMI), and body surface area (BSA), the latter being a parameter calculated based on height and weight [11]. BSA is positively correlated with blood pressure and serves as a fairly accurate indicator of total body water. It is commonly used to normalize cardiac output to cardiac index and to estimate the glomerular filtration rate (GFR) [12, 13, 14, 15]. In this context, we propose normalizing the end-diastolic volume by BSA to reduce its bias on LVEF. The resulting BSA-adjusted LVEF (bLVEF) is calculated as the stroke volume divided by the normalized end-diastolic volume (bLVEF(\%)=\frac{\text{Stroke volume}}{\text{End-diastolic volume}\left(\frac{1}{\mathrm{BSA}}\right)}). This study aims to investigate the roles of LVEF, BSA, and bLVEF in a special clinical setting and to evaluate whether their interactions following OPCABG can improve patient-centered care.

Consequently, we aimed to explore the distinct effects of LVEF, BSA, and bLVEF on early clinical outcomes for OPCABG patients using data from five leading Chinese cardiovascular centers. The objectives were to investigate (1) the impact of LVEF on perioperative outcomes, (2) the correlation between BSA and LVEF, and (3) the association of bLVEF with perioperative complications and mortality rates among.

2. Materials and Methods

2.1 Study Setting and Population

We analyzed data from 7927 patients across five prominent Chinese cardiovascular centers: Beijing Anzhen Hospital, Beijing Tongren Hospital, Beijing Hospital, Peking University People’s Hospital, and Beijing Xuanwu Hospital. The data, sourced from the Chinese Cardiac Surgery Registry database, cover admissions that occurred between December 2016 and January 2021 (refer to Supplementary Fig. 1). Clinical data were collected in accordance with the Society of Thoracic Surgeons National Adult Cardiac Database (http://www.sts.org). Data reliability and comprehensiveness were ensured through established procedures, as detailed in prior publications [16]. The study protocol was approved by the Ethics Committee of Fuwai Hospital (Approval No: 2017-943) and is registered at http://www.clinicaltrials.gov (NCT02400125). Patient confidentiality was maintained by pseudo-anonymizing all data, substituting patient names with identification codes and removing private information before analysis. A data committee from the Peking University Clinical Research Institute was responsible for assessing data quality and overseeing data quality and collection. All participants received standard care, with no additional interventions, as previously described [16]. Heart failure was classified according to the European Society of Cardiology, including heart failure with reduced ejection fraction (HFrEF, LVEF 40%) and heart failure with mid-range ejection fraction (HFmrEF, LVEF 40–49%).

2.2 Predictor and Outcomes

Patient demographics and clinical features were collected and assessed, including medical histories of peripheral vascular disease, cerebrovascular events, prior myocardial infarctions, previous percutaneous coronary interventions, and New York Heart Association (NYHA) classification. Peoperative test results for serum creatinine, total cholesterol, low-density lipoprotein, blood glucose levels, and estimated glomerular filtration rate (eGFR) were also recorded. Intraoperative echocardiogram data were examined for left ventricular end-diastolic volume (LVED), left ventricular end-diastolic diameter (LVEDd), and left atrial dimension (LAD). Data on concomitant cardiac medications—including nitrate lipid drugs, catecholamines, \beta-blockers, angiotensin-converting enzyme inhibitors (ACEI), angiotensin receptor blockers (ARB), statins, aspirin, clopidogrel, and ticagrelor—were meticulously documented. The primary outcome was in-hospital all-cause mortality within 30 days, while secondary outcomes included severe postoperative complications such as the postoperative use of extracorporeal membrane oxygenation (ECMO), multiorgan failure, intra-aortic balloon pump (IABP) postoperative usage, postoperative strokes, and myocardial infarctions (MIs). The BSA follows BSA\left({\mathrm{m}}^2\right)=\sqrt{\frac{Ht(\mathrm{cm})*Wt(\mathrm{kg})}{3600}} [11].

