[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Figure.
Kaplan-Meier Estimates After Isolated Coronary Artery Bypass Surgery According to Heart Failure (HF) Status and Preserved vs Reduced Ejection Fraction (EF) in Sweden Between 2001 and 2013
Kaplan-Meier Estimates After Isolated Coronary Artery Bypass Surgery According to Heart Failure (HF) Status and Preserved vs Reduced Ejection Fraction (EF) in Sweden Between 2001 and 2013

A, In 41 906 patients. B, In 39 395 patients. C, In 41 906 patients (note that the y-axis starts at 94%). pEF indicates preserved ejection fraction; rEF, reduced ejection fraction.

Table 1.  
Baseline Characteristicsa
Baseline Characteristicsa
Table 2.  
Event Rates and Relative Risks for Clinical Outcomesa
Event Rates and Relative Risks for Clinical Outcomesa
Table 3.  
Event Rates and Relative Risks for Death Within 30 Days of Surgerya
Event Rates and Relative Risks for Death Within 30 Days of Surgerya
Table 4.  
Event Rates and Relative Risks for Clinical Outcomes by Heart Failure (HF) and Ejection Fraction (EF)a
Event Rates and Relative Risks for Clinical Outcomes by Heart Failure (HF) and Ejection Fraction (EF)a
Table 5.  
Event Rates and Relative Risks for Death Within 30 Days of Surgery by Heart Failure (HF) and Ejection Fraction (EF)a
Event Rates and Relative Risks for Death Within 30 Days of Surgery by Heart Failure (HF) and Ejection Fraction (EF)a
1.
Lloyd-Jones  DM, Larson  MG, Leip  EP,  et al; Framingham Heart Study.  Lifetime risk for developing congestive heart failure: the Framingham Heart Study.  Circulation. 2002;106(24):3068-3072.PubMedGoogle ScholarCrossref
2.
Hillis  GS, Zehr  KJ, Williams  AW,  et al.  Outcome of patients with low ejection fraction undergoing coronary artery bypass grafting: renal function and mortality after 3.8 years.  Circulation. 2006;114(1)(suppl):I414-I419.PubMedGoogle Scholar
3.
Yoo  JS, Kim  JB, Jung  SH, Choo  SJ, Chung  CH, Lee  JW.  Coronary artery bypass grafting in patients with left ventricular dysfunction: predictors of long-term survival and impact of surgical strategies.  Int J Cardiol. 2013;168(6):5316-5322.PubMedGoogle ScholarCrossref
4.
Wrobel  K, Stevens  SR, Jones  RH,  et al.  Influence of baseline characteristics, operative conduct, and postoperative course on 30-day outcomes of coronary artery bypass grafting among patients with left ventricular dysfunction: results from the Surgical Treatment for Ischemic Heart Failure (STICH) trial.  Circulation. 2015;132(8):720-730.PubMedGoogle ScholarCrossref
5.
Bhatia  RS, Tu  JV, Lee  DS,  et al.  Outcome of heart failure with preserved ejection fraction in a population-based study.  N Engl J Med. 2006;355(3):260-269.PubMedGoogle ScholarCrossref
6.
Steinberg  BA, Zhao  X, Heidenreich  PA,  et al; Get With the Guidelines Scientific Advisory Committee and Investigators.  Trends in patients hospitalized with heart failure and preserved left ventricular ejection fraction: prevalence, therapies, and outcomes.  Circulation. 2012;126(1):65-75.PubMedGoogle ScholarCrossref
7.
Hwang  SJ, Melenovsky  V, Borlaug  BA.  Implications of coronary artery disease in heart failure with preserved ejection fraction.  J Am Coll Cardiol. 2014;63(25, pt A):2817-2827.PubMedGoogle ScholarCrossref
8.
Lund  LH, Donal  E, Oger  E,  et al; KaRen Investigators.  Association between cardiovascular vs. non-cardiovascular co-morbidities and outcomes in heart failure with preserved ejection fraction.  Eur J Heart Fail. 2014;16(9):992-1001.PubMedGoogle ScholarCrossref
9.
Marui  A, Nishiwaki  N, Komiya  T,  et al; CREDO-Kyoto CABG Registry Cohort-2 Investigators.  Comparison of 5-year outcomes after coronary artery bypass grafting in heart failure patients with versus without preserved left ventricular ejection fraction (from the CREDO-Kyoto CABG Registry Cohort-2).  Am J Cardiol. 2015;116(4):580-586.PubMedGoogle ScholarCrossref
10.
Donal  E, Thebault  C, Lund  LH,  et al.  Heart failure with a preserved ejection fraction additive value of an exercise stress echocardiography.  Eur Heart J Cardiovasc Imaging. 2012;13(8):656-665.PubMedGoogle ScholarCrossref
11.
Paulus  WJ, Tschöpe  C.  A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation.  J Am Coll Cardiol. 2013;62(4):263-271.PubMedGoogle ScholarCrossref
12.
Jernberg  T, Attebring  MF, Hambraeus  K,  et al.  The Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART).  Heart. 2010;96(20):1617-1621.PubMedGoogle ScholarCrossref
13.
