INTRODUCTION
A noncomprehensive list of factors that can assess the severity of heart failure (HF) and are related to outcomes in patients with chronic heart failure (CHF) include age; gender; ethnicity; etiology; comorbidity; New York Heart Association (NYHA class); exercise capacity; peak VO2; poor qua- lity of life; low body weight; left bundle branch block (LBBB); atrial fibrillation; nonsustained, sustained, and inducible ventricular tachycardia (VT); prolonged PR and QRS duration; T-wave alternans; QT dispersion; low heart rate variability; depressed baroreflex sensitivity; history of HF hospitalization; resuscitated death; hyponatremia; hypokalemia; raised serum creatinine and blood urea nitrogen (BUN); transaminases; bilirubin and urates; anemia; neuroendocrine activation; high serum brain natriuretic peptide (BNP); low left ventricular ejection fraction (LVEF); diastolic function parameters; raised serum levels of markers of extracellular matrix metabolism; viable myocardium; and central hemodynamic parameters.1
The above list of many predictive variables reflects the difficulties in choosing which prognostic variables to use for clinical purposes. This situation has been described by Jay Cohn as “Poverty amidst a wealth of variables.” Indeed, it appears that no single prognostic indicator is perfect.2,3

Prognostic stratification should differ in relation to the goal and must be useful for making therapeutic decisions. It may also be used for clinical trial design and more specifically for optimization of the risk level of the enrolled patient population, which is the main determinant of the trial population sample size.

Prognostic analyses have been predominantly carried out on populations with left ventricular systolic dysfunction (LVSD). Therefore, the following overview is mainly focused on prognostication in such patients. Much less data are available for HF with preserved systolic function.4

It must also be recognized that causal disease as well as comorbid conditions can strongly impact outcomes such as coronary etiology and diabetes.5–7 In addition, among demographic characteristics, advanced age, as one may expect, and Black ethnicity are consistently reported to negatively influence outcome.8

Nearly all predictors of prognosis are influenced by treatment, which can modify their prognostic weight over time. The influence of etiology, comorbidity, age, and ethnicity will not be discussed in this overview.

SYMPTOMS, FUNCTIONAL STATUS, AND THE SEVERITY OF HEART FAILURE
Using the NYHA classification and clinical judgment, patients may be classified into Class I–IV or alternatively into asymptomatic, mild, moderate, or severe. However, mild symptoms do not mean minor cardiac dysfunction. Indeed it should be emphasized that there is a poor relationship between symptoms and the severity of cardiac dysfunction or prognosis.9,10

Exercise Testing
Dyspnea and fatigue are the two main causes of limitation of functional capacity in patients with CHF; therefore, it makes sense to assess the severity of the disease by measuring its influence on exercise capacity.

Recommendations for exercise testing in HF patients have been released by the Working Group on Cardiac Rehabilitation and Exercise Physiology and the Working Group on Heart Failure of the ESC.11

Exercise capacity has proven to be a strong determinant of the risk profile in CHF. Oxygen uptake is a more stable and reliable measure of exercise tolerance than exercise time. A peak VO2 <10 mL/kg/min identifies high risk and a peak VO2 >18 mL/kg/min identifies low-risk patients.

Values between these cutoff limits define a zone of medium-risk patients without further possible stratification by VO2.11 Change of VO2 over time and following optimized therapy is more relevant than absolute values at one single assessment. A markedly reduced and continuously declining exercise capacity in patients with optimized therapy should warrant intensifying therapeutic management and is an indication for heart transplant.11,12

In patients with serious limitation of functional capacity, submaximal testing with the 6-minute walk test has been shown to provide useful prognostic information when walking distance is <300 m.13,14 However, the value of the 6-minute walk test has been established using mainly clinical trial databases and is unclear in the clinical setting. Cutoff values of VO2 = 14 mL/kg/min and distance walked in 6 minutes of 300 m are used for indicating heart transplantation, although this has never been properly validated.

