Journal of Cardiac Failure Vol. 22 No.

7 2016

Clinical Investigation

The Added Value of Exercise Variables in Heart Failure

Prognosis
ANA CAROLINA ALBA, MD, PhD, MATTHEW W. ADAMSON, HBSc, J. MACISAAC, HBSc, SPENCER D. LALONDE, HBSc,
WAI S. CHAN, BSc, DIEGO HERNAN DELGADO, MD, MSc, AND HEATHER JOAN ROSS, MD MHSc
Toronto, Ontario, Canada

ABSTRACT

Introduction: Diminished exercise capacity is a key symptom in heart failure (HF). Exercise predictors
(peak VO2, VE/VCO2 slope, and oxygen uptake efficiency slope [OUES]) are prognostic markers but
studied in isolation. We evaluated if these exercise variables offer additional prognostic value to clinical
predictors in HF.
Methods and Results: This was a single-institution retrospective cohort study of 517 consecutive HF
patients. We used Cox proportional hazards modeling to determine the additional prognostic value of
exercise variables on mortality, HF hospital admissions, and a composite outcome of ventricular assistance
device (VAD) implantation, heart transplantation (HT), and death. During a mean follow-up of 2.7 years,
52 deaths, 47 HTs, and 19 VAD implantations occurred. After adjusting for age, New York Heart
Association functional class, ejection fraction, body mass index, creatinine, and B-type natriuretic peptide,
peak VO2 (hazard ratio [HR] 0.91, 95% confidence interval [CI] 0.85–0.96), OUES (HR 0.92, 95% CI 0.87–
0.97), and VE/ VCO2 (HR 1.03, 95% CI 1.01–1.05) were independent predictors of the composite outcome.
Similar discriminatory capacity existed between the exercise variables (c-statistics 0.77, 0.78, and 0.78,
respectively). Only VE/VCO2 was an independent predictor of admissions (HR 1.04, 95% CI 1.01–1.07),
and only peak VO2 was an independent predictor of mortality (HR 0.90, 95% CI 0.84–0.98).
Conclusions: Peak VO2, OUES, and VE/VCO2 are independent predictors of HF prognosis over recognized
clinical variables. However, no single exercise variable was superior.
(J Cardiac Fail 2016;22:492–497)
Key Words: Peak oxygen consumption, exercise, prognosis, mortality, oxygen uptake efficiency slope.

Chronic heart failure (HF) is an increasingly common

disorder that carries a poor prognosis. The prevalence of HF
has increased dramatically owing to recent treatment
advances that have decreased death rates of other
cardiovascular diseases.1 In severe HF, cardiac
transplantation and mechanical circulatory support are the
treatments of choice. With the increasing number of cardiac
transplant candidates and the limited supply of donors, as
well as the costs associated with mechanical circulatory
support, the ability to estimate prognosis and determine the
timing of therapy is crucial.
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018.
For personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
From the Division of Cardiology and Heart Transplant, Toronto General
Hospital, Toronto, Ontario, Canada.
Manuscript received July 30, 2015; revised manuscript received January
2, 2016; revised manuscript accepted January 27, 2016.
Reprint requests: Ana Carolina Alba, Toronto General Hospital, 585
University Ave, PBM, 11th Floor Rm 153, Toronto, Ontario, Canada,
M5G 2N2. Tel: 1 416 340 3482; Fax: 1 406 340 4134. E-mail:
carolina.alba@ uhn.ca.
Funding: None.
See page 496 for disclosure information. 1071-
9164/$ - see front matter
© 2016 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.cardfail.2016.01.012
Diminished exercise capacity is a key symptom in HF,
varying directly with disease severity, so cardiopulmonary
exercise testing (CPXT) has become a tool for clinicians in
the management of HF patients. Numerous investigations
have been conducted attempting to elucidate the CPXT