Harnek  J, Nilsson  J, Friberg  O,  et al.  The 2011 outcome from the Swedish health care registry on heart disease (SWEDEHEART).  Scand Cardiovasc J. 2013;47(suppl 62):1-10.PubMedGoogle ScholarCrossref
14.
Ludvigsson  JF, Otterblad-Olausson  P, Pettersson  BU, Ekbom  A.  The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research.  Eur J Epidemiol. 2009;24(11):659-667.PubMedGoogle ScholarCrossref
15.
Dalén  M, Ivert  T, Holzmann  MJ, Sartipy  U.  Coronary artery bypass grafting in patients 50 years or younger: a Swedish nationwide cohort study.  Circulation. 2015;131(20):1748-1754.PubMedGoogle ScholarCrossref
16.
Dalén  M, Ivert  T, Holzmann  MJ, Sartipy  U.  Household disposable income and long-term survival after cardiac surgery: a Swedish nationwide cohort study in 100,534 patients.  J Am Coll Cardiol. 2015;66(17):1888-1897.PubMedGoogle ScholarCrossref
17.
Ludvigsson  JF, Andersson  E, Ekbom  A,  et al.  External review and validation of the Swedish National Inpatient Register.  BMC Public Health. 2011;11:450.PubMedGoogle ScholarCrossref
18.
Emilsson  L, Lindahl  B, Köster  M, Lambe  M, Ludvigsson  JF.  Review of 103 Swedish Healthcare Quality Registries.  J Intern Med. 2015;277(1):94-136.PubMedGoogle ScholarCrossref
19.
Ingelsson  E, Arnlöv  J, Sundström  J, Lind  L.  The validity of a diagnosis of heart failure in a hospital discharge register.  Eur J Heart Fail. 2005;7(5):787-791.PubMedGoogle ScholarCrossref
20.
White  IR, Royston  P, Wood  AM.  Multiple imputation using chained equations: issues and guidance for practice.  Stat Med. 2011;30(4):377-399.PubMedGoogle ScholarCrossref
21.
White  IR, Royston  P.  Imputing missing covariate values for the Cox model.  Stat Med. 2009;28(15):1982-1998.PubMedGoogle ScholarCrossref
22.
Dobrin  JS, Lebeche  D.  Diabetic cardiomyopathy: signaling defects and therapeutic approaches.  Expert Rev Cardiovasc Ther. 2010;8(3):373-391.PubMedGoogle ScholarCrossref
23.
Lund  LH, Benson  L, Dahlström  U, Edner  M, Friberg  L.  Association between use of β-blockers and outcomes in patients with heart failure and preserved ejection fraction.  JAMA. 2014;312(19):2008-2018.PubMedGoogle ScholarCrossref
24.
Thorvaldsen  T, Benson  L, Ståhlberg  M, Dahlström  U, Edner  M, Lund  LH.  Triage of patients with moderate to severe heart failure: who should be referred to a heart failure center?  J Am Coll Cardiol. 2014;63(7):661-671.PubMedGoogle ScholarCrossref
25.
Lee  DS, Gona  P, Vasan  RS,  et al.  Relation of disease pathogenesis and risk factors to heart failure with preserved or reduced ejection fraction: insights from the Framingham Heart Study of the National Heart, Lung, and Blood Institute.  Circulation. 2009;119(24):3070-3077.PubMedGoogle ScholarCrossref
26.
Nashef  SA, Roques  F, Sharples  LD,  et al. EuroSCORE II. Eur J Cardiothorac Surg. 2012;41(4):734-744. Medline:22378855
27.
Shahian  DM, O’Brien  SM, Filardo  G,  et al; Society of Thoracic Surgeons Quality Measurement Task Force.  The Society of Thoracic Surgeons 2008 cardiac surgery risk models, part 1: coronary artery bypass grafting surgery.  Ann Thorac Surg. 2009;88(1)(suppl):S2-S22.PubMedGoogle ScholarCrossref
28.
Dunlay  SM, Roger  VL, Weston  SA, Jiang  R, Redfield  MM.  Longitudinal changes in ejection fraction in heart failure patients with preserved and reduced ejection fraction.  Circ Heart Fail. 2012;5(6):720-726.PubMedGoogle ScholarCrossref
29.
Lund  LH, Aaronson  KD, Mancini  DM.  Validation of peak exercise oxygen consumption and the Heart Failure Survival Score for serial risk stratification in advanced heart failure.  Am J Cardiol. 2005;95(6):734-741.PubMedGoogle ScholarCrossref
30.
Lund  LH, Edwards  LB, Kucheryavaya  AY,  et al; International Society of Heart and Lung Transplantation.  The registry of the International Society for Heart and Lung Transplantation: thirty-first official adult heart transplant report: 2014: focus theme: retransplantation.  J Heart Lung Transplant. 2014;33(10):996-1008.PubMedGoogle ScholarCrossref
31.
Ferrari  R, Böhm  M, Cleland  JG,  et al.  Heart failure with preserved ejection fraction: uncertainties and dilemmas.  Eur J Heart Fail. 2015;17(7):665-671.PubMedGoogle ScholarCrossref
32.
Lund  LH, Jurga  J, Edner  M,  et al.  Prevalence, correlates, and prognostic significance of QRS prolongation in heart failure with reduced and preserved ejection fraction.  Eur Heart J. 2013;34(7):529-539.PubMedGoogle ScholarCrossref
Original Investigation
August 2016

