Determining the Preferred Percent-Predicted Equation for Peak Oxygen Consumption in Patients With Heart FailureCLINICAL PERSPECTIVE
Background— Peak oxygen consumption (Vo2) is routinely assessed in patients with heart failure undergoing cardiopulmonary exercise testing. The purpose of the present investigation was to determine the prognostic ability of several established peak Vo2 prediction equations in a large heart failure cohort.
Methods and Results— One thousand one hundred sixty-five subjects (70% males; age, 57.0±13.8 years; ischemic etiology, 43%) diagnosed with heart failure underwent cardiopulmonary exercise testing. Percent-predicted peak Vo2 was calculated according to normative values proposed by Wasserman and Hansen (equation), Jones et al (equation), the Cooper Clinic (below low fitness threshold), a Veteran’s Administration male referral data set (4 equations), and the St James Take Heart Project for women (equation). The prognostic significance of percent-predicted Vo2 values derived from the 2 latter, sex-specific equations were assessed collectively. There were 179 major cardiac events (117 deaths, 44 heart transplantations, and 18 left ventricular assist device implantations) during the 2-year tracking period (annual event rate, 10%). Measured peak Vo2 and all percent-predicted peak Vo2 calculations were significant univariate predictors of adverse events (χ2≥31.9, P<0.001) and added prognostic value to ventilatory efficiency (VE/Vco2 slope), the strongest cardiopulmonary exercise testing predictor of adverse events (χ2=150.7, P<0.001), in a multivariate regression. The Wasserman/Hansen prediction equation provided optimal prognostic information.
Conclusions— Actual peak Vo2 and the percent-predicted models included in this analysis all were significant predictors of adverse events. It seems that the percent-predicted peak Vo2 value derived from the Wasserman/Hansen equations may outperform other expressions of this cardiopulmonary exercise testing variable.
Received November 6, 2008; accepted December 22, 2008.
Peak oxygen consumption (Vo2) is a clinically accepted and important variable in the prognostic evaluation of patients with heart failure (HF) undergoing cardiopulmonary exercise testing (CPX).1 The actual value of peak Vo2, typically expressed relative to body weight, is the most common approach to reporting aerobic capacity in apparently healthy individuals and different patient populations, including HF.2 Reporting peak Vo2 as a percent-predicted value has also been advocated. Moreover, a number of different approaches to estimating normal aerobic capacity are readily available.3–9 These prediction equations have used various independent variables such as height, weight, and mode of exercise but the inclusion of age and consideration of sex is a shared commonality.
Clinical Perspective see p 113
The body of evidence demonstrating the prognostic utility of the actual peak Vo2 value is robust; collectively, these investigations have included thousands of patients and hundreds of adverse events, consistently demonstrating the ability of this CPX variable to identify those patients with HF at increased risk for poor outcomes.10 A limited number of investigations have also examined the prognostic value of percent-predicted peak Vo2 in patients with HF with mixed results. For example, using 2 different prediction equations, Aaronson and Mancini11 found neither percent-predicted peak Vo2 value was a superior prognostic marker compared with the actual value in 272 patients with HF. However, using one equation, Stelken et al12 reported percent-predicted peak Vo2 was superior to the actual value in predicting mortality in a separate group of 181 patients with HF.
We are unaware of any investigation that has simultaneously compared the prognostic utility of the most commonly used peak Vo2 prediction equations with each other and the actual aerobic capacity value in a large HF cohort undergoing CPX within the past 10 years. Moreover, a prognostic comparison of percent-predicted Vo2 values to other important CPX measures, such as ventilatory efficiency, has not been performed. The purpose of the present investigation was to address these issues in an effort to determine the clinical relevance of expressing peak Vo2 relative to a normative value in patients with HF.
This study was a multicenter analysis including patients with HF from the CPX laboratories at San Paolo Hospital (Milan, Italy), Wake Forest University Baptist Medical Center (Winston-Salem, NC), LeBauer Cardiovascular Research Foundation (Greensboro, NC) Veteran’s Administration (VA) Palo Alto Health Care System (Palo Alto, Calif), and Virginia Commonwealth University (Richmond, Va). A total of 1165 patients with chronic HF were included. Inclusion criteria consisted of a diagnosis of HF13 and evidence of left ventricular dysfunction by 2D echocardiography obtained within 1 month of data collection. All subjects completed a written informed consent and institutional review board approval was obtained at each institution. The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agree to the manuscript as written.