variables that best stratify risk and estimate
prognosis.2–14 The most extensively researched CPXT
variables include peak oxygen consumption (peak
VO2), minute ventilation to carbon dioxide production
slope (VE/VCO2 slope), and the oxygen uptake
efficiency slope (OUES). Studies comparing these
exercise variables to identify one as prognostically
superior have produced ambiguous results.
Eight of 9 studies comparing VE/VCO2 slope and
peak VO2 with the use of multivariable analysis
demonstrated that VE/ VCO2 slope is the most powerful
predictor of HF prognosis,2–9 whereas 1 showed that
both variables are of equal prognostic value.10 Three of
these studies showed that peak VO2 could add
prognostic value to VE/VCO2 slope,2,3,9 whereas three
others demonstrated that peak VO2 added no additional
value.4–6 Arena et al reported that the VE/VCO2 slope is
prognostically superior to the OUES,11 but Davies et
al12 reported that OUES
492
was the most powerful predictor of prognosis in chronic HF
patients. A meta-analysis by Poggio et al13 found that VE/
VCO2 slope and peak VO2 showed similar performance,
with VE/VO2 showing a trend toward superiority. That study
mentions that considering single variables independently is
of limited value and that this could be overcome using an
approach combining multiple variables derived from
clinical assessment and exercise testing. However, there are
few studies14–21 evaluating the additional prognostic
information of exercise-related variables after considering Exercise Parameters
differences in clinical variables.
What is of particular interest is that most studies
comparing the prognostic ability of CPXT variables
analyzed them predominately in isolation from each other
as opposed to in combination with other relevant clinical
variables; or they considered only some of the variables in
addition to other clinical parameters. For example, the
studies by Aaronson et al18 and O’Connor et al19 evaluated
only peak VO2, the studies by Agostini et al20 and Nakanishi
17
et al evaluated peak VO2 and VE/VCO2 slope within a score
without evaluating the added value of each exercise
variable, and the study by LaRovere et al21 evaluated only
6-minute walk test. The shortcoming of this approach is that
there are many other factors that are used by clinicians to
assess HF prognosis.22–24 Therefore, the present study
examined if CPXT variables offered additional prognostic The primary end point was a composite outcome of ventricular
value after taking into account differences in other assist device (VAD) implantation, heart transplantation, or death.
recognized clinical predictors of HF, and if so, which CPXT The secondary end points were death and hospital admission due
offered the most prognostic information.
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
Added Value of Exercise Variables in HF Prognosis • Alba et al 493
Methods
Study Population
This study was a retrospective cohort study of consecutive HF
patients seen in the Heart Function Clinic at Toronto General
Hospital. The inclusion criteria consisted of a diagnosis of HF
with reduced left ventricular ejection function (left ventricular
ejection fraction [LVEF] 40%), hemodynamically stable (no
HF admissions in the past 2 months), and completion of CPXT.
A total of 517 patients tested from January 2000 to July 2012
were included. The institutional Ethics Review Board approved
this study.
Data Collection
The patients’ charts were reviewed to obtain results from
laboratory testing, physical examination, and medications closest
to the patient’s CPXT date (within 6 months of visit). Variables
obtained included HF etiology (ischemic/ nonischemic), LVEF,
NewYork Heart Association (NYHA) functional class, body
mass index (BMI), type of medications, creatinine, lymphocyte
percentage, B-type natriuretic peptide (BNP), hemoglobin, and
sodium. In addition, other comorbidities (diabetes, hypertension,
and peripheral vascular disease) and implantable cardioverter-
defibrillator (ICD) or cardiac resynchronization therapy device
(CRT) status were obtained. In the event that the most recent
laboratory test was 6 months from the patient’s CPXT, the
laboratory values were imputed with the use of simple imputation
by means of multiple linear regression (occurred in 6% of the
population for lymphocyte percentage, BNP, and creatinine).
Patients performed CPXT on a cycle ergometer (Lode
MedGraphics) and a metabolic cart (MedGraphics
CardiO2Ultima) with the use of a ramp protocol. The testing
consisted of an initial 1 minute of seated rest, followed by 1 minute
of warm-up (0 W load), and full exercise by an individualized
ramp protocol with increments of 10 W/min. VE, VO2, and VCO2
data were collected continuously through breathby-breath analysis
of expired gases. Peak VO2 was calculated as the average of the
middle 5 of the last 7 breaths. The VE/ VCO2 slope was calculated
by least squares linear regression with the use of data from the
entire exercise period.25 The OUES was calculated using the
formula VO2 (mL/min) a (log10VE) b, where aOUES.
End Points
to HF exacerbation. In addition, a subgroup analysis was
performed on patients whose maximum respiratory exchange ratio
(RER) was 1.1 (suggesting a submaximal CPXT). This was
performed to investigate whether exercise variables obtained
during submaximal exercise (VE/VCO2 slope and OUES) provided
superior prognostic value compared with peak VO2 in ACEi, angiotensin converting enzyme inhibitor; ARB, angiotensin II receptor
patients who did not reach maximal exercise during their blocker; BNP, B-type natriuretic peptide; NYHA, New York Heart Association;
OUES, oxygen uptake efficiency slope; VE/VCO2 slope, minute ventilation to
CPXT. carbon dioxide production slope; VO2, oxygen consumption.