Survival After Coronary Artery Bypass Grafting in Patients With Preoperative Heart Failure and Preserved vs Reduced Ejection Fraction

Author Affiliations
  • 1Department of Cardiothoracic Surgery and Anesthesiology, Karolinska University Hospital, Stockholm, Sweden
  • 2Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden
  • 3Department of Cardiology, Karolinska University Hospital, Stockholm, Sweden
  • 4Department of Medicine, Karolinska Institutet, Stockholm, Sweden
  • 5Department of Emergency Medicine, Karolinska University Hospital, Stockholm, Sweden
 

Copyright 2016 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.

JAMA Cardiol. 2016;1(5):530-538. doi:10.1001/jamacardio.2016.1465
Abstract

Importance  Data on the prognostic consequence of heart failure (HF) with preserved ejection fraction in patients undergoing coronary artery bypass grafting (CABG) are limited and inconclusive.

Objective  To investigate the survival after CABG in patients with preoperative HF and preserved ejection fraction (pEF) vs reduced ejection fraction (rEF).

Design, Setting, and Participants  Swedish nationwide population-based cohort study that included all patients who underwent primary isolated CABG between January 1, 2001, and December 31, 2013, from the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) register, with follow-up for all-cause mortality in March 2014. Information regarding baseline characteristics, all-cause mortality, and readmissions for HF was obtained from national health data registers. Preserved EF was defined as at least 50%.

Main Outcomes and Measures  The primary outcome was all-cause mortality. A secondary outcome measure was a combination of all-cause mortality and readmission for HF.

Results  The study included 41 906 patients, 37 234 without known HF (27 165 with pEF and 10 069 with rEF) and 4672 with HF (1216 with pEF and 3456 with rEF). Their mean (SD) age was 67.4 (9.3) years, and 21.0% were female. During a mean (SD) follow-up time of 6.0 (3.3) years, 19.0% (7943 of 41 906) of patients died, including 13.2% (3574 of 27 165) with no HF and pEF, 24.6% (2476 of 10 069) with no HF and rEF, 33.9% (412 of 1216) with HFpEF, and 42.9% (1481 of 3456) with HFrEF. The multivariable-adjusted hazard ratios for death were 1.47 (95% CI, 1.40-1.56), 1.62 (95% CI, 1.46-1.80), and 2.29 (95% CI, 2.14-2.44) in patients with no HF and rEF, patients with HFpEF, and patients with HFrEF compared with patients with no HF and pEF. The findings were similar for the combined outcome of all-cause mortality and readmission for HF. The multivariable-adjusted hazard ratios for death within 30 days of surgery were 2.25 (95% CI, 1.86-2.73), 1.83 (95% CI, 1.26-2.66), and 2.52 (95% CI, 1.99-3.19) in patients with no HF and rEF, patients with HFpEF, and patients with HFrEF.

Conclusions and Relevance  A history of HF was an important risk factor for poor short-term and long-term outcomes after CABG regardless of preoperative EF. Reduced EF more than doubled the risk of early death after CABG.

Introduction

Coronary artery disease is the most common cause of heart failure (HF).1 In patients undergoing coronary artery bypass grafting (CABG), HF with reduced ejection fraction (HFrEF) is known to be associated with poor short-term and long-term prognosis. Ejection fraction (EF) is an important component of the preoperative risk assessment in patients scheduled for cardiac surgery.2-4 However, up to half of the patients with a history of HF have normal or near-normal left ventricular EF, called HF with preserved EF (HFpEF).5,6 Mortality in HFpEF may be as high as in HFrEF.5,6 Previous studies5,7,8 have demonstrated a variable prevalence of coronary artery disease among patients with HFpEF. However, the role of coronary artery disease and CABG in these patients is insufficiently characterized, and data on the prognostic effect of HFpEF in patients undergoing CABG are scarce and inconclusive.9 Patients with HFpEF may be subject to the same risk associated with anesthesia, inflammatory activation, and myocardial depression during CABG as patients with HFrEF. For example, although EF is preserved, contractility and systolic function are depressed, as demonstrated by reduced systolic strain,10 and global endothelial and coronary microvascular dysfunction11 may increase vulnerability to perioperative ischemic injury. Therefore, we investigated the prognosis after CABG in relation to HF with pEF or rEF in adults who underwent primary isolated CABG in Sweden.

Box Section Ref ID

Key Points

  • Question What is the prognostic consequence of heart failure with preserved ejection fraction in patients undergoing coronary artery bypass grafting?

  • Findings In this Swedish nationwide population-based cohort study, heart failure with preserved ejection fraction was significantly and independently associated with worse survival.

  • Meaning Ejection fraction adds short-term prognostic information essential for preoperative risk stratification, but the heart failure syndrome may be a stronger predictor of long-term outcomes regardless of ejection fraction.

Methods

This investigation was a nationwide population-based observational cohort study during a 13-year period (between January 1, 2001, and December 31, 2013). It was approved by the regional Human Research Ethics Committee, Stockholm, Sweden, and the need for informed consent was waived by the committee.

Study Population and Data Sources

We identified all patients who underwent primary isolated CABG in Sweden during the 13-year period from the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) register (eFigure 1 in the Supplement).12 During the study period, cardiac surgery was performed with similar and stable results in 8 hospitals in Sweden.13 Using unique personal identity numbers, which are assigned to all residents of Sweden,14 the data from the SWEDEHEART register were linked with data from several other nationwide health care registers, as previously described.15,16 The Swedish National Patient Register17 was used to determine relevant medical history, and the Longitudinal Integration Database for Health Insurance and Labor Market Studies,18 managed by Statistics Sweden, was used to obtain details regarding educational level, household disposable income, country of birth, and marital status. Record linkages were performed by the Swedish National Board of Health and Welfare. The International Classification of Diseases (ICD) codes (9th and 10th revisions) used for extraction of prior or concurrent medical conditions are listed in eTable 1 in the Supplement.