CPX Procedure and Data Collection
Symptom-limited CPX was performed on all subjects and pharmacological therapy was maintained during exercise testing. Conservative ramping protocols were used at all centers. Ventilatory expired gas analysis was performed using a metabolic cart at all 5 centers (Medgraphics CPX-D and Ultima, Minneapolis, Minn; Sensormedics Vmax29, Yorba Linda, Calif; or Parvomedics TrueOne 2400, Sandy, Utah). Before each test, the equipment was calibrated in standard fashion using reference gases and a 3-L syringe. Standard 12-lead electrocardiograms were monitored throughout exercise; blood pressure was measured using a standard cuff sphygmomanometer. Heart rate at rest and maximal exercise were obtained from the ECG. Heart rate recovery at 1 minute was the difference between heart rate at peak exercise and 1-minute post. All subjects performed an active cool-down, at the initial workload of the exercise protocol, for at least 1 minute after the cessation of exercise. Minute ventilation (VE), oxygen uptake (Vo2), carbon dioxide output (Vco2), and end-tidal carbon dioxide production (PETCO2) were acquired breath-by-breath, and averaged over 10-second intervals. Resting PETCO2 was expressed as a 2-minute averaged value in millimeter of mercury. Peak Vo2 and peak respiratory exchange ratio were expressed as the highest 10-second averaged samples obtained during the exercise test. VE and Vco2 values, acquired from the initiation of exercise to peak, were input into spreadsheet software (Microsoft Excel, Microsoft Corp, Bellevue, Wash) to calculate the VE/Vco2 slope via least squares linear regression (y=mx+b, m=slope).14,15 The oxygen uptake efficiency slope was also determined via least squares linear regression (Vo2=a log10VE+b)16 by spreadsheet software (Microsoft Excel, Seattle, Wash) using all of the exercise data. Percent-predicted peak Vo2 was calculated according to normative values proposed by Wasserman and Hansen4,5 (1 of 6 equations according to sex and bodyweight), Jones et al (equation),3 the Cooper Clinic (below low fitness threshold),8,9 a VA male referral data set (4 equations; A to D),6 and the St James Take Heart Project for women (equation).7 The prognostic significance of percent-predicted Vo2 values derived from the 2 latter, sex-specific prediction equations were assessed collectively (VA A to D/St James).
Subjects were followed for major cardiac events (mortality, left ventricular assist device implantation, heart transplantation) via hospital and outpatient medical chart review for a maximum of 2 years at all centers. Any death with a cardiac-related discharge diagnosis was considered an event. Clinicians conducting the exercise test were not involved in decisions regarding cause of death or heart transplant/left ventricular assist device implantation.
All continuous data are reported as mean±SD. Unpaired t testing assessed differences in continuous baseline and CPX variables between those subjects experienceing an event and those who were event free. χ2 analysis assessed differences in the distributions of HF etiology, gender, race, and pharmacological management between these groups. One-way ANOVA compared differences in the percent-predicted peak Vo2 values according to the various equations in the overall group. A mixed-model 2-way ANOVA assessed differences between percent-predicted peak Vo2 values (within subject factors) according to adverse event status (between subject factors). When a significant difference between percent-predicted peak Vo2 values was detected, post hoc analysis was performed by multiple paired testing with a Bonferroni correction (P<0.007 [0.05/7]). Receiver operating characteristic (ROC) curves were constructed for the prognostic classification schemes of all percent-predicted peak Vo2 values. A z test was used to assess for significance of differences among areas under the ROC curve for the prognostic models.17 Univariate Cox regression analysis assessed the prognostic value of percent-predicted peak Vo2 calculations for cardiac mortality alone and in subgroups according to sex, mode of exercise (treadmill versus lower extremity ergometer), peak respiratory exchange ratio (<versus ≥1.05), and age (<versus ≥50 years old). Separate multivariate Cox regression analyses (forward stepwise method, entry, and removal values 0.10 and 0.05, respectively) then assessed the combined prognostic value of the VE/Vco2 slope and each expression of peak Vo2 (actual and percent-predicted values). Multivariate Cox regression analysis was also used to assess the prognostic value of the VE/Vco2 slope, peak Vo2 expressions, and an expanded list of baseline variables including age, HF etiology, left ventricular ejection fraction, and New York Heart Association (NYHA) class. For the peak Vo2 expression demonstrating the highest prognostic value in the expanded multivariate Cox regression, Kaplan–Meier analysis was performed including all variables retained in that particular regression. The log-rank test determined statistical significance of the Kaplan–Meier analysis. With the exception of post hoc testing for the mixed-model, 2-way ANOVA, all statistical tests with a probability value <0.05 were considered significant.
