ABC of Heart Failure - R. Davis, F. Hobbs, G. Lip (BMJ) WW

ABC of heart failureHistory and epidemiologyR C Davis, F D R Hobbs, G Y H Lip

Heart failure is the end stage of all diseases of the heart and is amajor cause of morbidity and mortality. It is estimated toaccount for about 5% of admissions to hospital medical wards,with over 100 000 annual admissions in the United Kingdom.

The overall prevalence of heart failure is 320 per 1000population, although this exceeds 100 per 1000 in those aged65 years and over. The annual incidence of heart failure is 15per 1000, and the relative incidence doubles for each decade oflife after the age of 45 years. The overall incidence is likely toincrease in the future, because of both an ageing populationand therapeutic advances in the management of acutemyocardial infarction leading to improved survival in patientswith impaired cardiac function.

Unfortunately, heart failure can be difficult to diagnoseclinically, as many features of the condition are not organspecific, and there may be few clinical features in the earlystages of the disease. Recent advances have made the earlyrecognition of heart failure increasingly important as moderndrug treatment has the potential to improve symptoms andquality of life, reduce hospital admission rates, slow the rate ofdisease progression, and improve survival. In addition, coronaryrevascularisation and heart valve surgery are now regularlyperformed, even in elderly patients.

A brief historyDescriptions of heart failure exist from ancient Egypt, Greece,and India, and the Romans were known to use the foxglove asmedicine. Little understanding of the nature of the conditioncan have existed until William Harvey described the circulationin 1628. Röntgen’s discovery of x rays and Einthoven’sdevelopment of electrocardiography in the 1890s led toimprovements in the investigation of heart failure. The adventof echocardiography, cardiac catheterisation, and nuclearmedicine have since improved the diagnosis and investigationof patients with heart failure.

Blood letting and leeches were used for centuries, andWilliam Withering published his account of the benefits ofdigitalis in 1785. In the 19th and early 20th centuries, heartfailure associated with fluid retention was treated with Southey’stubes, which were inserted into oedematous peripheries,allowing some drainage of fluid.

“The very essence of cardiovascularpractice is the early detection of heartfailure”Sir Thomas Lewis, 1933

Some definitions of heart failure

“A condition in which the heart fails to discharge its contentsadequately” (Thomas Lewis, 1933)

“A state in which the heart fails to maintain an adequate circulationfor the needs of the body despite a satisfactory filling pressure”(Paul Wood, 1950)

“A pathophysiological state in which an abnormality of cardiacfunction is responsible for the failure of the heart to pump bloodat a rate commensurate with the requirements of the metabolisingtissues” (E Braunwald, 1980)

“Heart failure is the state of any heart disease in which, despiteadequate ventricular filling, the heart’s output is decreased or inwhich the heart is unable to pump blood at a rate adequate forsatisfying the requirements of the tissues with function parametersremaining within normal limits” (H Denolin, H Kuhn, H PKrayenbuehl, F Loogen, A Reale, 1983)

“A clinical syndrome caused by an abnormality of the heart andrecognised by a characteristic pattern of haemodynamic, renal,neural and hormonal responses” (Philip PooleWilson, 1985)

“[A] syndrome . . . which arises when the heart is chronically unable tomaintain an appropriate blood pressure without support” (PeterHarris, 1987)

“A syndrome in which cardiac dysfunction is associated with reducedexercise tolerance, a high incidence of ventricular arrhythmias andshortened life expectancy” (Jay Cohn, 1988)

“Abnormal function of the heart causing a limitation of exercisecapacity” or “ventricular dysfunction with symptoms” (anonymousand pragmatic)

“Symptoms of heart failure, objective evidence of cardiac dysfunctionand response to treatment directed towards heart failure” (TaskForce of the European Society of Cardiology, 1995)

The foxglove was used as a medicine in heartdisease as long ago as Roman times

Southey’s tubes were at one time used for removing fluid from oedematousperipheries in patients with heart failure

Clinical review

39BMJ VOLUME 320 1 JANUARY 2000 www.bmj.com

on 1 October 2006 bmj.comDownloaded from

1

It was not until the 20th century that diuretics weredeveloped. The early, mercurial agents, however, wereassociated with substantial toxicity, unlike the thiazide diuretics,which were introduced in the 1950s. Vasodilators were notwidely used until the development of angiotensin convertingenzyme inhibitors in the 1970s. The landmark CONSENSUSIstudy (first cooperative north Scandinavian enalapril survivalstudy), published in 1987, showed the unequivocal survivalbenefits of enalapril in patients with severe heart failure.

EpidemiologyStudies of the epidemiology of heart failure have beencomplicated by the lack of universal agreement on a definitionof heart failure, which is primarily a clinical diagnosis. Nationaland international comparisons have therefore been difficult,and mortality data, postmortem studies, and hospital admissionrates are not easily translated into incidence and prevalence.Several different systems have been used in large populationstudies, with the use of scores for clinical features determinedfrom history and examination, and in most cases chestradiography, to define heart failure.

The Task Force on Heart Failure of the European Society ofCardiology has recently published guidelines on the diagnosisof heart failure, which require the presence of symptoms andobjective evidence of cardiac dysfunction. Reversibility ofsymptoms on appropriate treatment is also desirable.Echocardiography is recommended as the most practicable wayof assessing cardiac function, and this investigation has beenused in more recent studies.

In the Framingham heart study a cohort of 5209 subjectshas been assessed biennially since 1948, with a further cohort(their offspring) added in 1971. This uniquely large dataset hasbeen used to determine the incidence and prevalence of heartfailure, defined with consistent clinical and radiographic criteria.

Several recent British studies of the epidemiology of heartfailure and left ventricular dysfunction have been conducted,including a study of the incidence of heart failure in one westLondon district (Hillingdon heart failure study) and largeprevalence studies in Glasgow (north Glasgow MONICA study)and the West Midlands ECHOES (echocardiographic heart ofEngland screening) study. It is important to note that

A brief history of heart failure

1628 William Harvey describes the circulation

1785 William Withering publishes an account of medicaluse of digitalis

1819 René Laennec invents the stethoscope

1895 Wilhelm Röntgen discovers x rays

1920 Organomercurial diuretics are first used

1954 Inge Edler and Hellmuth Hertz use ultrasound toimage cardiac structures

1958 Thiazide diuretics are introduced

1967 Christiaan Barnard performs first human hearttransplant

1987 CONSENSUSI study shows unequivocal survivalbenefit of angiotensin converting enzyme inhibitors insevere heart failure

1995 European Society of Cardiology publishes guidelinesfor diagnosing heart failure

The Framingham heart study has beenthe most important longitudinal source ofdata on the epidemiology of heart failure

Contemporary studies of the epidemiology of heart failurein United Kingdom

Study Diagnostic criteria

Hillingdon heart failure study(west London)

Clinical (for example, shortness ofbreath, effort intolerance, fluidretention), radiographic, andechocardiographic

ECHOES study (West Midlands) Clinical and echocardiographic(ejection fraction < 40%)

MONICA population(north Glasgow)

Clinical and echocardiographic(ejection fraction <30%)

In 1785 William Withering of Birmingham publishedan account of medicinal use of digitalis

12

Months

Cu

mu

lati

ve p

rob

abili

ty o

f d

eath

1086420 11975310

0.2

0.3

0.4

0.5

0.6

0.7

0.8Placebo

0.1

Enalapril

Mortality curves from the CONSENSUSI study

Clinical review

40 BMJ VOLUME 320 1 JANUARY 2000 www.bmj.com


2

epidemiological studies of heart failure have used differentlevels of ejection fraction to define systolic dysfunction. TheGlasgow study, for example, used an ejection fraction of 30% astheir criteria, whereas most other epidemiological surveys haveused levels of 4045%. Indeed, prevalence of heart failure seemssimilar in many different surveys, despite variation in the levelsof ejection fraction, and this observation is not entirelyexplained.

Prevalence of heart failureDuring the 1980s the Framingham study reported the ageadjusted overall prevalence of heart failure, with similar ratesfor men and women. Prevalence increased dramatically withincreasing age, with an approximate doubling in the prevalenceof heart failure with each decade of ageing.

In Nottinghamshire, the prevalence of heart failure in 1994was estimated from prescription data for loop diuretics andexamination of the general practice notes of a sample of thesepatients, to determine the number who fulfilled predeterminedcriteria for heart failure. The overall prevalence of heart failurewas estimated as 1.0% to 1.6%, rising from 0.1% in the 3039age range to 4.2% at 7079 years. This method, however, mayexclude individuals with mild heart failure and include patientstreated with diuretics who do not have heart failure.

Incidence of heart failureThe Framingham data show an age adjusted annual incidenceof heart failure of 0.14% in women and 0.23% in men. Survivalin the women is generally better than in the men, leading to thesame point prevalence. There is an approximate doubling in theincidence of heart failure with each decade of ageing, reaching3% in those aged 8594 years.

The recent Hillingdon study examined the incidence ofheart failure, defined on the basis of clinical and radiographicfindings, with echocardiography, in a population in westLondon. The overall annual incidence was 0.08%, rising from0.02% at age 4555 years to 1.2% at age 86 years or over. About80% of these cases were first diagnosed after acute hospitaladmission, with only 20% being identified in general practiceand referred to a dedicated clinic.

The Glasgow group of the MONICA study and theECHOES Group have found that coronary artery disease is themost powerful risk factor for impaired left ventricular function,either alone or in combination with hypertension. In thesestudies hypertension alone did not appear to contributesubstantially to impairment of left ventricular systoliccontraction, although the Framingham study did report a moresubstantial contribution from hypertension. This apparentdifference between the studies may reflect improvements in thetreatment of hypertension and the fact that some patients withhypertension, but without coronary artery disease, may developheart failure as a result of diastolic dysfunction.

Prevalence of left ventricular dysfunctionLarge surveys have been carried out in Britain in the 1990s, inGlasgow and the West Midlands, using echocardiography.

In Glasgow the prevalence of significantly impaired leftventricular contraction in subjects aged 2574 years was 2.9%;in the West Midlands, the prevalence was 1.8% in subjects aged45 and older.

The higher rates in the Scottish study may reflect the highprevalence of ischaemic heart disease, the main precursor ofimpaired left ventricular function in both studies. The numbersof symptomatic and asymptomatic cases, in both studies, wereabout the same.

Prevalence of heart failure (per 1000 population),Framingham heart study

Age (years) Men Women

5059 8 8

8089 66 79

All ages 7.4 7.7

Methods of assessing prevalence of heart failure inpublished studies

x Clinical and radiographic assessmentx Echocardiographyx General practice monitoringx Drug prescription data

Annual incidence of heart failure (per 1000 population),Framingham heart study

Age (years) Men Women

5059 3 2

8089 27 22

All ages 2.3 1.4

The MONICA study is an internationalstudy conducted under the auspices ofthe World Health Organisation tomonitor trends in and determinants ofmortality from cardiovascular disease

Prevalence (%) of left ventricular dysfunction, north Glasgow(MONICA survey)

Age group(years)

Asymptomatic Symptomatic

Men Women Men Women

4554 4.4 1.2 1.4 1.2

5564 3.2 0.0 2.5 2.0

6574 3.2 1.3 3.2 3.6

Clinical review



3

Ethnic differencesEthnic differences in the incidence of and mortality from heartfailure have also been reported. In the United States,AfricanAmerican men have been reported as having a 33%greater risk of being admitted to hospital for heart failure thanwhite men; the risk for black women was 50%.

A similar picture emerged in a survey of heart failureamong acute medical admissions to a city centre teachinghospital in Birmingham. The commonest underlyingaetiological factors were coronary heart disease in whitepatients, hypertension in black AfroCaribbean patients, andcoronary heart disease and diabetes in IndoAsians. Some ofthese racial differences may be related to the higher prevalenceof hypertension and diabetes in black people and coronaryartery disease and diabetes mellitus in IndoAsians.

Impact on health servicesHeart failure accounts for at least 5% of admissions to generalmedical and geriatric wards in British hospitals, and admissionrates for heart failure in various European countries (Sweden,Netherlands, and Scotland) and in the United States havedoubled in the past 1015 years. Furthermore, heart failureaccounts for over 1% of the total healthcare expenditure in theUnited Kingdom, and most of these costs are related to hospitaladmissions. The cost of heart failure is increasing, with anestimated UK expenditure in 1996 of £465m (£556m when thecosts of community health services and nursing homes areincluded).

Hospital readmissions and general practice consultationsoften occur soon after the diagnosis of heart failure. In elderlypatients with heart failure, readmission rates range from2947% within 3 to 6 months of the initial hospital discharge.Treating patients with heart failure with angiotensin convertingenzyme inhibitors can reduce the overall cost of treatment(because of reduced hospital admissions) despite increased drugexpenditure and improved long term survival.

The pictures of William Withering and of the foxglove are reproducedwith permission from the Fine Art Photographic Library. The box ofdefinitions of heart failure is adapted from PooleWilson PA et al, eds(Heart failure. New York: Churchill Livingstone, 1997:270). The table showing the prevalence of left ventricular dysfunction in north Glasgow isreproduced with permission from McDonagh TA et al (see key referencesbox). The table showing costs of heart failure is adapted from McMurray Jet al (Br J Med Econ 1993;6:99110).

The ABC of heart failure is edited by C R Gibbs, M K Davies, andG Y H Lip. CRG is research fellow and GYHL is consultantcardiologist and reader in medicine in the university department ofmedicine and the department of cardiology, City Hospital,Birmingham; MKD is consultant cardiologist in the department ofcardiology, Selly Oak Hospital, Birmingham. The series will bepublished as a book in the spring.

In the United States mortality from heartfailure at age <65 years has been reportedas being up to 2.5fold higher in blackpatients than in white patients

Cost of heart failure

Country Cost% Healthcare

costs% Of costs dueto admissions

UK, 19901 £360m 1.2 60

US, 1989 $9bn 1.5 71

France, 1990 FF11.4bn 1.9 64

New Zealand, 1990 $NZ73m 1.5 68

Sweden, 1996 Kr2.6m 2.0 75

Key references

x Clarke KW, Gray D, Hampton JR. Evidence of inadequateinvestigation and treatment of patients with heart failure. Br Heart J1994;71:5847.

x Cowie MR, Mosterd A, Wood DA, Deckers JW, PooleWilson PA,Sutton GC, et al. The epidemiology of heart failure. Eur Heart J1997;18:20825.

x Cowie MR, Wood DA, Coats AJS, Thompson SG, PooleWilson PA,Suresh V, et al. Incidence and aetiology of heart failure: apopulationbased study. Eur Heart J 1999;20:4218.

x Dries DL, Exner DV, Gersh BJ, Cooper HA, Carson PE, DomanskiMJ. Racial differences in the outcome of left ventricular dysfunction.N Engl J Med 1999;340:60916.

x Ho KK, Pinsky JL, Kannel WB, Levy D. The epidemiology of heartfailure: the Framingham study. J Am Coll Cardiol 1993;22:613A.

x Lip GYH, Zarifis J, Beevers DG. Acute admissions with heart failureto a district general hospital serving a multiracial population. Int JClin Pract 1997;51:2237.

x McDonagh TA, Morrison CE, Lawrence A, Ford I, TunstallPedoe H,McMurray JJV, et al. Symptomatic and asymptomatic leftventricularsystolic dysfunction in an urban population. Lancet 1997;350:82933.

x The Task Force on Heart Failure of the European Society ofCardiology. Guidelines for the diagnosis of heart failure. Eur Heart J1995;16:74151.

R C Davis is clinical research fellow and F D R Hobbs is professor inthe department of primary care and general practice, University ofBirmingham.

BMJ 2000;320:3942

One hundred years agoThe Bogey of Medical Etiquette.

There is a widespread opinion amongst the public that a rule ofconduct obtains in the medical profession the object of which isto protect the profession and individual members thereof againstthe consequences of their ignorance or mistakes. Probablyopinions differ as to the extent to which we are prepared to go inthis direction, and perhaps few believe that we would go so far asto commit perjury or sacrifice human life, but we certainly aresupposed to be capable of suppressing the truth in order to avoidexposing the mistakes of a colleague. We admit that there aremembers of the medical profession who regard their patients as

their property, and we believe that the petty tyranny sometimesexercised is responsible for the opinions upon medical etiquettewhich are undoubtedly entertained by the laity. But these extremeviews are not endorsed by any representative body in the medicalprofession, and we are quite certain that we are expressing thegeneral view when we say that the profession recognises no otherrules of medical etiquette than are consistent with the bestinterests of our patients and with courtesy and consideration forour colleagues. (BMJ 1900;i:156)

Clinical review



4

ABC of heart failureAetiologyG Y H Lip, C R Gibbs, D G Beevers

The relative importance of aetiological factors in heart failure isdependent on the nature of the population being studied, ascoronary artery disease and hypertension are common causesof heart failure in Western countries, whereas valvar heartdisease and nutritional cardiac disease are more common in thedeveloping world. Epidemiological studies are also dependenton the clinical criteria and relevant investigations used fordiagnosis, as it remains difficult, for example, to distinguishwhether hypertension is the primary cause of heart failure orwhether there is also underlying coronary artery disease.

Coronary artery disease and its riskfactorsCoronary heart disease is the commonest cause of heart failurein Western countries. In the studies of left ventriculardysfunction (SOLVD) coronary artery disease accounted foralmost 75% of the cases of chronic heart failure in male whitepatients, although in the Framingham heart study, coronaryheart disease accounted for only 46% of cases of heart failure inmen and 27% of chronic heart failure cases in women.Coronary artery disease and hypertension (either alone or incombination) were implicated as the cause in over 90% of casesof heart failure in the Framingham study.

Recent studies that have allocated aetiology on the basis ofnoninvasive investigations—such as the Hillingdon heart failurestudy—have identified coronary artery disease as the primaryaetiology in 36% of cases of heart failure. In the Hillingdonstudy, however, researchers were not able to identify theprimary aetiology in 34% of cases; this methodological failinghas been addressed in the current Bromley heart failure study,which uses coronary angiography as well as historical andnoninvasive findings.

Coronary risk factors, such as smoking and diabetesmellitus, are also risk markers of the development of heartfailure. Smoking is an independent and strong risk factor forthe development of heart failure in men, although the findingsin women are less consistent.

In the prevention arm of SOLVD diabetes was anindependent risk factor (about twofold) for mortality, the

Causes of heart failureCoronary artery diseasex Myocardial infarctionx Ischaemia

Hypertension

Cardiomyopathyx Dilated (congestive)x Hypertrophic/obstructivex Restrictive—for example, amyloidosis, sarcoidosis,

haemochromatosisx Obliterative

Valvar and congenital heart diseasex Mitral valve diseasex Aortic valve diseasex Atrial septal defect, ventricular septal defect

Arrhythmiasx Tachycardiax Bradycardia (complete heart block, the sick sinus

syndrome)x Loss of atrial transport—for example, atrial

fibrillation

Alcohol and drugsx Alcoholx Cardiac depressant drugs (â blockers, calcium

antagonists)

“High output” failurex Anaemia, thyrotoxicosis, arteriovenous fistulae,

Paget’s disease

Pericardial diseasex Constrictive pericarditisx Pericardial effusion

Primary right heart failurex Pulmonary hypertension—for example,

pulmonary embolism, cor pulmonalex Tricuspid incompetence

Relative risks for development of heart failure: 36 yearfollow up in Framingham heart study

Variable

Age (years)

Men Women

3564 6594 3564 6594

Serum cholesterol( > 6.3 mmol/l)

1.2 0.9 0.7 0.8

Hypertension( > 160/95 mm Hg orreceiving treatment)

4.0 1.9 3.0 1.9

Glucose intolerance 4.4 2.0 7.7 3.6

Electrocardiographic leftventricular hypertrophy 15.0 4.9 12.8 5.4

Epidemiological studies of aetiology of heart failure. Valuesare percentages

Aetiology

Teerlinket al

(31 studies198990)

Framinghamheart study*

HillingdonstudyMen Women

Ischaemic 50 59 48 36

Nonischaemic: 50 41 52 64

Hypertension 4 70 78 14

Idiopathic 18 0 0 0

Valvar 4 22 31 7

Other 10 7 7 10

“Unknown” 13 0 0 34

Because of rounding, totals may not equal 100%.*Total exceeds 100% as coronary artery disease and hypertension were notconsidered as mutually exclusive causes.