2.3 Statistical Analysis

Missing values and outliers were addressed using multiple imputations via the Multivariate Imputation by Chained Equations (MICE) package [17]. As the database was systematically monitored by a data committee, missing values and outliers represented less than 2% of all metrics. We assumed that missing data and misrecordings occurred randomly [18], and used predictive mean matching to generate five imputed datasets suitable for logistic model fitting [19]. For multivariate logistic regression, we adjusted for the following variables based on clinical expertise: age, gender, smoking within two weeks prior to surgery, diabetes, hypertension, hyperlipidemia, the last serum creatinine test before surgery, the last total cholesterol test, the last low-density lipoprotein test, the last blood glucose test, preoperative eGFR, and history of cerebrovascular events. We used weight-of-Evidence binning which is a technique for binning both continuous and categorical independent variables in a way that provides the most robust bifurcation of the data against the dependent variable. This technique was implemented by the woebin function from R. Continuous variables were categorized according to established cutoffs. The bLVEF was optimally binned using evidence-based segmentation via the scorecard package (https://CRAN.R-project.org/package=scorecard), and coefficients were calculated using Spearman correction. Continuous variables are presented as mean \pm standard deviation and were compared between groups using one-way Analysis of Variance (ANOVA). Categorical variables are presented as counts (percentages) and were compared between groups using Pearson’s chi-square test or Fisher’s exact test, as appropriate. The Cochran-Armitage trend test was used to assess trends across ordered LVEF categories. The area under the curve (AUC) for the receiver operating characteristics was assessed using the DeLong method. Using the generalized additive model to evaluate the nonlinear relationship. Sample size calculation indicated that 695 patients with bLVEF 85 would provide 99.50% power to detect a minimum clinically meaningful mortality rate of 5.18%, with a two-sided alpha of 0.05, compared to patients with bLVEF \ge85. All analyses were performed using R version 3.4.2 (The R Foundation for Statistical Computing, Vienna, Austria) (http://www.r-project.org/).

3. Results

3.1 Baseline

A total of 7927 patients were included in the final analysis, among whom 7093 (89.48%) had normal LVEF, 639 (8.06%) had HFmrEF, and 195 (2.46%) had HFrEF. The cohort’s mean age was 62.61 \pm 8.70 years consisting of 6051 (76.33%) males and 3573 (45.07%) current or former smokers (Table 1). As expected, HFrEF or HFmrEF exhibited higher rates of comorbidities, including diabetes mellitus, abnormal serum creatinine, glucose and eGFR (p 0.01 for trend), as well as higher previous myocardial infarction rates (Table 1, p for trend 0.01). Patients with low LVEF also had higher LVEDd and LAD, poorer New York Heart Association Functional Classification, and lower statins use (p for trend = 0.07). However, the use of Statins, ACEI/ARB did not differ significantly from patients with normal range LVEF (Table 1). These findings suggest that patients with reduced LVEF have a poorer preoperative baseline condition compared to those with normal LVEF. Baseline characteristics stratified by bLVEF are provided in the Supplementary Material.

3.2 HFrEF and HFmrEF LVEF Is a Negative Prognostic Factor

We further analyzed outcomes across LVEF categories. As shown in Table 2, HFrEF and HFmrEF patients experienced higher rates of multiorgan failure, acute kidney injury (AKI), and the cumulative mechanical ventilation time (Table 2, p for trend 0.01). Mortality, intensive care unit (ICU) length of stay, use of IABP, and use of ECMO increased with decreasing LVEF (p for trend = 0.05), while postoperative stroke (p for trend = 0.88), postoperative MI (p for trend = 0.52) and re-admission to ICU (p for trend = 0.51) did not show statistical differences (Table 2). Multivariable regression analysis identified HFrEF (adjusted odds ratio [OR] 6.50, 95% confidence interval [CI] 3.02–12.68, p 0.01, Table 3, Fig. 1B) and HFmrEF as significant negative predictors of survival, suggesting that an increase in the odds of mortality compared to the reference group (LVEF \ge50%). Likewise, patients with lower LVEF experienced a greater incidence of adverse events (adjusted OR 6.98, 95% CI 4.96–9.72, p 0.01, Table 3). These findings underscore low LVEF as a significant negative prognostic factor for both postoperative survival and adverse outcomes.

