Skip Navigation

European Journal of Heart Failure 2001 3(5):577-585; doi:10.1016/S1388-9842(01)00167-2
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (18)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Zugck, C.
Right arrow Articles by Haass, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zugck, C.
Right arrow Articles by Haass, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

© 2001 European Society of Cardiology

Risk stratification in middle-aged patients with congestive heart failure: prospective comparison of the Heart Failure Survival Score (HFSS) and a simplified two-variable model

Christian Zugck*, Carsten Krüger, Robert Kell, Sven Körber, Dieter Schellberg, Wolfgang Kübler and Markus Haass

Department of Cardiology, University of Heidelberg Bergheimerstr. 58, D-69115 Heidelberg, Germany

* Corresponding author. Tel.: +49-6221-568-611; fax: +49-6221-565-515. E-mail address: christian_zugck{at}med.uni-heidelberg.de (C. Zugck)


   Abstract
Top
 Abstract
1. Introduction
 2. Methods
 3. Results
 4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
 References
 
Aims: The performance of a US-American scoring system (Heart Failure Survival Score, HFSS) was prospectively evaluated in a sample of ambulatory patients with congestive heart failure (CHF). Additionally, it was investigated whether the HFSS might be simplified by assessment of the distance ambulated during a 6-min walk test (6'WT) instead of determination of peak oxygen uptake (peak VO2).

Methods and Results: In 208 middle-aged CHF patients (age 54 ± 10 years, 82% male, NYHA class 2.3 ± 0.7; follow-up 28 ± 14 months) the seven variables of the HFSS: CHF aetiology; heart rate; mean arterial pressure; serum sodium concentration; intraventricular conduction time; left ventricular ejection fraction (LVEF); and peak VO2, were determined. Additionally, a 6'WT was performed. The HFSS allowed discrimination between patients at low, medium and high risk, with mortality rates of 16, 39 and 50%, respectively. However, the prognostic power of the HFSS was not superior to a two-variable model consisting only of LVEF and peak VO2. The areas under the receiver operating curves (AUC) for prediction of 1-year survival were even higher for the two-variable model (0.84 vs. 0.74, P < 0.05). Replacing peak VO2 with 6'WT resulted in a similar AUC (0.83).

Conclusion: The HFSS continued to predict survival when applied to this patient sample. However, the HFSS was inferior to a two-variable model containing only LVEF and either peak VO2 or 6'WT. As the 6'WT requires no sophisticated equipment, a simplified two-variable model containing only LVEF and 6'WT may be more widely applicable, and is therefore recommended.

Key Words: Heart failure • Peak Vo2 • Prognostic model • 6-Min walk test

Received December 15, 2000; Revised February 28, 2001; Accepted May 9, 2001


   1. Introduction
Top
 Abstract
 1. Introduction
2. Methods
 3. Results
 4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
 References
 
Cardiac transplantation is an effective treatment option for patients with severe congestive heart failure (CHF). Though originally considered only for NYHA class IV patients the 18 and 26% 1- and 3-year mortality rates after cardiac transplantation compare favourably to the 15–20% annual mortality rate of NYHA class III patients [1]. As a result, an increasing number of ambulatory patients with advanced CHF are placed on transplant waiting lists. Given the limited number of donor organs, the time on the transplant waiting list has increased and is frequently more than 1 year [2]. On the other hand, Stevenson et al. [3] reported that CHF patients who survive on the transplant list for more than 6 months may no longer have a prognostic advantage from cardiac transplantation. Therefore, an accurate identification of patients most likely to benefit from cardiac transplantation is imperative. Ideally, risk stratification should allow reliable prediction of prognosis beyond the first year after primary assessment.

While there are numerous reports in the literature of single risk factors (for review [4]), clinically practicable models combining the prognostic power of multiple independent predictors are lacking. Recently, Aaronson et al. [5] proposed a prospectively validated clinical index, the Heart Failure Survival Score (HFSS), as an instrument to improve risk stratification in patients with advanced heart failure. The seven non-invasively obtained variables included in the HFSS incorporate multiple features of CHF pathophysiology: presence of ischemic cardiomyopathy (aetiology of CHF); systolic dysfunction and mean arterial pressure (hemodynamics); heart rate and serum sodium concentration (neurohumoral activation); intraventricular conduction delay (IVCD; myocardial injury/fibrosis); and peak VO2 (functional capacity).