Clinical review



5

development of heart failure, and admission to hospital forheart failure, whereas in the Framingham study diabetes andleft ventricular hypertrophy were the most significant riskmarkers of the development of heart failure. Body weight and ahigh ratio of total cholesterol concentration to high densitylipoprotein cholesterol concentration are also independent riskfactors for heart failure. Clearly, these risk factors may increasethe risks of heart failure through their effects on coronaryartery disease, although diabetes alone may induce importantstructural and functional changes in the myocardium, whichfurther increase the risk of heart failure.

HypertensionHypertension has been associated with an increased risk ofheart failure in several epidemiological studies. In theFramingham heart study, hypertension was reported as thecause of heart failure—either alone or in association with otherfactors—in over 70% of cases, on the basis of noninvasiveassessment. Other community and hospital based studies,however, have reported hypertension to be a less commoncause of heart failure, and, indeed, the importance ofhypertension as a cause of heart failure has been declining inthe Framingham cohort since the 1950s. Recent communitybased studies that have assessed aetiology using clinical criteriaand relevant noninvasive investigations have reportedhypertension to be the cause of heart failure in 1020%.However, hypertension is probably a more common cause ofheart failure in selected patient groups, including females andblack populations (up to a third of cases).

Hypertension predisposes to the development of heartfailure via a number of pathological mechanisms, including leftventricular hypertrophy. Left ventricular hypertrophy isassociated with left ventricular systolic and diastolic dysfunctionand an increased risk of myocardial infarction, and itpredisposes to both atrial and ventricular arrhythmias.Electrocardiographic left ventricular hypertrophy is stronglycorrelated with the development of heart failure, as it isassociated with a 14fold increase in the risk of heart failure inthose aged 65 years or under.

CardiomyopathiesCardiomyopathies are defined as the diseases of heart musclethat are not secondary to coronary disease, hypertension, orcongenital, valvar, or pericardial disease. As primary diseases ofheart muscle, cardiomyopathies are less common causes ofheart failure, but awareness of their existence is necessary tomake a diagnosis. Cardiomyopathies are separated into fourfunctional categories: dilated (congestive), hypertrophic,restrictive, and obliterative. These groups can include rare,specific heart muscle diseases (such as haemochromatosis (ironoverload) and metabolic and endocrine disease), in whichcardiac involvement occurs as part of a systemic disorder.Dilated cardiomyopathy is a more common cause of heartfailure than hypertrophic and restrictive cardiomyopathies;obliterative cardiomyopathy is essentially limited to developingcountries.

Dilated cardiomyopathyDilated cardiomyopathy describes heart muscle disease inwhich the predominant abnormality is dilatation of the leftventricle, with or without right ventricular dilatation. Myocardialcells are also hypertrophied, with increased variation in size andincreased extracellular fibrosis. Family studies have reported

Effective blood pressure lowering inpatients with hypertension reduces therisk of heart failure; an overview of trialshas estimated that effectiveantihypertensive treatment reduces theage standardised incidence of heartfailure by up to 50%

Causes of dilated cardiomyopathyFamilial

Infectiousx Viral (coxsackie B, cytomegalovirus, HIV)x Rickettsiax Bacteria (diphtheria)x Mycobacteriax Fungusx Parasites (Chagas’ disease, toxoplasmosis)x Alcoholx Cardiotoxic drugs (adriamycin, doxorubicin, zidovudine)x Cocainex Metals (cobalt, mercury, lead)x Nutritional disease (beriberi, kwashiorkor, pellagra)x Endocrine disease (myxoedema, thyrotoxicosis, acromegaly,

phaeochromocytoma)

Pregnancy

Collagen diseasex Connective tissue diseases (systemic lupus erythematosus,

scleroderma, polyarteritis nodosa)

Neuromuscularx Duchenne muscular dystrophy, myotonic dystrophy

Idiopathic

Two dimensional echocardiogram (top) and M mode echocardiogram(bottom) showing left ventricular hypertrophy. A=interventricularseptum; B=posterior left ventricular wall

Clinical review



6

that up to a quarter of cases of dilated cardiomyopathy have afamilial basis. Viral myocarditis is a recognised cause; connectivetissue diseases such as systemic lupus erythematosus, theChurgStrauss syndrome, and polyarteritis nodosa are rarercauses. Idiopathic dilated cardiomyopathy is a diagnosis ofexclusion. Coronary angiography will exclude coronary disease,and an endomyocardial biopsy is required to excludeunderlying myocarditis or an infiltrative disease.

Dilatation can be associated with the development of atrialand ventricular arrhythmias, and dilatation of the ventriclesleads to “functional” mitral and tricuspid valve regurgitation.

Hypertrophic cardiomyopathyHypertrophic cardiomyopathy has a familial inheritance(autosomal dominant), although sporadic cases may occur. It ischaracterised by abnormalities of the myocardial fibres, and inits classic form involves asymmetrical septal hypertrophy, whichmay be associated with aortic outflow obstruction (hypertrophicobstructive cardiomyopathy).

Nevertheless, other forms of hypertrophic cardiomyopathyexist—apical hypertrophy (especially in Japan) and symmetricalleft ventricular hypertrophy (where the echocardiographicdistinction between this and hypertensive heart disease may beunclear). These abnormalities lead to poor left ventricularcompliance, with high end diastolic pressures, and there is acommon association with atrial and ventricular arrhythmias, thelatter leading to sudden cardiac death. Mitral regurgitation maycontribute to the heart failure in these patients.

Restrictive and obliterative cardiomyopathiesRestrictive cardiomyopathy is characterised by a stiff and poorlycompliant ventricle, which is not substantially enlarged, and thisis associated with abnormalities of diastolic function (relaxation)that limit ventricular filling. Amyloidosis and other infiltrativediseases, including sarcoidosis and haemochromatosis, cancause a restrictive syndrome. Endomyocardial fibrosis is also acause of restrictive cardiomyopathy, although it is a rare causeof heart failure in Western countries. Endocardial fibrosis of theinflow tract of one or both ventricles, including the subvalvarregions, results in restriction of diastolic filling and cavityobliteration.

Valvar diseaseRheumatic heart disease may have declined in certain parts ofthe world, but it still represents an important cause of heartfailure in India and other developing nations. In theFramingham study rheumatic heart disease accounted for heartfailure in 2% of men and 3% of women, although the overallincidence of valvar disease has been steadily decreasing in theFramingham cohort over the past 30 years.

Mitral regurgitation and aortic stenosis are the mostcommon causes of heart failure, secondary to valvar disease.Mitral regurgitation (and aortic regurgitation) leads to volumeoverload (increased preload), in contrast with aortic stenosis,which leads to pressure overload (increased afterload). Theprogression of heart failure in patients with valvar disease isdependent on the nature and extent of the valvar disease. Inaortic stenosis heart failure develops at a relatively late stageand, without valve replacement, it is associated with a poorprognosis. In contrast, patients with chronic mitral (or aortic)regurgitation generally decline in a slower and moreprogressive manner.

Two dimensional (long axis parasternal view) echocardiogram (top) andM mode echocardiogram (bottom) showing severely impaired left ventricularfunction in dilated cardiomyopathy

Two dimensional, apical, four chamber echocardiogramshowing dilated cardiomyopathy. A=left ventricle; B=leftatrium; C=right atrium; D=right ventricle

Colour Doppler echocardiograms showing mitral regurgitation (left) andaortic regurgitation (right)

Clinical review



7

ArrhythmiasCardiac arrhythmias are more common in patients with heartfailure and associated structural heart disease, includinghypertensive patients with left ventricular hypertrophy. Atrialfibrillation and heart failure often coexist, and this has beenconfirmed in large scale trials and smaller hospital basedstudies. In the Hillingdon heart failure study 30% of patientspresenting for the first time with heart failure had atrialfibrillation, and over 60% of patients admitted urgently withatrial fibrillation to a Glasgow hospital had echocardiographicevidence of impaired left ventricular function.

Atrial fibrillation in patients with heart failure has beenassociated with increased mortality in some studies, althoughthe vasodilator heart failure trial (VHeFT) failed to show anincrease in major morbidity or mortality for patients with atrialfibrillation. In the stroke prevention in atrial fibrillation (SPAF)study, the presence of concomitant heart failure or leftventricular dysfunction increased the risk of stroke andthromboembolism in patients with atrial fibrillation. Ventriculararrhythmias are also more common in heart failure, leading toa sudden deterioration in some patients; such arrhythmias are amajor cause of sudden death in patients with heart failure.

Alcohol and drugsAlcohol has a direct toxic effect on the heart, which may lead toacute heart failure or heart failure as a result of arrhythmias,commonly atrial fibrillation. Excessive chronic alcoholconsumption also leads to dilated cardiomyopathy (alcoholicheart muscle disease). Alcohol is the identifiable cause ofchronic heart failure in 23% of cases. Rarely, alcohol misusemay be associated with general nutritional deficiency andthiamine deficiency (beriberi).

Chemotherapeutic agents (for example, doxorubicin) andantiviral drugs (for example, zidovudine) have been implicatedin heart failure, through direct toxic effects on the myocardium.

Other causesInfections may precipitate heart failure as a result of the toxicmetabolic effects (relative hypoxia, acid base disturbance) incombination with peripheral vasodilation and tachycardia,leading to increased myocardial oxygen demand. Patients withchronic heart failure, like patients with most chronic illnesses,are particularly susceptible to viral and bacterial respiratoryinfections. “High output” heart failure is most often seen inpatients with severe anaemia, although thyrotoxicosis may alsobe a precipitating cause in these patients. Myxoedema maypresent with heart failure as a result of myocardial involvementor secondary to a pericardial effusion.

The table of epidemiological studies of the aetiology of heart failure isadapted and reproduced with permission from Cowie MR et al (Eur Heart J1997;18:20825). The table showing relative risks for development of heartfailure (36 year follow up) is adapted and reproduced with permission fromKannel WB et al (Br Heart J 1994;72:S39).

D G Beevers is professor of medicine in the university department ofmedicine and the department of cardiology, City Hospital,Birmingham.


BMJ 2000;320:1047

Arrhythmias and heart failure: mechanisms

Tachycardiasx Reduce diastolic ventricular filling timex Increase myocardial workload and myocardial oxygen demand,

precipitating ischaemiax If they are chronic, with poor rate control, they may lead to

ventricular dilatation and impaired ventricular function(“tachycardia induced cardiomyopathy”)

Bradycardiasx Compensatory increase in stroke volume is limited in the presence

of structural heart disease, and cardiac output is reduced

Abnormal atrial and ventricular contractionx Loss of atrial systole leads to the absence of active ventricular filling,

which in turn lowers cardiac output and raises atrial pressure—forexample, atrial fibrillation

x Dissociation of atrial and ventricular activity impairs diastolicventricular filling, particularly in the presence of a tachycardia—forexample, ventricular tachycardia

Prevalence (%) of atrial fibrillation in major heart failuretrials

Trial NYHA class*Prevalence of

atrial fibrillation

SOLVD I–III 6

VHeFT I II–III 15

VHeFT II II–III 15

CONSENSUS III–IV 50

CONSENSUS = cooperative north Scandinavian enalapril survival study.*Classification of the New York Heart Association.

Key references

x Cowie MR, Wood DA, Coats AJS, Thompson SG, PooleWilson PA,Suresh V, et al. Incidence and aetiology of heart failure: apopulationbased study. Eur Heart J 1999;20:4218.

x Eriksson H, Svardsudd K, Larsson B, Ohlson LO, Tibblin G, WelinL, et al. Risk factors for heart failure in the general population: thestudy of men born in 1913. Eur Heart J 1989;10:64756.

x Levy D, Larson MG, Vasan RS, Kannel WB, Ho KKL. Theprogression from hypertension to congestive heart failure. JAMA1996;275:155762.

x Oakley C. Aetiology, diagnosis, investigation, and management ofcardiomyopathies. BMJ 1997;315:15204.

x Teerlink JR, Goldhaber SZ, Pfeffer MA. An overview ofcontemporary etiologies of congestive heart failure. Am Heart J1991;121:18523.

x Wheeldon NM, MacDonald TM, Flucker CJ, McKendrick AD,McDevitt DG, Struthers AD. Echocardiography in chronic heartfailure in the community. Q J Med 1993;86:1723.

I

II

III

aVR

aVL

aVF

V1

V2

V3

V4

V5

V6

II

Electrocardiogram showing atrial fibrillation with a rapid ventricular response

Clinical review



8

ABC of heart failurePathophysiologyG Jackson, C R Gibbs, M K Davies, G Y H Lip

Heart failure is a multisystem disorder which is characterised byabnormalities of cardiac, skeletal muscle, and renal function;stimulation of the sympathetic nervous system; and a complexpattern of neurohormonal changes.

Myocardial systolic dysfunctionThe primary abnormality in nonvalvar heart failure is animpairment in left ventricular function, leading to a fall incardiac output. The fall in cardiac output leads to activation ofseveral neurohormonal compensatory mechanisms aimed atimproving the mechanical environment of the heart. Activationof the sympathetic system, for example, tries to maintain cardiacoutput with an increase in heart rate, increased myocardialcontractility, and peripheral vasoconstriction (increasedcatecholamines). Activation of the reninangiotensinaldosterone system (RAAS) also results in vasoconstriction(angiotensin) and an increase in blood volume, with retentionof salt and water (aldosterone). Concentrations of vasopressinand natriuretic peptides increase. Furthermore, there may beprogressive cardiac dilatation or alterations in cardiac structure(remodelling), or both.

Neurohormonal activationChronic heart failure is associated with neurohormonalactivation and alterations in autonomic control. Although thesecompensatory neurohormonal mechanisms provide valuablesupport for the heart in normal physiological circumstances,they also have a fundamental role in the development andsubsequent progression of chronic heart failure.

Reninangiotensinaldosterone systemStimulation of the reninangiotensinaldosterone system leadsto increased concentrations of renin, plasma angiotensin II, andaldosterone. Angiotensin II is a potent vasoconstrictor of therenal (efferent arterioles) and systemic circulation, where itstimulates release of noradrenaline from sympathetic nerveterminals, inhibits vagal tone, and promotes the release ofaldosterone. This leads to the retention of sodium and waterand the increased excretion of potassium. In addition,angiotensin II has important effects on cardiac myocytes andmay contribute to the endothelial dysfunction that is observedin chronic heart failure.

Sympathetic nervous systemThe sympathetic nervous system is activated in heart failure, vialow and high pressure baroreceptors, as an early compensatorymechanism which provides inotropic support and maintainscardiac output. Chronic sympathetic activation, however, hasdeleterious effects, causing a further deterioration in cardiacfunction.

The earliest increase in sympathetic activity is detected inthe heart, and this seems to precede the increase in sympatheticoutflow to skeletal muscle and the kidneys that is present inadvanced heart failure. Sustained sympathetic stimulationactivates the reninangiotensinaldosterone system and otherneurohormones, leading to increased venous and arterial tone

Developments in our understanding ofthe pathophysiology of heart failure havebeen essential for recent therapeuticadvances in this area

Poor ventricular function/myocardial damage

(eg post myocardial infarction, dilated cardiomyopathy)

Decreased stroke volume and cardiac output

• Vasoconstriction: increased sympathetic tone, angiotensin II, endothelins,

impaired nitric oxide release

• Sodium and fluid retention: increased vasopressin and aldosterone

Neurohormonal response

Further stress on ventricular wall and dilatation (remodelling)

leading to worsening of ventricular function

Activation of sympathetic system Renin angiotensin aldosterone system

Heart failure

Further heart failure

Neurohormonal mechanisms and compensatory mechanisms in heartfailure

Liver Vessels

Renin substrate (angiotensinogen)

Angiotensin I

Angiotensin II

Renin

(kidney)

Angiotensin converting enzyme

(lungs and vasculature)

Aldosterone releaseVasoconstriction Enhanced sympathetic activity

Salt and water retention

Brain

Reninangiotensinaldosterone axis in heart failure

Clinical review



9

(and greater preload and afterload respectively), increasedplasma noradrenaline concentrations, progressive retention ofsalt and water, and oedema. Excessive sympathetic activity isalso associated with cardiac myocyte apoptosis, hypertrophy,and focal myocardial necrosis.

In the long term, the ability of the myocardium to respondto chronic high concentrations of catecholamines is attenuatedby a down regulation in â receptors, although this may beassociated with baroreceptor dysfunction and a further increasein sympathetic activity. Indeed, abnormalities of baroreceptorfunction are well documented in chronic heart failure, alongwith reduced parasympathetic tone, leading to abnormalautonomic modulation of the sinus node. Moreover, a reductionin heart rate variability has consistently been observed inchronic heart failure, as a result of predominantly sympatheticand reduced vagal modulation of the sinus node, which may bea prognostic marker in patients with chronic heart failure.

Natriuretic peptidesThere are three natriuretic peptides, of similar structure, andthese exert a wide range of effects on the heart, kidneys, andcentral nervous system.

Atrial natriuretic peptide (ANP) is released from the atria inresponse to stretch, leading to natriuresis and vasodilatation. Inhumans, brain natriuretic peptide (BNP) is also released fromthe heart, predominantly from the ventricles, and its actions aresimilar to those of atrial natriuretic peptide. Ctype natriureticpeptide is limited to the vascular endothelium and centralnervous system and has only limited effects on natriuresis andvasodilatation.

The atrial and brain natriuretic peptides increase inresponse to volume expansion and pressure overload of theheart and act as physiological antagonists to the effects ofangiotensin II on vascular tone, aldosterone secretion, andrenaltubule sodium reabsorption. As the natriuretic peptidesare important mediators, with increased circulatingconcentrations in patients with heart failure, interest hasdeveloped in both the diagnostic and prognostic potential ofthese peptides. Substantial interest has been expressed aboutthe therapeutic potential of natriuretic peptides, particularlywith the development of agents that inhibit the enzyme thatmetabolises atrial natriuretic peptide (neutral endopeptidase),and nonpeptide agonists for the A and B receptors.

Antidiuretic hormone (vasopressin)Antidiuretic hormone concentrations are also increased insevere chronic heart failure. High concentrations of thehormone are particularly common in patients receiving diuretictreatment, and this may contribute to the development ofhyponatraemia.

EndothelinsEndothelin is secreted by vascular endothelial cells and is apotent vasoconstrictor peptide that has pronouncedvasoconstrictor effects on the renal vasculature, promoting theretention of sodium. Importantly, the plasma concentration ofendothelin1 is of prognostic significance and is increased inproportion to the symptomatic and haemodynamic severity ofheart failure. Endothelin concentration is also correlated withindices of severity such as the pulmonary artery capillary wedgepressure, need for admission to hospital, and death.

In view of the vasoconstrictor properties of endothelin,interest has developed in endothelin receptor antagonists ascardioprotective agents which inhibit endothelin mediatedvascular and myocardial remodelling.

Other hormonal mechanisms in chronic heart failure

x The arachidonic acid cascade leads to increased concentrations ofprostaglandins (prostaglandin E2 and prostaglandin I2), whichprotect the glomerular microcirculation during renalvasoconstriction and maintain glomerular filtration by dilatingafferent glomerular arterioles

x The kallikrein kinin system forms bradykinin, resulting in bothnatriuresis and vasodilatation, and stimulates the production ofprostaglandins

x Circulating concentrations of the cytokine tumour necrosis factor(áTNF) are increased in cachectic patients with chronic heartfailure. áTNF has also been implicated in the development ofendothelial abnormalities in patients with chronic heart failure

Myocardial damage

Activation of sympathetic nervous system

Renin-angiotensin

system

Vasoconstriction

Fluid retention Increased wall stress

Increased heart rate

and contractility

Increased myocardial oxygen demand

Direct

cardiotoxicity

Myocardial hypertrophy Decreased contractility Myocyte damage

Sympathetic activation in chronic heart failure

100

90

80

70

60

50

40

30

20

10

00 6 12 18 24 30 36 42 48 54 60

Cu

mu

lati

ve m

ort

alit

y (%

)

Months

Concentrations of plasma norepinephrine

>5.32 nmol/l

>3.55 nmol/l and <5.32 nmol/l

<3.55 nmol/l

2 year P<0.0001

Overall P<0.0001

Norepinephrine concentrations and prognosis in chronic heart failure

AtriumAtrium

Ventricles

Stretch or increase in cardiac

chamber volume leads to release

of natriuretic peptides

Brain natriuretic

peptide

Atrial

natriuretic

peptide

N-terminal

atrial

natriuretic

peptide

Vasodilatation

Increased urinary sodium excretion

Effects of natriuretic peptides

Clinical review



10

Patterns of neurohormonal activationand prognosis

Asymptomatic left ventricular dysfunctionPlasma norepinephrine concentrations increase early in thedevelopment of left ventricular dysfunction, and plasma reninactivity usually increases in patients receiving diuretic treatment.Norepinephrine concentration in asymptomatic left ventriculardysfunction is a strong and independent predictor of thedevelopment of symptomatic chronic heart failure and longterm mortality. Plasma concentrations of Nterminal proatrialnatriuretic peptide and brain natriuretic peptide also seem to begood indicators of asymptomatic left ventricular dysfunctionand may be useful in the future as an objective blood test inthese patients.