3.3 LVEF Normalized for BSA

The mean BSA is 1.82 \pm 0.17 m², which showed a slightly negative relationship with the LVEF (R = –0.05, p 0.01, Supplementary Fig. 2) and this also reached statistical difference when compared to 3 LVEF groups (p for trend = 0.03). As anticipated, a high BSA was linked to decreased mortality (adjusted OR 0.76, 95% CI 0.62–0.91, p 0.01, Table 3, Fig. 1A) as well as secondary outcomes (adjusted OR 0.93, 95% CI 0.86–1, p = 0.05, Table 3, Fig. 1D). Given that LVEF has not been previously analyzed in conjunction with BSA, we hypothesized that bLVEF, which is defined as LVEF multiplied by BSA, may serve as a more effective predictor of postoperative prognosis. We found that the bLVEF decreased with LVEF (p for trend 0.01, Fig. 1, Supplementary Tables 2,3), but revealed a different relationship with the mortality than LVEF alone, that is, a plateau appeared near to the 85% of bLVEF.

3.4 The bLVEF Is a Better Predictor of Mortality

To evaluate bLVEF as a predictor of postoperative mortality and adverse events, as well as to compare its predictive efficacy with other outcome predictors, we performed univariate and multivariate logistic regression analyses. Our results indicated a significant correlation between bLVEF and postoperative mortality, with findings showing an adjusted OR of 0.97 (95% CI: 0.96–0.98, p 0.01) for mortality risk (refer to Table 3 and Fig. 1C). Additionally, bLVEF was strongly linked to secondary outcomes, revealing an adjusted OR of 0.97 (95% CI: 0.97–0.98, p 0.01 see Fig. 1D,E), which exhibited a consistent trend alongside decreases in bLVEF levels (see Table 3 and Fig. 1F). These results imply that bLVEF is a significant negative predictor for both mortality and adverse postoperative events.

To enhance the practicality of bLVEF in clinical applications, we employed a tree-based segmentation method to categorize bLVEF into discrete ranges. Our analysis revealed four distinct categories: [0, 85), [85, 120), [120, 135), and [135, \mathrm{\infty }) (adjusted K-S statistic of 0.42, p = 0.01, see Fig. 2A and Supplementary Fig. 3). The categorical version of bLVEF demonstrated comparable discrimination abilities to its numerical counterpart in predicting mortality, achieving an AUC of 0.70 (p 0.01). Notably, among the variables analyzed, bLVEF emerged as the most significant predictor of mortality, surpassing both BSA and LVEF. While BSA (AUC 0.63) and LVEF (AUC 0.64) showed some association with mortality risk, their AUC values were lower and less significant compared to bLVEF (DeLong test, p 0.01; see Fig. 2B,C). We propose to use the 85 as the threshold to categorize the patients and high-risk and low-risk groups. We added the relevant description to the manuscript. Importantly, patients with a bLVEF of less than 85 exhibited a substantially higher risk of mortality compared to those with a bLVEF of 85 or greater, with an OR of 4.65 (95% CI: 3.81–5.83, p 0.01; see Fig. 2B,C). This suggests that a bLVEF threshold of less than 85 is a more reliable indicator of high-risk patients than LVEF or BSA alone. For the secondary outcome, there was no statistically significant difference in the AUC between LVDD and BLVDD (DeLong test, p = 0.5661, see Fig. 2D,E).

4. Discussion

In this multi-center cohort investigation, we found that (1) low LVEF is a significant predictor of postoperative survival and adverse events; (2) BSA exhibited a slight negative correlation with LVEF; (3) bLVEF emerged as the most consistent predictor of mortality compared to BSA and LVEF; and (4) a bLVEF cutoff of 85 effectively identified patients at high risk of mortality.