There were 179 major cardiac events (117 deaths, 44 heart transplantations, and 18 left ventricular assist device implantations) during the 2-year tracking period (annual event rate, 10%). Baseline characteristics for the overall group as well as subgroups according to adverse event status are listed in Table 1. The percentages of subjects with an ischemic HF etiology and prescribed a diuretic were significantly higher in subjects experienceing a major cardiac event. In addition, mean NYHA class was significantly higher whereas left ventricular ejection fraction was significantly lower in the major cardiac event subgroup.
CPX results for the overall group and subgroups according to adverse event status are listed in Table 2. In the overall group, all percent-predicted peak Vo2 values were significantly different, with the exception of values derived from the VA-B/St James and Cooper clinic equations. Aside from mode of exercise characteristics and peak respiratory exchange ratio, all variables of interest were significantly different according to adverse event status. Peak Vo2, maximal heart rate, and all percent-predicted peak Vo2 calculations were significantly lower whereas the VE/Vco2 slope was significantly higher in subjects experienceing a major cardiac event. Moreover, all percent-predicted peak Vo2 calculations were significantly different between those with and without a major event. Resting PETCO2, the oxygen uptake efficiency slope, and heart rate recovery at 1 minute data were available in 737 (major events, 120), 452 (major events, 62), and 612 (major events, 82) subjects, respectively. For these subgroups, resting PETCO2 (34.3±4.5 versus 32.5±4.5 mm Hg), the oxygen uptake efficiency slope (1.8±0.9 versus 1.3±0.6), and heart rate recovery at 1 minute (19.2±11.8 versus 12.0±9.6 beats per minute) were all significantly higher (P<0.001) in subjects who did not experience a major cardiac event.
ROC curve analysis results for the different peak Vo2 prediction equations is listed in Table 3. All prognostic classification schemes were statistically significant. With the exception of the VA-D/St James equation, all optimal threshold values were within 5% points. The area under the ROC curve was greatest for the Wasserman/Hansen equation, although statistical significance in ROC areas was only reached in comparison between the VA-C and VA-D/St James equations. The Jones, Cooper Clinic, VA-A/St James, and VA-B/St James equations also demonstrated a significantly greater area under the ROC curve compared with VA-C/St James equation. All other area under the ROC curve comparisons did not reach statistical significance.
Table 4⇓ lists the prognostic value of percent-predicted Vo2 values according to cardiac mortality as the only end point and sex, mode of exercise, exercise effort, and age-based subgroups. The ability of all percent-predicted Vo2 values to predict major adverse events remained statistically significant when cardiac mortality was considered the only end point and in all subgroup analyses.
Multivariate Cox regression analysis including the VE/Vco2 slope and each expression of aerobic capacity is listed in Table 5. The VE/Vco2 slope was the superior predictor of major cardiac events in each analysis whereas measured peak Vo2 and each percent-predicted peak Vo2 calculation added predictive value and were thus retained in the regression. The residual χ2 value was greatest for the percent-predicted peak Vo2 value derived from the Wasserman/Hansen equation.
Two expanded multivariate Cox regression analyses that included the VE/Vco2 slope, either measured peak Vo2 or percent-predicted peak Vo2 according the Wasserman/Hansen equation and key baseline characteristics are presented in Table 6. The VE/Vco2 slope was again the superior prognostic marker in both assessments. Measured peak Vo2 and percent-predicted peak Vo2 according the Wasserman/Hansen equation were retained in the separate regression analyses in addition to NYHA class and left ventricular ejection fraction. On the basis of the residual χ2 values, the Wasserman/Hansen equation provided superior predictive value compared with measured peak Vo2 and other consistent demographic variables.
The Jones (residual χ2=10.8, P=0.001), VA-A/St James (residual χ2=8.0, P=0.005), VA-B/St James (residual χ2=7.9, P=0.005), VA-C/St James (residual χ2=5.2, P=0.02), VA-D/St James (residual χ2=8.0, P=0.005), and Cooper Clinic (residual χ2=7.9, P=0.005) equations were all retained in the same expanded multivariate Cox regression depicted in online supplemental data. The residual χ2 values were comparable with that found with measured peak Vo2 and below that provided by the Wasserman/Hansen equation. Moreover, as found with the analysis including measured peak Vo2, the residual χ2 values for left ventricular ejection fraction and NYHA class were higher (residual χ2=12.2, P<0.001) in each of these latter scenarios.