Statistical Analysis
variables (OUES and VE/VCO2 slope) were stronger
Descriptive statistics were performed. Continuous prognosticators during CPXT than peak VO2 in patients who did
variables are expressed as mean SD and categorical not reach an RER of 1.1, because an RER 1.1 is standardly used
variables are reported as percentages. Univariable Cox in VO2 max testing to verify a maximal effort.25 An interaction
regression analysis was performed to determine the effect between exercise variable and RER was added to each
association between the clinically important prognostic model to evaluate whether the association between the exercise
factors and exercise variables with the prespecified end variables and the composite outcomes differed according to RER.
points. Significant clinical variables were then
simultaneously entered in a multivariable Cox model and Results
removed by means of stepwise backward elimination, in
which the least significant variables were sequentially The study population comprised 517 patients with a mean age
removed according to a prespecified P value of 0.1. Age and of 52 12 years, of which 405 (78.3%) were male. The main cause
sex were forced in the model regardless of their P value. of cardiomyopathy was ischemic (34%). Eightyone percent of
Finally, the exercise variables were individually added to the patients were on beta-blockers, and 94% were on angiotensin-
final model to determine if they could significantly improve converting enzyme inhibitor or angiotensin receptor blocker; 46%
model performance, adding prognostic value. The outputs of had an ICD and/or CRT. The mean peak VO2, VE/VCO2 slope, and
the model were hazard ratios (HRs) and their 95% confidence OUES were 14.5 4.7 mL • kg1 • min1, 32.2 8.5, and 1416
intervals (CIs). We assessed goodness of fit of the model at
553 mL/
different stages with the use of the Akaike information Table 3. Goodness of Fit (Akaike Information Criteria [AIC]) and
criteria (AIC) statistic. Finally, Harrel c-statistic was used to Discriminatory Capacity of Each Model for the Composite
compare the discriminatory capacity of each model. A Outcome of Ventricular Assist Device Implantation, Heart
subgroup analysis was performed for the composite outcome Transplantation, or Death
in patients whose maximum RER was 1.1 and 1.1. This Model AIC c-Statistic* P Value 95% CI
was performed to assess whether submaximal 0.70–
494 Journal of Cardiac Failure Vol. 22 No. 7 July 2016 Base 1180.6 0.77 .001 0.84
Peak VO2 1170.5 0.77 .001 0.71–
Table 1. Patient Baseline Characteristics (n 517) 0.85
Variable Mean SD or n (%) OUES 1170.4 0.78 .001 0.71–
0.85
Male sex 405 (78.3)
VE/VCO2 1176.9 0.78 .001 0.71–
0.85
Age (y) 52 12
Abbreviations as in Table 1.