Cardiac surgery improves survival in patients with advanced left ventricular dysfunction compared to medical management [20]. However, low preoperative LVEF is a known risk factor for the adverse outcomes following CABG [21]. Since Benetti et al. [2] successfully performed OPCABG with the emphasis on myocardial protection and the surgeon’s increasing proficiency in bypass grafting, OPCABG has become a well-developed and safe procedure [22]. Recent years have seen increased interest in its use for severe coronary artery disease, with several studies highlighting its efficacy in patients with low LVEF [23]. Shennib et al. [24] reported favorable outcomes and low mortality rates in patients with impaired ventricular function undergoing off-pump surgical revascularization. Arom et al. [22] discovered that performing multivessel coronary artery bypass using the OPCAB technique is both suitable and feasible for patients with left ventricular function at or below 30%. In addition, Ueki et al. [10] reported that OPCABG is linked to significantly lower rates of early mortality and morbidity in patients with an ejection fraction of less than 30%. Despite these findings, comparative studies on OPCABG across different EF categories remain limited. In our cohort, 7093 (89.48%) patients had normal LVEF, 639 (8.06%) had HFmrEF, and 195 (2.46%) had HFrEF, consistent with previously reported population distributions [25]. Patients with HFrEF or HFmrEF exhibited higher rates of diabetes, hypertension, impaired kidney function, and prior myocardial infarctions, along with elevated LVEDd and LAD, worse NYHA Functional Classification scores, and greater statin use (p for trend 0.01). These findings highlight the poorer preoperative baseline condition of patients with low LVEF compared to those with normal LVEF.

CAD with reduced EF presents a significant challenge to OPCABG. Proper exposure of the anastomosis site during OPCABG necessitates cardiac manipulation, which can compress the left ventricular outflow tract and lead to hypotension. In patients with low LVEF, the limited cardiac reserve makes them highly susceptible to abrupt drops in blood pressure. Additionally, patients with left main trunk and three-vessel disease are at increased risk of malignant arrhythmias during cardiac manipulation [4, 26, 27]. Therefore, it is necessary to implement appropriate emergency measures and coordinate with anesthesiologists before repositioning the heart. The use of IABP helps to improve the success rate of OPCABG surgery [28]. Suzuki et al. [29] indicated that preoperative IABP therapy in high-risk coronary patients effectively prevents hemodynamic instability and yields surgical outcomes similar to those seen in moderate to low-risk patients. In our study, HFrEF and HFmrEF were identified as strong negative prognostic factors for mortality, suggesting that a decrease in the odds of mortality compared to the reference group (LVEF \ge50%). Similarly, adverse event rates rose by 63% with each LVEF category reduction. Surgeons must carefully assess intraoperative conditions to determine whether to proceed with OPCABG. If cardiac function deteriorates significantly during the procedure, transitioning to on-pump coronary artery bypass grafting should be considered to mitigate further risks.

