Cross sectional study of contribution of clinical assessment and simple cardiac investigations to diagnosis of left ventricular systolic dysfunction in patients admitted with acute dyspnoeaBMJ 1997; 314 doi: https://doi.org/10.1136/bmj.314.7085.936 (Published 29 March 1997) Cite this as: BMJ 1997;314:936
- Neil D Gillespie, research fellowa,
- Graeme McNeill, consultant cardiologistb,
- Terence Pringle, consultant cardiologistb,
- Simon Ogston, statisticianb,
- Allan D Struthers, professor of clinical pharmacologyc,
- Stuart D Pringle, consultant cardiologistc
- a Departments of Cardiology, Clinical Pharmacology and Epidemiology Ninewells Hospital Dundee DD1 9SY
- b Department of cardiology
- c Department of epidemiology
- Correspondence to: Dr N D Gillespie Department of Medicine (Section of Ageing and Health), Ninewells Hospital and Medical School, Dundee DD1 9SY.
- Accepted 16 January 1997
Objective: To assess the comparative contribution of clinical assessment, electrocardiography, and chest radiography to the diagnosis of left ventricular systolic dysfunction in patients admitted to a general medical ward with acute dyspnoea.
Design: Prospective cross sectional study.
Setting: Acute medical admissions ward of a teaching hospital.
Subjects: 71 randomly selected patients admitted with acute dyspnoea.
Main outcome measures: Sensitivity and specificity of each investigation and logistic regression analysis of each variable in identifying left ventricular systolic dysfunction.
Results: Clinical assessment in this cohort of patients with severe dyspnoea was generally sensitive (sensitivity 81%). Patients were divided into three groups on the basis of clinical assessment. In the first group (37 patients) the diagnosis of systolic dysfunction was clear, in the second (22) it was in doubt, and in the third (12) it was unlikely. The sensitivity of clinical assessment in identifying left ventricular systolic dysfunction was 81% and the specificity was 47%. The specificity of diagnosis was improved by electrocardiography (69%) and chest radiography (92%). Logistic regression analysis showed that isolated pulmonary crepitations were a comparatively poor predictor of left ventricular systolic dysfunction (2=10.215, P=0.0014) but that a full clinical examination had reasonable predictive value (2=24.82, P<0.00001). The combination of clinical assessment and chest radiography improved the accuracy of diagnosis (2=28.08, P<0.00001), as did the combination of clinical assessment and electrocardiography (2=32.41, P<0.00001).
Conclusion: Clinical assessment in patients admitted with acute dyspnoea is comparatively accurate. Patients with abnormal results on chest radiography, electrocardiography, and clinical examination have a high likelihood of having left ventricular systolic dysfunction. Echocardiography contributes little more to the diagnosis in these patients and may be more efficiently directed towards patients in whom the diagnosis is still in doubt after clinical assessment, chest radiography, and electrocardiography.
The availability of echocardiography is limited for patients admitted to hospital with acute dyspnoea
The presence of isolated lung crepitations is a poor predictor of left ventricular systolic dysfunction
Full clinical assessment is sensitive in detecting left ventricular systolic dysfunction, with specificity being added by either chest radiography or electrocardiography
Echocardiography should be reserved for cases with the most diagnostic doubt
Heart failure is increasingly common in an aging population.1 Emergency admissions are increasing,2 and patients with heart failure represent about 5% of all hospital admissions.3 Recent studies have emphasised the difficulty of diagnosing left ventricular systolic dysfunction without echocardiography in outpatients with suspected mild heart failure.4 5 However, whether echocardiography is as important in patients with severe disease–that is, those admitted as emergencies with acute dyspnoea–is unclear. An accurate diagnosis of left ventricular systolic dysfunction is important to start effective treatments such as angiotensin converting enzyme inhibitors.6
Most patients with suspected heart failure are investigated after clinical examination by electrocardiography, chest radiography, and echocardiography (if available). Although nearly all patients have electrocardiography and chest radiography, less than half of those admitted to hospital with suspected severe heart failure have access to echocardiography.7 8 A critical evaluation of the information supplied by clinical assessment, electrocardiography, and chest radiography may help streamline the selection of patients for echocardiography and use limited resources more efficiently. We assessed the relative contribution of these three investigations to the diagnosis of left ventricular systolic dysfunction in patients admitted to a general medical unit with acute dyspnoea.
Patients and methods
Seventy one randomly selected patients admitted with dyspnoea to the acute medical receiving ward at Ninewells Hospital in Dundee were recruited to the study. Patients were included if they presented with acute dyspnoea or had dyspnoea as a major component of their overall symptoms. Patients with obvious isolated pneumothorax, pneumonia, pulmonary emboli, cor pulmonale, or renal failure were excluded from the study.