NYHA functional class 2.3 0.8 *The c-statistic was not significantly different between models.
Left ventricular ejection fraction (%) 25 8
Ischemic cardiomyopathy 176 (34)
Body mass index (kg/m2) 28 5.6 min, respectively. The demographics, baseline characteristics, and
Lymphocyte percentage (%) 25 9 exercise testing results are described in Table 1.
BNP (pg/mL) 208 307
Creatinine (mg/dL) 1.2 0.5 Predictors of VAD, Heart Transplant, and Death
Diabetes 124 (24)
Peripheral vascular disease 31 (6)
Hypertension 191 (37)
During a mean follow-up period of 2.8 2.0 years, there were
Internal cardiac defibrillator 238 (46) 118 individual outcomes (58 deaths, 47 heart transplantations,
Cardiac resynchronization therapy 117 (23) and 19 VAD implantations).
Beta-blockers 476 (92) Independent predictors of the composite outcome were age,
ACEi/ARB 486 (94) NYHA functional class, LVEF, BMI, creatinine, and BNP, which
Spironolactone 227 (44)
Digoxin 165 (32)
were included in the final base model. After individually
Furosemide 352 (68) including exercise variables in the model, OUES, VO2 peak, and
Peak VO2 (mL • kg1 • min1) 14.5 5.0 VE/VCO2 slope were found to be independent predictors of the
Respiratory exchange ratio 1.12 0.14 composite outcome (Table 2). A 1-unit increase in peak VO2 was
OUES (mL/min) 1416 553 associated with a 9% reduction in the risk of the composite end
VE/VCO2 slope 32 8.5 point (HR 0.91, 95% CI 0.85–0.96). A 1-unit decrease in
VE/VCO2 slope was associated with a 3% increased risk of the
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
 Added Value of Exercise Variables in HF Prognosis • Alba et al 495
composite outcome (HR 1.03, 95% CI 1.01–1.05). And a whose maximum RER was 1.1 (mean RER 1.22 0.1). The base
100-unit increase in OUES was associated with an 8% risk model after multivariable analysis included NYHAfunctional
reduction (HR 0.92, 95% CI 0.87– class, BMI, creatinine, and BNP. All exercise variables, when
0.97). No significantly different discriminatory capacity included in the base model, were found to be independent
existed between OUES, VO2 peak, and VE/VCO2 slope on predictors of the composite outcome and improved the goodness
the primary outcome as assessed with the use of Harrell c- of fit of the model, as assessed by a reduced AIC in patients with
statistic (Table 3). A subgroup analysis was performed for RER 1.1. However, there was no significantly different
the composite outcome on the 257 patients whose maximum
RER was 1.1 (mean RER 1 0.08) and in 260 patients
Table 2. Results when exercise variables were included individually in the Base Model for the Composite Outcome of Ventricular Assist
Device Implantation, Heart Transplantation, or Death