Echocardiography is extensively utilized for diagnosing and managing cardiac conditions, particularly in the preoperative assessment of patients undergoing cardiac surgery [30, 31]. However, cardiac surgeons often rely on unstandardized absolute values when interpreting echocardiographic parameters, despite the influence of body mass factors on heart structure [32]. Assessing echocardiographic indicators solely based on absolute values does not enable precise diagnosis of cardiac conditions [33]. BSA, a metric for characterizing body size, is frequently employed to normalize mass and volume and to index physiological parameters related to cardiovascular disease [34, 35]. A study has indicated that left ventricle (LV) diameter provides a basic and simplified evaluation of a three-dimensional structure, failing to account for the more intricate variations in ventricular shape or size [36]. To address this, the American Society of Echocardiography and the European Association of Cardiovascular Imaging recommended indexing echocardiographic parameters, such as right and left ventricular sizes, to BSA [35]. Moreover, the Simpson’s biplane method, which uses orthogonal long-axis views, allows for a more accurate calculation of LV volume that can be indexed to BSA for improved precision [37]. LVEF is a parameter obtained from echocardiography that measures the amount of blood ejected from the heart’s left ventricle—the primary pumping chamber—with each contraction [38]. Furthermore, LVEF is acknowledged as an independent risk factor for CABG. A lower ejection fraction correlates with increased perioperative mortality and reduced five-year survival rates [39, 40]. While the conventional definition of the normal range for EF is derived from the general population, encompassing groups with varying clinical characteristics and prognoses. Thuijs DJFM et al. [25] found a U-shaped association between LVEF and all-cause mortality three years after CABG. In our study, we identified a J-shaped curve in LVEF when predicting perioperative mortality in patients undergoing OPCABG. This observation is of considerable clinical significance, as it suggests a potential for misinterpretation among cardiac clinicians. Specifically, clinicians may incorrectly assume that patients with an LVEF at the relatively “good” level have adequate cardiac function, thus deeming them to be at a low risk for perioperative complications. However, this oversimplification can obscure the true risks associated with the procedure. Given the complexity of heart failure and the multitude of factors influencing perioperative outcomes, it is imperative to consider all relevant risk factors beyond just LVEF when making surgical decisions for patients undergoing OPCABG. Relying solely on LVEF could lead to incomplete assessments of a patient’s functional status, potentially compromising patient safety. To mitigate this risk of bias, we advocate for the normalization of LVEF to BSA, a practice that enhances the reliability of prognostic assessments. By doing so, we established a critical numerical threshold of 85 as a ‘divider’ that serves to refine the prediction of perioperative mortality. This threshold represents a key parameter in stratifying patient risk, enabling more accurate and personalized surgical decision-making. It is worth noting that the observed patterns are hypothesis-generating and require validation in cohorts with harmonized analytical frameworks. Ultimately, our findings underscore the importance of incorporating LVEF normalization into clinical practice, ensuring that all relevant factors are weighed appropriately to optimize patient outcomes in OPCABG. This approach can significantly improve the predictive accuracy regarding mortality and enhance the overall management of patients undergoing coronary revascularization.

There are limitations in our study. Firstly, the retrospective cohort design relies on historical patient data, which restricts the availability of comprehensive preoperative activity tolerance information. Additionally, the extended follow-up period resulted in a considerable loss to follow-up. Secondly, common post-surgery complications such as severe heart failure and atrial fibrillation were not included due to challenges in accurately documenting them during follow-up assessments. Thirdly, the study did not discuss gender differences, which will be further explored in future research. Lastly, body surface area is calculated using a formula based on weight and height, which does not reflect the true surface area of an individual. Furthermore, factors like age, gender, and race may confound the BSA measurement, highlighting the need for future research to investigate the correlating variables associated with BSA.

5. Conclusions

bLVEF serves as a key predictor of mortality in OPCABG, effectively eliminating BSA-related bias and demonstrating a strong capacity to predict mortality.

References

[1] Grover FL, Shroyer AL, Hammermeister K, Edwards FH, Ferguson TB, Jr, Dziuban SW, Jr, et al. A decade’s experience with quality improvement in cardiac surgery using the Veterans Affairs and Society of Thoracic Surgeons national databases. Annals of Surgery. 2001; 234: 464–472; discussion 472–474. https://doi.org/10.1097/00000658-200110000-00006.

[2] Benetti FJ, Naselli G, Wood M, Geffner L. Direct myocardial revascularization without extracorporeal circulation. Experience in 700 patients. Chest. 1991; 100: 312–316. https://doi.org/10.1378/chest.100.2.312.

[3] Lamy A, Devereaux PJ, Prabhakaran D, Taggart DP, Hu S, Straka Z, et al. Five-Year Outcomes after Off-Pump or On-Pump Coronary-Artery Bypass Grafting. The New England Journal of Medicine. 2016; 375: 2359–2368. https://doi.org/10.1056/NEJMoa1601564.

[4] Shroyer AL, Hattler B, Wagner TH, Collins JF, Baltz JH, Quin JA, et al. Five-Year Outcomes after On-Pump and Off-Pump Coronary-Artery Bypass. The New England Journal of Medicine. 2017; 377: 623–632. https://doi.org/10.1056/NEJMoa1614341.