The purpose of the present study was to prospectively evaluate the HFSS in another sample of ambulatory CHF patients. Since the determination of submaximal exercise capacity by a 6-min walk test (6'WT) provides prognostic information similar to peak VO2 [6], it was additionally tested, whether the HFSS might be simplified by assessment of the distance ambulated during a 6'WT instead of determination of peak VO2.


   2. Methods
Top
 Abstract
 1. Introduction
 2. Methods
3. Results
 4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
 References
 
2.1. Study population
The study group included 208 middle-aged patients with chronic CHF (NYHA class I–III, age <70 years) and a left ventricular ejection fraction (LVEF) ≤40%. All patients were referred to the Department of Cardiology at the Medical Clinic of the University of Heidelberg for assessment of their heart failure status and evaluation of potential candidacy for cardiac transplant between November 23, 1995 and August 8, 1998. The aetiology of CHF (dilative or ischemic cardiomyopathy) was confirmed by cardiac catheterization prior to inclusion into the study. Patients with: neurologic, orthopaedic, peripheral vascular or severe pulmonary diseases; or with NYHA class IV CHF, that may have impaired successful completion of exercise testing were excluded from the study. Besides clinical history and current medication the following parameters were assessed in each patient within 48 h: heart rate; blood pressure; serum sodium concentration (routine laboratory); intraventricular conduction time (12-channel ECG with an IVCD being defined as a conduction delay >120 ms); LVEF (radionuclide ventriculography); and functional capacity (exercise testing). All patients had to be on stabilised medication for at least 4 weeks prior to inclusion into the study. The study conforms with the principles outlined in the Declaration of Helsinki [7] and was approved by the institutional Ethics Committee. All patients gave their written informed consent.

2.2. Exercise testing
All patients underwent a symptom-limited cardiopulmonary exercise test (CPX). The CPX equipment included a metabolic cart (Oxycon alpha, Jaeger, Würzburg, Germany) with an interfaced supine-positioned (30°) bicycle ergometer (Ergoline, Jaeger, Würzburg, Germany). Peak VO2 was defined as the highest oxygen consumption averaged over 10 s and measured during the last 30 s of symptom-limited exercise. Details of the test protocol have recently been published [6].

Within 24 h and at least 4 h before CPX a 6'WT was performed. Briefly, all patients were informed of the purpose, methods and use of the 6'WT results. Allowing the patients to set the pace of ambulation with rest and stops as needed, they were asked to ambulate as far as possible within 6 min. In order to guarantee a standardised and reproducible procedure, the time was called out after 3 and 5 min without offering additional encouragement during the test [6]. The total distance walked during the 6'WT was recorded using a tachograph [8].

2.3. Clinical follow-up
Patients were followed prospectively and were routinely seen in the outpatient clinic at the Department of Cardiology at the University of Heidelberg. Medical therapy was adjusted to maintain an oedema-free state. Death without transplantation was defined as an outcome event. All patients who received a cardiac transplant (UNOS 2) were considered as survivors until the date of their transplantation. A high urgency transplant (UNOS 1) was not registered. Follow-up information was available for all surviving and non-transplanted patients at the time of analysis (February 24, 2000).

2.4. Model validation and statistical analysis
The HFSS has been defined by the following seven variables: presence of ischemic cardiomyopathy; serum sodium concentration; resting heart rate; mean arterial pressure; presence of IVCD; LVEF; and peak VO2 [5]. The HFSS was calculated for each patient as the absolute value of the sum of the products of the aforementioned prognostic variables (β) and their previously published coefficients (c) (i.e. β1c1+β2c2+...+βncn) [5]. Subsequently, a multivariate Cox regression analysis was performed to redefine the coefficients of the aforementioned prognostic variables for the Heidelberg sample [9]. Additionally, the Cox regression analysis was repeated by substitution of the variable ‘peak VO2’ by ‘6'WT’, resulting in the so-called HD-HFSS. The discrimination of the applied scores were tested by Kaplan–Meier analysis and log rank test [10,11]. Spearman rank correlation coefficients were used as a measure of association.

