Letters

Investigation of left ventricular dysfunction in acute dyspnoea

BMJ 1997; 315 doi: https://doi.org/10.1136/bmj.315.7108.604 (Published 06 September 1997) Cite this as: BMJ 1997;315:604

Access to echocardiography facilitates informed management

  1. Pamela Crawford, Specialist registrara,
  2. Anne Hendry, Consultanta
  1. a Department of Medicine for the Elderly, West Glasgow Hospitals University NHS Trust, Glasgow G12 OYN
  2. b MRC Clinical Research Initiative in Heart Failure, University of Glasgow, Glasgow G12 8QQ
  3. c Department of Medicine, University of Auckland, Auckland, New Zealand
  4. d Ninewells Hospital and Medical School, Dundee DD1 9SY

    Editor—Neil D Gillespie and colleagues found that clinical assessment detected left ventricular systolic function (sensitivity 81%) and that chest radiography improved specificity from 47% to 95%.1 The authors therefore concluded that echocardiography is not essential for patients with acute dyspnoea and clear evidence of heart failure but no murmur. They are right, however, to be cautious in extrapolating their findings to other clinical settings. Patients with important respiratory disease were excluded (they were admitted to a separate chest unit), and their subjects were the younger patients, often with overt cardiac disease or disease of a single system, referred to an acute medical admissions unit. The incidence of heart failure increases with age, and the high prevalence of diastolic dysfunction2 and comorbidity presents difficulties for accurate diagnosis in elderly patients with heart failure.

    We determined the validity of a clinical diagnosis of systolic dysfunction and assessed the contribution made by echocardiography to patient management in a prospective study of consecutive patients with decompensated heart failure admitted to a geriatric assessment unit. For 61 patients (15 men; age 71-96 (mean 82)) with heart failure (two major or one major and two minor Framingham criteria) a consultant indicated proposed cardiac diagnosis—systolic heart failure, diastolic dysfunction, cor pulmonale, aortic/mitral valve disease—and management based on the patient's history, and clinical and radiographic findings.

    After echocardiography the main cardiac diagnoses were systolic dysfunction (n=42), mitral stenosis (6), diastolic dysfunction (5), cor pulmonale (3), aortic stenosis (1), constrictive pericarditis (1), and normal cardiac function (3). In identifying systolic dysfunction clinical diagnosis had a sensitivity of 93% and specificity of 32% (positive predictive value 0.75, negative predictive value 0.66); radiological features of pulmonary congestion or oedema had a sensitivity of 76% and a specificity of 42% (0.74, 0.44); a clinical or radiological diagnosis had a sensitivity of 95% and specificity of 16% (0.71, 0.60); and both clinical and radiological diagnosis had a sensitivity of 74% and specificity of 58% (0.79, 0.50).

    Patients with heart failure are often inadequately investigated and treatment may be suboptimal.3 The authors suggest that because of limited echocardiography resources patients in whom the diagnostic uncertainty is greatest should be targeted. Elderly patients presenting with heart failure in a geriatric medical setting have a range of cardiac diagnoses and comorbidity. In contrast to Gillespie and colleagues we found that a combined clinical and radiological assessment lacked predictive accuracy in this group. Access to echocardiography, however, facilitates informed management and may optimise the use of angiotensin converting enzyme inhibitors.

    References

    1. 1.
    2. 2.
    3. 3.

    A 100% sensitivity would be difficult to achieve

    1. Andrew Davie, Lecturer in cardiologyb,
    2. John McMurray, Consultant cardiologistb
    1. a Department of Medicine for the Elderly, West Glasgow Hospitals University NHS Trust, Glasgow G12 OYN
    2. b MRC Clinical Research Initiative in Heart Failure, University of Glasgow, Glasgow G12 8QQ
    3. c Department of Medicine, University of Auckland, Auckland, New Zealand
    4. d Ninewells Hospital and Medical School, Dundee DD1 9SY

      Editor—We disagree with the premise and at least one of the conclusions in the recent article by Neil D Gillespie and colleagues.1 The premise that the availability of inpatient echocardiography is limited is based on the observation that relatively few patients with acute dyspnoea have echocardiography at present.2 This is not so much an argument for more directed selection of patients for echocardiography as an argument for wider application of echocardiography. We cannot therefore agree with the conclusion that echocardiography should be reserved for cases in which the diagnostic doubt is greatest. This would serve merely to perpetuate the present lamentable state of affairs.