Quality of Life
Although quality of life as measured by appropriate questionnaires (the Minnesota Living with Heart Failure, the SF-36, and the Kansas City Cardiomyopathy Questionnaire) may be associated with prognosis, it is not used in the context of classification of severity.15–17

History of HF Hospitalization
The significant morbidity associated with HF is reflected in hospital readmission rates, which are higher than those observed for acute myocardial infarction. Estimates of risk of death or readmission vary, but the largest randomized trial in patients hospitalized with decompensated HF found the 60-day mortality rate to be 9.6% and the combined 60-day mortality or readmission rate to be 35.2%.18 The Euro Heart Failure Survey program found that 24% of patients with HF were readmitted within 12 weeks of discharge.19 Additionally, in a population-based survey in London, United Kingdom, 50% of patients with a new diagnosis of HF were subsequently hospitalized at least once over a period of 19 months.20 History of HF hospitalization is associated with a very high risk of mortality. Mortality rate reported by the Euro Heart Failure Survey (13% mortality at 12 weeks) is probably an underestimation.19 Prospective cohorts report 1-year mortality as high as 40%.21,22

ELECTROCARDIOGRAM
A normal electrocardiogram (ECG) may rule out LVSD (negative predictive 90%).23–26 LBBB and/or wide QRS are associated with LVSD and poor outcome.27

QRS duration may guide the use of resynchronization therapy.28 Atrial fibrillation or flutter and ventricular arrhythmia may be recorded using simple ECG. These arrhythmias are more frequent in severe left ventricular (LV) dysfunction, and are associated with poor outcome.

Asymptomatic ventricular arrhythmias on ambulatory electrocardiographic monitoring do not identify specific candidates for antiarrhythmic or device therapy. Contrary to findings from the GESICA trial, nonsustained VT was not found to be a specific predictor of mortality in a multivariate analysis of the CHF-STAT and PROMISE studies.29,30

In patients with symptomatic arrhythmias, Holter monitoring may detect and characterize atrial and ventricular arrhythmias, which could be causing or exacerbating symptoms of HF, and guide antiarrhythmic therapy.

Other parameters derived from Holter monitoring are those assessing heart rate variability. Heart rate variability is reduced in HF as a consequence of depressed autonomic balance.31–33 Time and frequency domain measures of heart rate variability correlate with clinical and hemodynamic measures of severity, and are independently associated with survival.30,31,34–37 However, so far, there has been no validation of specific management strategies based on heart rate variability assessment.

LEFT VENTRICULAR FUNCTION

Left Ventricular Ejection Fraction
The most important measurement of ventricular function is LVEF for distinguishing patients with cardiac systolic dysfunction from patients with preserved systolic function. In asymptomatic patients with LVSD, EF is an important prognostic marker for the development of manifest HF and death.38,39 Low EF with or without symptoms is the single risk factor considered as the basis for initiating treatment with angiotensinconverting enzyme (ACE) inhibitors.40 Low EF in symptomatic patients is an formal indication for b-blocker and angiotensin receptor blocker therapy.41,42 Patients with low EF that remain symptomatic on ACE inhibitor and b-blocker therapy should receive an aldosterone receptor blocker.43 Low EF in patients with ischemic CHF is an indication for an implantable cardioverter defibrillator (ICD), and a low EF and wide QRS in patients receiving optimal medical therapy is an indication for ICD with resynchronization therapy.28 Therefore, LVEF is the single most useful prognostic factor, because it is the basis of important therapeutic decisions. Although reproducibility of LVEF assessment with single photon emission computed tomography (SPECT) is better than with echocardiography, it is echo LVEF that is most extensively used for therapeutic decisions.