The ability of each model to predict survival at 1 year (excluding patients with follow-up of <1 year) was assessed by calculation of the area under the curve (AUC). AUCs for different models were compared by the area test for correlated test results [12]. The maximum likelihood estimated binormal receiver-operating curves (ROC) were constructed by means of plotting true-positive rates (sensitivity) against false-positive rates (1/specifity) [13]. Additionally, the results were confirmed (data not shown) by re-calculation of the AUC-comparison applying a non-parametric approach [14].

The data are presented as means ±S.D. except where otherwise specified. To test for significant differences between means, a two-sample Wilcoxon test was used and a value of P<0.05 was considered significant. Calculations were performed with SAS version 6.12.


   3. Results
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
 References
 
3.1. Patients’ characteristics
The clinical characteristics of the current patient sample (n=208, Heidelberg sample) are summarised in Table 1. In comparison to the original derivation sample the mean NYHA class was slightly lower in the Heidelberg sample (2.3±0.7 vs. 2.8±0.9), with 41 and 43% of the patients being in NYHA classes II and III, respectively. In half of the Heidelberg patients the LVEF was ≤20%. In both samples exercise capacity was markedly reduced, with 33 and 75 patients of the Heidelberg sample achieving a peak VO2 of less than 10 and 14 ml/min/kg, respectively. The mean distance ambulated during the 6'WT was 455±107 m (range: 170–692 m). Nineteen patients ambulated less than 300 m and 77 less than 450 m, with no patient requiring a rest stop.


View this table:
[in this window]
[in a new window]

  		Table 1 Patients’ profile of the model derivation [5] and the Heidelberg validation sample

 
Age, gender, resting heart rate, mean arterial pressure and serum sodium concentration were similar to the derivation sample. Slight differences were noted for the percentage of patients with ischemic cardiomyopathy and with an IVCD. In contrast to the derivation sample all Heidelberg patients were treated with an ACE-inhibitor or an AT1-receptor blocker; additionally 132 patients (64%) received both a digitalis glycoside and a diuretic, and 62 patients (30%) a β-blocker (Table 1).

3.2. Survival analysis
After a follow-up of 28.3±14.1 months (range: 0.2–50.8) 52 patients (25%) of the Heidelberg sample died, all of cardiovascular causes, and 55 patients (26%) were hospitalised due to worsening heart failure. Non-cardiac deaths and urgent cardiac transplants (UNOS 1) were not registered. Cardiac transplants (UNOS 2) were performed in 24 patients, seven of them within 1 year. A subgroup analysis of the 35 patients with an ICD (follow-up 28.9±12.7 months) revealed no significant difference in comparison to the whole group: nine ICD patients (26%) died and seven ICD patients (20%) were hospitalised due to worsening heart failure.

As compared to survivors (n=156), non-surviving patients (n=52) ambulated a shorter distance during the 6'WT (417±119 vs. 477±99 m, P=0.003) and had a lower peak VO2 (12.4±4.8 vs. 15.7±5.2 ml/min/kg, P<0.001). Additionally, non-survivors were characterised by a lower LVEF (17±7 vs. 23±8%, P<0.001), an older age (57±9 vs. 53±10 years, P=0.011), and a higher frequency of an IVCD (60 vs. 45%, P<0.001). The aetiology of CHF, gender, resting heart rate, mean arterial pressure and serum sodium concentration were not significantly different between both groups.

3.3. Risk stratification by use of the HFSS model
By applying the HFSS model to the Heidelberg sample a score of 8.61±1.05 (range 5.71–11.86) was calculated. The HFSS of the surviving and non-surviving patients were significantly different, both for the total follow-up period (8.81±1.02 vs. 8.01±0.94, P<0.001) and for 1-year survival (8.76±1.04 vs. 7.94±0.73, P<0.001). However, when only potential candidates for cardiac transplantation (NYHA class III patients, n=90) were analysed, the HFSS showed a difference between surviving and non-surviving patients only for long-term survival (8.31±0.87 vs. 7.84±1.01, P=0.025), but just failed to predict 1-year survival (8.28±0.99 vs. 7.75±0.75, P=ns).