Univariable Analysis Multivariable analysis*

Variable HR 95% CI P Value HR 95% CI P
Value
Peak VO2 (1–mL • k1 • min1 increase) 0.88 0.84–0.93 .001 0.91 0.85–0.96 .001
OUES (100-unit increase) 0.88 0.85–0.92 .001 0.92 0.87–0.97 .001
VE/VCO2 slope (1-unit increase) 1.06 1.04–1.08 .001 1.03 1.01–1.05 .014
CI, confidence interval; HR, hazard ratio; other abbreviations as in Table 1.
*Each of the exercise variables was tested in individual models including age, NYHA functional class, left ventricular ejection fraction, body mass index,
creatinine, and BNP as covariates.
Table 4. Results When Exercise Variables Were Included Individually in the Base Model for Admission

Univariable Analysis Multivariable Analysis*

Variable HR 95% CI P Value HR 95% CI P
Value
Peak VO2 (1–mL • kg1 • min1 increase) 0.91 0.86–0.96 .001 0.95 0.89–1.02 .134
OUES (100-unit increase) 0.92 0.88–0.97 .001 0.95 0.90–1.00 .064
VE/VCO2 slope (1-unit increase) 1.06 1.04–1.09 .001 1.04 1.01–1.07 .003
Abbreviations as in Table 1.
*Each of the exercise variables was tested in individual models including age, NYHA functional class, creatinine, and BNP as covariates.
discriminatory capacity between OUES, VO2 peak, and VE/ Predictors of Mortality
VCO2 slope. In 260 patients with RER 1.1, only peak VO2
There were a total of 58 deaths during the follow-up
and OUES remained as independent predictors of the
period. The results of the univariable analysis showed that
composite outcome improving goodness of fit of the model,
female sex, NYHAfunctional class, LVEF, lymphocyte
as assessed by a reduced AIC, without significantly
percentage, BNP, creatinine, sodium, and hemoglobin were
increased discrimination capacity. However, the test of
significantly associated with mortality. Independent
subgroup interaction showed that the association between
predictors of mortality were female sex, LVEF, creatinine,
the exercise variables and the composite outcome was not
and BNP. After individually including the exercise
statistically different between patients with RER or 1.1 variables into the base model, only peak VO2 was found to
(P values 0.91 for peak VO2, 0.82 for VE/VCO2 slope and be an independent predictor of mortality (Table 5).
0.57 for OUES). However, there was no statistically significant difference in
the discriminatory capacity of the models, including any of
Predictors of HF Hospital Admissions
the exercise variables.
There were a total of 80 hospital admissions due to Discussion
decompensated HF. The results of the univariable analysis
showed that NYHA functional class, LVEF, lymphocyte In clinical practice, CPXT is conventionally used to
percentage, BNP, creatinine, sodium, and hemoglobin were supplement other clinical data in the management of patients
significantly associated with HF admissions. Independent with HF. So although there have been studies showing the
predictors of HF admission were age, NYHA functional prognostic value of exercise variables by comparing them
class, creatinine, and BNP. After individually including the against each other11 or by determining that one adds
exercise variables into the base model, only VE/VCO2 slope prognostic value to another,2,3,9 it was of significance to
was found to be an independent predictor of hospital determine whether these variables continued to offer
admission (Table 4). prognostic value when considered in conjunction with
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
clinical predictors of HF prognosis. The present study
showed that peak VO2, OUES, and VE/ VCO2 slope are
independent predictors of HF prognosis when analyzed in
association with recognized clinical predictors. Although no
single exercise variable was shown to be superior to the
others, all 3 variables offered some additional prognostic
value to other clinically relevant parameters of prognosis in
HF.
The incorporation of exercise variables added prognostic
value by demonstrating an independent association with
adverse outcomes and improving model calibration or
goodness of fit. Discrimination capacity was not improved
after adding exercise variables to clinical parameters. The c-
statistic has been criticized for being insensitive to changes