Congestive heart failureIn severe untreated chronic heart failure, concentrations ofrenin, angiotensin II, aldosterone, noradrenaline, and atrialnatriuretic peptide are all increased. Plasma concentrations ofvarious neuroendocrine markers correlate with both theseverity of heart failure and the long term prognosis. Forexample, raised plasma concentrations of Nterminal andCterminal atrial natriuretic peptide and of brain natriureticpeptide are independent predictors of mortality in patients withchronic heart failure. Patients with congestive heart failure andraised plasma noradrenaline concentrations also have a worseprognosis.

Other noncardiac abnormalities inchronic heart failure

VasculatureThe vascular endothelium has an important role in theregulation of vascular tone, releasing relaxing and contractingfactors under basal conditions or during exercise. The increasedperipheral resistance in patients with chronic heart failure isrelated to the alterations in autonomic control, includingheightened sympathetic tone, activation of thereninangiotensinaldosterone system, increased endothelinconcentrations, and impaired release of endothelium derivedrelaxing factor (or nitric oxide). There is emerging evidence thatimpaired endothelial function in chronic heart failure may beimproved with exercise training and drug treatment, such asangiotensin converting enzyme inhibitors.

Skeletal muscle changesConsiderable peripheral changes occur in the skeletal muscle ofpatients with chronic heart failure. These include a reduction inmuscle mass and abnormalities in muscle structure,metabolism, and function. There is also reduced blood flow toactive skeletal muscle, which is related to vasoconstriction andthe loss in muscle mass. All these abnormalities in skeletalmuscles, including respiratory muscles, contribute to thesymptoms of fatigue, lethargy, and exercise intolerance thatoccur in chronic heart failure.

Diastolic dysfunctionDiastolic dysfunction results from impaired myocardialrelaxation, with increased stiffness in the ventricular wall andreduced left ventricular compliance, leading to impairment ofdiastolic ventricular filling. Infiltrations, such as amyloid heartdisease, are the best examples, although coronary artery

After myocardial infarction

x Plasma noradrenaline is of prognostic value in patients early aftermyocardial infarction, predicting subsequent changes in leftventricular volume

x Natriuretic peptides have also been shown to predict outcome aftermyocardial infarction, although it is not clear whether the predictivevalue is additive to measurements of ventricular function

Activation of

neurohormonal

systems

Myocardial or

valvar damage

Ventricular

performance

Cardiac

output

Arrhythmias

Angiotensin converting

enzyme inhibitors

Left ventricular

afterload

Vascular

resistance plus

sodium retention

+

–

Effect of angiotensin converting enzyme inhibitors in heart failure

Contrast left ventriculogram in patient with poor systolic function (diastolic(left) and systolic (right) views)

Two dimensional echocardiogram in patient with hypertrophiccardiomyopathy showing asymmetrical septal hypertrophy

Clinical review



11

disease, hypertension (with left ventricular hypertrophy), andhypertrophic cardiomyopathy are more common causes.

The incidence and contribution of diastolic dysfunctionremains controversial, although it has been estimated that3040% of patients with heart failure have normal ventricularsystolic contraction. Indices of diastolic dysfunction can beobtained noninvasively with Doppler echocardiography orinvasively with cardiac catheterisation and measurement of leftventricular pressure changes. There is no agreement as to themost accurate index of left ventricular diastolic dysfunction, butthe Doppler mitral inflow velocity profile is probably the mostwidely used.

Although pure forms exist, in most patients with heartfailure both systolic and diastolic dysfunction can be present.Knowing about diastolic dysfunction, however, has little effecton management of most patients with chronic heart failure, asthere are still many uncertainties over its measurement andoptimal management strategies.

Myocardial remodelling, hibernation,and stunningAfter extensive myocardial infarction, cardiac contractility isfrequently impaired and neurohormonal activation leads toregional eccentric and concentric hypertrophy of thenoninfarcted segment, with expansion (regional thinning anddilatation) of the infarct zone. This is known as remodelling.Particular risk factors for this development of progressiveventricular dilatation after a myocardial infarction include alarge infarct, anterior infarctions, occlusion (or nonreperfusion)of the artery related to the infarct, and hypertension.

Myocardial dysfunction may also occur in response to“stunning” (postischaemic dysfunction), which describes delayedrecovery of myocardial function despite restoration of coronaryblood flow, in the absence of irreversible damage. This is incontrast to “hibernating” myocardium, which describespersistent myocardial dysfunction at rest, secondary to reducedmyocardial perfusion, although cardiac myocytes remain viableand myocardial contraction may improve with revascularisation.

When stunning or hibernation occurs, viable myocardiumretains responsiveness to inotropic stimulation, which can thenbe identified by resting and stress echocardiography, thalliumscintigraphy and positron emission tomography.Revascularisation may improve the overall left ventricularfunction with potential beneficial effects on symptoms andprognosis.

The graph showing mortality curves is adapted from Cohn et al (N Engl JMed 1984;311:81923); the diagram of the process of ventricularremodelling is adapted from McKay et al (Circulation 1986;74:693702).

G Jackson is consultant cardiologist in the department of cardiology,Guy’s and St Thomas’s Hospital, London.


BMJ 2000;320:16770

Contrast left ventriculogram in patient with hypertrophic cardiomyopathy(diastolic (left) and systolic (right) views)

Ris

k o

f h

eart

fai

lure

Age

Acute myocardial

infarction

Risk of heart failure and relation with age and history of myocardialinfarction

Myocardial infarction

Reduced systolic function

Increased left ventricular end

diastolic volume and pressure

Neurohormonal

activation

Increased cardiac

output

Increased wall stress

Heart failure

Death

Non-infarcted segment:

regional hypertrophy

Infarcted segment:

infarct expansion

Process of ventricular remodelling

Key references

x Grossman W. Diastolic dysfunction in congestive heart failure.N Engl J Med 1991;325:155764.

x Love MP, McMurray JJV. Endothelin in heart failure: a promisingtherapeutic target. Heart 1997;77:934.

x McDonagh TA, Robb SD, Murdoch DR, Morton JJ, Ford I, MorrisonCE, et al. Biochemical detection of left ventricular systolicdysfunction. Lancet 1998;351:913.

x Rahimtoola SH. The hibernating myocardium. Am Heart J1989;117:21121.

x Wilkins MR, Redondo J, Brown LA. The natriureticpeptide family.Lancet 1997;349:130710.

x Packer M. The neurohormonal hypothesis: a theory to explain themechanisms of disease progression in heart failure. J Am CollCardiol 1992;20:24854.

Clinical review



12

ABC of heart failureClinical features and complicationsR D S Watson, C R Gibbs, G Y H Lip

Clinical featuresPatients with heart failure present with a variety of symptoms,most of which are nonspecific. The common symptoms ofcongestive heart failure include fatigue, dyspnoea, swollenankles, and exercise intolerance, or symptoms that relate to theunderlying cause. The accuracy of diagnosis by presentingclinical features alone, however, is often inadequate, particularlyin women and elderly or obese patients.

Symptoms

DyspnoeaExertional breathlessness is a frequent presenting symptom inheart failure, although it is a common symptom in the generalpopulation, particularly in patients with pulmonary disease.Dyspnoea is therefore moderately sensitive, but poorly specific,for the presence of heart failure. Orthopnoea is a more specificsymptom, although it has a low sensitivity and therefore haslittle predictive value. Paroxysmal nocturnal dyspnoea resultsfrom increased left ventricular filling pressures (due tonocturnal fluid redistribution and enhanced renal reabsorption)and therefore has a greater sensitivity and predictive value.Nocturnal ischaemic chest pain may also be a manifestation ofheart failure, so left ventricular systolic dysfunction should beexcluded in patients with recurrent nocturnal angina.

Fatigue and lethargyFatigue and lethargy in chronic heart failure are, in part, relatedto abnormalities in skeletal muscle, with premature musclelactate release, impaired muscle blood flow, deficient endothelialfunction, and abnormalities in skeletal muscle structure andfunction. Reduced cerebral blood flow, when accompanied byabnormal sleep patterns, may occasionally lead to somnolenceand confusion in severe chronic heart failure.

OedemaSwelling of ankles and feet is another common presentingfeature, although there are numerous noncardiac causes of thissymptom. Right heart failure may manifest as oedema, righthypochondrial pain (liver distension), abdominal swelling(ascites), loss of appetite, and, rarely, malabsorption (boweloedema). An increase in weight may be associated with fluidretention, although cardiac cachexia and weight loss areimportant markers of disease severity in some patients.

Physical signsPhysical examination has serious limitations as many patients,particularly those with less severe heart failure, have fewabnormal signs. In addition, some physical signs are difficult tointerpret and, if present, may occasionally be related to causesother than heart failure.

Oedema and a tachycardia, for example, are too insensitiveto have any useful predictive value, and although pulmonarycrepitations may have a high diagnostic specificity they have alow sensitivity and predictive value. Indeed, the commonestcause of lower limb oedema in elderly people is immobility, andpulmonary crepitations may reflect poor ventilation withinfection, or pulmonary fibrosis, rather than heart failure.Jugular venous distension has a high specificity in diagnosing

Symptoms and signs in heart failure

SymptomsDyspnoeaOrthopnoeaParoxysmal nocturnal dyspnoeaReduced exercise tolerance, lethargy, fatigueNocturnal coughWheezeAnkle swellingAnorexia

SignsCachexia and muscle wastingTachycardiaPulsus alternansElevated jugular venous pressureDisplaced apex beatRight ventricular heaveCrepitations or wheezeThird heart soundOedemaHepatomegaly (tender)Ascites

Common causes of lower limb oedema

x Gravitational disorder—for example, immobilityx Congestive heart failurex Venous thrombosis or obstruction, varicose veinsx Hypoproteinaemia—for example, nephrotic syndrome, liver diseasex Lymphatic obstruction

Sensitivity, specificity, and predictive value of symptoms,signs, and chest x ray findings for presence of heart failure(ejection fraction <40%) in 1306 patients with coronaryartery disease undergoing cardiac catheterisation

Clinical featuresSensitivity

(%)Specificity

(%)

Positivepredictivevalue (%)

History:

Shortness of breath 66 52 23

Orthopnoea 21 81 2

Paroxysmal nocturnaldyspnoea 33 76 26

History of oedema 23 80 22

Examination:

Tachycardia ( > 100beats/min) 7 99 6

Crepitations 13 91 27

Oedema (onexamination) 10 93 3

Gallop (S3) 31 95 61

Neck vein distension 10 97 2

Chest x ray examination:

Cardiomegaly 62 67 32

Clinical review



13

heart failure in patients who are known to have cardiac disease,although some patients, even with documented heart failure, donot have an elevated venous pressure. The presence of adisplaced apex beat in a patient with a history of myocardialinfarction has a high positive predictive value. A third heartsound has a relatively high specificity, although its universalvalue is limited by a high interobserver variability, withinterobserver agreement of less than 50% in nonspecialists.

In patients with preexisting chronic heart failure, otherclinical features may be evident that point towards precipitatingcauses of acute heart failure or deteriorating heart failure.Common factors that may be obvious on clinical assessmentand are associated with relapses in congestive heart failureinclude infections, arrhythmias, continued or recurrentmyocardial ischaemia, and anaemia.

Clinical diagnosis and clinical scoring systemsSeveral epidemiological studies, including the Framinghamheart study, have used clinical scoring systems to define heartfailure, although the use of these systems is not recommendedfor routine clinical practice.

In a patient with appropriate symptoms and a number ofphysical signs, including a displaced apex beat, elevated venouspressure, oedema, and a third heart sound, the clinical diagnosisof heart failure may be made with some confidence. However,the clinical suspicion of heart failure should also be confirmedwith objective investigations and the demonstration of cardiacdysfunction at rest. It is important to note that, in some patients,exerciseinduced myocardial ischaemia may lead to a rise inventricular filling pressures and a fall in cardiac output, leadingto symptoms of heart failure during exertion.

ClassificationSymptoms and exercise capacity are used to classify the severityof heart failure and monitor the response to treatment. Theclassification of the New York Heart Association (NYHA) isused widely, although outcome in heart failure is bestdetermined not only by symptoms (NYHA class) but also byechocardiographic criteria. As the disease is progressive, theimportance of early treatment, in an attempt to preventprogression to more severe disease, cannot be overemphasised.

Complications

Arrhythmias

Atrial fibrillationAtrial fibrillation is present in about a third (range 1050%) ofpatients with chronic heart failure and may represent either acause or a consequence of heart failure. The onset of atrialfibrillation with a rapid ventricular response may precipitateovert heart failure, particularly in patients with preexistingventricular dysfunction. Predisposing causes should beconsidered, including mitral valve disease, thyrotoxicosis, andsinus node disease. Importantly, sinus node disease may beassociated with bradycardias, which might be exacerbated byantiarrhythmic treatment.

Atrial fibrillation that occurs with severe left ventriculardysfunction following myocardial infarction is associated with apoor prognosis. In addition, patients with heart failure andatrial fibrillation are at particularly high risk of stroke and otherthromboembolic complications.

Ventricular arrhythmiasMalignant ventricular arrhythmias are common in end stageheart failure. For example, sustained monomorphic ventricular

Precipitating causes of heart failure

x Arrhythmias, especially atrial fibrillationx Infections (especially pneumonia)x Acute myocardial infarctionx Angina pectoris or recurrent myocardial ischaemiax Anaemiax Alcohol excessx Iatrogenic cause—for example, postoperative fluid replacement or

administration of steroids or nonsteroidal antiinflammatory drugsx Poor drug compliance, especially in antihypertensive treatmentx Thyroid disorders—for example, thyrotoxicosisx Pulmonary embolismx Pregnancy

European Society of Cardiology’s guidelines for diagnosis ofheart failure

Essential featuresSymptoms of heart failure (for example, breathlessness, fatigue, ankleswelling)andObjective evidence of cardiac dysfunction (at rest)

Nonessential featuresResponse to treatment directed towards heart failure (in cases wherethe diagnosis is in doubt)

NYHA classification of heart failure

Class I: asymptomaticNo limitation in physical activity despite presence of heart disease.This can be suspected only if there is a history of heart disease whichis confirmed by investigations—for example, echocardiography

Class II: mildSlight limitation in physical activity. More strenuous activity causesshortness of breath—for example, walking on steep inclines andseveral flights of steps. Patients in this group can continue to have analmost normal lifestyle and employment

Class III: moderateMore marked limitation of activity which interferes with work. Walkingon the flat produces symptoms

Class IV: severeUnable to carry out any physical activity without symptoms. Patientsare breathless at rest and mostly housebound

Gross oedema of ankles,including bullae with serousexudate

Clinical review



14

tachycardia occurs in up to 10% of patients with advanced heartfailure who are referred for cardiac transplantation. In patientswith ischaemic heart disease these arrhythmias often havereentrant mechanisms in scarred myocardial tissue. An episodeof sustained ventricular tachycardia indicates a high risk forrecurrent ventricular arrhythmias and sudden cardiac death.

Sustained polymorphic ventricular tachycardia and torsadesde pointes are more likely to occur in the presence ofprecipitating or aggravating factors, including electrolytedisturbance (for example, hypokalaemia or hyperkalaemia,hypomagnesaemia), prolonged QT interval, digoxin toxicity,drugs causing electrical instability (for example, antiarrhythmicdrugs, antidepressants), and continued or recurrent myocardialischaemia. â Blockers are useful for treating arrhythmias, andthese agents (for example, bisoprolol, metoprolol, carvedilol)are likely to be increasingly used as a treatment option inpatients with heart failure.

Stroke and thromboembolismCongestive heart failure predisposes to stroke andthromboembolism, with an overall estimated annual incidenceof approximately 2%. Factors contributing to the increasedthromboembolic risk in patients with heart failure include lowcardiac output (with relative stasis of blood in dilated cardiacchambers), regional wall motion abnormalities (includingformation of a left ventricular aneurysm), and associated atrialfibrillation. Although the prevalence of atrial fibrillation in someof the earlier observational studies was between 12% and36%—which may have accounted for some of thethromboembolic events—patients with chronic heart failurewho remain in sinus rhythm are also at an increased risk ofstroke and venous thromboembolism. Patients with heart failureand chronic venous insufficiency may also be immobile, and thiscontributes to their increased risk of thrombosis, including deepvenous thrombosis and pulmonary embolism.

Recent observational data from the studies of left ventriculardysfunction (SOLVD) and vasodilator heart failure trials(VHeFT) indicate that mild to moderate heart failure isassociated with an annual risk of stroke of about 1.5%(compared with a risk of less than 0.5% in those without heartfailure), rising to 4% in patients with severe heart failure. Inaddition, the survival and ventricular enlargement (SAVE) studyrecently reported an inverse relation between risk of stroke andleft ventricular ejection fraction, with an 18% increase in risk forevery 5% reduction in left ventricular ejection fraction; thisclearly relates thromboembolism to severe cardiac impairmentand the severity of heart failure. As thromboembolic risk seemsto be related to left atrial and left ventricular dilatation,echocardiography may have some role in the risk stratificationof thromboembolism in patients with chronic heart failure.

PrognosisMost long term (more than 10 years of follow up) longitudinalstudies of heart failure, including the Framingham heart study(1971), were performed before the widespread use ofangiotensin converting enzyme inhibitors. In the Framinghamstudy the overall survival at eight years for all NYHA classes was30%, compared with a one year mortality in classes III and IV of34% and a one year mortality in class IV of over 60%. Theprognosis in patients whose left ventricular dysfunction isasymptomatic is better than that in those whose left ventriculardysfunction is symptomatic. The prognosis in patients withcongestive heart failure is dependent on severity, age, and sex,with a poorer prognosis in male patients. In addition, numerousprognostic indices are associated with an adverse prognosis,

Predisposing factors for ventricular arrhythmias

x Recurrent or continued coronary ischaemiax Recurrent myocardial infarctionx Hypokalaemia and hyperkalaemiax Hypomagnesaemiax Psychotropic drugs—for example, tricyclic antidepressantsx Digoxin (leading to toxicity)x Antiarrhythmic drugs that may be cardiodepressant (negative

inotropism) and proarrhythmic

Complications of heart failure

Arrhythmias—Atrial fibrillation; ventricular arrhythmias (ventriculartachycardia, ventricular fibrillation); bradyarrhythmias

Thromboembolism—Stroke; peripheral embolism; deep venousthrombosis; pulmonary embolism

Gastrointestinal—Hepatic congestion and hepatic dysfunction;malabsorption

Musculoskeletal—Muscle wastingRespiratory—Pulmonary congestion; respiratory muscle weakness;

pulmonary hypertension (rare)

Morbidity and mortality for all grades ofsymptomatic chronic heart failure arehigh, with a 2030% one year mortality inmild to moderate heart failure and agreater than 50% one year mortality insevere heart failure. These prognosticdata refer to patients with systolic heartfailure, as the natural course of diastolicdysfunction is less well defined

24 Hour Holter tracing showing frequent ventricular extrasystoles

Clinical review



15

including NYHA class, left ventricular ejection fraction, andneurohormonal status.

Survival can be prolonged in chronic heart failure thatresults from systolic dysfunction if angiotensin convertingenzyme inhibitors are given. Longitudinal data from theFramingham study and the Mayo Clinic suggest, however, thatthere is still only a limited improvement in the one year survivalrate of patients with newly diagnosed symptomatic chronicheart failure, which remains at 6070%. In these studies only aminority of patients with congestive heart failure wereappropriately treated, with less than 25% of them receivingangiotensin converting enzyme inhibitors, and even amongtreated patients the dose used was much lower than doses usedin the clinical trials.

Treatment with angiotensin converting enzyme inhibitorsprevents or delays the onset of symptomatic heart failure inpatients with asymptomatic, or minimally symptomatic, leftventricular systolic dysfunction. The increase in mortality withthe development of symptoms suggests that the optimal timefor intervention with these agents is well before the onset ofsubstantial left ventricular dysfunction, even in the absence ofovert clinical symptoms of heart failure. This benefit has beenconfirmed in several large, well conducted, postmyocardialinfarction studies.