In the original article by Aaronson et al. [5], distinct HFSS scores were proposed as indicators of a high (HFSS <7.19), medium (HFSS 7.20–8.09) and low (HFSS >8.10) risk. Within 1 year, 26 deaths occurred and seven UNOS-2 transplants were registered in the Heidelberg sample, with the latter being censored for the following analysis. One-year survival rates for the remaining 201 non-transplanted patients were 92% for the low (11 deaths out of 134 patients), 77% for the medium (12 deaths out of 52 patients) and 80% for the high risk strata (three deaths out of 15 patients), respectively. A better discrimination was achieved for the total follow-up period with survival rates of 84, 61 and 50% for the low, medium and high risk strata, respectively. When only NYHA class III patients were analysed, 1-year survival rates of 83, 73 and 84% and overall survival rates of 77, 59 and 54% were obtained for the low, medium and high risk strata, respectively.

3.4. Predictors of survival
Although all non-invasively obtained variables (aetiology of CHF, serum sodium concentration, resting heart rate, mean arterial pressure, IVCD, LVEF and peak VO2) were independent predictors of survival in the derivation sample [5], only LVEF and peak VO2 persisted to predict outcome when the multivariable Cox regression analysis was re-calculated for the Heidelberg sample (Table 2). When the Cox regression analysis was repeated with substitution of the variable ‘peak VO2’ by ‘6'WT’, resulting in the so-called HD-HFSS, again only LVEF and 6'WT remained as independent variables. Additionally, two-variable models containing the combination of LVEF with either peak VO2 or 6'WT were tested. Although the model coefficient of the serum sodium concentration was negative in the original derivation sample, HD-HFSS and the two-variable models showed (after z-transformation) a strong correlation to the HFSS of the formerly published derivation sample (Fig. 1).


View this table:
[in this window]
[in a new window]

  		Table 2 Multivariable cox regression analysis of the derivation [5] and the Heidelberg sample

 


Figure 1
View larger version (17K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 1 After z-transformation in order to compare the applied multivariable scores, the HD-HFSS (r=0.82), as well as the two-variable models with combination of LVEF plus peak VO2 (r=0.80), or LVEF plus 6'WT (r=0.76), were highly related to HFSS (for all P<0.001).

 
Differences in event-free survival were then assessed by Kaplan–Meier analysis, with stratification by the median. In each model, survival was significantly lower in patients with a lower score than in those with a higher score (all P<0.0001, Fig. 2).


Figure 2
View larger version (13K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 2 The survival curves for the HFSS, HD-HFSS and the two-variable models with combination of LVEF plus peak VO2, or LVEF plus 6'WT were stratified according to medians of each score by Kaplan–Meier survival analysis. The survival curves of each stratification were significantly different (P<0.05).

 
Additionally, the prediction of 1-year survival was calculated for the uni- and multivariable models. Both in the original derivation sample and the Heidelberg sample, all of the aforementioned variables exhibited a significantly worse discrimination than the multivariable model, with the exception of LVEF, peak VO2 and 6'WT, which all performed similar to the HFSS (Table 3). The interplay between sensitivity and specificity for prediction of 1-year survival was analysed by calculation of receiver operating curves (ROC). The area under the ROC was significantly lower for the HFSS (AUC 0.73) than for the simple combination of either LVEF and peak VO2 (AUC 0.84, P=0.0007 vs. HFSS), or LVEF and 6'WT (AUC 0.83, P=0.0114 vs. HFSS), respectively (Fig. 3).


View this table:
[in this window]
[in a new window]

  		Table 3 Discrimination of uni- and multivariable models in the derivation [5] and Heidelberg samples by AUCs, that indicate freedom from death without transplantation at 1 year

 


Figure 3
View larger version (11K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		Fig. 3 The areas under the ROC curves for 1-year survival were significantly greater for two-variable models with combination of LVEF plus peak VO2 (AUC: 0.84, P=0.0007), or LVEF plus 6'WT (AUC: 0.83, P=0.0114) compared to the HFSS (AUC: 0.73). The two-variable models showed no significant difference.

 

   4. Discussion
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
5. Limitations of the...
 6. Conclusions and future...
 References
 
The HFSS, a multivariable US-American scoring system [5], continued to discriminate survivors and non-survivors when applied to the present sample of 208 ambulatory middle-aged CHF patients with a LVEF <40%. However, a simple two-variable model based only on LVEF and peak VO2 or — more widely applicable — on LVEF and 6'WT was superior to the seven variables in the HFSS.