HF and EF

Patients without information regarding EF were excluded (eFigure 1 in the Supplement). A preoperative diagnosis of HF was ascertained from the Swedish National Patient Register using the following ICD codes (revision 10 codes I50-I50.9, I42-I43.9, I25.5, K76.1, I11.0, I13.0, and I13.2 and revision 9 codes 425 and 428). The frequency of each diagnostic code is listed in eTable 2 in the Supplement. The accuracy of an HF diagnosis from the Swedish National Patient Register was validated previously.19 In the validation study by Ingelsson et al,19 it was found that 82% of the patients with an HF diagnosis in the Swedish National Patient Register were classified as having HF according to the European Society of Cardiology definition that was used in the study. However, the validity increased to 88% for patients undergoing an echocardiographic examination and further increased to 91% for patients treated in the cardiology department. In patients with a primary diagnosis of HF who were treated in the cardiology department, the validity was 96%. We performed a sensitivity analysis in which HF was defined only from the primary diagnosis code. Preoperative EF was obtained from the SWEDEHEART register and was categorized as follows: at least 50% (good), 30% to less than 50% (moderate), and less than 30% (poor). For the purpose of this study, we defined pEF as at least 50% and rEF as less than 50%. However, we repeated all analyses with 3 EF categories (EF <30%, 30% to <50%, and ≥50%).

All-Cause Mortality and Readmission for HF

The Cause of Death Register was used to confirm vital status and ascertain dates of death. An HF hospitalization episode was defined as a hospital admission with a primary discharge ICD-10 code of I50 to I50.9 from the Swedish National Patient Register.

Statistical Analysis

Patients were categorized into 4 groups according to a prior diagnosis of HF (yes or no) and pEF or rEF as follows: no HF and pEF (reference category), no HF and rEF, HFpEF, and HFrEF. Baseline characteristics were described with frequencies and percentages for categorical variables and with means (SDs) for continuous variables. The primary outcome measure was death from any cause. Person-time in days was counted from the date of surgery until the date of death or the end of follow-up (March 24, 2014). We reported crude incidence rates and 95% CIs, and the Kaplan-Meier method was used to calculate cumulative survival. We used Cox proportional hazards regression with and without multivariable adjustment to model survival. The association between patient categories (according to HF status and pEF or rEF) and all-cause mortality was estimated by hazard ratios (HRs) (95% CIs). Patients in the no HF and pEF group were used as the reference category. The multivariable Cox proportional hazards model included all variables listed in Table 1 and was stratified by calendar year of surgery and hospital. Patient age, body mass index, and estimated glomerular filtration rate were modeled using restricted cubic splines, and all other variables were included as categorical terms.

Missing data (educational level [3.0%], renal function [5.3%], body mass index [7.6%], and emergent operation [7.3%]) were handled by multiple imputation by chained equations.20 The imputation models included all variables listed in Table 1, as well as the event indicator and the Nelson-Aalen estimator of the cumulative baseline hazard.21 Ten data sets were imputed, and estimates from these data sets were combined. Missing data regarding birth region (53 patients) and civil status (77 patients) were substituted with the most common category. Data management and statistical analyses were performed using software programs (Stata, version 14.0; StataCorp LP and R, version 3.2.2; R Foundation for Statistical Computing).

Results
Study Population and Patient Characteristics

Patient characteristics according to HF status and pEF or rEF are listed in Table 1. A total of 41 906 patients (mean age, 67.4 years) were included, and 21.0% were female. Patients were categorized into the following 4 groups: 27 165 (64.8%) with no HF and pEF, 10 069 (24.0%) with no HF and rEF, 1216 (2.9%) with HFpEF, and 3456 (8.2%) with HFrEF (eFigure 1 in the Supplement). Patients with HF were older, were more often female, and had more diabetes, worse renal function, and more frequent atrial fibrillation than patients without HF. Patients in the HFpEF group were more often female and had more hypertension, prior percutaneous coronary intervention (but not myocardial infarction), more atrial fibrillation, or previous stroke compared with the other 3 groups. One or 2 bypass grafts were used in 19.7% of the patients, 3 to 4 bypass grafts were used in 69.7%, and more than 4 bypass grafts were used in 10.6%. Off-pump CABG was performed in 3.7% of the patients, and multiple arterial grafts were used in 4.4%. There was little variation among the groups (eTable 3 in the Supplement).

Follow-up and Long-term Mortality

The mean (SD) follow-up time was 6.0 (3.3) years (median, 6.1 years), and the total follow-up time was 251 382 person-years. In total, 19.0% (7943 of 41 906) of patients died, including 13.2% (3574 of 27 165) with no HF and pEF, 24.6% (2476 of 10 069) with no HF and rEF, 33.9% (412 of 1216) with HFpEF, and 42.9% (1481 of 3456) with HFrEF. The unadjusted Kaplan-Meier estimated survival according to HF status and pEF or rEF is shown in the Figure, A. In patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF, respectively, 1-year survival was 98.2%, 94.6%, 94.1%, and 91.1%, and 5-year survival was 92.3%, 84.0%, 78.5%, and 70.5%. The incidence rates per 1000 person-years for death were 21 (95% CI, 21-22), 42 (95% CI, 41-44), 58 (95% CI, 52-64), and 82 (95% CI, 78-86) in patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF, respectively (Table 2). The crude and age and sex–adjusted relative risks for all-cause mortality in each of the 4 patient groups are summarized in Table 2. After multivariable adjustment for all variables listed in Table 1, the HRs for all-cause mortality were 1.47 (95% CI, 1.40-1.56), 1.62 (95% CI, 1.46-1.80), and 2.29 (95% CI, 2.14-2.44) in patients with no HF and rEF, HFpEF, and HFrEF, respectively, compared with patients in the reference category (no HF and pEF). In a separate analysis that was adjusted for EF and all baseline characteristics listed in Table 1, we found a strong and statistically significant association between HF and mortality (HR, 1.45; 95% CI, 1.37-1.54; P < .001) that was independent of EF.