Sudden deathThe mode of death in heart failure has been extensivelyinvestigated, and progressive heart failure and sudden deathseem to occur with equal frequency. Some outstandingquestions still remain, however. Although arrhythmias arecommon in patients with heart failure and are indicators ofdisease severity, they are not powerful independent predictorsof prognosis. Sudden death may be related to ventriculararrhythmias, although asystole is a common terminal event insevere heart failure. It has not been firmly established whetherthese arrhythmias are primary arrhythmias or whether someare secondary to acute coronary ischaemia or indicate in situcoronary thrombosis. The cause of death is often uncertain,especially as the patient may die of a cardiac arrest outsidehospital or while asleep.

R D S Watson is consultant cardiologist in the university departmentof medicine and the department of cardiology, City Hospital,Birmingham.

BMJ 2000;320:2369

Some predictors of poor outcome in chronic heart failure

x High NYHA functional classx Reduced left ventricular ejection fractionx Low peak oxygen consumption with maximal exercise (% predicted

value)x Third heart soundx Increased pulmonary artery capillary wedge pressurex Reduced cardiac indexx Diabetes mellitusx Reduced sodium concentrationx Raised plasma catecholamine and natriuretic peptide

concentrations

Key references

x Doval HC, Nul DR, Grancelli HO, Perrone SV, Bortman GR,Curiel R, et al. Randomised trial of lowdose amiodarone in severecongestive heart failure. Lancet 1994;334:4938.

x Gradman A, Deedwania P, Cody R, Massie B, Packer M, Pitt B, et al.Predictors of total mortality and sudden death in mild to moderateheart failure. J Am Coll Cardiol 1989;14:56470.

x Guidelines for the diagnosis of heart failure. The Task Force onHeart Failure of the European Society of Cardiology. Eur Heart J1995;16:74151.

x Rodeheffer RJ, Jacobsen SJ, Gersh BJ, Kottke TE, McCann HA,Bailey KR, et al. The incidence and prevalence of congestive heartfailure in Rochester, Minnesota. Mayo Clin Proc 1993;68:114350.

x The SOLVD Investigators. Effect of enalapril on mortality and thedevelopment of heart failure in asymptomatic patients with reducedleft ventricular ejection fractions. N Engl J Med 1992;327:68591.

x The CONSENSUS Trial Study Group. Effects of enalapril onmortality in severe congestive heart failure: results of thecooperative north Scandinavian enalapril survival study(CONSENSUS). N Engl J Med 1987;316:142935.

The table on the sensitivity, specificity, and predictive value of symptoms,signs, and chest x ray findings is adapted with permission from Harlan et al(Ann Intern Med 1977;86:1338).


Cardiac mortality in placebo controlled heart failure trials

TrialPatients’

characteristics

Ischaemicheart

disease (%) Treatment

Cardiovascular mortality

Follow up(years)

Treatment(%)

Placebo(%)

CONSENSUS NYHA IV (cardiomegaly) 73 Enalapril 38 54 1SOLVDP Asymptomatic (EF < 35%) 83 Enalapril 13 14 4SOLVDT Symptomatic (EF < 35%) 71 Enalapril 31 36 4SAVE Postmyocardial infarction (EF < 40%) 100 Captopril 17 21 4VHeFT I NYHA IIIII (EF < 45%) 44 HISDN 37 41 5VHeFT II NYHA IIIII (EF < 45%) 52 Enalapril 28 34* 5PRAISE NYHA IIIIV (EF < 30%) 63 Amlodipine 28 33 1.2

EF ejection fraction. SOLVDP, SOLVDT = studies of left ventricular dysfunction prevention arm (P) and treatment arm (T).HISDN = hydralazine and isosorbide dinitrate.*Treatment with HISDN.

Clinical review



16

ABC of heart failureInvestigationM K Davies, C R Gibbs, G Y H Lip

Clinical assessment is mandatory before detailed investigationsare conducted in patients with suspected heart failure, althoughspecific clinical features are often absent and the condition canbe diagnosed accurately only in conjunction with moreobjective investigation, particularly echocardiography. Althoughopen access echocardiography is now increasingly available,appropriate prereferral investigations include chestradiography, 12 lead electrocardiography, and renal chemistry.

Chest x ray examinationThe chest x ray examination has an important role in theroutine investigation of patients with suspected heart failure,and it may also be useful in monitoring the response totreatment. Cardiac enlargement (cardiothoracic ratio > 50%)may be present, but there is a poor correlation between thecardiothoracic ratio and left ventricular function. The presenceof cardiomegaly is dependent on both the severity ofhaemodynamic disturbance and its duration: cardiomegaly isfrequently absent, for example, in acute left ventricular failuresecondary to acute myocardial infarction, acute valvarregurgitation, or an acquired ventricular septal defect. Anincreased cardiothoracic ratio may be related to left or rightventricular dilatation, left ventricular hypertrophy, andoccasionally a pericardial effusion, particularly if the cardiacsilhouette has a globular appearance. Echocardiography isrequired to distinguish reliably between these different causes,although in decompensated heart failure other radiographicfeatures may be present, such as pulmonary congestion orpulmonary oedema.

In left sided failure, pulmonary venous congestion occurs,initially in the upper zones (referred to as upper lobe diversionor congestion). When the pulmonary venous pressure increasesfurther, usually above 20 mm Hg, fluid may be present in thehorizontal fissure and Kerley B lines in the costophrenic angles.In the presence of pulmonary venous pressures above 25 mmHg, frank pulmonary oedema occurs, with a “bats wing”appearance in the lungs, although this is also dependent on therate at which the pulmonary oedema has developed. Inaddition, pleural effusions occur, normally bilaterally, but if theyare unilateral the right side is more commonly affected.Nevertheless, it is not possible to distinguish, when viewed inisolation, whether pulmonary congestion is related to cardiac ornoncardiac causes (for example, renal disease, drugs, therespiratory distress syndrome).

Rarely, chest radiography may also show valvar calcification,a left ventricular aneurysm, and the typical pericardialcalcification of constrictive pericarditis. Chest radiography mayalso provide valuable information about noncardiac causes ofdyspnoea.

12 lead electrocardiographyThe 12 lead electrocardiographic tracing is abnormal in mostpatients with heart failure, although it can be normal in up to10% of cases. Common abnormalities include Q waves,abnormalities in the T wave and ST segment, left ventricularhypertrophy, bundle branch block, and atrial fibrillation. It is a

Investigations if heart failure is suspected

Initial investigationsx Chest radiographyx Electrocardiographyx Echocardiography, including Doppler studiesx Haematology testsx Serum biochemistry, including renal function and glucose

concentrations, liver function tests, and thyroid function testsx Cardiac enzymes (if recent infarction is suspected)

Other investigationsx Radionuclide imagingx Cardiopulmonary exercise testingx Cardiac catheterisationx Myocardial biopsy—for example, in suspected myocarditis

Chest radiographs showing gross cardiomegaly in patientwith dilated cardiomyopathy (top); cardiomegaly andpulmonary congestion with fluid in horizontal fissure(bottom)

Clinical review



17

useful screening test as a normal electrocardiographic tracingmakes it unlikely that the patient has heart failure secondary toleft ventricular systolic dysfunction, since this test has highsensitivity and a negative predictive value. The combination of anormal chest x ray finding and a normal electrocardiographictracing makes a cardiac cause of dyspnoea very unlikely.

In patients with symptoms (palpitations or dizziness), 24 hourelectrocardiographic (Holter) monitoring or a Cardiomemodevice will detect paroxysmal arrhythmias or other abnormalities,such as ventricular extrasystoles, sustained or nonsustainedventricular tachycardia, and abnormal atrial rhythms(extrasystoles, supraventricular tachycardia, and paroxysmal atrialfibrillation). Many patients with heart failure, however, showcomplex ventricular extrasystoles on 24 hour monitoring.

EchocardiographyEchocardiography is the single most useful noninvasive test inthe assessment of left ventricular function; ideally it should beconducted in all patients with suspected heart failure. Althoughclinical assessment, when combined with a chest x rayexamination and electrocardiography, allows a preliminarydiagnosis of heart failure, echocardiography provides anobjective assessment of cardiac structure and function. Leftventricular dilatation and impairment of contraction isobserved in patients with systolic dysfunction related toischaemic heart disease (where a regional wall motionabnormality may be detected) or in dilated cardiomyopathy(with global impairment of systolic contraction).

A quantitative measurement can be obtained fromcalculation of the left ventricular ejection fraction. This is thestroke volume (the difference between the end diastolic and endsystolic volumes) expressed as a percentage of the leftventricular end diastolic volume. Measurements, and theassessment of left ventricular function, are less reliable in thepresence of atrial fibrillation. The left ventricular ejectionfraction has been correlated with outcome and survival inpatients with heart failure, although the assessment may beunreliable in patients with regional abnormalities in wallmotion. Regional abnormalities can also be quantified into awall motion index, although in practice the assessment ofsystolic function is often based on visual assessment and theobserver’s experience of normal and abnormal contractilefunction. These abnormalities are described as hypokinetic(reduced systolic contraction), akinetic (no systolic contraction)and dyskinetic (abnormalities of direction or timing ofcontraction, or both), and refer to universally recognisedsegments of the left ventricle. Echocardiography may also showother abnormalities, including valvar disease, left ventricularaneurysm, intracardiac thrombus, and pericardial disease.

Mitral incompetence is commonly identified onechocardiography in patients with heart failure, as a result ofventricular and annular dilatation (“functional” mitralincompetence), and this must be distinguished from mitralincompetence related to primary valve disease. Twodimensional echocardiography allows the assessment of valvestructure and identifies thickening of cusps, leaflet prolapse,cusp fusion, and calcification. Doppler echocardiography allowsthe quantitative assessment of flow across valves and theidentification of valve stenosis, in addition to the assessment ofright ventricular systolic pressures and allowing the indirectdiagnosis of pulmonary hypertension. Doppler studies havebeen used in the assessment of diastolic function, althoughthere is no single reliable echocardiographic measure ofdiastolic dysfunction. Colour flow Doppler techniques are

Value of electrocardiography* in identifying heart failureresulting from left ventricular systolic dysfunction

Sensitivity 94%

Specificity 61%

Positive predictive value 35%

Negative predictive value 98%

*Electrocardiographic abnormalities are defined as atrial fibrillation, evidence ofprevious myocardial infarction, left ventricular hypertrophy, bundle branchblock, and left axis deviation.

Who should have an echocardiogram?

x Almost all patients with symptoms or signs of heart failurex Symptoms of breathlessness in association with signs of a murmurx Dyspnoea associated with atrial fibrillationx Patients at “high risk” for left ventricular dysfunction—for example,

those with anterior myocardial infarction, poorly controlledhypertension, or arrhythmias

Echocardiography as a guide to management

x Identification of impaired systolic function for decision ontreatment with angiotensin converting enzyme inhibitors

x Identification of diastolic dysfunction or predominantly rightventricular dysfunction

x Identification and assessment of valvar diseasex Assessment of embolic risk (severe left ventricular impairment with

mural thrombus)

aVRI

II

V1 V4

aVL V2 V5

III

II

aVF V3 V6

aVRI

II

V1 V4

aVL V2 V5

III

II

aVF V3 V6

Electrocardiograms showing previous anterior myocardial infarction with Qwaves in anteroseptal leads (top) and left bundle branch block (bottom)

Clinical review



18

particularly sensitive in detecting the direction of blood flowand the presence of valve incompetence.

Advances in echocardiography include the use of contrastagents for visualisation of the walls of the left ventricle in moredetail, especially as in about 10% of patients satisfactory imagescannot be obtained with standard transthoracicechocardiography. Transoesophageal echocardiography allowsthe detailed assessment of the atria, valves, pulmonary veins,and any cardiac masses, including thrombi.

The logistic and health economic aspects of large scalescreening with echocardiography have been debated, but thedevelopment of open access echocardiography heart failureservices for general practitioners and the availability of provedtreatments for heart failure that improve prognosis, such asangiotensin converting enzyme inhibitors, highlight theimportance of an agreed strategy for the echocardiographicassessment of these patients.

Haematology and biochemistryRoutine haematology and biochemistry investigations arerecommended to exclude anaemia as a cause of breathlessnessand high output heart failure and to exclude importantpreexisting metabolic abnormalities. In mild and moderateheart failure, renal function and electrolytes are usually normal.In severe (New York Heart Association, class IV) heart failure,however, as a result of reduced renal perfusion, high dosediuretics, sodium restriction, and activation of theneurohormonal mechanisms (including vasopressin), there is aninability to excrete water, and dilutional hyponatraemia may bepresent. Hyponatraemia is, therefore, a marker of the severity ofchronic heart failure.

A baseline assessment of renal function is important beforestarting treatment, as the renal blood flow and the glomerularfiltration rate fall in severe congestive heart failure. Baselineserum creatinine concentrations are important: increasingcreatinine concentrations may occur after the start of treatment,particularly in patients who are receiving angiotensinconverting enzyme inhibitors and high doses of diuretics and inpatients with renal artery stenosis. Proteinuria is a commonfinding in severe congestive heart failure.

Hypokalaemia occurs when high dose diuretics are usedwithout potassium supplementation or potassium sparingagents. Hyperkalaemia can also occur in severe congestive heartfailure with a low glomerular filtration rate, particularly with theconcurrent use of angiotensin converting enzyme inhibitorsand potassium sparing diuretics. Both hypokalaemia andhyperkalaemia increase the risk of cardiac arrhythmias;hypomagnesaemia, which is associated with long term diuretictreatment, increases the risk of ventricular arrhythmias. Liverfunction tests (serum bilirubin, aspartate aminotransferase, andlactate dehydrogenase) are often abnormal in advancedcongestive heart failure, as a result of hepatic congestion.Thyroid function tests are also recommended in all patients, inview of the association between thyroid disease and the heart.

Radionuclide methodsRadionuclide imaging—or multigated ventriculography—allowsthe assessment of the global left and right ventricular function.Images may be obtained in patients where echocardiography isnot possible. The most common method labels red cells withtechnetium99m and acquires 16 or 32 frames per heart beat bysynchronising (“gating”) imaging with electrocardiography. Thisallows the assessment of ejection fraction, systolic filling rate,diastolic emptying rate, and wall motion abnormalities. These

Natriuretic peptides

x Biochemical markers are being sought for the diagnosis ofcongestive heart failure

x Brain natriuretic peptide concentrations correlate with the severityof heart failure and prognosis

x These could, in the future, be used to distinguish between patientsin whom heart failure is extremely unlikely and those in whom theprobability of heart failure is high

x At present, however, the evidence that blood natriuretic peptideconcentrations are valuable in identifying important left ventricularsystolic dysfunction is conflicting, and their use in routine practice isstill limited

x Further studies are necessary to determine the most convenientand cost effective methods of identifying patients with heart failureand asymptomatic left ventricular dysfunction

Transthoracic echocardiograms: two dimensional apical view (top) andDoppler studies (bottom) showing severe calcific stenosis, with an estimatedaortic gradient of over 70 mm Hg (A=left ventricle, B=aortic valve, andC=left atrium)

Multigated ventriculography scan in patient withhistory of extensive myocardial infarction andcoronary bypass grafting (left ventricular ejectionfraction of 30%)

Clinical review



19

variables can be assessed, if necessary, during rest and exercise;this method is ideal for the serial reassessment of ejectionfraction, but these methods do expose the patient to radiation.

Radionuclide studies are also valuable for assessingmyocardial perfusion and the presence or extent of coronaryischaemia, including myocardial stunning and hibernatingmyocardium.

Angiography, cardiac catheterisation,and myocardial biopsyAngiography should be considered in patients with recurrentischaemic chest pain associated with heart failure and in thosewith evidence of severe reversible ischaemia or hibernatingmyocardium. Cardiac catheterisation with myocardial biopsycan be valuable in more difficult cases where there is diagnosticdoubt—for example, in restrictive and infiltratingcardiomyopathies (amyloid heart disease, sarcoidosis),myocarditis, and pericardial disease. Left ventricularangiography can show global or segmental impairment offunction and assess end diastolic pressures, and right heartcatheterisation allows an assessment of the right sided pressures(right atrium, right ventricle, and pulmonary arteries) andpulmonary artery capillary wedge pressure, in addition tooxygen saturations.

Pulmonary function testsObjective measurement of lung function is useful in excludingrespiratory causes of breathlessness, although respiratory andcardiac disease commonly coexist. Peak expiratory flow rate andforced expiratory volume in one second are reduced in heartfailure, although not as much as in severe chronic obstructivepulmonary disease. In patients with severe breathlessness andwheeze, a peak expiratory flow rate of < 200 l/min suggestsreversible airways disease, not acute left ventricular failure.

Stress studies use graded physicalexercise or pharmacological stress withagents such as adenosine, dipyridamole,and dobutamine. Stress echocardiographyis emerging as a useful technique forassessing myocardial reversibility inpatients with coronary artery disease

Coronary angiography is essential foraccurate assessment of the coronaryarteries

Cardiopulmonary exercise testing

x Exercise tolerance is reduced in patients with heart failure,regardless of method of assessment

x Assessment methods include a treadmill test, cycle ergometry, a 6minute walking test, or pedometry measurements

x Exercise testing is not routinely performed for all patients withcongestive heart failure, but it may be valuable in identifyingsubstantial residual ischaemia, thus leading to more detailedinvestigation

x Respiratory physiological measurements may be made duringexercise, and most cardiac transplant centres use data obtained atcardiopulmonary exercise testing to aid the selection of patients fortransplantation

x The maximum oxygen consumption is the value at whichconsumption remains stable despite increasing exercise, and itrepresents the upper limit of aerobic exercise tolerance

x The maximum oxygen consumption and the carbon dioxideproduction correlate well with the severity of heart failure

x The maximum oxygen consumption has also been independentlyrelated to long term prognosis, especially in patients with severe leftventricular dysfunction

Assessments for the investigation and diagnosis of heart failure

Assessments

Diagnosis of heart failure Suggestsalternative or

additional diseaseNecessary Supports Opposes

Symptoms of heartfailure

+ + + +(if absent)

Signs of heart failure + + +(if absent)

Response to treatment + + + +(if absent)

Electrocardiography + +(if normal)

Chest radiography(cardiomegaly orcongestion)

+ + +(if normal)

Pulmonary

Echocardiography(cardiac dysfunction)

++ + +(if absent)

Haematology Anaemia

Biochemistry (renal,liver function, andthyroid function tests)

Renal, liver,thyroid

Urine analysis Renal

Pulmonary functiontests

Pulmonary

+ + = Great importance; + = some importance.

Further reading

x Cheeseman MG, Leech G, Chambers J,Monaghan MJ, Nihoyannopoulos P. Centralrole of echocardiography in the diagnosisand assessment of heart failure. Heart1998;80(suppl 1):S15.

x Dargie HJ, McMurray JVV. Diagnosis andmanagement of heart failure. BMJ1994;308:3218.

x Schiller NB, Foster E. Analysis of leftventricular systolic function. Heart1996;75(suppl 2):1726.

The table showing the value of electrocardiography isadapted from Davie et al (BMJ 1996;312:222). Thetable of assessments for the investigation anddiagnosis of heart failure is adapted with permissionfrom the Task Force on Heart Failure of the EuropeanSociety of Cardiology (Eur Heart J 1995;16:74151).

The ABC of heart failure is edited by C R Gibbs,M K Davies, and G Y H Lip. CRG is researchfellow and GYHL is consultant cardiologist andreader in medicine in the university department ofmedicine and the department of cardiology, CityHospital, Birmingham; MKD is consultantcardiologist in the department of cardiology, SellyOak Hospital, Birmingham. The series will bepublished as a book in the spring.

BMJ 2000;320:297300

Clinical review



20

ABC of heart failureNondrug managementC R Gibbs, G Jackson, G Y H Lip

Approaches to the management of heart failure can be bothnonpharmacological and pharmacological; each approachcomplements the other. This article will discussnonpharmacological management.

Counselling and education of patientsEffective counselling and education of patients, and of therelatives or carers, is important and may enhance long termadherence to management strategies. Simple explanationsabout the symptoms and signs of heart failure, including detailson drug and other treatment strategies, are valuable. Emphasisshould be placed on self help strategies for each patient; theseshould include information on the need to adhere to drugtreatment. Some patients can be instructed how to monitortheir weight at home on a daily basis and how to adjust the doseof diuretics as advised; sudden weight increases ( > 2 kg in 13days), for example, should alert a patient to alter his or hertreatment or seek advice.