Since physicians have great difficulty predicting survival of an individual CHF patient [15], an optimised prediction of mortality risk is needed to manage transplant waiting lists, but few multivariable indices for risk stratification in advanced CHF have been published so far. Campana et al. [16] proposed a scoring system that includes a mix of catheterization-derived variables, which limit its use in clinical routine. Alla et al. [17] recently published a non-invasive algorithm to differentially predict survival in dilated and ischemic CHF. However, both studies have not yet been validated in an independent patient sample.

In contrast, the HFSS proposed by Aaronson et al. [5] performed well when used in a separate (validation) sample of CHF patients in the US [5], and continued to identify CHF patients at high risk in a recently published study [18] from Germany. However, in the latter study, two out of the seven parameters were completely (IVCD) or partially (peak VO2 existed in only 16% of the patients) missing, and the single risk predictors were not evaluated for their independence. Therefore, further analysis had to be performed to evaluate this multivariable scoring system in another cohort of patients with a complete set of data.

In the present unicenter study the HFSS continued to predict event-free overall survival (survival rates of 50, 61 and 84% for the high, medium and low risk strata, respectively), but failed to reliably differentiate 1-year survival in NYHA class III patients, the potential candidates for cardiac transplantation.

In contrast to the derivation sample [5], all patients in the Heidelberg sample received an ACE inhibitor or an angiotensin 1-receptor antagonist, and a three times higher percentage were treated with a β-blocker. Therefore, a therapy-related reduction in the event-rate may partially explain the poorer performance of the HFSS in the present study. But more significant is the fact that, in the present study, multivariable Cox regression analysis revealed only LVEF and peak VO2 as strong independent predictors of prognosis, while aetiology of CHF, serum sodium concentration, heart rate, mean arterial pressure and IVCD — the other five variables included in the HFSS — failed to improve the prediction of survival in the Heidelberg sample. This is consistent with the prognostic analysis performed on the patients studied in the V-HeFT trial [19], and in agreement with the current guidelines for selection of potential candidates for cardiac transplantation in which LVEF and peak VO2 are recommended as the most reliable prognostic parameters [20,21].

Apart from choosing the relevant parameters for a prognostic score, the most useful clinical model should have at least one distinct threshold value at which the outcome of interest markedly changes. Since even the originally published derivation and validation samples varied in respect to their HFSS range for low, medium and high risk patients [5], sensitivity vs. specificity analyses (ROC curves) were applied as a way of comparing findings of different institutions in different countries. Whereas aetiology of CHF, serum sodium concentration, heart rate, mean arterial pressure and IVCD failed to improve the prediction of survival in the Heidelberg sample, a two-variable model containing both LVEF and peak VO2 was even superior to the HFSS (AUC 0.84 vs. 0.74; P<0.05). Further simplification of the prognostic score by determination of LVEF in combination with the distance ambulated during a 6'WT similarly improved risk stratification HFSS (AUC 0.83). Since the 6'WT is a much simpler and less expensive clinical tool to assess functional capacity than determination of peak VO2 by cardiopulmonary exercise testing [6,8], a two-variable model including the 6'WT may be more widely applicable in clinical practice.

Although the HFSS continued to discriminate mortality risk, and the combination of LVEF with either peak VO2 or 6'WT improved the prediction of 1-year survival, sensitivity and specificity (both approx. 70%) of each model are still not satisfactory, especially in the selection process for cardiac transplant candidates. The weaker performance of the previously identified HFSS thresholds, may result from the fact that only two (LVEF and 6'WT or peak VO2) out of seven variables continued to predict survival after multivariable Cox regression analysis. Furthermore, the HFSS assumes a linear relationship between six of the seven risk predictors and their impact on prognosis. However, the prognostic relevance of most risk predictors, such as peak VO2 [19,23] or LVEF [18,21], is based on a curvilinear relationship. Thus, the HFSS may lose prognostic power in the ‘healthier’ CHF cohorts, such as in the present study (NYHA 2.3±0.7 vs. 2.8±0.9 in the derivation sample).


   5. Limitations of the study
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 5. Limitations of the...
6. Conclusions and future...
 References
 
Although extensive work has been devoted to identifying multiple risk markers [24], the present study was — with the exception of the 6'WT — limited to the prognostic variables identified by Aaronson et al. [5]. Data from Holter recordings [25,26] and complementary risk factors such as VE/CO2 [22,27], impaired respiratory muscle strength [28], or inclusion of novel, easily obtainable independent circulating risk predictors, such as brain natriuretic peptides [29–31] or cytokines [32], might have further improved the predictive model. Whether such measurements could provide prognostic information in addition to a two-variable model consisting of LVEF in combination with either 6'WT or peak VO2 requires further investigation.