Diastolic Function
Staging of diastolic dysfunction may be performed during a routine echocardiographic examination assessing transmitral blood flow velocities and mitral annular velocities. Three abnormal left ventricular filling patterns have been described corresponding to mild, moderate, and severe diastolic dysfunction, respectively.44,45 Mild diastolic dysfunction is characterized by a reduction of peak transmitral E-velocity and an increase in the atrial-induced

(A) velocity. Therefore, the E/A ratio is reduced and usually <1. In moderate diastolic dysfunction, E/A ratio and E-deceleration time may be normal. Only tissue Doppler imaging may depict reduced peak E velocity.45 Patients with severe diastolic dysfunction have a pattern of “restrictive filling,” with a short isovolumic relaxation time (IVRT), an elevated peak E-velocity, a short E-deceleration time, and a markedly increased E/A ratio.46–49

Beyond these distinctive patterns, a restrictive filling pattern, characterized by short transmitral E-deceleration time (115–150 milliseconds) and increased E/A flow velocity ratio (>1.5–2), is associated with increased mortality.50,51

Although assessment of diastolic function may be clinically useful in determining prognosis in HF patients, so far there is no prospective validation of therapeutic management strategies based on the assessment of diastolic function.

Other Measurements of Cardiac Function
Ventricular volume changes over time, and the onset or worsening of mitral regurgitation have important decisional implications because it should lead to further diagnostic investigations and/or intensification of therapy.52,53
Other measurements include fractional shortening, myocardial performance index, and left ventricular wall motion index.54–57 Cardiac magnetic resonance imaging is a highly accurate and reproducible technique for the assessment of left and right ventricular volumes and function.58,59

MYOCARDIAL VIABILITY
Although this is controversial, results of nonrandomized trials in ischemic CHF indicate that revascularization can improve clinical status and survival in patients with hibernating myocardium.60–62

Therefore, in patients with CHF and coronary artery disease, exercise or pharmacological stress echocardiography may be useful for detecting ischemia as a cause of reversible or persistent dysfunction and in determining the viability of akinetic myocardium.63,64 Viable myocardium (stunned or nontransmural infarction) has preserved flow reserve and sustained contractile improvement. Hibernating myocardium has blunted flow reserve and a biphasic contractile response.

Cardiac magnetic resonance imaging after an injection of gadolinium can identify areas of delayed hyperenhancement in areas of stunning or hibernation.65,66

RISK OF SUDDEN DEATH
A review of the measures of risk of sudden arrhythmic death has been reported by Huikuri and others.67 Several observational studies have shown that low EF predicts both sudden and nonsudden cardiac death.68,69 MADIT II SCD-HeFT showed that ICDs reduce mortality among patients with an EF 30% with ischemic HF.70 SCD-HeFT has shown that in Class II or III CHF patients with EF <35%, simple shock ICD may decrease mortality.71 The COMPANION trial has shown that ICD combined with resynchronization with biventricular pacing decreases mortality in patients with EF <35% and ECG measured QRS duration >120 milliseconds.28

Resuscitated sudden death, when it occurs out of context of an acute ischemic event, is a strong predictor of recurrence of sudden death and warrants implantation of an ICD.72

Electrical programmed stimulation is useful only in patients with low EF and nonsustained VT. Patients with nonsustained VT, low EF, and inducible VT benefit from ICD therapy, according to two randomized trials.72–74 Sustained VT warrants antitachycardia ICD implantation.

The majority of other variables occasionally reported to be associated with the risk of sudden death in observational and case-control studies have not received a prospective validation and are not included in therapeutic strategy algorithms. Therefore, this is an area where prognostication is most needed. A large number of patients may be implanted with an ICD, which would never fire for a life-threatening arrhythmia. Robust and reliable risk factors may limit the implantation of ICDs to patients who would need it most.

HEMATOLOGY AND BIOCHEMISTRY

Anemia
A growing body of literature from observational databases and clinical trials suggests that anemia is an independent risk factor for adverse outcomes in patients with HF.

Anemia has recently been recognized as an important comorbid condition and potentially novel therapeutic target in patients with HF. It is common in CHF patients, with a prevalence ranging from 4% to 55% depending on the population studied. Lower hemoglobin (Hb) is associated with greater disease severity, and higher hospitalization and mortality rates.