Kaplan–Meier analysis curves are illustrated in the Figure⇓. Dichotomous thresholds of < or ≥36.0, < or ≥47%, ≤ or >25%, and I/II versus III/IV were set for the VE/Vco2 slope,18 percent-predicted peak Vo2 according to the Wasserman/Hansen equation (determined by ROC curve analysis in the present investigation), left ventricular ejection fraction,13 and the NYHA class, respectively. Using these thresholds, there was a significant difference in adverse event rates between subgroups according to the number of abnormal characteristics.
A number of equations designed to estimate the percentage of normal aerobic capacity achieved during exercise testing are presently available for clinical application. The present study demonstrates that several well-known methods for the determination of percent-predicted aerobic capacity: (1) provide values that are for the most part significantly different from one another, limiting the portability of a given equation to different populations of patients with HF; (2) are all prognostically significant in a large HF cohort from a univariate perspective; (3) are all prognostically significant when only cardiac mortality was considered as an end point; (4) all provide prognostic value in subgroups according to sex, mode of exercise, exercise effort and age; and (5) all provide additional prognostic value in a multivariate model including other clinically established exercise and resting variables.
In the prognostic comparison between equations, it seems the Wasserman/Hansen calculations provided better resolution, although differences in areas under the ROC curve were not significantly different compared with most other equations assessed. The potential value of the Wasserman/Hansen equations is more so apparent in the multivariate Cox regression analysis that included the VE/Vco2 slope and key resting variables (Tables 5 and 6⇑). In these analyses, percent-predicted values derived from the Wasserman/Hansen approach possessed the highest univariate χ2 value (compared with measured peak Vo2 and other percent-predicted calculations), was retained in the multivariate regression and outperformed both left ventricular ejection fraction and NYHA class. Although measured peak Vo2 and all other percent-predicted calculations were also retained in their respective multivariate Cox regression analyses, their added prognostic value was not as powerful as compared with that derived from inclusion of the Wasserman/Hansen equations. The findings of the present study are consistent with the previous investigation by Stelken et al12 in that percent-predicted peak Vo2 according to the Wasserman/Hansen equations prognostically outperformed an abbreviated list of other equations and measured peak Vo2 in 181 patients with HF. Similalry, Osada et al19 found percent-predicted peak Vo2, again derived from the Wasserman/Hansen equations, prognostically outperformed measured peak Vo2 in 500 patients with HF. Aaronson and Mancini,11 however, found that the Wasserman/Hansen percent-predicted Vo2 calculations and measured peak Vo2 performed similarly in a HF cohort comprised of 272 patients. An advantage of the present investigation compared to these previous studies is the larger sample size, possibly lending more credence to our findings and supporting the previous investigations by Stelken et al12 and Osada et al.19
Previous investigations12,19 have solely used a percent-predicted peak Vo2 threshold of 50% in their dichotomous prognostic assessments. Determination of this threshold seems to have been more arbitrary than the optimal sensitivity/specificity determination via ROC curve analysis. Despite differences in determination of the dichotomous threshold used in the current and previous investigations, the cut point was for the most part similar. This concordance of research indicates that patients with a percent-predicted Vo2 value below ≈50% have poorer outcomes compared with those surpassing this threshold, although the optimal cut point may slightly deviate from 50% for a given predicted peak Vo2 equation.
There is considerable variation in how presently available prediction equations are defined. Although all equations considered age and sex, the Wasserman/Hansen male/female equations have taken the greatest number of additional factors into consideration, including body weight (underweight/normal weight/overweight), mode of exercise (treadmill/lower extremity ergometer), and sedentary lifestyle. Moreover, the Wasserman/Hansen, Jones, VA-A, and VA-B equations, used peak Vo2 determined by ventilatory expired gas analysis to develop their normative values. The VA-C, VA-D, St James, and Cooper Clinic equations all estimated aerobic capacity from treadmill speed/grade or exercise time. Given the differences in equation development, the improved prognostic utility of the Wasserman/Hansen equations may be the result of their ability to account for more explanatory variables and provide a truer depiction of aerobic capacity for an apparently healthy, but sedentary individual. It should be noted, however, that despite the potential limitations of the other prediction equations assessed in the present investigation, they all provided significant prognostic value both independently and in combination with clinically important variables.