      We agree that the presence of isolated lung crepitations is a poor predictor of left ventricular systolic dysfunction. The sensitivity of this sign can be as low as 13%3 and the specificity can be as low as 35%.4 Our own examination of signs and symptoms in heart failure has shown that the best predictor of left ventricular systolic dysfunction is a displaced apex beat on examination (sensitivity 66%, specificity 96%, positive predictive value 75%, negative predictive value 94%).5

      We are impressed with the apparent sensitivity of clinical examination in detecting left ventricular systolic dysfunction and unsurprised by the greater specificity conferred by the addition of chest radiography or electrocardiography in Gillespie and colleagues' study. We are puzzled, however, by table 2, that shows greater sensitivity for these combinations. This is not credible. Greater specificity is achieved by greater exclusivity, and greater sensitivity is achieved by greater inclusivity. It is impossible for the combination of clinical examination and electrocardiography and chest radiography to achieve 100% sensitivity when each of them alone has less than 100% sensitivity. It is possible that clinical examination or electrocardiography or chest radiography (or a combination thereof) might achieve 100% sensitivity, but this would clearly be less specific. There is an error here.

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.

      Echocardiography is important in the management of heart failure

      1. Warwick Bagg, Research fellowc,
      2. Robert N Doughty, Senior research fellowc,
      3. Gillian A Whalley, Research fellowc,
      4. Gary Gordon, Consultant cardiologistc,
      5. Norman Sharpe, Headc
      1. a Department of Medicine for the Elderly, West Glasgow Hospitals University NHS Trust, Glasgow G12 OYN
      2. b MRC Clinical Research Initiative in Heart Failure, University of Glasgow, Glasgow G12 8QQ
      3. c Department of Medicine, University of Auckland, Auckland, New Zealand
      4. d Ninewells Hospital and Medical School, Dundee DD1 9SY

        Editor—Neil D Gillespie and colleagues conclude that clinical examination and simple cardiac investigations are adequate to diagnose left ventricular systolic dysfunction in patients with acute dyspnoea and that, unless a cardiac murmur is detected, echocardiography should be reserved for cases in which the diagnosis is most doubtful.1 This recommendation is based on a small study and is contrary to the guidelines of the European Society of Cardiology.2 In contradiction, Gillespie and colleagues point out that the management of heart failure requires echocardiographic assessment of left ventricular function. This was shown by the 17 of the 59 patients (group 1 and 2) who did not have systolic dysfunction as their main pathology.

        Patients admitted to hospital with heart failure are mostly severely affected and are at high risk of subsequent readmission. Consequently, discharging these patients back to primary care without echocardiography would deny them and their general practitioners clarification of the type and degree of left ventricular impairment, the aetiology of the heart failure, and guidance for treatment.

        We assessed the usefulness of echocardiography in a prospective audit of 80 patients (mean age 68) at Auckland Hospital. Echocardiography was requested in all cases to assess left ventricular systolic function. Clinical data were obtained from the patients' notes. Heart failure (n=36) and ischaemic heart disease (n=17) were the predominant clinical problems prompting a referral for echocardiography.