Multiple potential mechanisms of interaction exist between anemia and the clinical syndrome of CHF, including hemodilution, inflammatory activation, renal insufficiency, and malnutrition. Although correction of anemia appears to be an attractive therapeutic approach, it is still unclear whether it is useful. Multiple ongoing studies will provide data on the balance of risks and benefits of anemia treatment in chronic HF.5,75

Kidney Function
Kidney function has a significant impact on clinical outcomes of patients with congestive heart failure. Impaired renal function, whether mild or severe, is an independent predictor of a worse prognosis.76,77 CHF patients with elevated serum creatinine levels of 1.5–2 mg/dL have a 41% higher death rate, and those with a glomerular filtration rate (GFR) of <44 mL/min have a threefold increase in mortality.78,79

Renal function may be a primary determinant of disease progression in patients with HF, rather than simply being a marker of the severity of underlying disease.

Renal function is an independent prognostic indicator of mortality, possibly due to the direct effects of renal dysfunction on survival. The fluid retention resulting from failing kidneys may lead to neurohormonal activation, ventricular dilatation, and arrhythmias. In addition, electrolyte and metabolic abnormalities—hematologic effects and consequences on the immune system that commonly occur in renal failure— could conceivably lead to a worse outcome. Renal dysfunction may also prevent the use of medications known to improve clinical outcome and survival such as ACE inhibitors.80

Preserving or improving kidney function in patients with CHF is one of the major unmet needs in CHF management and an area where research should be intensified. The majority of clinical trials that have led to our current evidence-based management excluded patients with renal failure. Therefore, apart from the usual recommendation of caution of use of a number of drugs, there is no specific guideline of management based on the presence of renal failure in patients with CHF.

Serum Sodium
Serum sodium concentration is a powerful predictor of cardiovascular mortality. Although hyponatremia is thought to be an indicator of the level of stimulation of the renin-angiotensin system (RAS), it is among the most consistent prognostic factor in advanced HF,81 even in the contemporary patients receiving RAS inhibitor therapy.

NEUROENDOCRINE EVALUATIONS

Natriuretic Peptides
An excellent recent overview has summarized the role of BNP in HF.82 BNP and N-terminal probrain natriuretic peptide (NT-proBNP) have considerable prognostic potential, although evaluation of their role in treatment decision and monitoring remains to be determined.

Several clinical and epidemiological studies have demonstrated a direct relationship between increasing plasma concentrations of BNP (and its precursor NT-proBNP) and decreasing cardiac function.83–85 There is also evidence that their elevation in CHF patients with preserved systolic function can indicate that diastolic dysfunction is present.86,87

In patients at high risk for developing new HF, an elevated atrial natriuretic peptides (ANP) was 85% sensitive and 66% specific for the development of a subsequent HF episode during 1 year of follow-up.88 A Framingham community study has shown that a strategy of combining EF and BNP (or NT-proBNP) improved risk stratification beyond using either alone.89

Elevated ANP and BNP have been shown to be predictive of poor long-term survival and of sudden cardiac death.90–94 In those with elevated BNP, rate-corrected QT interval was a strong and independent predictor of total mortality, as well as sudden death mortality.95

Change in BNP is also useful to predict clinical outcomes. A decrease in BNP during treatment was associated with fewer adverse events after discharge for acutely decompensated HF.96 Outpatients with the greatest increase in BNP despite therapy had the poorest outcome.97

At present, the relative merits of the available assays for BNP and NT-proBNP, the value of BNP for therapeutic decision making and in monitoring therapy are important areas for continuing investigation. The natriuretic peptide assays commercially available are fluorescence or radioimmunoassay for BNP, and electrochemiluminescent assay for NT-proBNP. In general, the plasma BNP concentration rises with age and may be slightly higher in women than in men. A suggested “normal” range for BNP is 0.5–30 pg/mL (0.15–8.7 pmol/L) and 68–112 pg/mL (8.2–13.3 pmol/L) for NT-proBNP.98

Neuroendocrine Evaluations Other than Natriuretic Peptides
In large cohorts of patients, there is good evidence that circulating levels of noradrenaline, renin, angiotensin II, aldosterone, vasopressin, endothelin-1, and adrenomedullin are related to the severity and prognosis of HF, but in individual patients these predictors are inaccurate and difficult to interpret. Degree of neurohormonal activation cannot guide treatment with reninangiotensin and aldosterone inhibitors.