in absolute risk estimates26: an important factor may add
clinically significant value, yet the c-statistic may not
increase substantially. This may partially explain the
unchanged model discrimination.
Peak VO2 is currently the exercise variable recommended
when assessing need for heart transplantation/VAD
implantation.27 This may exaggerate the relationship between
peak VO2 and the probability of undergoing heart
transplantation or VAD implantation. This is not the case for
the analysis of the association between OUES or VE/VCO2
slope and the composite outcomes. However, when mortality
was analyzed in isolation, only peak VO2 was shown to be an
independent predictor. This demonstrates that its value
Table 5. Results When Exercise Variables Were Included Individually in the Base Model for Mortality
Univariable Analysis*
Variable HR
Peak VO2 (1–mL • kg1 • min1 increase) 0.85
OUES (100-unit increase) 0.92
VE/VCO2 slope (1-unit increase) 1.04
Abbreviations as in Table 1.
*Each of the exercise variables was tested in individual models including age, sex, left ventricular ejection fraction, creatinine, and BNP as covariates.
extends
496 Journal of Cardiac Failure Vol. 22 No. 7 July 2016
outside the scope of the outcomes that depend on clinical
judgment (ie, heart transplantation and VAD implantation).
Previous studies16,17 have suggested that peak VO2 has
higher prognostic power in patients reaching higher exercise
effort as assessed by RER; similarly, we found that peak VO2
and OUES but not VEVCO2 slope were independent
predictors of the composite outcome of death, heart
transplantation, and VAD implantation in patients reaching
an RER of 1.1. The lack of prognostic power of VE/VCO2
slope in patients with higher levels of exercise remains
unclear. However, in this study, all of the data acquired
during exercise was used to calculate VE/VCO2 slope, which
could lead to unreliable measurement of VE/VCO2 slope in
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
patients with high RER. At RER 1, VE/VCO2 slope
increases dramatically owing to disproportionally increased
ventilation due to higher production of lactic acid.25
Because of physical limitations (eg, deconditioning,
peripheral vascular disease, aging, etc.), some HF patients
may not be able to perform symptom-limited exercise
testing, but submaximal testing may still be appropriate.
Based on the findings of this study, even during submaximal
exercise (determined by a maximal RER 1.1), peak VO2,
OUES, and VE/VCO2 slope were all independent predictors
of the composite outcome and there was no significantly
different discriminatory capacity among the exercise
variables; therefore, peak VO2 still offers valuable prognostic
information, even in submaximal studies, and pushing
patients to reach a maximal exercise level may not be
necessary. In addition, peak VO2 and VE/VCO2 slope are
incorporated into most standard CPXT reports and OUES is
not. To calculate OUES, the raw data from the CPXT must
be obtained, which can be time consuming for the technician,
threatening the utility of OUES in clinical practice.
Numerous studies have investigated the prognostic value
of variables obtained from CPXT, with some demonstrating
that OUES or peak VO2 are prognostically superior and
others claiming that VE/VCO2 slope is superior.28–30 The
present study shows that these exercise variables are
equivalent in predicting prognosis and add value when
included with relevant clinical predictors of HF in patients
Multivariable Analysis*
95% CI P Value HR 95% CI P
Value
0.79–0.92 .001 0.90 0.84–0.98 .010
0.87–0.97 .001 0.94 0.88–1.00 .064
1.01–1.07 .005 1.00 0.97–1.04 .869
with poor to fair level of exercise. However, peak VO 2 and
OUES, and not VE/ VCO2 slope, remained as independent
predictors with similar discriminatory capacity. A more
accurate prognosis can be performed through assessment of
a combination of clinical variables and variables attained
from CPXT.
Study Limitations
One of the limitations of the present study is due to the
retrospective design, which limits the completeness of the
data. To minimize the impact of this limitation, instead of