[5] Mele D, Nardozza M, Ferrari R. Left ventricular ejection fraction and heart failure: an indissoluble marriage? European Journal of Heart Failure. 2018; 20: 427–430. https://doi.org/10.1002/ejhf.1071.

[6] Antunes MJ. Commentary: Low left ventricular ejection fraction in coronary artery bypass grafting: Accept or control factors determining survival? The Journal of Thoracic and Cardiovascular Surgery. 2022; 163: 120–121. https://doi.org/10.1016/j.jtcvs.2020.03.111.

[7] Alderman EL, Fisher LD, Litwin P, Kaiser GC, Myers WO, Maynard C, et al. Results of coronary artery surgery in patients with poor left ventricular function (CASS). Circulation. 1983; 68: 785–795. https://doi.org/10.1161/01.cir.68.4.785.

[8] Maile MD, Mathis MR, Habib RH, Schwann TA, Engoren MC. Association of Both High and Low Left Ventricular Ejection Fraction With Increased Risk After Coronary Artery Bypass Grafting. Heart, Lung & Circulation. 2021; 30: 1091–1099. https://doi.org/10.1016/j.hlc.2020.11.005.

[9] Magee MJ, Coombs LP, Peterson ED, Mack MJ. Patient selection and current practice strategy for off-pump coronary artery bypass surgery. Circulation. 2003; 108: II9–II14. https://doi.org/10.1161/01.cir.0000089187.51855.77.

[10] Ueki C, Miyata H, Motomura N, Sakaguchi G, Akimoto T, Takamoto S. Off-pump versus on-pump coronary artery bypass grafting in patients with left ventricular dysfunction. The Journal of Thoracic and Cardiovascular Surgery. 2016; 151: 1092–1098. https://doi.org/10.1016/j.jtcvs.2015.11.023.

[11] Mosteller RD. Simplified calculation of body-surface area. The New England Journal of Medicine. 1987; 317: 1098. https://doi.org/10.1056/NEJM198710223171717.

[12] Hume R, Weyers E. Relationship between total body water and surface area in normal and obese subjects. Journal of Clinical Pathology. 1971; 24: 234–238. https://doi.org/10.1136/jcp.24.3.234.

[13] Heerspink HJL, Stefánsson BV, Correa-Rotter R, Chertow GM, Greene T, Hou FF, et al. Dapagliflozin in Patients with Chronic Kidney Disease. The New England Journal of Medicine. 2020; 383: 1436–1446. https://doi.org/10.1056/NEJMoa2024816.

[14] Verbraecken J, Van de Heyning P, De Backer W, Van Gaal L. Body surface area in normal-weight, overweight, and obese adults. A comparison study. Metabolism: Clinical and Experimental. 2006; 55: 515–524. https://doi.org/10.1016/j.metabol.2005.11.004.

[15] Lang RM, Badano LP, Mor-Avi V, Afilalo J, Armstrong A, Ernande L, et al. Recommendations for cardiac chamber quantification by echocardiography in adults: an update from the American Society of Echocardiography and the European Association of Cardiovascular Imaging. Journal of the American Society of Echocardiography. 2015; 28: 1–39.e14. https://doi.org/10.1016/j.echo.2014.10.003.

[16] Rao C, Zhang H, Gao H, Zhao Y, Yuan X, Hua K, et al. The Chinese Cardiac Surgery Registry: Design and Data Audit. The Annals of Thoracic Surgery. 2016; 101: 1514–1520. https://doi.org/10.1016/j.athoracsur.2015.09.038.

[17] Zhang Z. Multiple imputation with multivariate imputation by chained equation (MICE) package. Annals of Translational Medicine. 2016; 4: 30. https://doi.org/10.3978/j.issn.2305-5839.2015.12.63.

[18] Ibrahim JG, Chu H, Chen MH. Missing data in clinical studies: issues and methods. Journal of Clinical Oncology. 2012; 30: 3297–3303. https://doi.org/10.1200/JCO.2011.38.7589.