SCORING AND PROGNOSTIC ALGORITHMS
An interesting prognostic approach can lie in integrated strategies based on an initial screening of patients of different disease severity and then the application of specific algorithms to selected subjects.95 A review of prognostic algorithms has been reported by Bouvy and others.99 None is widely accepted for routine use in HF. Available prognostic algorithms are usually limited because they are mainly derived prospectively from nonrepresentative cohorts or clinical trial databases. Very few have received a formal external validation in independent prospective cohorts.

ACUTE HEART FAILURE SYNDROMES
In acute heart failure syndromes (AHFS), predictive models for mortality and rehospitalization can aid clinical decision making. Patients at low risk could potentially be treated and discharged from the hospital early, whereas those at high risk may benefit from intensive specialized care.

In this setting, stratification based on the degree of neurohormonal activation and the severity of LVSD become less discriminant.6 Central hemodynamic patterns (pulmonary capillary pressure and right ventricular function) are relatively more important.100–102 Invasive hemodynamic monitoring variables by means of a pulmonary arterial catheter may be useful to monitor therapy and for decision making in patients with AHFS not responding promptly to appropriate treatment and when it must be decided whether to use ventricular assistance or replacement therapies.6,103

Three main prognostic indicators are emerging as most important in AHFS. These include myocardial injury (quantified by troponin), renal dysfunction (as measured by increases in serum creatinine or BUN, and decreases in serum sodium), and hemodynamic congestion (as measured by increased pulmonary capillary wedge pressure [PCWP] or BNP/NT-proBNP). While previous research has largely focused on correction of altered hemodynamics (e.g., increasing cardiac output), new data suggest that the above three factors may be more important for long-term prognosis as well as for therapeutic targets. While current standard therapies may improve symptoms (e.g., diuretics), they may worsen renal function. Similarly, while some positive inotropes may improve hemodynamics (increase cardiac index), they may promote myocardial injury.

Several observational studies have shown that 30–50% of patients hospitalized with AHFS have detectable plasma levels of cardiac troponin at the time of admission in the absence of an acute coronary event. These patients have a two-fold increase in 60-day postdischarge mortality and a three-fold increase in the rehospitalization rate during the same time period.104

The severity of renal dysfunction in hospitalized patients with AHFS provides important prognostic information for in-hospital and post-discharge mortality. Aggravated renal dysfunction (defined as ≥25% increase in serum creatinine concentration to ≥2 mg/dL) occurs in at least 20–30% of patients undergoing intensive treatment for HF.105 The development of worsening renal function is associated with an increased risk of death, a significantly longer length of stay, and higher in-hospital cost.106 Impaired renal function also increases the likelihood of readmission after discharge from hospitalization for HF.107,108 A1 adenosine antagonism might preserve renal function while simultaneously promoting natriuresis during treatment for HF and is currently under clinical investigation.109

Approximately 20% of patients hospitalized with AHFS and systolic dysfunction have hyponatremia (serum sodium <136 mEq/L), and these patients have a twofold increase in in-hospital and postdischarge mortality, and a 30% increase in the combined readmission or mortality rate.110,111 Vasopressin antagonist therapy (such as Tolvaptan) is being tested with the aim of improving post-discharge outcomes in patients with AHFS.112 Though Tolvaptan initiated for acute treatment in AHFS patients did not have an effect on long-term moratality or heart failure morbidity,113 it did relieve acute symptoms.114

CONCLUSION
It is likely that different sets of variables will prove useful in different settings. In primary care, it is important to be able to stratify risk on the basis of simple, readily available clinical or laboratory variables to identify patients who should be referred for specialist advice. Serum BNP monitoring may emerge as a useful tool in this setting. Meanwhile, primary care physicians rely mainly on functional capacity, signs, and symptoms.