It could be argued that the additional steps required for the Wasserman/Hansen equations are not worth the relatively modest increment gained in prognostic information. However, these equations are easily incorporated into the software packages that operate CPX systems. The percent-predicted peak Vo2 value according to the Wasserman/Hansen equation can, therefore, be automatically generated by manually inputting age, sex, height, weight, and mode of exercise, a procedure which is already commonplace in preparation for a CPX. The ease by which all percent-predicted peak Vo2 calculations are derived by presently available software packages eliminates the need for consideration of a given equations complexity.
There is a wealth of research supporting the prognostic strength of the VE/Vco2 slope in HF, which commonly outperforms measured peak Vo2.18,20–22 Although the prognostic value of the VE/Vco2 slope is gaining clinical recognition, peak Vo2 continues to be the sole or primary CPX variable considered in the prognostic assessment of patients with HF undergoing this procedure. The results of the present study indicate measured or percent-predicted peak Vo2 should continue to be considered as a secondary CPX variable, complimenting the insight gained from the VE/Vco2 slope. Irrespective of which variable provides the highest level of prognostic information, it seems clear from the recent literature that the ability to predict adverse events improves with multivariate modeling that includes CPX variables and key resting measures such as NYHA class and left ventricular ejection fraction.21
Subjects included in the present investigation were referred for CPX at their respective institution, creating the potential for selection bias. Caution must, therefore, be taken in extrapolating our findings to the population with HF as a whole or to CPX laboratories assessing patients with HF with differing characteristics, such as, for example, younger individuals with a congenital heart defect being considered for transplantation. Moreover, although the overall number of subjects in the present investigation exceeded 1000, the majority were male. Although all percent-predicted Vo2 equations were prognostic in the female subgroup, particular caution should be taken in extrapolating our findings to the female population with HF. Given the heterogeneous nature of this disease process, future investigations should be performed in other HF cohorts to determine whether the prognostic value of percent-predicted peak Vo2 is universally applicable to this chronic disease population. Although peak Vo2 and the VE/Vco2 slope are well established, other variables, such as resting PETCO2,23 the oxygen uptake efficiency slope,24 and heart rate recovery at 1 minute25 have demonstrated prognostic value. Although these variables were unfortunately not available for the entire cohort in the present investigation, subgroup analysis revealed all 3 demonstrated significantly better characteristics in subjects who did not experience a major cardiac event. In addition, peak Vo2 adjusted for body fat has demonstrated prognostic value in patients with HF.26,27 Analysis of this expression of aerobic capacity could not be performed in the present study as body fat assessment was not performed in any of the subjects. Future research should be directed toward a more comprehensive multivariate survival analysis to determine all clinically relevant CPX variables and optimal expression.
In conclusion, variables obtained from CPX provide important prognostic insight in patients with HF. The findings of the present investigation further confirm the prognostic superiority of ventilatory efficiency (VE/Vco2 slope) and suggest equations used to determine percent-predicted peak Vo2 provide similar and in some instances better predictive information compared with the measured value obtained from CPX. Although many laboratories conducting CPX in patients with HF report percent-predicted peak Vo2 values in their written report, they do not commonly consider its prognostic significance. Our results suggest that (1) peak Vo2 should be expressed as a percentage of the predicted normal value and this should be a routine part of the summary report, and (2) the Wasserman/Hansen equation is superior to other equations in terms of prognostic power in patients with HF.
Sources of Funding
This work was supported, in part, by NIH grants R37AG18915, P30AG21332, and MO1RR07122.
Gibbons RJ, Balady GJ, Timothy BJ, Chaitman BR, Fletcher GF, Froelicher VF, Mark DB, McCallister BD, Mooss AN, O'Reilly MG, Winters WL, Gibbons RJ, Antman EM, Alpert JS, Faxon DP, Fuster V, Gregoratos G, Hiratzka LF, Jacobs AK, Russell RO, Smith SC. ACC/AHA 2002 guideline update for exercise testing: summary article. A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee to Update the 1997 Exercise Testing Guidelines). J Am Coll Cardiol. 2002; 40: 1531–1540.
Arena R, Myers J, Williams MA, Gulati M, Kligfield P, Balady GJ, Collins E, Fletcher G. Assessment of functional capacity in clinical and research settings: a scientific statement from the American Heart Association Committee on Exercise, Rehabilitation, and Prevention of the Council on Clinical Cardiology and the Council on Cardiovascular Nursing. Circulation. 2007; 116: 329–343.