[19] Schafer JL, Graham JW. Missing data: our view of the state of the art. Psychological Methods. 2002; 7: 147–177.

[20] Mahesh B, Peddaayyavarla P, Ong LP, Gardiner S, Nashef SAM. Cardiac surgery improves survival in advanced left ventricular dysfunction: multivariate analysis of a consecutive series of 4491 patients over an 18-year period. European Journal of Cardio-Thoracic Surgery. 2016; 50: 857–866. https://doi.org/10.1093/ejcts/ezw134.

[21] Shapira I, Isakov A, Yakirevich V, Topilsky M. Long-term results of coronary artery bypass surgery in patients with severely depressed left ventricular function. Chest. 1995; 108: 1546–1550. https://doi.org/10.1378/chest.108.6.1546.

[22] Arom KV, Flavin TF, Emery RW, Kshettry VR, Petersen RJ, Janey PA. Is low ejection fraction safe for off-pump coronary bypass operation? The Annals of Thoracic Surgery. 2000; 70: 1021–1025. https://doi.org/10.1016/s0003-4975(00)01761-6.

[23] Cao J, Yu M, Xiao Y, Dong R, Wang J. Effects of different surgical strategies and left ventricular remodelling on the outcomes of coronary artery bypass grafting in heart failure patients with reduced ejection fraction. Frontiers in Cardiovascular Medicine. 2024; 11: 1398700. https://doi.org/10.3389/fcvm.2024.1398700.

[24] Shennib H, Endo M, Benhamed O, Morin JF. Surgical revascularization in patients with poor left ventricular function: on- or off-pump? The Annals of Thoracic Surgery. 2002; 74: S1344–S1347. https://doi.org/10.1016/s0003-4975(02)03966-8.

[25] Thuijs DJFM, Milojevic M, Stone GW, Puskas JD, Serruys PW, Sabik JF, 3rd, et al. Impact of left ventricular ejection fraction on clinical outcomes after left main coronary artery revascularization: results from the randomized EXCEL trial. European Journal of Heart Failure. 2020; 22: 871–879. https://doi.org/10.1002/ejhf.1681.

[26] Møller CH, Perko MJ, Lund JT, Andersen LW, Kelbæk H, Madsen JK, et al. Three-year follow-up in a subset of high-risk patients randomly assigned to off-pump versus on-pump coronary artery bypass surgery: the Best Bypass Surgery trial. Heart (British Cardiac Society). 2011; 97: 907–913. https://doi.org/10.1136/hrt.2010.211680.

[27] Benedetto U, Puskas J, Kappetein AP, Brown WM, 3rd, Horkay F, Boonstra PW, et al. Off-Pump Versus On-Pump Bypass Surgery for Left Main Coronary Artery Disease. Journal of the American College of Cardiology. 2019; 74: 729–740. https://doi.org/10.1016/j.jacc.2019.05.063.

[28] Craver JM, Murrah CP. Elective intraaortic balloon counterpulsation for high-risk off-pump coronary artery bypass operations. The Annals of Thoracic Surgery. 2001; 71: 1220–1223. https://doi.org/10.1016/s0003-4975(00)02685-0.

[29] Suzuki T, Okabe M, Handa M, Yasuda F, Miyake Y. Usefulness of preoperative intraaortic balloon pump therapy during off-pump coronary artery bypass grafting in high-risk patients. The Annals of Thoracic Surgery. 2004; 77: 2056–2059; discussion 2059–2060. https://doi.org/10.1016/j.athoracsur.2003.12.027.

[30] Metkus TS, Thibault D, Grant MC, Badhwar V, Jacobs JP, Lawton J, et al. Transesophageal Echocardiography in Patients Undergoing Coronary Artery Bypass Graft Surgery. Journal of the American College of Cardiology. 2021; 78: 112–122. https://doi.org/10.1016/j.jacc.2021.04.064.