For the specialists, prognostic variables are required to direct the intensity of therapy. The main selection criteria in clinical trials were symptoms, NYHA class, LVEF and, in some trials, history of recent hospitalization for HF. Therefore, only these variables are prognostic factors strongly validated and widely accepted and applied in evidence-based patient management.

In more advanced HF, prognostic stratification may guide the need for or urgency of device therapy and/or surgery, including transplantation. Peak VO2 and the 6-minute walk test are among the most used assessments for the management of transplant waiting list. They should also be recommended as decision tools for indicating cardiac resynchronization therapy in patients with optimized drug therapy. VT inducibility during ventricular programmed stimulation is useful (only) in patients with asymptomatic VT, in order to direct ICD indication.

New validated information and more integrated approaches may offer, in the future, prognostic algorithms that are more robust prognostication in HF. Genomics and proteomics may offer novel disease markers and risk (or protecting) factors. However, to date, no tests can overcome the clinical judgment in grading risk and guiding therapy in HF patients.

REFERENCES
1. Cowburn P, Cleland J, Coats A, et al. Risk stratification in chronic heart failure. Eur Heart J. 1998;19:696–710.
2. Cohn J. Prognostic factors in heart failure: poverty amidst a wealth of variables. J Am Coll Cardiol. 1989;14:571–572.
3. Eichhorn E. Prognosis determination in heart failure. Am J Med. 2001;110:S14–S36.
4. Gustafsson F, Torp-Pedersen C, Brendorp B, et al. Long-term survival in patients hospitalized with congestive heart failure: relation to preserved and reduced left ventricular systolic function. Eur Heart J. 2003;24:863–870.
5. Felker G, Shaw L, O’Connor C. A standardized definition of ischemic cardiomyopathy for use in clinical research. J Am Coll Cardiol. 2002;39:210–218.
6. Nohria A, Tsang S, Fang J, et al. Clinical assessment identifies hemodynamic profiles that predict outcomes in patients admitted with heart failure. J Am Coll Cardiol. 2003;41:1797–1804.
7. Dries D, Sweitzer N, Drazner M, et al. Prognostic impact of diabetes mellitus in patients with heart failure according to the etiology of left ventricular systolic dysfunction. J Am Coll Cardiol. 2001;38:421–428.
8. Pulignano G, Del Sindaco D, Tavazzi L, et al. Clinical features and outcomes of elderly outpatients with heart failure followed up in hospital cardiology units: data from a large nationwide cardiology database (IN-CHF Registry). Am Heart J. 2002;143:45–55.
9. Marantz P, Tobin J, Wassertheil-Smoller S, et al. The relationship between left ventricular systolic function and congestive heart failure diagnosed by clinical criteria. Circulation. 1988;77:607–612.
10. Van den Broek S, Van Veldhuisen D, De Graeff P, et al. Comparison between New York Heart Association classification and peak oxygen consumption in the assessment of functional status and prognosis in patients with mild to
moderate chronic congestive heart failure secondary to either ischemic or idiopathic. Am J Cardiol. 1992;70:359–363.
1. Working Group on Cardiac Rehabilitation and Exercise Physiology and Working Group on Heart Failure of the European Society of Cardiology: Recommendations for exercise testing in chronic heart failure patients. Eur Heart J. 2001;22:37–45.
2. Mancini D, Eisen H, Kussmaul W, et al. Value of peak exercise oxygen consumption for optimal timing of cardiac transplantation in ambulatory patients with heart failure. Circulation. 1991;83:778–786.