Additionally, as ACE inhibitors [33], β-blockers [34,35] and aldosterone antagonists [36] all markedly improve prognosis, one might assume that optimised treatment may be associated with a better outcome for a given HFSS score. The present study was too small to substantiate this notion. Thus, in future attempts to develop a prognostic scoring system for CHF patients, type and dosage of prognostically relevant medications may have to be integrated as further variables.


   6. Conclusions and future perspectives
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
References
 
This study confirms the important prognostic value of LVEF and peak VO2. In addition, the easily obtainable 6'WT was identified as an independent predictor of prognosis. Furthermore, the 6'WT was able to replace peak VO2 in a two-variable model (combination with LVEF) which itself outperformed the HFSS. However, sensitivity/specificity of the two-variable models still require further improvement and validation in independent patient cohorts. In future investigations, non-invasively obtainable independent risk factor(s) have to be identified, to optimise the proposed two-variable model(s), in order to improve the selection process for patients most likely to benefit from cardiac transplantation.


   Acknowledgements
 
This study was supported by a grant from the faculty for clinical medicine of the University of Heidelberg (project 32/95). The assistance of Karin Hornig, RN, is gratefully acknowledged. We thank Dr Bertram Krumm (Department of Biostatistics, Central Institute of Mental Health, Mannheim, Germany) for valuable comments and the critical review of this manuscript as well as Dr Daniel-L. Clarke-Pearson (Department of Oncology, Duke University, Durham, NC, USA) for providing the SAS-code for non-parametric area comparisons of correlated ROC curves.


   References
Top
 Abstract
 1. Introduction
 2. Methods
 3. Results
 4. Discussion
 5. Limitations of the...
 6. Conclusions and future...
 References
 