Lifestyle measuresUrging patients to alter their lifestyle is important in themanagement of chronic heart failure. Social activities should beencouraged, however, and care should be taken to ensure thatpatients avoid social isolation. If possible, patients shouldcontinue their regular work, with adaptations to accommodate areduced physical capacity where appropriate.

Contraceptive adviceAdvice on contraception should be offered to women ofchildbearing potential, particularly those patients with advancedheart failure (class IIIIV in the New York Heart Association’sclassification), in whom the risk of maternal morbidity andmortality is high with pregnancy and childbirth. Currenthormonal contraceptive methods are much safer than in thepast: low dose oestrogen and third generation progestogenderivatives are associated with a relatively low thromboembolicrisk.

SmokingCigarette smoking should be strongly discouraged in patientswith heart failure. In addition to the well established adverseeffects on coronary disease, which is the underlying cause in asubstantial proportion of patients, smoking has adversehaemodynamic effects in patients with congestive heart failure.For example, smoking tends to reduce cardiac output, especiallyin patients with a history of myocardial infarction.

Other adverse haemodynamic effects include an increase inheart rate and systemic blood pressure (double product) andmild increases in pulmonary artery pressure, ventricular fillingpressures, and total systemic and pulmonary vascular resistance.

The peripheral vasoconstriction may contribute to theobserved mild reduction in stroke volume, and thus smokingincreases oxygen demand and also decreases myocardialoxygen supply owing to reduced diastolic filling time (withfaster heart rates) and increased carboxyhaemoglobinconcentrations.

Nonpharmacological measures for the management of heartfailure

x Compliance—give careful advice about disease, treatment, and selfhelp strategies

x Diet—ensure adequate general nutrition and, in obese patients,weight reduction

x Salt—advise patients to avoid high salt content foods and not to addsalt (particularly in severe cases of congestive heart failure)

x Fluid—urge overloaded patients and those with severe congestiveheart failure to restrict their fluid intake

x Alcohol—advise moderate alcohol consumption (abstinence inalcohol related cardiomyopathy)

x Smoking—avoid smoking (adverse effects on coronary disease,adverse haemodynamic effects)

x Exercise—regular exercise should be encouragedx Vaccination—patients should consider influenza and pneumococcal

vaccinations

Intrauterine devices are a suitable form ofcontraception, although these may be aproblem in patients with primary valvardisease, in view of the risks of infectionand risks associated with oralanticoagulation

Menopausal women with heart failure

x Observational data indicate that hormone replacement therapyreduces the risk of coronary events in postmenopausal women

x However, there is limited prospective evidence to advise the use ofsuch therapy in postmenopausal women with heart failure

x Nevertheless, there may be an increased risk of venous thrombosisin postmenopausal women taking hormone replacement therapy,which may exacerbate the risk associated with heart failure

Daily weighing

Compliance with

medication

Self management

of diuretics

Self help strategies

Regular

exercise

Self help strategies for patients with heart failure

Clinical review

366 BMJ VOLUME 320 5 FEBRUARY 2000 www.bmj.com


21

AlcoholIn general, alcohol consumption should be restricted tomoderate levels, given the myocardial depressant properties ofalcohol. In addition to the direct toxic effects of alcohol on themyocardium, a high alcohol intake predisposes to arrhythmias(especially atrial fibrillation) and hypertension and may lead toimportant alterations in fluid balance. The prognosis in alcoholinduced cardiomyopathy is poor if consumption continues, andabstinence should be advised. Abstinence can result in markedimprovements, with echocardiographic studies showingsubstantial clinical benefit and improvements in left ventricularfunction. Resumed alcohol consumption may subsequently leadto acute or worsening heart failure.

Immunisation and antiobiotic prophylaxisChronic heart failure predisposes to and can be exacerbated bypulmonary infection, and influenza and pneumococcalvaccinations should therefore be considered in all patients withheart failure. Antibiotic prophylaxis, for dental and othersurgical procedures, is mandatory in patients with primary valvedisease and prosthetic heart valves.

Diet and nutritionAlthough controlled trials offer only limited information on dietand nutritional measures, such measures are as important inheart failure, as in any other chronic illness, to ensure adequateand appropriate nutritional balance. Poor nutrition maycontribute to cardiac cachexia, although malnutrition is notlimited to patients with obvious weight loss and muscle wasting.

Patients with chronic heart failure are at an increased riskfrom malnutrition owing to (a) a decreased intake resultingfrom a poor appetite, which may be related to drug treatment(for example, aspirin, digoxin), metabolic disturbance (forexample, hyponatraemia or renal failure), or hepaticcongestion; (b) malabsorption, particularly in patients withsevere heart failure; and (c) increased nutritional requirements,with patients who have congestive heart failure having anincrease of up to 20% in basal metabolic rate. These factors maycontribute to a net catabolic state where lean muscle mass isreduced, leading to an increase in symptoms and reducedexercise capacity. Indeed, cardiac cachexia is an independentrisk factor for mortality in patients with chronic heart failure. Aformal nutritional assessment should thus be considered inthose patients who appear to have a poor nutritional state.

Weight loss in obese patients should be encouraged asexcess body mass increases cardiac workload during exercise.Weight reduction in obese patients to within 10% of the optimalbody weight should be encouraged.

Salt restrictionNo randomised studies have addressed the role of saltrestriction in congestive heart failure. Nevertheless restriction toabout 2 g of sodium a day may be useful as an adjunct totreatment with high dose diuretics, particularly if the conditionis advanced.

In general, patients should be advised that they should avoidfoods that are rich in salt and not to add salt to their food at thetable.

Fluid intakeFluid restriction (1.52 litres daily) should be considered inpatients with severe symptoms, those requiring high dosediuretics, and those with a tendency towards excessive fluidintake. High fluid intake negates the positive effects of diureticsand induces hyponatraemia.

Community and social support

x Community support is particularly important for elderly orfunctionally restricted patients with chronic heart failure

x Support may help to improve the quality of life and reduceadmission rates

x Social services support and community based interventions, withadvice and assistance for close relatives, are also important

Managing cachexia in chronic heart failure

Combined management by physician and dietician is recommendedx Alter size and frequency of mealsx Ensure a higher energy dietx Supplement diet with (a) water soluble vitamins (loss associated

with diuresis), (b) fat soluble vitamins (levels reduced as a result ofpoor absorption), and (c) fish oils

Commonly consumed processed foods that have a highsodium content

x Cheesex Sausagesx Crisps, salted peanutsx Milk and white chocolatex Tinned soup and tinned vegetablesx Ham, bacon, tinned meat (eg corned beef)x Tinned fish (eg sardines, salmon, tuna)x Smoked fish

Fresh produce, such as fruit, vegetables,eggs, and fish, has a relatively low saltcontent

Date

Pulse

BP (lying)

BP (standing)

Urine

Weight

Drug 1

Drug 2

Drug 3

Drug 4

Drug 5

Drug 6

Serum urea/creatinine

Serum potassium

Other investigations

Next visit

Doctor's signature

Heart failure cooperation card: patients and doctors are able to monitorchanges in clinical signs (including weight), drug treatment, and baselineinvestigations. Patients should be encouraged to monitor their weightbetween clinic visits

Clinical review

367BMJ VOLUME 320 5 FEBRUARY 2000 www.bmj.com


22

Exercise training and rehabilitationExercise training has been shown to benefit patients with heartfailure: patients show an improvement in symptoms, a greatersense of wellbeing, and better functional capacity. Exercise doesnot, however, result in obvious improvement in cardiac function.

All stable patients with heart failure should be encouragedto participate in a supervised, simple exercise programme.Although bed rest (“armchair treatment”) may be appropriatein patients with acute heart failure, regular exercise should beencouraged in patients with chronic heart failure. Indeed,chronic immobility may result in loss of muscle mass in thelower limb and generalised physical deconditioning, leading toa further reduction in exercise capacity and a predisposition tothromboembolism. Deconditioning itself may be detrimental,with peripheral alterations and central abnormalities leading tovasoconstriction, further deterioration in left ventricularfunction, and greater reduction in functional capacity.

Importantly, regular exercise has the potential to slow orstop this process and exert beneficial effects on the autonomicprofile, with reduced sympathetic activity and enhanced vagaltone, thus reversing some of the adverse consequences of heartfailure. Large prospective clinical trials will establish whetherthese beneficial effects improve prognosis and reduce theincidence of sudden death in patients with chronic heart failure.

Regular exercise should therefore be advocated in stablepatients as there is the potential for improvements in exercisetolerance and quality of life, without deleterious effects on leftventricular function. Cardiac rehabilitation services offer benefitto this group, and patients should be encouraged to developtheir own regular exercise routine, including walking, cycling,and swimming. Nevertheless, patients should know their limits,and excessive fatigue or breathlessness should be avoided. Inthe first instance, a structured walking programme would be theeasiest to adopt.

Treatment of underlying diseaseTreatment should also be aimed at slowing or reversing anyunderlying disease process.

HypertensionGood blood pressure control is essential, and angiotensinconverting enzyme inhibitors are the drugs of choice in patientswith impaired systolic function, in view of their beneficial effectson slowing disease progression and improving prognosis. Incases of isolated diastolic dysfunction, either â blockers orcalcium channel blockers with rate limiting properties—forexample, verapamil, diltiazem—have theoretical advantages. Ifsevere left ventricular hypertrophy is the cause of diastolicdysfunction, however, an angiotensin converting enzymeinhibitor may be more effective at inducing regression of leftventricular hypertrophy. Angiotensin II receptor antagonistsshould be considered as an alternative if cough that is inducedby angiotensin converting enzyme inhibitors is problematic.

Effects of deconditioning in heart failure

Peripheral alterations Increased peripheral vascular resistance;impaired oxygen utilisation duringexercise

Abnormalities ofautonomic control

Enhanced sympathetic activation; vagalwithdrawal; reduced baroreflex sensitivity

Skeletal muscleabnormalities

Reduced mass and composition

Reduced functionalcapacity

Reduced exercise tolerance; reducedpeak oxygen consumption

Psychological effects Reduced activity; reduced overall sense ofwellbeing

Exercise class for group of patients with heart failure (published withpermission of participants)

M mode echocardiogram showing left ventricular hypertrophy inhypertensive patient (A=interventricular septum; B=posterior wall of leftventricle)

Beneficial effects of exercise in chronic heart failure

Has positive effects on:x Skeletal musclex Autonomic functionx Endothelial functionx Neurohormonal functionx Insulin sensitivity

No positive effects on survival have been shown

Clinical review



23

SurgeryIf coronary heart disease is the underlying cause of chronicheart failure and if cardiac ischaemia is present, the patient maybenefit from coronary revascularisation, including coronaryangioplasty or coronary artery bypass grafting.Revascularisation may also improve the function of previouslyhibernating myocardium. Valve replacement or valve repairshould be considered in patients with haemodynamicallyimportant primary valve disease.

Cardiac transplantation is now established as the treatmentof choice for some patients with severe heart failure whoremain symptomatic despite intensive medical treatment. It isassociated with a one year survival of about 90% and a 10 yearsurvival of 5060%, although it is limited by the availability ofdonor organs. Transplantation should be considered in youngerpatients (aged < 60 years) who are without severe concomitantdisease (for example, renal failure or malignancy).

Bradycardias are managed with conventional permanentcardiac pacing, although a role is emerging for biventricularcardiac pacing in some patients with resistant severe congestiveheart failure. Implantable cardiodefibrillators are wellestablished in the treatment of some patients with resistant lifethreatening ventricular arrhythmias. New surgical approachessuch as cardiomyoplasty and ventricular reduction surgery(Batista procedure) are rarely used owing to the high associatedmorbidity and mortality and the lack of conclusive trialevidence of substantial benefit.

The box about managing cachexia is based on recommendations from theScottish Intercollegiate Guidelines Network (SIGN) (publication No 35,1999).

G Jackson is consultant cardiologist in the department of cardiology,Guy’s and St Thomas’s Hospital, London.


BMJ 2000;320:3669

Role of surgery in heart failure

Type of surgery Reason

Coronary revascularisation(PTCA, CABG)

Angina, reversible ischaemia,hibernating myocardium

Valve replacement (or repair) Significant valve disease (aorticstenosis, mitral regurgitation)

Permanent pacemakers andimplantable cardiodefibrillators

Bradycardias; resistant ventriculararrhythmias

Cardiac transplantation End stage heart failure

Ventricular assist devices Short term ventricularsupport—eg awaitingtransplantation

Novel surgical techniques Limited role (high mortality,limited evidence of substantialbenefit)

PTCA = percutaneous transluminal coronary angioplasty; CABG = coronaryartery bypass graft.

Key references

x Demakis JG, Proskey A, Rahimtoola SH, Jamil M, Sutton GC, RosenKM, et al. The natural course of alcoholic cardiomyopathy. AnnIntern Med 1974;80:2937.

x The Task Force of the Working Group on Heart Failure of theEuropean Society of Cardiology. Guidelines on the treatment ofheart failure. Eur Heart J 1997;18:73653.

x Kostis JB, Rosen RC, Cosgrove NM, Shindler DM, Wilson AC.Nonpharmacologic therapy improves functional and emotionalstatus in congestive heart failure. Chest 1994;106:9961001.

x McKelvie RS, Teo KK, McCartney N, Humen D, Montague T, YusufS. Effects of exercise training in patients with congestive heartfailure: a critical review. J Am Coll Cardiol 1995;25:78996.

aVRI

II

II

V1 V4

aVL V2 V5

III aVF V3 V6

Electrocardiogram showing left ventricular hypertrophy on voltage criteria, with associated T wave and ST changes in thelateral leads (“strain pattern”)

Clinical review



24

ABC of heart failureManagement: diuretics, ACE inhibitors, and nitratesM K Davies, C R Gibbs, G Y H Lip

In the past 15 years several large scale, randomised controlledtrials have revolutionised the management of patients withchronic heart failure. Although it is clear that some drugsimprove symptoms, others offer both symptomatic andprognostic benefits, and the management of heart failureshould be aimed at improving both quality of life and survival.

Diuretics and angiotensin converting enzyme (ACE)inhibitors, when combined with nonpharmacologicalmeasures, remain the basis of treatment in patients withcongestive heart failure. Digoxin has a possible role in some ofthese patients, however, and the potential benefits of â blockersand spironolactone (an aldosterone antagonist) in chronic heartfailure are now increasingly recognised.

DiureticsDiuretics are effective in providing symptomatic relief andremain the first line treatment, particularly in the presence ofoedema. Nevertheless, there is no direct evidence that loop andthiazide diuretics confer prognostic benefit in patients withcongestive heart failure.

Loop diureticsLoop diuretics—frusemide (furosemide) and bumetanide—havea powerful diuretic action, increasing the excretion of sodiumand water via their action on the ascending limb of the loop ofHenle. They have a rapid onset of action (intravenously 5minutes, orally 12 hours; duration of action 46 hours). Oralabsorption of frusemide may be reduced in congestive heartfailure, although the pharmacokinetics of bumetanide mayallow improved bioavailability.

Patients receiving high dose diuretics (frusemide >80 mg orequivalent) should be monitored for renal and electrolyteabnormalities. Hypokalaemia, which may precipitatearrhythmias, should be avoided, and potassiumsupplementation, or concomitant treatment with a potassiumsparing agent, should generally be used unlesscontraindicated—for example, in renal dysfunction withpotassium retention. Acute gout is a relatively common adverseeffect of treatment with high dose intravenous diuretics.

Thiazide diureticsThiazides—such as bendrofluazide (bendroflumethiazide)—acton the cortical diluting segment of the nephron. They are oftenineffective in elderly people, owing to the age related and heartfailure mediated reduction in glomerular filtration rate.Hyponatraemia and hypokalaemia are commonly associatedwith higher doses of thiazide diuretics, and potassiumsupplementation, or concomitant treatment with a potassiumsparing agent, is usually needed with high dose thiazide therapy.

In some patients with chronic severe congestive heartfailure, particularly in the presence of chronic renalimpairment, oedema may persist despite conventional oraldoses (frusemide 40160 mg daily) of loop diuretics. In thesepatients, a thiazide diuretic (for example, bendrofluazide) or athiazidelike diuretic (for example, metolazone) may becombined with a loop diuretic. This combination blocksreabsorption of sodium at different sites in the nephron

Aims of heart failure management

To achieve improvement in symptomsx Diureticsx Digoxinx ACE inhibitors

To achieve improvement in survivalx ACE inhibitorsx â blockers (for example, carvedilol and bisoprolol)x Oral nitrates plus hydralazinex Spironolactone

In general, diuretics should be introducedat a low dose and the dose increasedaccording to the clinical response. Thereare dangers, however, in eitherundertreating or overtreating patientswith diuretics, and regular review isnecessary

How to use diuretics in advanced heart failure

x Optimise diuretic dosex Consider combination diuretic treatment with a loop and thiazide

(or thiazidelike) diureticx Consider combining a low dose of spironolactone with an ACE

inhibitor, provided that there is no evidence of hyperkalaemiax Administer loop diuretics (either as a bolus or a continuous

infusion) intravenously

Cortical collecting

duct

Medullary collecting

ductLoop of

Henle

Proximal

tubule

Thick

ascending

limb

Distal tubule2

1

3

Early Late

Diagram of nephron showing sites of action of different diuretic classes:1=loop (eg frusemide); 2=thiazide (eg bendrofluazide); and 3=potassiumsparing (eg amiloride)

Clinical review



25

(“double nephron blockade”), and this synergistic action leadsto a greater diuretic effect. The incidence of associatedmetabolic abnormalities is, however, increased, and suchtreatment should be started only under close supervision. Insome patients, a large diuretic effect may occur soon after acombination regimen (loop diuretic plus either thiazide ormetalozone) has been started. It is advisable, therefore, toconsider such a combination treatment on a twice weekly basis,at least initially.

Potassium sparing diureticsAmiloride acts on the distal nephron, while spironolactone is acompetitive aldosterone inhibitor. Potassium sparing diureticshave generally been avoided in patients receiving ACEinhibitors, owing to the potential risk of hyperkalaemia.Nevertheless, a recent randomised placebo controlled study, therandomised aldactone evaluation study (RALES), reported thathyperkalaemia is uncommon when low dose spironolactone(<25 mg daily) is combined with an ACE inhibitor. Risk factorsfor developing hyperkalaemia include spironolactone dose> 50 mg/day, high doses of ACE inhibitor, or evidence of renalimpairment. It is recommended that measurement of the serumcreatinine and potassium concentrations is performed within57 days of the addition of a potassium sparing diuretic to anACE inhibitor until the levels are stable, and then every one tothree months.

ACE inhibitorsACE inhibitors have consistently shown beneficial effects onmortality, morbidity, and quality of life in large scale, prospectiveclinical trials and are indicated in all stages of symptomaticheart failure resulting from impaired left ventricular systolicfunction.

Mechanisms of actionACE inhibitors inhibit the production of angiotensin II, apotent vasoconstrictor and growth promoter, and increaseconcentrations of the vasodilator bradykinin by inhibiting itsdegradation. Bradykinin has been shown to have beneficialeffects associated with the release of nitric oxide andprostacyclin, which may contribute to the positivehaemodynamic effects of the ACE inhibitors. Bradykinin mayalso be responsible, however, for some of the adverse effects,such as dry cough, hypotension, and angiooedema.

ACE inhibitors also reduce the activity of the sympatheticnervous system as angiotensin II promotes the release ofnoradrenaline and inhibits its reuptake. In addition, they alsoimprove â receptor density (causing their up regulation),variation in heart rate, baroreceptor function, and autonomicfunction (including vagal tone).

Clinical effectsSymptomatic left ventricular dysfunctionACE inhibitors, when added to diuretics, improve symptoms,exercise tolerance, and survival and reduce hospital admissionrates in chronic heart failure.

These beneficial effects are apparent in all grades of systolicheart failure—that is, mild to moderate chronic heart failure (asevident, for example, in the Munich mild heart failure study, thevasodilator heart failure trials (VHeFT), and the studies of leftventricular dysfunction treatment trial (SOLVDT)) and severechronic heart failure (as, for example, in the first cooperativenorth Scandinavian enalapril survival study (CONSENSUS I).