Clinical assessment and preliminary investigations
On admission each patient underwent assessment by the receiving medical registrar with chest radiography, electrocardiography, and routine blood testing. Each patient was reviewed by a consultant physician on the following morning and a presumptive diagnosis was given, pending the results of further investigations. The clinical diagnosis of left ventricular systolic dysfunction was based on the clinical decision of both the receiving medical registrar and the consultant physician–that is, it was based on standard routine clinical assessment by the admitting medical team. This diagnosis was given independently of the research fellow (NDG) performing the echocardiography. The following details were recorded from the case notes for subsequent analysis: the presence of pulmonary crepitations, pulse, blood pressure, presence of murmur, jugular venous pulse, presence of oedema (ankle or lung), drug treatment, and findings on electrocardiography.
Each patient had echocardiography performed on the morning after admission by an experienced echocardiographer whose assessments had been previously validated by comparison with radionuclide ventriculography. The echocardiographer was unaware of the full clinical diagnosis while performing echocardiography. Assessments were performed on a Challenge Sim 7000 device, and each patient was studied on the ward on the morning after their admission. Three measurements were taken while patients were lying in the left lateral decubitus position: M mode assessment of left ventricular dimensions at the tip of the mitral valve leaflets,9 calculation of fractional shortening, and assessment of regional wall motion index.10 In addition, Doppler studies of the mitral and aortic valves were performed.
Left ventricular systolic dysfunction was diagnosed before discharge from the information obtained at clinical assessment, echocardiography, and electrocardiography and from radiological findings.
The contribution of each investigation to the overall final diagnosis of left ventricular systolic dysfunction was assessed. For the purposes of analysis the patients were categorised into three groups on the basis of the original clinical assessment. Those in the first group had a clear diagnosis of heart failure on the basis of clinical history, raised jugular venous pressure, crepitations, and the presence of a third heart sound. In the second group the findings of the clinical assessment were less clear, and in the third group the probability of heart failure was considered to be low and the main cause of the dyspnoea was non-cardiac.
For the purposes of interpretation electrocardiograms were either normal or abnormal, with minor abnormalities being included in the normal group. Findings on chest radiography were categorised as being within normal limits, as isolated cardiomegaly, as pulmonary oedema or cardiomegaly, or both, or as other findings. Most findings on chest radiography were easily interpreted by the receiving medical team, but cases of doubt were assessed by a radiologist.
Left ventricular systolic dysfunction was diagnosed echocardiographically on the basis of a fractional shortening of less than 20% or the presence of one or more regional wall motion abnormalities, a dilated left ventricle on M mode, or a subjectively determined reduction in left ventricular systolic function assessed by an experienced echocardiographer.
The contribution of each investigation to the final diagnosis of left ventricular systolic dysfunction was assessed by performing logistic regression analysis on the clinical assessment and on the findings of electrocardiography and chest radiography. The sensitivity, specificity, positive predictive accuracy, and negative predictive accuracy were calculated for each investigation in determining left ventricular systolic dysfunction.
The mean age of the patients was 73 (range 33-95). Thirty patients were men. Twenty three patients had a history of myocardial infarction and 18 a history of hypertension. Most patients admitted to this medical receiving ward had predominantly cardiac causes for their dyspnoea as patients with lung disease are admitted to a separate chest unit.
A total of 45 patients had impaired left ventricular systolic function. Three patients had isolated left ventricular diastolic dysfunction.
Table 1) shows the three groups of patients according to the findings at initial clinical assessment. Twenty two of the 23 patients with a previous myocardial infarction were in groups 1 or 2. Twelve of the patients had left ventricular hypertrophy at echocardiography and 11 of these 12 patients were in groups 1 or 2. Those with severe left ventricular dysfunction identified readily at clinical examination were the largest proportion (37/71 (52%)), but despite the comparative severity of symptoms of the whole group, there were many patients in which the diagnosis was not clear after initial clinical assessment. Thirty three of the 37 patients (89%) in group 1 had radiological evidence of heart failure, while 17 of the 22 patients (77%) in group 2 had normal radiological findings.
Table 2) shows the sensitivities, specificities, and predictive accuracy of combinations of the investigations after initial clinical assessment in diagnosing left ventricular systolic dysfunction.
Logistic regression analysis showed the relative contributions of both electrocardiography and chest radiography to the diagnosis of left ventricular systolic dysfunction (table 3). In the model for clinical failure and electrocardiography the odds ratio for clinical failure was 4.39 and 24.02 for an abnormal electrocardiogram. In the model for clinical failure and chest radiography the odds ratio was 1.72 for clinical failure and 4.34 for chest radiography.