  1. Hosenpud J.D., Bennett L.E., Keck B.M., Boucek M.M., Novick R.J. The registry of the international society for heart and lung transplantation: seventeenth official report: 2000. J Heart Lung Transplant (2000) 19:909–931.[CrossRef][Web of Science][Medline]
  2. Haywood G.A., Rickenbacher P.R., Trindade P.T., et al. Analysis of deaths in patients awaiting heart transplantation: impact on patient selection criteria. Heart (1996) 75:455–462.[Abstract/Free Full Text]
  3. Stevenson L.W., Steimle A.E., Fonarow G., et al. Improvement in exercise capacity of candidates awaiting heart transplantation. J Am Coll Cardiol (1995) 25:163–170.[Abstract]
  4. Rector T.S., Cohn J.N. Prognosis in congestive heart failure. Annu Rev Med (1994) 45:341–350.[CrossRef][Web of Science][Medline]
  5. Aaronson K., Schwartz J., Chen T., Wong K., Goin J., Mancini D. Development and prospective validation of a clinical index to predict survival in ambulatory patients referred for cardiac transplant evaluation. Circulation (1997) 95:2660–2667.[Abstract/Free Full Text]
  6. Zugck C., Krüger C., Dürr S., et al. Is the 6-minute walk test a reliable substitute for peak oxygen uptake in patients with dilated cardiomyopathy? Eur Heart J (2000) 21:540–549.[Abstract/Free Full Text]
  7. Declaration of Helsinki. Br Med J 1964; ii: 177.
  8. Haass M., Zugck C., Kübler W. The 6 minute walking test: a cost-effective alternative to spiroergometry in patients with chronic heart failure? Z Kardiol (2000) 89:72–80.[CrossRef][Web of Science][Medline]
  9. Cox D. Regression models and life tables with discussion. J R Stat Soc (1972) 34:187–202.
  10. Kaplan E., Meier P. Nonparametric estimation from incomplete observations. Am J Stat Assoc (1958) 53:457–481.[CrossRef]
  11. Peto R., Peto L. Asymptomatically efficient rank invariant procedures. J R Stat Soc (1972) 135:185–207.
  12. Metz C., Wang P., Kronman H. A new approach for testing the significance of differences between ROC curves measured from correlated data. In: Information Processing in Medical Imaging—Deconinck F., ed. (1984) The Hague: Martinus Nijhoff. 432–445.
  13. Metz C., Herman B., Roe C. Statistical comparison of two ROC curves estimates obtained from partially paired datasets. Med Decis Mak (1998) 18:110–121.[CrossRef]
  14. Delong E., Delong D., Clarke-Pearson D. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics (1988) 44:837–845.[CrossRef][Web of Science][Medline]
  15. Poses R., Smith W., McClish D., et al. Physicians’ survival predictions for patients with acute congestive heart failure. Arch Intern Med (1997) 157:1001–1007.[Abstract/Free Full Text]
  16. Campana C., Gavazzi A., Berzuini C., et al. Predictors of prognosis in patients awaiting heart transplantation. J Heart Lung Transplant (1993) 12:756–765.[Web of Science][Medline]
  17. Alla F., Briancon S., Juilliere Y., Mertes P., Villemot J., Zannad F. Differential clinical prognostic classifications in dilated and ischemic advanced heart failure: the EPICAL study. Am Heart J (2000) 139:895–904.[Web of Science][Medline]
  18. Deng M.C., De Meester J.M.J., Smits J.M.A., et al. Effect of receiving a heart transplant: analysis of a national cohort entered on to a waiting list, stratified by heart failure severity. Br Med J (2000) 321:540–545.[Abstract/Free Full Text]
  19. The V-HeFT VA Cooperative Studies Group. Smith R.F., Johnson G., Ziesche S., Bhat G., Blankenship K., Cohn J.N. Functional capacity in heart failure. Comparison of methods for assessment and their relation to other indexes of heart failure. Circulation (1993) 87 (Suppl. VI):VI88–93.[Medline]
  20. Task force on Heart Failure of the European Society of Cardiology. Guidelines for the diagnosis of heart failure. Eur Heart J (1995) 16:741–751.[Free Full Text]
  21. American College of Cardiology/American Heart Association Task Force Report. Guidelines for the evaluation and management of heart failure. Circulation 1995; 92: 2764–2784.
  22. Francis D., Shamim W., Davies L., 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) 21:154–161.[Abstract/Free Full Text]
  23. Myers J., Gullestad L., Vagelos R., et al. Cardiopulmonary exercise testing and prognosis in severe heart failure: 14 ml/kg/min revisited. Am Heart J (2000) 139:78–84.[Web of Science][Medline]
  24. Gradman A., Deedwania P. Predictors of mortality in patients with heart failure. Cardiol Clin (1994) 12:25–35.[Medline]
  25. Rockman H., Juneau C., Chatterjee K., Rouleau J. Long-term predictors of sudden and low output death in chronic congestive heart failure secondary to coronary artery disease. Am J Cardiol (1989) 64:1344–1348.[CrossRef][Web of Science][Medline]
  26. Meinertz T., Hofmann T., Kasper W., et al. Significance of ventricular arrhythmias in idiopathic dilated cardiomyopathy. Am J Cardiol (1984) 53:902–907.[CrossRef][Web of Science][Medline]
  27. Kleber F., Vietzke G., Wernecke K., et al. Impairment of ventilatory efficiency in heart failure: prognostic impact. Circulation (2000) 101:2803–2809.[Abstract/Free Full Text]
  28. Meyer F., Borst M., Zugck C., et al. Respiratory muscle dysfunction in congestive heart failure: clinical correlation and prognostic significance. Circulation (2001) 103:2153–2158.[Abstract/Free Full Text]
  29. Gottlieb S., Kukin M., Ahern D., Packer M. Prognostic importance of atrial natriuretic peptide in patients with chronic heart failure. J Am Coll Cardiol (1989) 13:1534–1539.[Abstract]
  30. Cowie M. BNP: soon to become a routine measure in the care of patients with heart failure? Heart (2000) 83:617–618.[Free Full Text]
  31. Haass M, Kell R, Zugck C, Feist A, Chwallek F, Kübler W. N-terminal pro-brain natriuretic peptide in congestive heart failure. J Heart Failure (2000) 6:147 (Abstract).
  32. Levine B., Kalman J., Mayer L., Fillit H., Packer M. Elevated circulating levels of tumor necrosis factor in severe chronic heart failure. N Engl J Med (1990) 323:236–241.[Abstract]
  33. The SOLVD Investigators. Effet of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions. N Engl J Med (1992) 327:685–691.[Abstract]
  34. Carson P.E. Beta blocker treatment in heart failure. Prog Cardiovasc Dis (1999) 41:301–322.[CrossRef][Web of Science][Medline]
  35. Heidenreich P.A., Lee T., Massie B.M. Effect of beta-blockade on mortality in patients with heart failure: a meta-analysis of randomized clinical trials. J Am Coll Cardiol (1997) 30:27–34.[Abstract]
  36. Randomized Aldactone Evaluation Study Investigators. Pitt B., Zannad F., Remme W.J., et al. The effect of spironolactone on morbidity and mortality in patients with severe heart failure. N Engl J Med (1999) 341:709–717.[Abstract/Free Full Text]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?