The two main potassium sparingdiuretics, amiloride and spironolactone,have a weak diuretic action when usedalone; amiloride is most commonly usedin fixed dose combinations with a loopdiuretic—for example, coamilofruse

Guidelines for using ACE inhibitors

x Stop potassium supplements and potassium sparing diureticsx Omit (or reduce) diuretics for 24 hours before first dosex Advise patient to sit or lie down for 24 hours after first dosex Start low doses (for example, captopril 6.25 mg twice daily,

enalapril 2.5 mg once daily, lisinopril 2.5 mg once daily)x Review after 12 weeks to reassess symptoms, blood pressure, and

renal chemistry and electrolytesx Increase dose unless there has been a rise in serum creatinine

concentration (to > 200 ìmol/l) or potassium concentration (to> 5.0 mmol/l)

x Titrate to maximum tolerated dose, reassessing blood pressure andrenal chemistry and electrolytes after each dose titration

If patient is “high risk” consider hospital admission to start treatment

Front view and side view of woman with angiooedema related to treatmentwith ACE inhibitors (published with permission of patient)

1.0

0.9

0.8

0.7

0.6

0.5

030 6 9 12 15 18 21 24 27 30 33 36

Months

Pro

bab

ility

of

surv

ival

Spironolactone

Placebo

Survival curve for randomised aldactone evaluation study (RALES) showing30% reduction in all cause mortality when spironolactone (up to 25 mg) wasadded to conventional treatment in patients with severe (New York HeartAssociation class IV) chronic heart failure

Clinical review



26

Asymptomatic left ventricular dysfunctionACE inhibitors have also been shown to be effective inasymptomatic patients with left ventricular systolic dysfunction.The studies of left ventricular dysfunction prevention trial(SOLVDP) confirmed the benefit of ACE inhibitors inasymptomatic left ventricular systolic dysfunction, whereenalapril reduced the development of heart failure and relatedhospital admissions.

Left ventricular dysfunction after myocardial infarctionLarge scale, randomised controlled trials—for example, theacute infarction ramipril efficacy (AIRE) study, the survival andventricular enlargement (SAVE) study, and the trial oftrandolapril cardiac evaluation (TRACE)—have shown lowermortality in patients with impaired systolic function aftermyocardial infarction, irrespective of symptoms.

Slowing disease progressionACE inhibitors also seem to influence the natural course ofchronic heart failure. The Munich mild heart failure studyshowed that ACE inhibitors combined with standard treatmentslowed the progression of heart failure in patients with mildsymptoms, with significantly fewer patients in the activetreatment group developing severe heart failure.

Doses and tolerabilityACE inhibitors should be started at a low dose and graduallytitrated to the highest tolerated maintenance level. The recentprospective assessment trial of lisinopril and survival (ATLAS)randomised patients with symptomatic heart failure to low dose(2.55 mg daily) or high dose (32.535 mg daily) lisinopril, and,although there was no significant mortality difference, highdose treatment was associated with a significant reduction in thecombined end point of all cause mortality and all causeadmissions to hospital. Adverse effects of the ACE inhibitorsinclude cough, dizziness, and a deterioration in renal function,although the overall incidence of hypotension and renalimpairment in the CONSENSUS and SOLVD studies was only5%. Angiooedema related to ACE inhibitors is rare, althoughmore common in patients of AfroCaribbean origin than inother ethnic groups.

ACE inhibitors can therefore be regarded as thecornerstone of treatment in patients with all grades ofsymptomatic heart failure and in patients with asymptomaticleft ventricular dysfunction. Every attempt should be made toprovide this treatment for patients, unless it is contraindicated,and to use adequate doses.

Angiotensin receptor antagonistsOrally active angiotensin II type 1 receptor antagonists, such aslosartan, represent a new class of agents that offer an alternativemethod of blocking the reninangiotensin system. The effects ofangiotensin II receptor antagonists on haemodynamics,neuroendocrine activity, and exercise tolerance resemble thoseof ACE inhibitors, although it still remains to be establishedfully whether these receptor antagonists are an effectivesubstitute for ACE inhibitors in patients with chronic heartfailure.

The evaluation of losartan in the elderly (ELITE) studycompared losartan with captopril in patients aged 65 or overwith mild to severe congestive heart failure. Although theELITE study was designed as a tolerability study, and survivalwas not the primary end point, it did report a reduction in allcause mortality (4.8% v 8.7%) in patients treated with losartan.Important limitations of the ELITE study included the limited

Metaanalysis of effects of ACE inhibitors on mortality andadmissions in chronic heart failure

Nooftrials

Total Noof

patientsPlacebo

(%)

Activetreatment

(%)

Riskreduction

(%) P value

32 7105 32.6 22.4 35 < 0.001

ACE inhibitors: high risk patients warranting hospitaladmission for start of treatment

x Severe heart failure (NYHA class IV) or decompensated heartfailure

x Low systolic blood pressure ( < 100 mm Hg)x Resting tachycardia > 100 beats/minutex Low serum sodium concentration ( < 130 mmol/l)x Other vasodilator treatmentx Severe chronic obstructive airways disease and pulmonary heart

disease (cor pulmonale)

Doses of ACE inhibitors used in large controlled trials

Trial ACE inhibitorTarget

dose (mg)Mean dailydose (mg)

CONSENSUS Enalapril 20* 18.4

VHeFT II Enalapril 10* 15.0

SOLVD Enalapril 10* 16.6

SAVE Captopril 50† NA

*Twice daily; †three times daily. NA = information not available.

Recommended maintenance doses of ACE inhibitors

Drug Starting dose (mg) Maintenance dose (mg)

Captopril 6.25† 2550†

Enalapril 2.5‡ 10*

Lisinopril 2.5‡ 520‡

Quinapril 2.55‡ 510*

Perindopril 2‡ 4‡

Ramipril 1.252.5‡ 2.55*

Trandolapril 0.5‡ 24‡

*Twice daily; †three times daily; ‡once daily.

1.2

No of patients to betreated per year toprevent one death

Mortality in controlgroup (per 100

patients/year)5.1 8.1 11.5 19.1 123.3

330 66 76 22 2

1.0

0.8

0.6

0.4

0.2

Rel

ativ

e ri

sk

SOLVD

(prevention)

SAVE SOLVD

(treatment)

AIRE CONSENSUS

ACE inhibitors in left ventricular dysfunction: best benefit for ACEinhibitors in higher risk group

Clinical review



27

size and the relatively short follow up. However, the recentlyreported ELITE II mortality study failed to show that treatmentwith losartan was superior to captopril, although it confirmedimproved tolerability with losartan.

ACE inhibitors, therefore, remain the treatment of choice inpatients with left ventricular systolic dysfunction, althoughangiotensin II receptor antagonists are an appropriatealternative in patients who develop intolerable side effects fromACE inhibitors.

Oral nitrates and hydralazineThe VHeFT trials showed a survival benefit from combinedtreatment with nitrates and hydralazine in patients withsymptomatic heart failure (New York Heart Association classIIIII). The VHeFT II trial also showed a modest improvementin exercise capacity, although the nitrate and hydralazinecombination was less well tolerated than enalapril, owing to thedose related adverse effects (dizziness and headaches). There isno reproducible evidence of symptomatic improvement fromother randomised placebo controlled trials, however, andsurvival rates were higher with ACE inhibitors than with thenitrate and hydralazine combination (VHeFT II trial).

In general, oral nitrates should be considered in patientswith angina and impaired left ventricular systolic function. Thecombination of nitrates and hydralazine is an alternativeregimen in patients with severe renal impairment, in whomACE inhibitors and angiotensin II receptor antagonists arecontraindicated. Although it is rational to consider the additionof a combination of nitrates and hydralazine in patients whocontinue to have severe symptoms despite optimal doses ofACE inhibitors, no large scale trials have shown an additiveeffect of these combinations.

Other vasodilatorsLong acting dihydropyridine calcium channel blockersgenerally have neutral effects in heart failure, although others,such as diltiazem and verapamil, have negatively inotropic andchronotropic properties, with the potential to exacerbate heartfailure. Two recent trials of the newer calcium channel blockersamlodipine (the prospective randomised amlodipine survivalevaluation (PRAISE) trial) and felodipine (VHeFT III) inpatients with heart failure suggest that long acting calciumantagonists may have beneficial effects in nonischaemic dilatedcardiomyopathy, although further studies are in progress—forexample, PRAISE II. Importantly, these studies indicate thatamlodipine and felodipine seem to be safe in patients withcongestive heart failure and could therefore be used to treatangina and hypertension in this group of patients.

The two tables on recommended doses of ACE inhibitors are adapted andreproduced with permission from Remme WJ (Eur Heart J 1997;18:73653). The metaanalysis table is adapted and used with permission fromGarg R et al (JAMA 1995;273:14506). The graph showing the benefit ofACE inhibitors in left ventricular dysfunction is adapted from Davey Smithet al (BMJ 1994;308:734).


BMJ 2000;320:42831

Vasodilator heart failure (VHeFT) studies

Study ComparisonNYHAclass* Outcome

VHeFT I Hydralazine plusisosorbidedinitrate v placebo

II, III Improved mortalitywith active treatment

VHeFT II Hydralazine plusisosorbidedinitrate venalapril

II, III Enalapril superior tohydralazine plusisosorbide dinitratefor survival

*I = asymptomatic, II = mild, III = moderate, IV = severe.

Key references

x Cohn JN, Johnson G, Ziesche S, Cobb F, Francis G, Tristani F, et al.A comparison of enalapril with hydralazineisosorbide dinitrate inthe treatment of chronic congestive heart failure. N Engl J Med1991;325:30310.

x Cohn JN, Ziesche S, Smith R, Anand I, Dunkman WB, Loeb H, et al.Effect of the calcium antagonist felodipine as supplementaryvasodilator therapy in patients with chronic heart failure treatedwith enalapril. VHeFT III. Circulation 1997;96:85663.

x Packer M, O’Connor CM, Ghali JK, Pressler ML, Carson PE, BelkinRN, et al. Effect of amlodipine on morbidity and mortality in severechronic heart failure. N Engl J Med 1996;335:110714.

x Pitt B, Segal R, Martinez FA, Meurers G, Cowley AJ, Thomas I, et al.Randomised trial of losartan versus captopril in patients over 65with heart failure. Lancet 1997;349:74752.

x Remme WJ. The treatment of heart failure. The Task Force of theWorking Group on Heart Failure of the European Society ofCardiology. Eur Heart J 1997;18:73653.

x Pitt B, Zannad F, Remme WJ, Cody R, Castaigne A, Perez A, et al.The effect of spironolactone on morbidity and mortality in patientswith severe heart failure. N Engl J Med 1999;341:70917.

60

Months

Cu

mu

lati

ve m

ort

alit

y

302418126 5448423600

0.50

0.75

0.25

Enalapril

Hydralazine plus isosorbide dinitrate

Cumulative mortality in VHeFT II trial: enalapril v hydralazine plusisosorbide dinitrate in patients with congestive heart failure (mild tomoderate)

ELITE II study: the losartan heart failure survival study

x Multicentre, randomised, parallel group trial of captopril v losartanin chronic stable heart failure

x 3152 patients; age > 60 years (mean age 71.5 years); NYHA classIIIV heart failure; mean follow up of 2 years

x No significant difference in all cause mortality between thecaptopril group (15.9%) and losartan group (17.7%)

x Better tolerability with losartan (withdrawal rate 9.4%) than withcaptopril (14.5%)

Clinical review



28

ABC of heart failureManagement: digoxin and other inotropes, â blockers, andantiarrhythmic and antithrombotic treatmentC R Gibbs, M K Davies, G Y H Lip

Digoxin

Use of digoxin for heart failure varies between countries acrossEurope, with high rates in Germany and low rates in the UnitedKingdom. It is potentially invaluable in patients with atrialfibrillation and coexistent heart failure, improving control of theventricular rate and allowing more effective filling of theventricle. Digoxin is also used in patients with chronic heartfailure secondary to left ventricular systolic impairment, in sinusrhythm, who remain symptomatic despite optimal doses ofdiuretics and angiotensin converting enzyme inhibitors, whereit acts as an inotrope.

Evidence of symptomatic benefit from digoxin in patientswith chronic heart failure in sinus rhythm has been reported inseveral randomised placebo controlled trials and several smallertrials. The RADIANCE and PROVED trials, for example,reported that the withdrawal of digoxin from patients withcongestive heart failure who had already been treated with thedrug was associated with worsening heart failure and increasedhospital readmission rates. The Digitalis Investigation Group’slarge study found that digoxin was associated with asymptomatic improvement in patients with congestive heartfailure, when added to treatment with diuretics and angiotensinconverting enzyme inhibitors. Importantly, there were greaterabsolute and relative benefits in the patients who had resistantsymptoms and more severe impairment of left ventricularsystolic function. However, although there was a reduction inthe combined end points of admission and mortality resultingfrom heart failure, there was no significant improvement inoverall survival. â Blockers were used rarely in the DigitalisInvestigation Group’s study, and as a result it is not clearwhether digoxin is additive to both the â blockers andangiotensin converting enzyme inhibitors in congestive heartfailure.

Digoxin should be considered in patientswith sinus rhythm plus (a) continuedsymptoms of heart failure despite optimaldoses of diuretics and angiotensinconverting enzyme inhibitors; (b) severeleft ventricular systolic dysfunction withcardiac dilatation; or (c) recurrent hospitaladmissions for heart failure

Digoxin: practical aspects

x Ensure a maintenance dose of 125375 ìg (0.1250.375 mg) dailyx Give a reduced maintenance dose in elderly people, when renal

impairment is present, and when used with drugs that increasedigoxin concentrations (amiodarone, verapamil)

x Concentrations should be monitored especially in cases ofuncertainty about whether therapeutic levels have been achieved(range 6 hours after dose: 1.21.9 ng/ml)

x Monitor potassium concentrations (avoid hypokalaemia) and renalfunction

x Digoxin toxicity may be associated with: (a) adverse symptoms (forexample, nausea, vomiting, headache, confusion, visual symptoms);and (b) arrhythmias (for example, atrioventricular junctionalrhythms, atrial tachycardia, atrioventricular block, ventriculartachycardia)

x Serious toxicity should be treated by correcting potassiumconcentrations and with drugs such as â blockers and glycosidebinding agents (cholestyramine), and in severe cases specificdigoxin antibodies (Digibind)

Source of information: Uretsky et al (J Am Coll Cardiol 1993;22:955) and Packeret al (N Engl J Med 1993;329:1)

Study of effect of digoxin on mortality and morbidity inpatients with heart failure*

Number of participants: 6800

Design: prospective, randomised, double blind, placebo controlled

Participants: left ventricular ejection fraction < 45%

Intervention: randomised to digoxin (0.1250.500 mg) or placebo;follow up at 37 months

Results:x Reduced admissions to hospital owing to heart failure (greater

absolute and relative benefits in the patients with resistantsymptoms and more severe impairment of left ventricular systolicfunction)

x No effect on overall survival

*The Digitalis Investigation Group’s study (see key references box)

50PlaceboDigoxin

P< 0.001

40

30

20

10

00 4 8 12 16 20 24 28 32 36 40 4844

Months

Dea

th o

r ad

mis

sion

to h

ospi

tal

due

to w

orse

ning

hea

rt fa

ilure

(%

)

52

Incidence of death or admission to hospital due to worsening heart failurein two groups of patients: those receiving digoxin and those receivingplacebo (Digitalis Investigation Group’s study—see key references box at endof article)

Clinical review


29

Other inotropesThe potential role of inotropic agents other than digoxin inchronic heart failure has been addressed in several studies.Although these drugs seem to improve symptoms in somepatients, most have been associated with an increase inmortality.

For example, the PRIME II trial (a prospective randomisedstudy) examined ibopamine, a weak inotrope, in patients withchronic heart failure who were already receiving standardtreatment. An excess mortality was shown, however,particularly in those with severe symptoms; this was possiblyrelated to an excess of arrhythmias. In addition, a previous trialevaluating intermittent dobutamine infusions in patients withchronic heart failure was stopped prematurely because ofexcess mortality in the group taking dobutamine. Xamoterol, aâ receptor antagonist with mild agonist inotropic effects, alsofailed to show any positive benefits in patients with heartfailure.

In overall terms, no evidence exists at present to support theuse of oral catecholamine receptor (or postreceptor pathway)stimulants in the treatment of chronic heart failure. Digoxinremains the only (albeit weak) positive inotrope that is valuablein the management of chronic heart failure.

â Blockersâ Adrenoceptor blockers have traditionally been avoided inpatients with heart failure due to their negative inotropic effects.However, there is now considerable clinical evidence to supportthe use of â blockers in patients with chronic stable heart failureresulting from left ventricular systolic dysfunction. Recentrandomised controlled trials in patients with chronic heartfailure have reported that combining â blockers withconventional treatment with diuretics and angiotensinconverting enzyme inhibitors results in improvements in leftventricular function, symptoms, and survival, as well as areduction in admissions to hospital.

Recently, two randomised controlled trials have studied theeffects of carvedilol, a â blocker with á blocking and vasodilatorproperties, in patients with symptomatic heart failure. The USmulticentre carvedilol study programme was stopped earlybecause of a highly significant (65%) mortality benefit inpatients receiving carvedilol, when compared to placebo, andthe Australia/New Zealand heart failure study reported a 41%reduction in the combined primary end point of all causehospital admission and mortality. Bisoprolol has also beenstudied, and, although the first cardiac insufficiency bisoprololstudy (CIBIS I) reported a trend towards a reduction inmortality and need for cardiac transplantation, there was noconclusive survival benefit. The recent CIBIS II study, however,was stopped prematurely because of the beneficial effects ofactive treatment on both morbidity and mortality. Metoprololhas also shown similar prognostic advantages in the metoprololrandomised intervention trial in heart failure (MERITHF),which was also stopped early. In summary, evidence is now

Inotropic drugs associated with increased mortality inchronic heart failure

Drug ClassInotropicactivity

Xamoterol â Receptor antagonist Mild

Dobutamine Dopamine, á, and â receptorantagonist

Strong

Ibopamine Dopamine, á, and â receptorantagonist

Weak

Amrinone Phosphodiesterase inhibitor Strong

Enoximone Phosphodiesterase inhibitor Strong

Flosequinan Attenuates inositol triphosphate Weak

Milrinone Phosphodiesterase inhibitor Strong

Vesnarinone Phosphodiesterase inhibitor Mild

Potential mechanisms and benefits ofâ blockers: improved left ventricularfunction; reduced sympathetic tone;improved autonomic nervous systembalance; up regulation of â adrenergicreceptors; reduction in arrhythmias,ischaemia, further infarction, myocardialfibrosis, and apoptosis

Randomised, placebo controlled â blocker trials incongestive heart failure

Study TreatmentNYHAclass* Outcome

MDC Metoprolol II, III Improved clinical statewithout effect onsurvival. Reduction inneed for transplantationin patients with dilatedcardiomyopathy

CIBIS I Bisoprolol II, III Trend (nonsignificant)towards improvedsurvival

ANZ trial Carvedilol I, II Carvedilol superior toplacebo for morbidityand mortality

Carvedilol(US)

Carvedilol II, III Carvedilol superior toplacebo for morbidityand mortality

CIBIS II Bisoprolol III, IV Bisoprolol superior toplacebo for morbidityand mortality

MERITHF Metoprolol II, III Metoprolol superior toplacebo for morbidityand mortality

Placebo groups in all trials received appropriate conventional treatment(diuretics alone; diuretics plus either digoxin or angiotensin converting enzymeinhibitors; or diuretics plus digoxin and angiotensin converting enzymeinhibitors). Trials still in progress: COMET (carvedilol v metoprolol) andCOPERNICUS (carvedilol in severe chronic heart failure).*Classification of the New York Heart Association (I = no symptoms, II = mild,III = moderate, IV = severe).

Metaanalysis of effects of â blockers on mortality andadmissions to hospital in chronic heart failure

No of trials(total No ofpatients)

% receivingplacebo

% receivingactive

treatment

Riskreduction

(%)P

value

18 (3023) 24.6 15.8 38 < 0.001

Clinical review


30

available to support the use of â blockers in chronic heartfailure, as the benefits supplement those already obtained fromangiotensin converting enzyme inhibitors.

Carvedilol is now licensed in the United Kingdom for use inmild to moderate chronic stable heart failure, although atpresent its use is still not recommended in patients with severesymptoms (New York Heart Association class IV). This lattergroup has been underrepresented in the trials to date.