        A good agreement was shown between left ventricular dysfunction and clinical signs. These signs, however, were inconsistently recorded: jugular venous pressure was noted in 74 of the patients, while the position of the apex beat and presence or absence of a third heart sound were recorded in only 24 and 34 respectively. In 18/34 (53%) cases with a third heart sound present 11 (61%) had moderate to severe left ventricular dysfunction while 6 (33%) had normal left ventricular function. An abnormal apex beat was infrequently recorded in the case notes. Twenty seven moderate or severe valve lesions were detected by echocardiography. Of these, 8 were not clinically documented. Regurgitant murmurs were more frequently undetected (6/8) than stenotic murmurs (2/8). Apex beat and a third heart sound were not mentioned in the analysis by Gillespie and colleagues.

        As in most countries, inaccessibility of echocardiography and funding restrictions occur in New Zealand. Despite this echocardiography is a useful procedure and the cost of performing it is low in the context of total healthcare expenditure on the management of heart failure. We support the recommendation of current guidelines regarding the use of echocardiography in the assessment of heart failure.

        References

        1. 1.
        2. 2.

        Authors' reply

        1. Neil D Gillespie, Lecturerd,
        2. Allan D Struthers, Professord,
        3. Stuart D Pringle, Consultantd
        1. a Department of Medicine for the Elderly, West Glasgow Hospitals University NHS Trust, Glasgow G12 OYN
        2. b MRC Clinical Research Initiative in Heart Failure, University of Glasgow, Glasgow G12 8QQ
        3. c Department of Medicine, University of Auckland, Auckland, New Zealand
        4. d Ninewells Hospital and Medical School, Dundee DD1 9SY

          Editor—We agree with Warwick Bagg and colleagues that in an ideal world where resources are unlimited all patients with suspected heart failure should have immediate access to echocardiography, in accordance with the European Society of Cardiology's guidelines.1 A recent study in Britain in outpatients with chronic heart failure showed that as many as two thirds were denied access to echocardiography. Elderly patients were particularly disadvantaged.2 We believe that in these days of escalating healthcare costs echocardiography should be reserved for patients with acute dyspnoea in whom the diagnostic doubt is greatest. The cohort of patients in our study represent patients with disease at the severe end of the spectrum, in whom a diagnosis of left ventricular systolic dysfunction was relatively clear on clinical grounds. This is in contrast to patients with heart failure in the community, in whom the diagnosis is less clear and may be incorrect in as many as half of suspected cases. In those cases in which a diagnosis is unclear, echocardiography is essential to obtain an accurate early diagnosis of left ventricular systolic dysfunction so that treatment may be optimised with angiotensin converting enzyme inhibitors to prevent progression to more advanced heart failure.

          Bagg and colleagues seem to base their argument on the fact that their junior doctors do not record physical signs in the case notes. Our view is that technology should never be used to replace proper clinical assessment of patients. In addition, Bagg and colleagues are concerned that treatment of these patients may be affected by a lack of echocardiography. In the acute infarction ramipril efficacy study the investigators randomised patients who had had myocardial infarction to treatment with ramipril on the basis of clinical signs, a third heart sound, and radiological evidence of heart failure, with a major benefit on mortality.3 Although our cohort represents a different population of patients, many of our patients were started on an angiotensin converting enzyme inhibitor, and this decision could easily have been taken without the knowledge gained from an echocardiogram.

          Bagg and colleagues comment that a third heart sound was not mentioned in the analysis. We refer them to table 3 in our paper.

          In conclusion, we believe that, with the current limited resources for echocardiography, requests should be reserved for cases in which there is the most diagnostic doubt. If there is a specific clinical concern in a particular patient then echocardiography should be performed. If a patient is breathless; clinical examination shows a third heart sound, raised jugular venous pressure, and pulmonary crepitations; and the electrocardiogram and chest x ray film are abnormal, then echocardiography may not be essential to confirm left ventricular systolic dysfunction.

          References

          1. 1.
          2. 2.
          3. 3.
          View Abstract

          Sign in

          Log in through your institution

          Free trial

          Register for a free trial to thebmj.com to receive unlimited access to all content on thebmj.com for 14 days.
          Sign up for a free trial

          Subscribe