HF Hospitalizations and Death

The mean (SD) follow-up time was 5.1 (3.1) years (median, 5.1 years), and the total follow-up time was 200 588 person-years. In total, 20.9% (8244 of 39 395) of patients had an HF hospitalization or death, including 13.9% (3533 of 25 493) with no HF and pEF, 29.4% (2742 of 9341) with no HF and rEF, 25.5% (421 of 1185) with HFpEF, and 45.9% (1548 of 3376) with HFrEF. The rates of hospital readmission for HF or mortality within 30 days of surgery were 2.0%, 6.5%, 8.1%, and 8.6% in patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF, respectively. The unadjusted Kaplan-Meier estimated cumulative incidence of HF hospitalization or death according to HF status and pEF or rEF is shown in the Figure, B. Freedom from HF or death was 96.6%, 89.4%, 89.0%, and 83.4%, respectively, after 1 year in patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF and 89.4%, 76.0%, 70.6%, and 59.7%, respectively, after 5 years. The incidence rates per 1000 person-years for HF hospitalization or death were 26 (95% CI, 25-27), 60 (95% CI, 58-63), 74 (95% CI, 67-82), and 113 (95% CI, 108-119) in patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF, respectively (Table 2). The crude and age and sex–adjusted relative risks for a combination of all-cause mortality and readmission for HF in each of the 4 patient groups are listed in Table 2. After multivariable adjustment for all variables listed in Table 1, the HRs for a combination of all-cause mortality and readmission for HF were 1.71 (95% CI, 1.62-1.80), 1.64 (95% CI, 1.47-1.82), and 2.44 (95% CI, 2.29-2.61) in patients with no HF and rEF, HFpEF, and HFrEF, respectively, compared with patients in the reference category (no HF and pEF).

Death Within 30 Days of Surgery

Thirty days after surgery, all-cause mortality was 0.8%, 2.9%, 2.8%, and 4.0%, in patients with no HF and pEF, no HF and rEF, HFpEF, and HFrEF, respectively. These results are summarized in Table 3 and Figure, C.

The crude and age and sex–adjusted relative risks for early mortality in each of the 4 patient groups are summarized in Table 3. After multivariable adjustment for all variables listed in Table 1, the HRs for early mortality were 2.25 (95% CI, 1.86-2.73), 1.83 (95% CI, 1.26-2.66), and 2.52 (95% CI, 1.99-3.19) in patients with no HF and rEF, HFpEF, and HFrEF, respectively, compared with patients in the reference category (no HF and pEF) (Table 3).

EF According to 3 Categories

We repeated the analyses for all outcomes according to HF status and EF as the following 3 categories: less than 30%, 30% to less than 50%, and at least 50%. The risk for each outcome increased with decreasing EF in patients with and without HF (Table 4 and Table 5).

Subgroup Analyses

We investigated the risk for all-cause mortality in patients with no HF and rEF, HFpEF, and HFrEF vs no HF and pEF in selected clinically relevant subgroups (eFigures 2, 3, and 4 in the Supplement). The risk for death was similar in selected subgroups, with the exception of those patients without a prior myocardial infarction among whom the risk for death was significantly higher in those with HFpEF and HFrEF compared with patients with no HF and rEF.

Sensitivity Analyses

We performed a sensitivity analysis in which HF was defined only from the primary diagnosis code. The results of the sensitivity analysis were consistent with our main analysis and are summarized in eTable 4 and eTable 5 in the Supplement.

During the 13-year study period, 11 053 patients had a missing EF and were excluded (eFigure 1 in the Supplement). However, 90.7% of the patients with a missing EF were operated on during 2001 and 2002 because registration of EF was not mandatory in the registry at that time (eTable 6 in the Supplement). Patients were followed up for all-cause mortality in March 2014. To investigate if the exclusion of patients with a missing EF would bias our main results, we performed a sensitivity analysis in which patients who were operated on during 2001 and 2002 were excluded. The main findings were essentially unchanged: multivariable-adjusted HRs for all-cause mortality were 1.47 (95% CI, 1.39-1.56), 1.66 (95% CI, 1.48-1.86), and 2.29 (95% CI, 2.13-2.46) in patients with no HF and rEF, HFpEF, and HFrEF, respectively, compared with patients in the reference category (no HF and pEF). Therefore, we conclude that the effect of a missing EF was negligible in our analysis.

Discussion

Preoperative assessment before CABG includes echocardiography and assessment of EF. However, HFpEF is increasingly common, and this diagnosis may be neglected. Herein, we show that in patients with HF, a reduced EF was associated with worse long-term outcomes. Most important, in patients without confirmed HF, an rEF was more strongly associated with worse outcomes, especially in the short-term. Therefore, echocardiography adds important preoperative information. However, we also show that a history of HF was an important independent risk factor for poor short-term and long-term outcomes after CABG regardless of preoperative EF. In crude analysis, HFpEF was associated with even greater risk than no HF and rEF, and the risk was similar after adjustment. The HF syndrome itself may be a stronger predictor of long-term outcomes than EF and should be carefully considered in preoperative assessment and postoperative follow-up.