In general, â blockers should be started at very low doses,with the dose being slowly increased, under expert supervision,to the target dose if tolerated. In the short term there may be adeterioration in symptoms, which may improve with alterationsin other treatment, particularly diuretics.

Antithrombotic treatment

In patients with chronic heart failure the incidence of strokeand thromboembolism is significantly higher in the presence ofatrial and left ventricular dilatation, particularly in severe leftventricular dysfunction. Nevertheless, there is conflictingevidence of benefit from routine treatment of patients withheart failure who are in sinus rhythm with antithrombotictreatment, although anticoagulation should be considered inthe presence of mobile ventricular thrombus, atrial fibrillation,and severe cardiac impairment. Large scale, prospectiverandomised controlled trials of antithrombotic treatment inheart failure are in progress, such as the WATCH study (a trialof warfarin and antiplatelet therapy); the full results are awaitedwith interest.

The combination of atrial fibrillation and heart failure (orevidence of left ventricular systolic dysfunction onechocardiography) is associated with a particularly high risk ofthromboembolism, which is reduced by long term treatmentwith warfarin. Aspirin seems to have little effect on the risk ofthromboembolism and overall mortality in such patients.

Antiarrhythmic treatment

Chronic heart failure and atrial fibrillationRestoration and long term maintenance of sinus rhythm is lesssuccessful in the presence of severe structural heart disease,particularly when the atrial fibrillation is longstanding. Inpatients with a deterioration in symptoms that is associated withrecent onset atrial fibrillation, treatment with amiodaroneincreases the long term success rate of cardioversion. Digoxin isotherwise appropriate for controlling ventricular rate in mostpatients with heart failure and chronic atrial fibrillation, with theaddition of amiodarone in resistant cases.

Summary of the cardiac insufficiency bisoprolol study II(CIBIS II)*

x Randomised, double blind, parallel group studyx 2647 participants (class IIIIV (moderate to severe) according to

classification of the New York Heart Association)x Bisoprolol, increased in dose to a maximum of 10 mg a dayx Trial stopped because of significant mortality benefit in patients

treated with bisoprolol:

(a) 32% reduction in all cause mortality(b) 32% reduction in admissions to hospital for worsening heart

failure(c) 42% reduction in sudden death

*CIBIS II Investigators and Committee (Lancet 1999;353:913)

Dose and titration of â blockers in large, placebo controlled heart failure trials

â BlockerInitial dose

(mg)

Weekly titration schedule: total daily dose (mg) Targettotal dailydose (mg)1 2 3 4 5 6 7 8–11 12–15

Metoprolol (MDC trial) 5 10 15 20 50 75 100 150 NI NI 100–150

Carvedilol (US trials) 3.125 6.25 NI 12.5 NI 25 NI 50 NI NI 50

Bisoprolol (CIBIS II) 1.25 1.25 2.5 3.75 5 5 5 5 7.5 10 10

References: Waagstein F et al (Lancet 1993;342:14426), Packer M et al (N Engl J Med 1996;334:134955), and CIBIS II Investigators and Committee (Lancet1999;353:913).NI = no increase in dose.

The use of class I antiarrhythmic agentsin patients with atrial fibrillation andchronic heart failure substantiallyincreases the risk of mortality

Echocardiogram showing thrombus at left ventricular apex in patient withdilated cardiomyopathy (A=thrombus, B=left ventricle, C=left atrium)

Clinical review


31

Chronic heart failure and ventricular arrhythmiasVentricular arrhythmias are a common cause of death in severeheart failure. Precipitating or aggravating factors should thus beaddressed, including electrolyte disturbance (for example,hypokalaemia, hypomagnesaemia), digoxin toxicity, drugscausing electrical instability (for example, antiarrhythmic drugs,antidepressants), and continued or recurrent myocardialischaemia.

Amiodarone is effective for the symptomatic control ofventricular arrhythmias in chronic heart failure, although moststudies have reported that long term antiarrhythmic treatmentwith amiodarone has a neutral effect on survival. AnArgentinian trial (the GESICA study) of empirical amiodaronein patients with chronic heart failure reported, however, thatactive treatment was associated with a 28% reduction in totalmortality, although this trial included a high incidence ofpatients with nonischaemic heart failure. In contrast, in thesurvival trial of antiarrhythmic therapy in congestive heartfailure (CHFSTAT), amiodarone did not improve overallsurvival, although there was a significant (46%) reduction incardiac death and admission to hospital in the patients withnonischaemic chronic heart failure.

In general, amiodarone should probably be reserved forpatients with chronic heart failure who also have symptomaticventricular arrhythmias. Interest has also developed inimplantable cardioverter defibrillators, which reduce the risk ofsudden death in high risk patients with ventricular arrhythmias(MADIT and AVID studies), although the role of these devicesin patients with chronic heart failure still remains to beestablished.

Summary of drug management in chronic heart failure

Drug class Potential therapeutic role

Diuretics Symptomatic improvement of congestion.Spironolactone improves survival in severe(NYHA class IV) heart failure

Angiotensinconverting enzyme(ACE) inhibitors

Improved symptoms, exercise capacity, andsurvival in patients with asymptomatic andsymptomatic systolic dysfunction

Digoxin Improved symptoms, exercise capacity, andfewer admissions to hospital

Angiotensin IIreceptor antagonists

Treatment of symptomatic heart failure inpatients intolerant to ACE inhibitors*

Nitrates andhydralazine

Improved survival in symptomatic patientsintolerant to ACE inhibitors or angiotensin IIreceptor antagonists*

â Blockers Improved symptoms and survival in stablepatients who are already receiving ACEinhibitors

Amiodarone Prevention of arrhythmias in patients withsymptomatic ventricular arrhythmias

*Recommendations of when these agents might be considered (the use of theseagents has not been addressed in randomised trials of patients intolerant toACE inhibitors).

Key references

x Australia/New Zealand Heart Failure Research CollaborativeGroup. Randomized, placebocontrolled trial of carvedilol inpatients with congestive heart failure due to ischaemic heartdisease. Lancet 1997;349:37580.

x Lip GYH. Intracardiac thrombus formation in cardiac impairment:investigation and the role of anticoagulant therapy. Postgrad Med J1996;72:7318.

x Massie BM, Fisher SG, Radford M, Deedwania PC, Singh BN,Fletcher RD, et al for the CHFSTAT Investigators. Effect ofamiodarone on clinical status and left ventricular function inpatients with congestive heart failure. Circulation 1996;93:212834.

x MERITHF Study Group. Effect of metoprolol CR/XL in chronicheart failure: metoprolol CR/XL randomised intervention trial incongestive heart failure (MERITHF). Lancet 1999;353:20017.

x Doval HC, Nul DR, Grancelli HO, Perrone SV, Bortman GR, CurielR, et al. Randomised trial of lowdose amiodarone in severecongestive heart failure [GESICA trial]. Lancet 1994;344:4938.

x Packer M, Bristow MR, Cohn JN, Colucci WS, Fowler MB, GilbertEM, et al. Effect of carvedilol on morbidity and mortality in patientswith chronic heart failure. N Engl J Med 1996;334:134955.

x Digitalis Investigation Group. The effect of digoxin on mortality andmorbidity in patients with heart failure. N Engl J Med 1997;336:52533.

The survival graph is adapted with permission from Doval et al (Lancet1994;344:4938). The table of inotropic drugs is adapted with permissionfrom Niebauer et al (Lancet 1997;349:966). The table of results of ametaanalysis of effects of â blockers is adapted with permission fromLechat P et al (Circulation 1998;98:118491). The table on doses and titration of â blockers is adapted with permission from Remme WJ (Eur Heart J1997;18:73653).


BMJ 2000;320:4958

1.00

0.9

0.8

0.7

0.6

0.5

0.40 90 180 270 360 450 540 630 720

Days from randomisation

Per

cen

tag

e o

f p

atie

nts

aliv

e

Amiodarone

Control

Survival curves from GESICA trial (see key references box), showingdifference between patients taking amiodarone and controls

Clinical review


32

ABC of heart failureAcute and chronic management strategiesT Millane, G Jackson, C R Gibbs, G Y H Lip

Acute and chronic management strategies in heart failure areaimed at improving both symptoms and prognosis, althoughmanagement in individual patients will depend on theunderlying aetiology and the severity of the condition. It isimperative that the diagnosis of heart failure is accompanied byan urgent attempt to establish its cause, as timely interventionmay greatly improve the prognosis in selected cases—forexample, in patients with severe aortic stenosis.

Management of acute heart failure

AssessmentCommon presenting features include anxiety, tachycardia, anddyspnoea. Pallor and hypotension are present in more severecases: the triad of hypotension (systolic blood pressure < 90mm Hg), oliguria, and low cardiac output constitutes adiagnosis of cardiogenic shock. Severe acute heart failure andcardiogenic shock may be related to an extensive myocardialinfarction, sustained cardiac arrhythmias (for example, atrialfibrillation or ventricular tachycardia), or mechanical problems(for example, acute papillary muscle rupture or postinfarctionventricular septal defect).

Severe acute heart failure is a medical emergency, andeffective management requires an assessment of the underlyingcause, improvement of the haemodynamic status, relief ofpulmonary congestion, and improved tissue oxygenation.Clinical and radiographic assessment of these patients providesa guide to severity and prognosis: the Killip classification hasbeen developed to grade the severity of acute and chronic heartfailure.

TreatmentBasic measures should include sitting the patient in an uprightposition with high concentration oxygen delivered via a facemask. Close observation and frequent reassessment arerequired in the early hours of treatment, and patients with acutesevere heart failure, or refractory symptoms, should bemonitored in a high dependency unit. Urinary catheterisationfacilitates accurate assessment of fluid balance, while arterialblood gases provide valuable information about oxygenationand acidbase balance. The “base excess” is a guide to actualtissue perfusion in patients with acute heart failure: a worsening(more negative) base excess generally indicates lactic acidosis,which is related to anaerobic metabolism, and is a poorprognostic feature. Correction of hypoperfusion will correct themetabolic acidosis; bicarbonate infusions should be reserved foronly the most refractory cases.

Intravenous loop diuretics, such as frusemide (furosemide),induce transient venodilatation, when administered to patientswith pulmonary oedema, and this may lead to symptomaticimprovement even before the onset of diuresis. Loop diureticsalso increase the renal production of vasodilator prostaglandins.This additional benefit is antagonised by the administration ofprostaglandin inhibitors, such as nonsteroidalantiinflammatory drugs, and these agents should be avoidedwhere possible. Parenteral opiates or opioids (morphine ordiamorphine) are an important adjunct in the management ofsevere acute heart failure, by relieving anxiety, pain, and distress

Survival rates (%) compared with chronic heart failure

At 1 year At 2 years At 3 years

Breast cancer 88 80 72

Prostate cancer 75 64 55

Colon cancer 56 48 42

Heart failure 67 41 24

Killip classification

Class Clinical features

Hospitalmortality

(%)

Class I No signs of left ventricular dysfunction 6

Class II S3 gallop with or without mild tomoderate pulmonary congestion

30

Class III Acute severe pulmonary oedema 40

Class IV Shock syndrome 8090

Chest x ray film in patient with acute pulmonary oedema

Basic measures

Sit patient upright

High dose oxygen

Initial drug treatment

Intravenous loop diuretics

Intravenous opiates/opioids

(morphine/diamorphine)

Intravenous, buccal, or

sublingual nitrates

Corrects hypoxia

Cause venodilatation and diuresis

Reduce anxiety and preload

(venodilatation)

Reduce preload and afterload, ischaemia

and pulmonary artery pressures

Acute heart failure: basic measures and initial drug treatment

Clinical review



33

and reducing myocardial oxygen demand. Intravenous opiatesand opioids also produce transient venodilatation, thusreducing preload, cardiac filling pressures, and pulmonarycongestion.

Nitrates (sublingual, buccal, and intravenous) may alsoreduce preload and cardiac filling pressures and are particularlyvaluable in patients with both angina and heart failure. Sodiumnitroprusside is a potent, directly acting vasodilator, which isnormally reserved for refractory cases of acute heart failure.

Short term inotropic supportIn cases of severe refractory heart failure in which the cardiacoutput remains critically low, the circulation can be supportedfor a critical period of time with inotropic agents. For example,dobutamine and dopamine have positive inotropic actions,acting on the â1 receptors in cardiac muscle. Phosphodiesteraseinhibitors (for example, enoximone) are less commonly used,and long term use of these agents is associated with increasedmortality. Intravenous aminophylline is now rarely used fortreating acute heart failure. Inotropic agents in general increasethe potential for cardiac arrhythmias.

Chronic heart failureChronic heart failure can be “compensated” or“decompensated.” In compensated heart failure, symptoms arestable, and many overt features of fluid retention andpulmonary oedema are absent. Decompensated heart failurerefers to a deterioration, which may present either as an acuteepisode of pulmonary oedema or as lethargy and malaise, areduction in exercise tolerance, and increasing breathlessnesson exertion. The cause or causes of decompensation should beconsidered and identified; they may include recurrentischaemia, arrhythmias, infections, and electrolyte disturbance.Atrial fibrillation is common, and poor control of ventricularrate during exercise despite adequate control at rest should beaddressed.

Common features of chronic heart failure includebreathlessness and reduced exercise tolerance, andmanagement is directed at relieving these symptoms andimproving quality of life. Secondary but important objectivesare to improve prognosis and reduce hospital admissions.

Initial managementNonpharmacological and lifestyle measures should beaddressed. Loop diuretics are valuable if there is evidence offluid overload, although these may be reduced once salt andwater retention has been treated. Angiotensin convertingenzyme inhibitors should be introduced at an early stage, in theabsence of clear contraindications. Angiotensin II receptorantagonists are an appropriate alternative in patients who areintolerant to angiotensin converting enzyme inhibitors. â

Blockers (carvedilol, bisoprolol, metoprolol) are increasinglyused in stable patients, although these agents require low doseinitiation and cautious titration under specialist supervision.Oral digoxin has a role in patients with left ventricular systolicimpairment, in sinus rhythm, who remain symptomatic despiteoptimal doses of diuretics and angiotensin converting enzymeinhibitors. Warfarin should be considered in patients with atrialfibrillation.

Severe congestive heart failureDespite conventional treatment with diuretics and angiotensinconverting enzyme inhibitors, hospital admission may benecessary in severe congestive heart failure. Fluid restriction is

Intravenous inotropes and circulatory assist devices

x Short term support with intravenous inotropes or circulatory assistdevices, or with both, may temporarily improve haemodynamicstatus and peripheral perfusion

x Such support can act as a bridge to corrective valve surgery orcardiac transplantation in acute and chronic heart failure

Management of chronic heart failure

General advicex Counselling—about symptoms and compliancex Social activity and employmentx Vaccination (influenza, pneumococcal)x Contraception

General measuresx Diet (for example, reduce salt and fluid intake)x Stop smokingx Reduce alcohol intakex Take exercise

Treatment options—pharmacologicalx Diuretics (loop and thiazide)x Angiotensin converting enzyme inhibitorsx â Blockersx Digoxinx Spironolactonex Vasodilators (hydralazine/nitrates)x Anticoagulationx Antiarrhythmic agentsx Positive inotropic agents

Treatment options—devices and surgeryx Revascularisation (percutaneous transluminal coronary angioplasty

and coronary artery bypass graft)x Valve replacement (or repair)x Pacemaker or implantable cardiodefibrillatorx Ventricular assist devicesx Heart transplantation

Supervised exercise programmes are ofproved benefit, and regular exerciseshould be encouraged in patients withchronic stable heart failure

Advanced management

Assisted ventilation

Circulatory assist devices

Second line drug treatment

Inotropes: β agonists

(dobutamine)

Dopamine (low dose)

Inotropes: phosphodiesterase

inhibitors (enoximone)

Intravenous aminophylline

Reduces myocardial oxygen demand;

improves alveolar ventilation

Give mechanical support

Increase myocardial contractility

Increases renal perfusion, sodium

excretion, and urine flow

Increase myocardial contractility and

venodilatation

Weak inotropic effect, diuretic effect,

bronchodilating effect

Acute heart failure: second line drug treatment and advanced management

Clinical review



34

important—fluid intake should be reduced to 11.5 litres/24 h,and dietary salt restriction may be helpful.

Short term bed rest is valuable until signs and symptomsimprove: rest reduces the metabolic demand and increasesrenal perfusion, thus improving diuresis. Although bed restpotentiates the action of diuretics, it increases the risk of venousthromboembolism, and prophylactic subcutaneous heparinshould be considered in immobile inpatients. Fullanticoagulation is not advocated routinely unless concurrentatrial fibrillation is present, although it may be considered inpatients with very severe impairment of left ventricular systolicfunction, associated with significant ventricular dilatation.Intravenous loop diuretics may be administered to overcomethe short term problem of gut oedema and reduced absorptionof tablets, and these may be used in conjunction with an oralthiazide or thiazidelike diuretic (metolazone). Low dosespironolactone (25 mg) improves morbidity and mortality insevere (New York Heart Association class IV) heart failure,when combined with conventional treatment (loop diureticsand angiotensin converting enzyme inhibitors). Potassiumconcentrations should be closely monitored after the additionof spironolactone.

Special procedures

Intraaortic balloon pumping and mechanical devicesIntraaortic balloon counterpulsation and left ventricular assistdevices are used as bridges to corrective valve surgery, cardiactransplantation, or coronary artery bypass surgery in thepresence of poor cardiac function. Mechanical devices areindicated if (a) there is a possibility of spontaneous recovery (forexample, peripartum cardiomyopathy, myocarditis) or (b) as abridge to cardiac surgery (for example, ruptured mitralpapillary muscle, postinfarction ventricular septal defect) ortransplantation. Intraaortic balloon counterpulsation is themost commonly used form of mechanical support.

Weighing the patient daily is valuable inmonitoring the response to treatment

Education, counselling, and support

x A role is emerging for heart failure liaison nurses in educating andsupporting patients and their families, promoting long termcompliance, and supervising treatment changes in the community

x Depression is common, underdiagnosed, and often undertreated;counselling is therefore important for patients and families, and thenewer antidepressants (particularly the selective serotonin reuptakeinhibitors) seem to be well tolerated and are useful in selectedpatients

Left ventricularassist device

Symptomatic Asymptomatic

Add loop diuretic (eg frusemide)

Consider β blocker* in patients with chronic, stable condition

Persisting clinical features of heart failure

Options

Treatment of left ventricular systolic dysfunction

• Confirm diagnosis by echocardiography

• If possible, discontinue aggravating drugs (eg non-steroidal anti-inflammatory drugs)

• Address non-pharmacological and lifestyle measures

Angiotensin converting

enzyme inhibitorAngiotensin converting enzyme inhibitor

• Optimise dose of loop diuretic• Low dose spironolactone (25mg once a day)• Digoxin• Combine loop and thiazide diuretics• Oral nitrates/ hydralazine

• Digoxin• β blocker (if not already given)• Warfarin

Atrial fibrillation

Options • β blocker (if not already given)• Oral nitrates• Calcium antagonist (eg amlodipine)

Angina

Options

* Initial low dose (eg carvedilol, bisoprolol, metoprolol) with cautious titration under expert supervision

Consider specialist referral in patients with atrial fibrillation (electrical cardioversion or other antiarrhythmic agents - eg amiodarone - may be indicated), angina (coronary angiography and revascularisation may be indicated), or persistent or severe symptoms

In the United Kingdom carvedilol is licensedfor mild to moderate symptoms and bisoprololfor moderate to severe congestive heart failure

Example of management algorithm for left ventricular dysfunction

Clinical review



35

Revascularisation and other operative strategiesImpaired ventricular function in itself is not an absolutecontraindication to cardiac surgery, although the operative risksare increased. Ischaemic heart disease is the most commonprecursor of chronic heart failure in Britain: coronaryischaemia should be identified and revascularisation consideredwith coronary artery bypass surgery or occasionallypercutaneous coronary angioplasty. The concept of“hibernating” myocardium is increasingly recognised, althoughthe most optimal and practical methods of identifyinghibernation remain open to debate. Revascularisation ofhibernating myocardium may lead to an improvement in theoverall left ventricular function.

Correction of valve disease, most commonly in severe aorticstenosis or mitral incompetence (not secondary to leftventricular dilatation), relieves a mechanical cause of heartfailure; closure of an acute ventricular septal defect or mitralvalve surgery for acute mitral regurgitation, complicating amyocardial infarction, may be lifesaving. Surgical excision of aleft ventricular aneurysm (aneurysectomy) is appropriate inselected cases. Novel surgical procedures such as extensiveventricular reduction (Batista operation) and cardiomyoplastyhave been associated with successful outcome in a smallnumber of patients, although the high mortality, and the limitedevidence of substantial benefit, has restricted the widespread useof these procedures.