3. Guyatt G, Sullivan M, Thompson P, et al. The 6-minute walk: a new measure of exercise capacity in patients with chronic heart failure. CMAJ. 1985;132:919–923.
4. Bittner V, Weiner D, Yusuf S, et al. Prediction of mortality and morbidity with a 6-minute walk test in patients with left ventricular dysfunction. JAMA. 1993;270:1702–1707.
5. Rector T, Cohn J, Pimobendan Multicenter Research Group. Assessment of patient outcome with the Minnesota Living with Heart Failure questionnaire: reliability and validity during a randomized, double-blind, placebo-controlled trial of pimobendan. Am Heart J. 1992;124:1017–1025.
6. McHorney C, Ware J, Raczek A. The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs. Med Care. 1993;31:247–263.
7. Green C, Porter C, Bresnahan D, et al. Development and evaluation of the Kansas City Cardiomyopathy Questionnaire: a new health status measure for heart failure. J Am Coll Cardiol. 2000;35:1245–1255.
8. Cuffe M, Califf R, Adams K J, et al. Short-term intravenous milrinone for acute exacerbation of chronic heart failure: a randomized controlled trial. JAMA. 2002;287:1541–1547.
9. Cleland J, Swedberg K, Follath F, et al. The EuroHeart Failure survey programme— a survey on the quality of care among patients with heart failure in Europe. Part 1: patient characteristics and diagnosis. Eur Heart J. 2003; 24:442–463.
10. Cowie M, Fox K, Wood D, et al. Hospitalization of patients with heart failure: a population-based study. Eur Heart J. 2002;23:877–885.
1. Zannad F, Cohen-Solal A, Desnos M, et al. Clinical and etiological features, management and outcomes of acute heart failure: the EFICA cohort study. Eur Heart J. 2002;4 (Suppl):579.
2. Zannad F, Briançon S, Juillière Y, et al. Incidence, clinical and etiologic features, and outcomes of advanced chronic heart failure: the EPICAL Study. Epidemiologie de l’Insuffisance Cardiaque Avancee en Lorraine. J Am Coll Cardiol. 1999;33:734–42.    3. Rihal C, Davis K, Kennedy J, et al. The utility of clinical, electrocardiographic, and roentgenographic variables in the prediction of left ventricular function. Am J Cardiol. 1995;75:220–223.
4. Gillespie N, McNeill G, Pringle T, et al. Cross sectional study of contribution of clinical assessment and simple cardiac investigations to diagnosis of left ventricular systolic dysfunction in patients admitted with acute dyspnoea. BMJ. 1997;314:936–340.
5. Mosterd A, de Bruijne M, Hoes A, et al. Usefulness of echocardiography in detecting left ventricular dysfunction in population-based studies (The Rotterdam Study). Am J Cardiol. 1997;79:103–104.
6. Badgett R, Lucey C, Mulrow C. Can the clinical examination diagnose left-sided heart failure in adults? JAMA. 1997;277:1712–1719.
7. Baldasseroni S, Opasich C, Gorini M, et al. Left bundle-branch block is associated with increased 1-year sudden and total mortality rate in 5517 outpatients with congestive heart failure: a report from the Italian network on congestive heart failure. Am Heart J. 2002;143:398–405.
8. Bristow M, Saxon L, Boehmer J, et al. Cardiacresynchronization therapy with or without an implantable defibrillator in advanced chronic heart failure. N Engl J Med. 2004;350:2140–2150.
9. Doval H, Nul D, Grancelli H, et al. Nonsustained ventricular tachycardia in severe heart failure. Independent marker of increased mortality due to sudden death. GESICA-GEMA Investigators. Circulation. 1996;94:3198–3203.
10. Teerlink J, Jalaluddin M, Anderson S, et al. Ambulatory ventricular arrhythmias in patients with heart failure do not specifically predict an increased risk of sudden death. PROMISE (Prospective Randomized Milrinone Survival Evaluation) Investigators. Circulation. 2000;101:40–46.