[31] Patel MR, Calhoon JH, Dehmer GJ, Grantham JA, Maddox TM, Maron DJ, et al. ACC/AATS/AHA/ASE/ASNC/SCAI/SCCT/STS 2017 Appropriate Use Criteria for Coronary Revascularization in Patients With Stable Ischemic Heart Disease: A Report of the American College of Cardiology Appropriate Use Criteria Task Force, American Association for Thoracic Surgery, American Heart Association, American Society of Echocardiography, American Society of Nuclear Cardiology, Society for Cardiovascular Angiography and Interventions, Society of Cardiovascular Computed Tomography, and Society of Thoracic Surgeons. Journal of the American College of Cardiology. 2017; 69: 2212–2241. https://doi.org/10.1016/j.jacc.2017.02.001.

[32] Zhu Z, Li Y, Zhang F, Steiger S, Guo C, Liu N, et al. Prediction of Male Coronary Artery Bypass Grafting Outcomes Using Body Surface Area Weighted Left Ventricular End-diastolic Diameter: Multicenter Retrospective Cohort Study. Interactive Journal of Medical Research. 2023; 12: e45898. https://doi.org/10.2196/45898.

[33] Vitale G, Galderisi M, Pivonello R, Spinelli L, Ciccarelli A, de Divitiis O, et al. Prevalence and determinants of left ventricular hypertrophy in acromegaly: impact of different methods of indexing left ventricular mass. Clinical Endocrinology. 2004; 60: 343–349. https://doi.org/10.1111/j.1365-2265.2004.01985.x.

[34] Ristow B, Ali S, Na B, Turakhia MP, Whooley MA, Schiller NB. Predicting heart failure hospitalization and mortality by quantitative echocardiography: is body surface area the indexing method of choice? The Heart and Soul Study. Journal of the American Society of Echocardiography. 2010; 23: 406–413. https://doi.org/10.1016/j.echo.2010.01.019.

[35] Zafrir B, Salman N, Crespo-Leiro MG, Anker SD, Coats AJ, Ferrari R, et al. Body surface area as a prognostic marker in chronic heart failure patients: results from the Heart Failure Registry of the Heart Failure Association of the European Society of Cardiology. European Journal of Heart Failure. 2016; 18: 859–868. https://doi.org/10.1002/ejhf.551.

[36] Bellenger NG, Burgess MI, Ray SG, Lahiri A, Coats AJ, Cleland JG, et al. Comparison of left ventricular ejection fraction and volumes in heart failure by echocardiography, radionuclide ventriculography and cardiovascular magnetic resonance; are they interchangeable? European Heart Journal. 2000; 21: 1387–1396. https://doi.org/10.1053/euhj.2000.2011.

[37] Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. Journal of the American Society of Echocardiography. 2005; 18: 1440–1463. https://doi.org/10.1016/j.echo.2005.10.005.

[38] Triposkiadis F, Giamouzis G, Boudoulas KD, Karagiannis G, Skoularigis J, Boudoulas H, et al. Left ventricular geometry as a major determinant of left ventricular ejection fraction: physiological considerations and clinical implications. European Journal of Heart Failure. 2018; 20: 436–444. https://doi.org/10.1002/ejhf.1055.

[39] Shah S, Benedetto U, Caputo M, Angelini GD, Vohra HA. Comparison of the survival between coronary artery bypass graft surgery versus percutaneous coronary intervention in patients with poor left ventricular function (ejection fraction <30%): a propensity-matched analysis. European Journal of Cardio-Thoracic Surgery. 2019; 55: 238–246. https://doi.org/10.1093/ejcts/ezy236.

[40] Yan P, Liu T, Zhang K, Cao J, Dang H, Song Y, et al. Development and Validation of a Novel Nomogram for Preoperative Prediction of In-Hospital Mortality After Coronary Artery Bypass Grafting Surgery in Heart Failure With Reduced Ejection Fraction. Frontiers in Cardiovascular Medicine. 2021; 8: 709190. https://doi.org/10.3389/fcvm.2021.709190.

Funding

Beijing Science and Technology Commission of China(D171100002917001)

Beijing Science and Technology Commission of China(D171100002917003)

Beijing Anzhen Hospital High Level Research Funding(2024AZC1003)