HFpEF

Previous studies5,6 of the overall HF population have shown that approximately half of the patients with HF could be classified as HFpEF and that these patients have similar adverse outcomes compared with patients with HFrEF. Diabetes and hypertension, which cause left ventricular hypertrophy and diastolic dysfunction with inability of the heart to relax between contractions,22 were frequent findings among our patients with HFpEF. In the present study, the number of patients with HFpEF was considerably lower than the number of patients with HFrEF, which may reflect underdiagnosis of HF in patients with pEF or a lower prevalence of coronary artery disease among patients with HFpEF vs HFrEF. In an analogous register, the Swedish Heart Failure Registry (SwedeHF) (http://www.swedehf.se), the prevalence of ischemic heart disease is considerably lower in HFpEF23 compared with HFrEF.24 In our study, short-term and long-term outcomes after CABG differed between patients with HFpEF and HFrEF. Previously, HFpEF has been associated with hypertension, atrial fibrillation, and female sex, whereas HFrEF has been associated with prior myocardial infarction.5,8,25 The present study is in line with these previous studies, with a larger proportion of women and a higher prevalence of hypertension but a lower incidence of prior myocardial infarction among patients with HFpEF.

HFrEF and CABG

In patients undergoing CABG, the grade of left ventricular EF impairment, reflecting reduced amount of contracting myocardium, is a well-known strong risk factor for poor short-term and long-term prognosis.2-4 The magnitude of increased risk for early and late mortality among patients with impaired left ventricular EF in the present study was comparable to risks previously reported for this patient category.2-4

HFpEF and CABG

Although considerable amounts of research have been conducted regarding HFpEF in recent years, the role of coronary artery disease in HFpEF remains poorly understood.7 One small study9 has investigated the prognostic effect of HFpEF in patients undergoing CABG. The authors included 1877 patients who had undergone isolated CABG, of whom 1489 patients had a left ventricular EF exceeding 50% and no history of HF, 152 patients had HFpEF, and 236 patients had HFrEF. The authors stated the following conclusions: (1) as expected, HFpEF was associated with poorer 5-year outcomes (including all-cause mortality and readmission for HF) than left ventricular EF exceeding 50% and no history of HF (HR, 1.42; 95% CI. 1.02-1.97); (2) the 5-year outcomes were similar between patients with HFpEF and HFrEF (HR, 0.88; 95% CI, 0.61-1.29); and (3) the 30-day mortality was highest in patients with HFrEF, although the study was not powered to study this. The study did not report data regarding patients with no HF and rEF. Our results support that HFpEF is associated with poorer long-term outcomes than left ventricular EF of at least 50% and no history of HF. However, owing to a larger study population, we also demonstrated that HFrEF was a stronger risk factor for poor long-term survival than HFpEF and that rEF was more important than HF status for early mortality.

Early Mortality Risk Models

Although current models used to stratify early mortality risk after cardiac surgery—such as European System for Cardiac Operative Risk Evaluation (EuroSCORE) II26 and The Society of Thoracic Surgeons cardiac surgery risk model27—include left ventricular EF and New York Heart Association (NYHA) functional classification, they do not include a history of HF. It should be noted that NYHA function is a classification of the grade of limitation during physical activity and cannot be considered equal to a previous HF diagnosis. Therefore, a major implication of our study is that referring cardiologists and cardiac surgeons should be aware that a history of HF is an additional, independent, and important risk factor for early death that the currently used risk models do not take into consideration.

Prognostic Effect of Revascularization in HFpEF

The finding that survival after CABG in patients with HFpEF was better than that for patients with HFrEF is notable because previous investigations in the overall HF population have shown equally poor survival in patients with HFpEF and HFrEF.6 It has been demonstrated in patients with HFpEF with coronary artery disease that complete revascularization is associated with less deterioration in left ventricular EF and lower mortality compared with patients with incomplete revascularization independent of other factors.7 It could be argued that CABG identified a healthier and lower-risk subset of the overall HFpEF population, which could explain the lower mortality among patients with HFpEF compared with patients with HFrEF. However, this explanation is unlikely because it has been demonstrated that patients with HFpEF and coronary artery disease have increased mortality compared with patients with HFpEF without coronary artery disease.7,28 It could also be argued that CABG identified a higher-risk subset of patients with HFrEF. Indeed, underlying ischemic heart disease in HFrEF is one of several factors portending a worse prognosis,29 and patients with ischemic heart disease undergoing heart transplantation have worse posttransplant prognosis as well.30

Limitations

The strengths of our study included the large study population and the complete and accurate follow-up and survival ascertainment owing to the high-quality national Swedish registers, as well as the external validity and generalizability of our findings due to the nationwide complete coverage. In epidemiological studies, exposure and clinical end points are often assessed using health data registers. The reliability is dependent on the quality of the register data. As discussed by Ludvigsson et al,17 the proportion of valid diagnoses in the Swedish National Patient Register is possibly higher in patients with severe disease compared with patients with mild disease. Moreover, the validity of a diagnosis may be higher in patients with causally related complications. Therefore, it is plausible that the validity of an HF diagnosis in our study may be higher than indicated by the prior validation study19 because all patients were evaluated by cardiologists owing to severe cardiac disease that subsequently warranted cardiac surgery.

There are some important limitations of the study. First, although symptoms or signs of HF combined with a normal or near-normal left ventricular EF are generally considered as suggestive of HFpEF, there is an absence of a clear diagnostic criteria for HFpEF.31 Elevated concentration of natriuretic peptides, signs of abnormal left ventricular relaxation or filling, left ventricular hypertrophy, left atrial enlargement, and diastolic dysfunction have been used to support the HFpEF diagnosis.31 The national registries used did not provide information regarding these characteristics. Conversely, in patients with no HF and rEF, the HF diagnosis may have been missed. In generalizable HF populations, more than 10% are in NYHA class I (asymptomatic).32 In our population, the event rate in patients with rEF was only half in those without an HF diagnosis compared with those with an HF diagnosis, suggesting that, while perhaps not perfectly accurate, the HF diagnosis in our nationwide Cardiac Surgery Register is consistent with the clinical risk profile. Another limitation was the possibility of residual confounding due to unmeasured or unknown factors that may have influenced the results. We did not have information regarding medical treatment. Other limitations include the following: (1) information whether revascularization was considered complete or not was not available, and information regarding EF was only available as a categorical variable from the register; (2) NYHA class was not available from the register; and (3) we lacked information regarding the time between EF assessment and CABG.