This article has been cited by other articles:


Home page
 CirculationHome page
E. Hsich, E. Z. Gorodeski, R. C. Starling, E. H. Blackstone, H. Ishwaran, and M. S. Lauer
Importance of Treadmill Exercise Time as an Initial Prognostic Screening Tool in Patients With Systolic Left Ventricular Dysfunction
Circulation, June 30, 2009; 119(25): 3189 - 3197.
[Abstract] [Full Text] [PDF]

Home page
 QJMHome page
B. Whelan, K. Bennett, D. O'riordan, and B. Silke
Serum sodium as a risk factor for in-hospital mortality in acute unselected general medical patients
QJM, March 1, 2009; 102(3): 175 - 182.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
J. A. Spertus
Evolving Applications for Patient-Centered Health Status Measures
Circulation, November 11, 2008; 118(20): 2103 - 2110.
[Abstract] [Full Text] [PDF]

Home page
 Journals of Gerontology Series A: Biological Sciences and Medical SciencesHome page
C. M. Boyd, C. O. Weiss, J. Halter, K. C. Han, W. B. Ershler, and L. P. Fried
Framework for Evaluating Disease Severity Measures in Older Adults With Comorbidity
J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2007; 62(3): 286 - 295.
[Abstract] [Full Text] [PDF]

Home page
 CirculationHome page
W. C. Levy, D. Mozaffarian, D. T. Linker, S. C. Sutradhar, S. D. Anker, A. B. Cropp, I. Anand, A. Maggioni, P. Burton, M. D. Sullivan, et al.
The Seattle Heart Failure Model: Prediction of Survival in Heart Failure
Circulation, March 21, 2006; 113(11): 1424 - 1433.
[Abstract] [Full Text] [PDF]

Home page
 Eur J Heart FailHome page
J. Junger, D. Schellberg, T. Muller-Tasch, G. Raupp, C. Zugck, A. Haunstetter, S. Zipfel, W. Herzog, and M. Haass
Depression increasingly predicts mortality in the course of congestive heart failure
Eur J Heart Fail, March 2, 2005; 7(2): 261 - 267.
[Abstract] [Full Text] [PDF]

Home page
 JAMAHome page
G. C. Fonarow, K. F. Adams Jr, W. T. Abraham, C. W. Yancy, W. J. Boscardin, and for the ADHERE Scientific Advisory Committee, Stud
Risk Stratification for In-Hospital Mortality in Acutely Decompensated Heart Failure: Classification and Regression Tree Analysis
JAMA, February 2, 2005; 293(5): 572 - 580.
[Abstract] [Full Text] [PDF]

Home page
 HeartHome page
M R Cowie
Estimating prognosis in heart failure: time for a better approach
Heart, June 1, 2003; 89(6): 587 - 588.
[Full Text] [PDF]

Home page
 HeartHome page
M L Bouvy, E R Heerdink, H G M Leufkens, and A W Hoes
Predicting mortality in patients with heart failure: a pragmatic approach
Heart, June 1, 2003; 89(6): 605 - 609.
[Abstract] [Full Text] [PDF]

Home page
 J Am Coll CardiolHome page
C. Zugck, A. Haunstetter, C. Kruger, R. Kell, D. Schellberg, W. Kubler, and M. Haass
Impact of beta-blocker treatment on the prognostic value of currently used risk predictors in congestive heart failure
J. Am. Coll. Cardiol., May 15, 2002; 39(10): 1615 - 1622.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (18)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Zugck, C.
Right arrow Articles by Haass, M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zugck, C.
Right arrow Articles by Haass, M.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?