Wasserman K, Hansen JE, Sue DY, Stringer W, Whipp BJ. Normal values. In: Weinberg R, ed. Principles of Exercise Testing and Interpretation. 4th ed. Philadelphia, Pa: Lippincott Williams and Wilkins; 2005: 160–182.
Gulati M, Pandey DK, Arnsdorf MF, Lauderdale DS, Thisted RA, Wicklund RH, Al-Hani AJ, Black HR. Exercise capacity and the risk of death in women: the St. James Women Take Heart Project. Circulation. 2003; 108: 1554–1559.
Barlow CE, LaMonte MJ, Fitzgerald SJ, Kampert JB, Perrin JL, Blair SN. Cardiorespiratory fitness is an independent predictor of hypertension incidence among initially normotensive healthy women. Am J Epidemiol. 2006; 163: 142–150.
Stelken AM, Younis LT, Jennison SH, Miller DD, Miller LW, Shaw LJ, Kargl D, Chaitman BR. Prognostic value of cardiopulmonary exercise testing using percent achieved of predicted peak oxygen uptake for patients with ischemic and dilated cardiomyopathy. J Am Coll Cardiol. 1996; 27: 345–352.
Radford MJ, Arnold JM, Bennett SJ, Cinquegrani MP, Cleland JGF, Havranek EP, Heidenreich PA, Rutherford JD, Spertus JA, Stevenson LW, Heidenreich PA, Goff DC, Grover FL, Malenka DJ, Peterson ED, Radford MJ, Redberg RF. ACC/AHA key data elements and definitions for measuring the clinical management and outcomes of patients with chronic heart failure: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Data Standards (Writing Committee to Develop Heart Failure Clinical Data Standards): developed in collaboration with the American College of Chest Physicians and the International Society for Heart and Lung Transplantation: endorsed by the Heart Failure Society of America. Circulation. 2005; 112: 1888–1916.
Baba R, Nagashima M, Goto M, Nagano Y, Yokota M, Tauchi N, Nishibata K. Oxygen intake efficiency slope: a new index of cardiorespiratory functional reserve derived from the relationship between oxygen consumption and minute ventilation during incremental exercise. Nagoya J Med Sci. 1996; 59: 55–62.
Arena R, Myers J, Abella J, Peberdy MA, Bensimhon D, Chase P, Guazzi M. Development of a ventilatory classification system in patients with heart failure. Circulation. 2007; 115: 2410–2417.
Francis DP, Shamim W, Davies LC, Piepoli MF, Ponikowski P, Anker SD, Coats AJ. 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: 154–161.
Davies LC, Wensel R, Georgiadou P, Cicoira M, Coats AJ, Piepoli MF, Francis DP. Enhanced prognostic value from cardiopulmonary exercise testing in chronic heart failure by non-linear analysis: oxygen uptake efficiency slope. Eur Heart J. 2006; 27: 684–690.
Arena R, Guazzi M, Myers J, Peberdy MA. Prognostic value of heart rate recovery in patients with heart failure. Am Heart J. 2006; 151: 851.
Previous investigations have consistently demonstrated that cardiopulmonary exercise testing is a valuable tool in the clinical and prognostic assessment of patients with heart failure. Peak oxygen consumption (Vo2) is one of the primary variables obtained from such testing and is typically assessed as an actual value relative to body weight. A number of prediction equations have been developed to estimate normal aerobic capacity and are readily available to clinicians. While documenting a percent-predicted peak Vo2 value on the cardiopulmonary exercise testing report is typically advocated, it is frequently not afforded any consideration by clinicians assessing prognosis or weighing treatment options based on the exercise response. The present study demonstrates that the percent-predicted peak Vo2 value derived from several established equations provide prognostic value in patients with heart failure. In particular, the prediction equation established by Wasserman and Hansen seems to provide optimal prognostic value, potentially outperforming the predictive resolution obtained from the actual peak Vo2 value. This study may provide healthcare professionals performing cardiopulmonary exercise testing with important information regarding which peak Vo2 prediction equation to use and its potential clinical value in patients with heart failure. In conclusion, clinicians responsible for the interpretation of cardiopulmonary exercise testing data in patients with heart failure should consider the clinical utility of all information that is gained from this valuable assessment technique.
The online-only Data Supplement is available at http://circheartfailure.ahajournals.org/cgi/rapidpdf/CIRCHEARTFAILURE.108.834168/DCI.