Conclusions

The syndrome of HF regardless of EF is an important risk factor for poor short-term and long-term outcomes. While EF adds prognostic information in preoperative risk stratification, the HF syndrome may be a stronger predictor of long-term outcomes and should be carefully considered in preoperative assessment and postoperative follow-up.

Back to top
Article Information

Accepted for Publication: April 15, 2016.

Corresponding Author: Ulrik Sartipy MD, PhD, Department of Cardiothoracic Surgery and Anesthesiology, Karolinska University Hospital, SE-171 76 Stockholm, Sweden (ulrik.sartipy@karolinska.se).

Published Online: July 13, 2016. doi:10.1001/jamacardio.2016.1465.

Author Contributions: Dr Sartipy had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: All authors.

Acquisition, analysis, or interpretation of data: Dalén, Lund, Holzmann, Sartipy.

Drafting of the manuscript: Dalén, Ivert, Sartipy.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Sartipy.

Obtained funding: Holzmann, Sartipy.

Administrative, technical, or material support: Dalén, Lund, Holzmann, Sartipy.

Study supervision: Lund, Ivert, Sartipy.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest, and none were reported.

Funding/Support: This study was supported by grants from the Swedish Society of Medicine (Drs Lund and Holzmann), Karolinska Institutet Foundations and Funds (Drs Holzmann and Sartipy), the Mats Kleberg Foundation (Dr Sartipy), the Swedish Research Council (Dr Lund), and the Swedish Heart-Lung Foundation (Drs Lund and Sartipy).

Role of the Funder/Sponsor: The funders had no influence on the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; or the decision to submit the manuscript for publication.

Additional Contributions: We thank the steering committee of the Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART) register for providing data for this study.