There has been a steady increase in patients admitted to acute medical wards over the past 10 years.2 Many patients have left ventricular systolic dysfunction, and treatment with angiotensin converting enzyme inhibitors is beneficial for them.6 Such treatment is not without its hazards, however, and a clear diagnosis of left ventricular systolic dysfunction is necessary before treatment begins. Ideally, all patients should have echocardiography performed before treatment begins to confirm the diagnosis and to exclude significant valvar disease, but up to 50% of patients undergoing assessment for suspected left ventricular dysfunction are not investigated by echocardiography.7 8
There was a good correlation between clinical assessment and objective assessments of left ventricular function in this cohort of patients. This is likely to be because these patients had disease severe enough to necessitate hospital admission. It may therefore be possible to bypass echocardiography in patients with little clinical doubt about the diagnosis so long as they do not have an associated heart murmur. We still believe that all patients with a heart murmur should have echocardiography. Otherwise, our proposed approach could possibly result in cases of aortic valve disease being missed, although a previous study in our department showed no cases of unsuspected aortic valve disease in an audit of 400 requests for echocardiography.11 However, our findings have to be put in perspective as our unit is in a large teaching hospital and the accuracy of clinical assessment may vary from centre to centre and in primary care.
Most patients with obvious clinical heart failure had abnormal chest radiographs. Consequently, little additional information was provided by echocardiography in patients who did not have a heart murmur.
In acute medical admissions, heart failure is usually the result of systolic dysfunction due to coronary artery disease or hypertension. Although, echocardiography would be able to identify patients who have additional diastolic dysfunction, there are still no established treatments for isolated diastolic dysfunction.12 As a result, the limited resources for echocardiography may be better directed to those with an uncertain clinical diagnosis and in whom the radiological findings are inconclusive.
In patients with possible heart failure chest radiography provided diagnostic information in only five of them (22%), while echocardiography confirmed systolic dysfunction in 11 (50%), with a further three having other abnormalities identified. In these patients a clear diagnosis is essential as appropriate treatment of systolic dysfunction and hypertension may prevent progression to overt heart failure, with possible reductions in subsequent hospital admissions.2
In the third group–patients with non-cardiac causes for their dyspnoea–echocardiography may be more appropriate than chest radiography in those with abnormal electrocardiograms as many of these patients often undergo both investigations with little additional information provided by chest radiography.
Electrocardiography gave abnormal results in most patients with left ventricular dysfunction. This agrees with the findings of other studies13 and suggests that the diagnosis of left ventricular systolic dysfunction should be reconsidered in patients with a normal electrocardiogram.14
On the basis of these findings, we propose that the diagnosis and treatment of left ventricular systolic function in this setting could be based on the following approaches.
If the diagnosis is clinically obvious act on the clinical diagnosis and obtain an echocardiogram only if there is a heart murmur.
When the diagnosis is not so obvious add chest radiography and electrocardiography, with echocardiography later.
When the diagnosis is not clear use echocardiography as the initial investigation.
The mode of delivery of echocardiography in this study–that is, performed at the bedside with a portable device–may permit a more rapid assessment of patients to enable a quicker diagnosis and a shorter duration of admission. The increased demand for and availability of echocardiography has clear financial implications, but the cost of smaller simpler devices continues to fall. The British Society of Echocardiography has now issued training guidelines which will help non-cardiologists gain expertise in simple echocardiography.15 The more widespread use of these devices could result in more patients with heart failure being assessed for treatment with angiotensin converting enzyme inhibitors. Currently, only 20-30% of patients with heart failure receive such treatment.16
In conclusion, this study illustrates the contribution of clinical assessment and simple non-invasive cardiac investigations to the diagnosis of left ventricular systolic dysfunction in a cohort of patients admitted with acute dyspnoea. Ideally, all patients admitted to hospital with acute dyspnoea should have access to echocardiography, but current resources prevent this. We suggest that echocardiography is not essential in patients with acute dyspnoea and clear clinical evidence of cardiac failure but no heart murmur. Clinical assessment is sensitive and specificity can be added by either chest radiography or electrocardiography. Echocardiography should be reserved for those cases in which there is the most diagnostic doubt. The observations made in this study may help streamline the selection of patients for echocardiography services, enabling more efficient use of currently available resources.
We thank the medical staff of ward 15, particularly Dr R Duffy, in Ninewells Hospital, Dundee, and Charge Nurse Connolly and the other nursing staff.
Funding: Merck Sharpe and Dohme provided an educational grant to fund the study.
Conflict of interest: None.