Cardiac transplantationThe outcome in cardiac transplantation is now good, with longterm improvements in survival and quality of life in patientswith severe heart failure. However, although the demand forcardiac transplantation has increased over recent years, thenumber of transplant operations has remained stable, owingprimarily to limited availability of donor organs.

The procedure now carries a perioperative mortality of lessthan 10%, with approximate one, five, and 10 year survival ratesof 92%, 75%, and 60% respectively (much better outcomes thanwith optimal drug treatment, which is associated with a one yearmortality of 3050% in advanced heart failure). Cardiactransplantation should be considered in patients with anestimated one year survival of < 50%. Well selected patientsover 5560 years have a survival rate comparable to those ofyounger patients. Patients need strong social and psychologicalsupport; transplant liaison nurses are valuable in this role.

The long term survival of the transplanted human heart iscompromised by accelerated graft atherosclerosis which resultsin small vessel coronary artery disease and an associateddeterioration in left ventricular performance. This can occur asearly as three months and is the major cause of graft loss afterthe first year. The antirejection regimens currently used mayresult in an acceleration of preexisting atherosclerotic vasculardisease—hence the exclusion of patients who already havesignificant peripheral vascular disease. Rejection is now a lessserious problem, with the use of cyclosporin and otherimmunosuppressant agents.

Nevertheless, the supply of donors limits the procedure. TheEurotransplant database (19905) indicates that 25% of patientslisted for transplantation die on the waiting list, with 60%receiving transplants at two years (most within 12 months).Although ventricular assist devices may be valuable during thewait for transplantation, the routine use of xenotransplants isunlikely in the short or medium term.

The graph showing cardiac transplantations worldwide is adapted withpermission from Hosenpud et al (J Heart Lung Transplant 1998;17:6568).The table showing survival rates is adapted from Hobbs (Heart 1999;82(suppl IV):IV810).

Indications and contraindications to cardiac transplantationin adults

Indicationsx End stage heart failure—for example, ischaemic heart disease and

dilated cardiomyopathyx Rarely, restrictive cardiomyopathy and peripartum cardiomyopathyx Congenital heart disease (often combined heartlung

transplantation required)

Absolute contraindicationsx Recent malignancy (other than basal cell and squamous cell

carcinoma of the skin)x Active infection (including HIV, Hepatitis B, Hepatitis C with liver

disease)x Systemic disease which is likely to affect life expectancyx Significant pulmonary vascular resistance

Relative contraindicationsx Recent pulmonary embolismx Symptomatic peripheral vascular diseasex Obesityx Severe renal impairmentx Psychosocial problems—for example, lack of social support, poor

compliance, psychiatric illnessx Age (over 6065 years)

Key references

x Dargie HJ, McMurray JJ. Diagnosis and management of heartfailure. BMJ 1994;308:3218.

x ACC/AHA Task Force Report. Guidelines for the evaluation andmanagement of heart failure. J Am Coll Cardiol 1995;26:137698.

x Hunt SA. Current status of cardiac transplantation. JAMA1998;280:16928.

x Remme WJ. The treatment of heart failure. The Task Force of theWorking Group on Heart Failure of the European Society ofCardiology. Eur Heart J 1997;18:73653.

T Millane is consultant cardiologist in the department of cardiology,City Hospital, Birmingham; G Jackson is consultant cardiologist in thedepartment of cardiology at Guy’s and St Thomas’s Hospital, London.


BMJ 2000;320:55962

Year

No

of

tran

spla

nta

tio

ns

0

1000

1500

2000

2500

3000

3500

4000

4500Transplantations

500

Mea

n a

ge

of

do

no

rs (

year

s)

22

24

25

26

27

28

29

30

31

23

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

Age of donors

Number of heart transplantations worldwide and mean age of donors

Clinical review



36

ABC of heart failureHeart failure in general practiceF D R Hobbs, R C Davis, G Y H Lip

Management of heart failure in general practice has beenhampered by difficulties in diagnosing the condition and byperceived difficulties in starting and monitoring treatment inthe community. Nevertheless, improved access to diagnostictesting and increased confidence in the safety of treatmentshould help to improve the primary care management of heartfailure. With improved survival and reduced admission rates(achieved by effective treatment) and a reduction in numbers ofhospital beds, the community management of heart failure islikely to become increasingly important and the role of generalpractitioners even more crucial.

Diagnostic accuracyHeart failure is a difficult condition to diagnose clinically, andhence many patients thought to have heart failure by theirgeneral practitioners may not have any demonstrableabnormality of cardiac function on objective testing.

A study from Finland reported that only 32% of patientssuspected of having heart failure by primary care doctors haddefinite heart failure (as determined by a clinical andradiographic scoring system). A recent study in the UnitedKingdom showed that only 29% of 122 patients referred to a“rapid access” clinic with a new diagnosis of heart failure fullymet the definition of heart failure approved by the EuropeanSociety of Cardiology—that is, appropriate symptoms, objectiveevidence of cardiac dysfunction, and response to treatment ifdoubt remained.

Similar findings have been reported in theechocardiographic heart of England screening (ECHOES)study, in which only about 22% of the patients with a diagnosisof heart failure in their general practice records had definiteimpairment of left ventricular systolic function onechocardiography, with a further 16% having borderlineimpairment. In addition, 23% had atrial fibrillation, with overhalf of these patients having normal left ventricular systoliccontraction. Finally, a minority of patients may have clinicalheart failure with normal systolic contraction and abnormaldiastolic function; management of such patients with diastolicdysfunction is very different from those with impaired systolicfunction.

Open access echocardiography anddiagnosisOwing to the noninvasive nature of echocardiography, its highacceptability to patients, and its usefulness in assessingventricular size and function, as well as valvar heart disease,many general practitioners now want direct access toechocardiography services for their patients. Although openaccess echocardiography services are available in some districtsin Britain, many specialists still have reservations aboutintroducing such services because of financial and staffing issuesand concern that general practitioners would have difficultyinterpreting technical reports. The cost of echocardiography(£50 to £70 per patient) is relatively small, however, comparedwith the cost of expensive treatment for heart failure that maynot be needed. The cost is also small compared with the costs of

Heart failure affects at least 20 patientson the average general practitioner’s list

Recent studies have shown that withappropriate education of generalpractitioners the workload of an openaccess echocardiography service can bemanageable

Clinical assessment of patient,

history, and hospital records

together suggest heart failure

Echocardiography shows

moderate or severe left

ventricular dysfunction?

Heart failure: start

angiotensin converting

enzyme inhibitor

Probability of heart

failure high: are you

confident of diagnosis?

Refer for further

investigation

Heart failure

unlikely

Electrocardiogram abnormal?

(Q waves, left bundle

branch block)

Chest x ray film shows

pulmonary congestion

or cardiomegaly?

Documented previous

myocardial infarction?

Remaining unexplained

indication of heart failure?

Not available

No, inconclusive,

or not known

No, inconclusive,

or not known

No, inconclusive,

or not known

Yes

Yes

No Yes

Yes

Yes

Yes

No

No

Diagnostic algorithm for suspected heart failure in primary care. Based onguidance from the north of England evidence based guideline developmentproject (see key references box)

Clinical review

626 BMJ VOLUME 320 4 MARCH 2000 www.bmj.com


37

hospital admission, which may be avoided by appropriate, earlytreatment of heart failure.

One approach may be to refer only patients with abnormalbaseline investigations as heart failure is unlikely if theelectrocardiogram and chest x ray examination are normal andthere are no predisposing factors for heart failure—for example,previous myocardial infarction, angina, hypertension, anddiabetes mellitus. Requiring general practitioners to performelectrocardiography and arrange chest radiography, as acomplement to careful assessment of the risk factors for heartfailure, is likely to reduce substantially the number ofinappropriate referrals to an open access echocardiographyservice.

Role of natriuretic peptidesGiven the difficulties in diagnosing heart failure on clinicalgrounds alone, and current limited access to echocardiographyand specialist assessment, the possibility of using a blood test ingeneral practice to diagnose heart failure is appealing.Determining plasma concentrations of brain natriureticpeptide, a hormone found at an increased level in patients withleft ventricular systolic dysfunction, may be one option. Such ablood test has the potential to screen out patients in whomheart failure is extremely unlikely and identify those in whomthe probability of heart failure is high—for example, in patientswith suspected heart failure who have low plasmaconcentrations of brain natriuretic peptide, the heart is unlikelyto be the cause of the symptoms, whereas those who havehigher concentrations warrant further assessment.

Primary prevention and early detectionGeneral practitioners have a vital role in the early detection andtreatment of the main risk factors for heart failure—namely,hypertension and ischaemic heart disease—and othercardiovascular risk factors, such as smoking andhyperlipidaemia. The Framingham study has shown a decline inhypertension as a risk factor for heart failure over the years,which probably reflects improvements in treatment. Ischaemicheart disease, however, remains very common. Aspirin, â

blockers, and lipid lowering treatment, as well as smokingcessation, can reduce progression to myocardial infarction inpatients with angina, and â blockers may also reduce ischaemicleft ventricular dysfunction. Early detection of left ventriculardysfunction in “high risk” asymptomatic patients—for example,those who have already had a myocardial infarction or whohave hypertension or atrial fibrillation—and treatment withangiotensin converting enzyme inhibitors can minimise theprogression to symptomatic heart failure.

Startingandmonitoringdrug treatmentBoth hospital doctors and general practitioners used to beconcerned about the initiation of angiotensin convertingenzyme inhibitors outside hospital. It is now accepted, however,that most patients with heart failure can safely be established onsuch treatment without needing hospital admission. Theprevious concern—over first dose hypotension—was heightenedby the initial experience of large doses of captopril, especially inthose with severe heart failure, who are at greater risk ofproblems. Patients with mild or moderate heart failure, whohave normal renal function and a systolic blood pressure over100 mm Hg and who have stopped taking diuretics for at least24 hours rarely have problems, especially if the first dose of an

Open access services have provedpopular and are likely to become evenmore common; indeed,echocardiographic screening of patientsin the high risk categories may well bejustified and cost effective

Sensitivity and specificity of brain natriuretic peptides indiagnosis of heart failure

New diagnosis ofheart failure

(primary care)Left ventricular

systolic dysfunction

Sensitivity 97% 77%

Specificity 84% 87%

Positive predictivevalue 70% 16%

Starting angiotensin converting enzyme inhibitors in chronicheart failure in general practice

x Measure blood pressure and determine electrolytes and creatinineconcentrations before treatment

x Consider referring “high risk” patients to hospital for assessmentand supervised start of treatment

x Angiotensin converting enzyme inhibitors should be used withsome caution in patients with severe peripheral vascular diseasebecause of the possible association with atherosclerotic renal arterystenosis

x Doses should be gradually increased over two to three weeks,aiming to reach the doses used in large clinical trials

x Blood pressure and electrolytes or renal chemistry should bemonitored after start of treatment, initially at one week then lessfrequently depending on the patient and any abnormalitiesdetected

Detect and treat hypertensionOther cardiovascular disease

prevention strategies

(eg avoid smoking, lipid lowering)

Angiotensin converting enzyme

inhibitors in asymptomatic left

ventricular dysfunction

Prevent progression to

symptomatic heart failure

Strategies for preventing progression to symptomatic heart failure in highrisk asymptomatic patients

Clinical review

627BMJ VOLUME 320 4 MARCH 2000 www.bmj.com


38

angiotensin converting enzyme inhibitor is taken at night,before going to bed.

Heart failure clinicsDedicated heart failure clinics within general practices, run by adoctor or nurse with an interest in the subject, have thepotential to improve the care of patients with the condition, asthey have for other chronic conditions, such as diabetes.

Blood should be taken for electrolytes and renal chemistryat least every 12 months, but more frequently in new cases andwhen drug treatment has been changed or results have beenabnormal. The clinics should be used to educate patients abouttheir condition, particularly in relation to their treatment, withmessages being reinforced and drug treatment simplified andrationalised where appropriate. Patients whose condition isdeteriorating may be referred for specialist opinion.

Variables that should be monitored in patients withestablished heart failure comprise changes in symptoms andseverity (New York Heart Association classification); weight;blood pressure; and signs of fluid retention or excessive diuresis.

Impact of heart failure on thecommunityAfter a patient is diagnosed as having heart failure, substantialmonitoring by the general practitioner is required. In oursurvey of heart failure in three general practices from the westof Birmingham, 44% of general practice consultations (average2.6 visits per patient) took place within three months of the firstdiagnosis of heart failure, 23% were at three to six months (1.4visits per patient), and 33% were at six to 12 months (2.0 visitsper patient). Such management requires regular supervisionand audit.

Relevance to hospital practiceIn our survey of acute hospital admissions of patients with heartfailure to a city centre hospital, the median duration of stay was8 (range 196) days, with 20% inpatient mortality. Clinicalvariables associated with an adverse prognosis include thepresence of atrial fibrillation, poor exercise tolerance, electrolyteabnormalities, and the presence of coronary artery disease.Angiotensin converting enzyme inhibitors were prescribed inonly 51% of heart failure patients on discharge; after the firstdiagnosis of heart failure, the average number of hospitalattendances (inpatient and outpatient) in the first 12 monthswas 3.2 visits per patient, with an average of 6.0 general practiceconsultations per patient. However, 44% of hospital attendances(1.4 visits per patient) took place within three months ofdiagnosis, 33% were at three to six months (1.0 visits perpatient), and 23% were at 612 months (0.74 visits per patient).

These figures represent the collective burden of heart failureon hospital practice. Indeed, about 200 000 people in theUnited Kingdom require admission to hospital for heart failureeach year.

Specialist nurse supportThe important role of nurses in the management of heartfailure has been relatively neglected in Britain. In the UnitedStates the establishment of a nurse managed heart failure clinicin South Carolina resulted in a reduction in readmissions of 4%

Conditions indicating that referral to a specialist is necessary

x Diagnosis in doubt or when specialist investigation andmanagement may help

x Significant murmurs and valvar heart diseasex Arrhythmias—for example, atrial fibrillationx Secondary causes—for example, thyroid diseasex Severe left ventricular impairment—for example, ejection fraction

< 20%x Preexisting (or developing) metabolic abnormalities—for example,

hyponatraemia (sodium < 130 mmol/l) and renal impairmentx Severe associated vascular disease—for example, caution with

angiotensin converting enzyme inhibitors in case of coexistingrenovascular disease

x Relative hypotension (systolic blood pressure < 100 mm Hg beforestarting angiotensin converting enzyme inhibitors)

x Poor response to treatment

Examples of topics for audit of heart failure management ingeneral practice

Means of diagnosisHas left ventricular function been assessed, by echocardiography orother means?

Appropriateness of treatmentAre all appropriate patients taking angiotensin converting enzymeinhibitors (unless there is a documented contraindication)? Havedoses been increased where possible to those used in the large clinicaltrials?

Monitoring treatmentWere blood pressure and renal function recorded before and afterstart of angiotensin converting enzyme inhibitors, and at intervalssubsequently?

Causes of readmission in patients with heart failure

x Anginax Infectionsx Arrhythmiasx Poor compliancex Inadequate drug treatmentx Iatrogenic factorsx Inadequate discharge planning or follow upx Poor social support

Admissions with heart failure over six months to a districtgeneral hospital serving a multiracial population

Presentation (%) Associated medical history (%)

Pulmonary oedema (52) Ischaemic heart disease (54)

Congestive heart failure,with fluid overload (32)

Hypertension (34)

Myocardial infarctionand heart failure (9)

Valve disease (12); previous stroke (10)

Associated atrialfibrillation (29)

Diabetes mellitus (19); peripheral vasculardisease (13); cardiomyopathy (1)

Population of 300 000 (7451 admissions; 348 (5%) had heart failure (mean age73 years)).

Clinical review

628 BMJ VOLUME 320 4 MARCH 2000 www.bmj.com


39

and in length of hospital stay of almost two days. In anotherNorth American study a comprehensive, multidisciplinaryapproach to heart failure management, including supervisionby nurses, resulted in a significant (56%) reduction inreadmissions and hospital stay, with a trend towards reducedmortality. Quality of life scores also improved in theintervention group. A more dramatic result was obtained in astudy from Adelaide, Australia, where multidisciplinaryintervention resulted in a 20% reduction in mortality.

Nurse management of heart failure has implications for theprovision of care in patients with chronic heart failure, sharingthe increasing burden of heart failure. Specialist nurses wouldprovide advice, information, and support to patients with heartfailure and to their families and would ensure that the besttreatment is given. The potential benefits are substantial, withreduced hospital admission rates, improved quality of life, andlower costs.

Economic considerationsWith an increasingly elderly population, the prevalence of heartfailure could have increased by as much as 70% by the year2010. Heart failure currently accounts for 12% of totalspending on health care in Europe and in the United States. In1993 in the United Kingdom, heart failure cost the NHS£360m a year; the figure now is probably closer to £600m,equivalent to 12% of the total NHS budget, and hospitaladmissions account for 6070% of this expenditure. Admissionsfor heart failure have been increasing and are expected toincrease further. Preventing disease progression, hencereducing the frequency and duration of admissions, is thereforean important objective in the treatment of heart failure in thefuture.

The table on sensitivity and specificity is based on information in Cowie etal (Lancet 1997;350:134953) and McDonagh et al (Lancet 1998;351:913).The table showing admissions with heart failure to a district generalhospital is adapted with permission from Lip et al (Int J Clin Prac 1997;51:2237). The table showing the economic costs of heart failure is publishedwith permission from McMurray et al (Eur Heart J 1993;14(suppl):133).

R C Davis is clinical research fellow and F D R Hobbs is professor inthe department of primary care and general practice, University ofBirmingham.


BMJ 2000;320:6269

Economic cost of heart failure to NHS in UK, 19901

Total cost (£m) % of total cost

General practice visits 8.3 2.5

Referrals to hospital fromgeneral practice

8.2 2.4

Other outpatient attendances 31.8 9.4

Inpatient stay 213.8 63.5

Diagnostic tests 45.6 13.5

Drugs 22.1 6.6

Surgery 7.2 2.1

Total 337.0 100

Heart failure is likely to continue tobecome a major public health problem inthe coming decades; new and bettermanagement strategies are necessary,including risk factor interventions, forpatients at risk of developing heartfailure

Key references

x Eccles M, Freemantle N, Mason J, for the North of EnglandGuideline Development Group. North of England evidence baseddevelopment project: guideline for angiotensin converting enzymeinhibitors in primary care management of adults with symptomaticheart failure. BMJ 1998;316:136975.

x Francis CM, Caruana L, Kearney P, Love M, Sutherland GR, StarkeyIR, et al. Open access echocardiography in the management ofheart failure in the community. BMJ 1995;310:6346.

x Lip GYH, Sarwar S, Ahmed I, Lee S, Kapoor V, Child D, et al. Asurvey of heart failure in general practice. The west Birminghamheart failure project. Eur J Gen Pract 1997;3:859.

x Remes J, Miettinen H, Reunanen A, Pyorala K. Validity of clinicaldiagnosis of heart failure in primary health care. Eur Heart J1991;12:31521.

x Rich MW, Beckham V, Wittenberg C, Leven CL, Freedland KE,Carney RM, et al. A multidisciplinary intervention to prevent thereadmission of elderly patients with congestive heart failure. N EnglJ Med 1995:333:11905.

x Stewart S, Vandenbroek AJ, Pearson S, Horowitz JD. Prolongedbeneficial effects of homebased intervention on unplannedreadmissions and mortality among patients with congestive heartfailure. Arch Intern Med 1999;159:25761.

100Home based intervention (n=49)Usual care (n=48)

P= 0.049

90

80

70

60

500 4 8 12 16 20 24 28 32 36 40 4844

Week of study follow up

Sur

viva

l (%

)

52 56 60 64 68 7672 80

Cumulative survival curves from the Adelaide nurse intervention study: 18month follow up (see Stewart et al, key references box at end of article)

Nurse specialising in heart failure

Educating patient and family

Monitoring weight and blood tests(renal chemistry and electrolytes)

Promoting long term compliance

Implementing treatment algorithms

Role of specialist nurse in management of patients with heart failure

Clinical review

629BMJ VOLUME 320 4 MARCH 2000 www.bmj.com


40

Documents

ABC of Heart Failure - R. Davis, F. Hobbs, G. Lip (BMJ) WW