References
1.
Lloyd-Jones  DM, Larson  MG, Leip  EP,  et al; Framingham Heart Study.  Lifetime risk for developing congestive heart failure: the Framingham Heart Study.  Circulation. 2002;106(24):3068-3072.PubMedGoogle ScholarCrossref
2.
Hillis  GS, Zehr  KJ, Williams  AW,  et al.  Outcome of patients with low ejection fraction undergoing coronary artery bypass grafting: renal function and mortality after 3.8 years.  Circulation. 2006;114(1)(suppl):I414-I419.PubMedGoogle Scholar
3.
Yoo  JS, Kim  JB, Jung  SH, Choo  SJ, Chung  CH, Lee  JW.  Coronary artery bypass grafting in patients with left ventricular dysfunction: predictors of long-term survival and impact of surgical strategies.  Int J Cardiol. 2013;168(6):5316-5322.PubMedGoogle ScholarCrossref
4.
Wrobel  K, Stevens  SR, Jones  RH,  et al.  Influence of baseline characteristics, operative conduct, and postoperative course on 30-day outcomes of coronary artery bypass grafting among patients with left ventricular dysfunction: results from the Surgical Treatment for Ischemic Heart Failure (STICH) trial.  Circulation. 2015;132(8):720-730.PubMedGoogle ScholarCrossref
5.
Bhatia  RS, Tu  JV, Lee  DS,  et al.  Outcome of heart failure with preserved ejection fraction in a population-based study.  N Engl J Med. 2006;355(3):260-269.PubMedGoogle ScholarCrossref
6.
Steinberg  BA, Zhao  X, Heidenreich  PA,  et al; Get With the Guidelines Scientific Advisory Committee and Investigators.  Trends in patients hospitalized with heart failure and preserved left ventricular ejection fraction: prevalence, therapies, and outcomes.  Circulation. 2012;126(1):65-75.PubMedGoogle ScholarCrossref
7.
Hwang  SJ, Melenovsky  V, Borlaug  BA.  Implications of coronary artery disease in heart failure with preserved ejection fraction.  J Am Coll Cardiol. 2014;63(25, pt A):2817-2827.PubMedGoogle ScholarCrossref
8.
Lund  LH, Donal  E, Oger  E,  et al; KaRen Investigators.  Association between cardiovascular vs. non-cardiovascular co-morbidities and outcomes in heart failure with preserved ejection fraction.  Eur J Heart Fail. 2014;16(9):992-1001.PubMedGoogle ScholarCrossref
9.
Marui  A, Nishiwaki  N, Komiya  T,  et al; CREDO-Kyoto CABG Registry Cohort-2 Investigators.  Comparison of 5-year outcomes after coronary artery bypass grafting in heart failure patients with versus without preserved left ventricular ejection fraction (from the CREDO-Kyoto CABG Registry Cohort-2).  Am J Cardiol. 2015;116(4):580-586.PubMedGoogle ScholarCrossref
10.
Donal  E, Thebault  C, Lund  LH,  et al.  Heart failure with a preserved ejection fraction additive value of an exercise stress echocardiography.  Eur Heart J Cardiovasc Imaging. 2012;13(8):656-665.PubMedGoogle ScholarCrossref
11.
Paulus  WJ, Tschöpe  C.  A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation.  J Am Coll Cardiol. 2013;62(4):263-271.PubMedGoogle ScholarCrossref
12.
Jernberg  T, Attebring  MF, Hambraeus  K,  et al.  The Swedish Web-system for Enhancement and Development of Evidence-based care in Heart disease Evaluated According to Recommended Therapies (SWEDEHEART).  Heart. 2010;96(20):1617-1621.PubMedGoogle ScholarCrossref
13.
Harnek  J, Nilsson  J, Friberg  O,  et al.  The 2011 outcome from the Swedish health care registry on heart disease (SWEDEHEART).  Scand Cardiovasc J. 2013;47(suppl 62):1-10.PubMedGoogle ScholarCrossref
14.
Ludvigsson  JF, Otterblad-Olausson  P, Pettersson  BU, Ekbom  A.  The Swedish personal identity number: possibilities and pitfalls in healthcare and medical research.  Eur J Epidemiol. 2009;24(11):659-667.PubMedGoogle ScholarCrossref
15.
Dalén  M, Ivert  T, Holzmann  MJ, Sartipy  U.  Coronary artery bypass grafting in patients 50 years or younger: a Swedish nationwide cohort study.  Circulation. 2015;131(20):1748-1754.PubMedGoogle ScholarCrossref
16.
Dalén  M, Ivert  T, Holzmann  MJ, Sartipy  U.  Household disposable income and long-term survival after cardiac surgery: a Swedish nationwide cohort study in 100,534 patients.  J Am Coll Cardiol. 2015;66(17):1888-1897.PubMedGoogle ScholarCrossref
17.
Ludvigsson  JF, Andersson  E, Ekbom  A,  et al.  External review and validation of the Swedish National Inpatient Register.  BMC Public Health. 2011;11:450.PubMedGoogle ScholarCrossref
18.
Emilsson  L, Lindahl  B, Köster  M, Lambe  M, Ludvigsson  JF.  Review of 103 Swedish Healthcare Quality Registries.  J Intern Med. 2015;277(1):94-136.PubMedGoogle ScholarCrossref
19.
Ingelsson  E, Arnlöv  J, Sundström  J, Lind  L.  The validity of a diagnosis of heart failure in a hospital discharge register.  Eur J Heart Fail. 2005;7(5):787-791.PubMedGoogle ScholarCrossref
20.
White  IR, Royston  P, Wood  AM.  Multiple imputation using chained equations: issues and guidance for practice.  Stat Med. 2011;30(4):377-399.PubMedGoogle ScholarCrossref
21.
White  IR, Royston  P.  Imputing missing covariate values for the Cox model.  Stat Med. 2009;28(15):1982-1998.PubMedGoogle ScholarCrossref
22.
Dobrin  JS, Lebeche  D.  Diabetic cardiomyopathy: signaling defects and therapeutic approaches.  Expert Rev Cardiovasc Ther. 2010;8(3):373-391.PubMedGoogle ScholarCrossref
23.
Lund  LH, Benson  L, Dahlström  U, Edner  M, Friberg  L.  Association between use of β-blockers and outcomes in patients with heart failure and preserved ejection fraction.  JAMA. 2014;312(19):2008-2018.PubMedGoogle ScholarCrossref
24.
Thorvaldsen  T, Benson  L, Ståhlberg  M, Dahlström  U, Edner  M, Lund  LH.  Triage of patients with moderate to severe heart failure: who should be referred to a heart failure center?  J Am Coll Cardiol. 2014;63(7):661-671.PubMedGoogle ScholarCrossref
25.
Lee  DS, Gona  P, Vasan  RS,  et al.  Relation of disease pathogenesis and risk factors to heart failure with preserved or reduced ejection fraction: insights from the Framingham Heart Study of the National Heart, Lung, and Blood Institute.  Circulation. 2009;119(24):3070-3077.PubMedGoogle ScholarCrossref
26.
Nashef  SA, Roques  F, Sharples  LD,  et al. EuroSCORE II. Eur J Cardiothorac Surg. 2012;41(4):734-744. Medline:22378855
27.
Shahian  DM, O’Brien  SM, Filardo  G,  et al; Society of Thoracic Surgeons Quality Measurement Task Force.  The Society of Thoracic Surgeons 2008 cardiac surgery risk models, part 1: coronary artery bypass grafting surgery.  Ann Thorac Surg. 2009;88(1)(suppl):S2-S22.PubMedGoogle ScholarCrossref
28.
Dunlay  SM, Roger  VL, Weston  SA, Jiang  R, Redfield  MM.  Longitudinal changes in ejection fraction in heart failure patients with preserved and reduced ejection fraction.  Circ Heart Fail. 2012;5(6):720-726.PubMedGoogle ScholarCrossref
29.
Lund  LH, Aaronson  KD, Mancini  DM.  Validation of peak exercise oxygen consumption and the Heart Failure Survival Score for serial risk stratification in advanced heart failure.  Am J Cardiol. 2005;95(6):734-741.PubMedGoogle ScholarCrossref
30.
Lund  LH, Edwards  LB, Kucheryavaya  AY,  et al; International Society of Heart and Lung Transplantation.  The registry of the International Society for Heart and Lung Transplantation: thirty-first official adult heart transplant report: 2014: focus theme: retransplantation.  J Heart Lung Transplant. 2014;33(10):996-1008.PubMedGoogle ScholarCrossref
31.
Ferrari  R, Böhm  M, Cleland  JG,  et al.  Heart failure with preserved ejection fraction: uncertainties and dilemmas.  Eur J Heart Fail. 2015;17(7):665-671.PubMedGoogle ScholarCrossref
32.
Lund  LH, Jurga  J, Edner  M,  et al.  Prevalence, correlates, and prognostic significance of QRS prolongation in heart failure with reduced and preserved ejection fraction.  Eur Heart J. 2013;34(7):529-539.PubMedGoogle ScholarCrossref
×