excluding cases with missing data (6% of the patients), we
used multiple regression analysis to impute values of
lymphocyte percentage, BNP, and creatinine. To reduce
 Added Value of Exercise Variables in HF Prognosis
missing data, we also extended the window of laboratory
values to 6 months from the enrollment date. This may
introduce some bias by attenuating a relationship between
laboratory values and end points. However, the use of simple
imputation can underestimate uncertainty when missing data
is 5%.31
The results of this study may be generalizable to patients
with similar clinical characteristics. The population in this
study represent relatively young, mostly moderately
symptomatic heart failure patients on optimal medical
therapy.
Conclusion
This study solidifies the importance of evaluating
exercise variables to prognosticate HF patients. Information
obtained from CPXT continues to demonstrate prognostic
value in HF patients when considered with other clinically
relevant prognostic variables. Although no exercise variable
outperformed the others, it was demonstrated that during
submaximal exercise peak VO2 yielded equivalent
predictive results to other submaximal exercise variables.
Disclosures
None.
References
1. Ansari M, Massie BM. Heart failure: How big is the problem? Who
are the patients? What does the future hold? Am Heart J
2003;146:121–8.
2. Arena R, Myers J, Aslam SS, Varughese EB, Peberdy MA. Peak VO2
and VE/VCO2 slope in patients with heart failure: a prognostic
comparison. Am Heart J 2004;147:354–60.
3. Francis DP, SHamin W, Davies LC, Piepoli MF, Ponikowski P, Anker
SD, et al. Cardiopulmonary exercise testing for prognosis in chronic
heart failure: continuous and independent prognostic value from
VE/VCO2 slope and peak VO2. Eur Heart J 2000;21:154–61.
4. Robbins M, Francis G, Pashkow FJ, Snader CE, Hoercher K, Young
JB, et al. Ventilatory and heart rate responses to exercise: better
predictors of heart failure mortality than peak oxygen consumption.
Circulation 1999;100:2411–7.
5. Kleber FX, Vietzke G, Wernecke KD, Bauer U, Optiz C, Wensel R, et
al. Impairment of ventilatory efficiency in heart failure: prognostic
impact. Circulation 2000;101:2803–9.
6. Ponikowski P, Francis DP, Piepoli MF, Davies LC, Chua TP, Davos
CH, et al. Enhanced ventilatory response to exercise in patients with
chronic heart failure and preserved exercise tolerance: marker of
abnormal cardiorespiratory reflex control and predictor of poor
prognosis. Circulation 2001;103:967–72.
7. Corra U, Mezzani A, Bosimini E, Scapellato F, Imparato A, Giannuzzi
P. Ventilatory response to exercise improves risk stratification in
patients with chronic heart failure and intermediate functional
capacity. Am Heart J 2002;143:418–26.
8. Guazzi M, Myers J, Arena R. Cardiopulmonary exercise testing in the
clinical and prognostic assessment of diastolic heart failure. J Am Coll
Cardiol 2005;46:1883–90.
Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.
• Alba et al 497
9. Bard RL, Gillespie BW, Clarke NS, Egan TG, Nicklas JM.
Determining the best ventilatory efficiency measure to predict
mortality in patients with heart failure. J Heart Lung Transpl
2006;25:589–95.
10. Chua TP, Ponikowski P, Harrington D, Anker SD, Webb-Peploe K,
Clark AL, et al. Clinical correlates and prognostic significance of the
ventilatory response to exercise in chronic heart failure. J Am Coll
Cardiol 1997;29:1585–90.
11. Arena R, Myers J, Hsu L, Peberdy MA, Pinkstaff S, Bensimhon D, et
al. The minute ventilation/carbon dioxide production slope is
prognostically superior to the oxygen uptake efficiency slope. J
Cardiac Fail 2007;13:462–9.
12. Davies LC, Wensel R, Georgiadou P, Cicoira M, Coats AJS, Piepoli
MF, et al. Enhanced prognostic value from cardiopulmonary exercise
testing in chronic heart failure by non-linear analysis: oxygen uptake
efficiency slope. Eur Heart J 2006;27:684–90.
13. Poggio R, Arazi HC, Giorgi M Miriuka SG. Prediction of severe
cardiovascular events by VE/VCO2 slope versus peak VO2 in systolic
heart failure: a meta-analysis of the published literature. Am Heart J
2010;160:1004–12.
14. Levy WC, Arena R, Wagoner LE, Dardas T, Abraham WT. Prognostic
impact of the addition of ventilatory efficiency to the Seattle Heart
Failure Model in patients with heart failure. J Card Fail 2012;18:614–
9.
15. Dardas T, Li Y, Reed SD, O’Connor CM, Whellan DJ, Ellis SJ, et al.
Incremental and independent value of cardiopulmonary exercise test
measures and the Seattle Heart Failure Model for prediction of risk in
patients with heart failure. J Heart Lung Transplant 2015;34:1017–23.
16. Arena R, Guazzi M, Myers J, Chase P, Bensimhon D, Cahalim LP, et
al. The prognostic utility of cardiopulmonary exercise testing stands
the test of time in patients with heart failure. J Cardiopulm Rehabil
Prev 2012;32:198–202.
17. Nakanishi M, Takaki H, Kumasaka R, Arakawa T, Noguchi T,
Sugimachi M, et al. Targeting of high peak respiratory exchange ratio
is safe and enhances the prognostic power of peak oxygen uptake for
heart failure patients. Circ J 2014;78:2268–75.
18. Aaronson KD, Schwartz JS, Chen T-M, Wong K-L, Goin JE, Mancini
DM. Development and prospective validation of a clinical index to
predict survival in ambulatory patients referred for cardiac transplant
evaluation. Circulation 1997;95:2660–7.
19. O’Connor CM, Whellan DJ, Wojdyla D, Leifer E, Clare RM, Ellis SJ,
et al. Factors related to morbidity and mortality in patients with chronic
heart failure with systolic dysfunction: the HF-ACTION predictive
risk score model. Circ Heart Fail 2012;5:63–71.
20. Agostoni P, Corrà U, Cattadori G, Veglia F, la Gioia R, Scardovi AB,
et al, MECKI Score Research Group. Metabolic exercise test data
combined with cardiac and kidney indexes, the MECKI score: a
multiparametric approach to heart failure prognosis. Int J Cardiol
2013;167:2710–8.
21. la Rovere MT, Maestri R, Caporotondi A, Corbellini D, Guazzotti G,
Pinna GD, et al. Pre-discharge evaluation in heart failure—additive
predictive value of the 6-minute walking test to clinical scores. Circ J
2015;79:1756–63.
22. Doust JA, Pietrzak E, Dobson A, Glasziou PP. How well does B-type
natriuretic peptide predict death and cardiac events in patients with
heart failure: systematic review. BMJ 2005;330:625.
23. Cohn JN, Johnson GR, Shabetai R, Loeb H, Tristani F, Rector T, et al.,
V-HeFT VACooperative Studies Group. Ejection fraction, peak
exercise oxygen consumption, cardiothoracic ratio, ventricular
arrhythmias, and plasma norepinephrine as determinants of prognosis
in heart failure. Circulation 1993;87(6 Suppl):V15–6.
24. Mohammed AA, van Kimmenade RRJ, Richards M, Bayes-Genis A,
Pinto Y, Moora SA, et al. Hyponatremia, natriuretic peptides, and
outcomes in acutely decompensated heart failure. Circ Heart Fail
2010;3:354–61.
25. Milani R, Lavie CJ, Mehra MR, Ventura HO. Understanding the basics
of cardiopulmonary exercise testing. Mayo Clin. Proc. 2006;81:1603–
11.
26. Cook NR. Advances in measuring the effect of individual predictors
of cardiovascular risk: the role of reclassification measures. Ann Intern
Med 2009;150:795–802.
27. Mehra MR, Kobashigawa J, Starling R, Russel S, Uber PA,
Parameshwar J, et al. Listing criteria for heart transplantation:
International Society for Heart and Lung Transplantation guidelines
for the care of cardiac transplant patients. J Heart Lung Transplant
2006;25:1024–42.
28. MacGowan GA, Janosko K, Cecchetti A, Murali S. Exercise-related
ventilatory abnormalities and survival in congestive heart failure. Am
J Cardiol 1997;29:1585–90.
29. Cicoria M, Davos CH, Florea V, Shamim W, Doehner W, Coats AJ,
et al. Chronic heart failure in the very elderly: clinical status, survival
and prognostic factors in 188 patients more than 70 years old. Am
Heart J 2001;142:174–80.
30. Scharf C, Merz T, Kiowski W,
Oechslin E, Schalcher C,
Brunner-La Rocca HP. Noninvasive assessment of cardiac pumping
capacity during exercise predicts prognosis in patients with congestive
heart failure. Eur Heart J 2002;23:806–14.
31. Dong Y, Peng CY. Principled missing data methods for researchers.
Springerplus 2013;2:222.

Downloaded for Anonymous User (n/a) at Universitas Kristen Duta Wacana from ClinicalKey.com by Elsevier on January 30, 2018. For
personal use only. No other uses without permission. Copyright ©2018. Elsevier Inc. All rights reserved.