• 
• 

Three Part Question

In a [patient with acute breathlessness] does a [normal ECG] rule out the [diagnosis of heart failure]?

Clinical Scenario

A 65y/o gentleman presents to the Emergency Department complaining of acute onset of shortness of breath. You suspect on clinical grounds that this may be due to heart failure. His ECG shows sinus rhythm with a rate of 96bpm and no abnormalities that you can detect. You wonder if this suggests that there is another cause for his symptoms.

Search Strategy

Medline 1966 - Nov 2005 using OVID interface.
Embase 1974 - Nov 2005 using Dialog Datastar interface.
Medline search: [Exp Heart Failure, Congestive/ OR heart failure.mp. OR exp Cardiac Output, Low/ OR exp Heart Failure, Congestive/ OR cardiac insufficiency.mp. OR exp Ventricular Dysfunction, Left/ OR lvf.mp. OR left ventricular failure.mp.] AND [exp "Sensitivity and Specificity"/ or exp Diagnostic Tests, Routine/ or diagnos$.mp. OR sensitivit$.mp. OR specificit$.mp. OR diagnos$.m_titl.] AND [exp Electrocardiography/ or electrocardio$.mp. OR ecg.mp. OR ekg.mp.] LIMIT to humans and English language.
Embase search:[heart adj failure OR heart-failure#.DE. OR congestive-heart-failure#.de OR CCF OR ventricular adj dysfunction OR heart-left-ventricle-failure#.de OR heart-left-ventricle-function#.de or heart-right-ventricle-function#.de OR heart-right-ventricle-failure#.de OR LVF OR cardiac adj insufficiency OR failing adj heart OR heart-failure#.de OR congestive-heart-failure#.de OR heart-edema.de OR pulmonary adj edema OR lung-edema#.de or pulmonary adjj oedema] AND [sensitiv$.TI OR specific$.TI OR diagnos$.TI OR utilit$.TI] AND [ecg OR ekg OR electrocardiog$] LIMIT to humans and English Language.

Search Outcome

1803 papers were found in the Medline search and 334 papers were found in the Embase search. Eight papers from the Medline search had relevant results, no further relevant papers were found from the Embase search.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Davie A et al 1996 UK	534 patients referred to open access echocardiography clinic by the GP with a suspected diagnosis of heart failure.	ECGs reported as normal or abnormal and compared against findings of echocardiogram to determine predictive value of left ventricular systolic function.	Prediction of abnormal left ventricular systolic function	Sensitivity 94%, specificity 61%, PPV 35%, NPV 98%, PLR 2.4, NLR 0.10	Does not specifiy which ECG abnormalities have the best predictive value. Does not quantify left ventricular systolic function.
Gillespie, ND et al 1997 UK	71 randomly selected patients who had been admitted with acute dyspnoea to the acute medical receiving ward. Patients with obvious pulmonary disease or renal failure were excluded.	ECG classified as normal(including minor abnormalities) or abnormal prior to an echocardiogram. LV systolic dysfunction defined as fractional shortening of less than 20%, 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 as defined by the echocardiographer.	Predictive value of ECG of left ventricular systolic dysfunction	Sensitivity 98%(95% CI 88-99), specificity 69(95% CI 48-85), PPV 0.94(0.80-0.99), NPV 0.95(0.74-0.99)	Method of randomisation not explained. Not clear what constituted major/minor abnormalities of ECGs. Unable to confirm sensitivity / specificity etc from data provided.
Houghton, AR; et al 1997 UK	200 patients chosen at random from patients referred for investigation at a heart failure clinic. LVSD present if ejection fraction <40%, fractional shortening <25% or global left ventricular systolic function felt to be impaired by the echocardiographer. ECG examined by blinded cardiologist and judged to be normal or abnormal.	Comparison of ECG findings with an echocardiographic assessment of left ventricular systolic function.	Comparison of results from echocardiogram versus electrocardiogram.	Sensitivity 89.1%, specificity 45.7%m PPV 0.89, NPV 0.46, positive LR 1.64.	Retrospective. Method of randomisation not explained.
Murkofsky, RL; et al 1998 USA	270 patients referred to the Nuclear Cardiology Laboratory at Mount Sinai Hospital in New York. Patients with known cause of QRS prolongation excluded including patients on anti-arrhythmic drugs, recent MI, typical RBBB or LBBB, pacemaker or patients with insufficient data, eg those with AF.	All patients had and ECG and and radionuclide ventriculography. The ejection fraction was used to divide the patients into two groups with EF <45% and EF >45%. ECGs were examined and the QRS complexes measured and divided into groups of <0.10s or >0.10s. The R-wave score was also calculated by adding the R-waves in mV in leads aVL, aVF and V1-6 from each ECG.	Prediction of decreased LVEF by R-wave score <4mV	Sensitivity 33%, specificity 95%, positive LR 6.6, negative LR 1.4	Many patients excluded from this study so unable to calculate predicitive value for more general population.
			Prediction of decreased LVEF by QRS>0.10s	Sensitivity 43.8%, specificity 83.6%, positive LR 2.67
			Prediction of decreased LVEF by QRS>0.12s	Sensitivity 13.8%, specificity 99.3%, positive LR 19.7
			Prediction of decreased LVEF by combination of R-wave score <4 and QRS>0.10s	Sensitivity 15.3%, specificity 99%, positive LR 15 and negative LR of 1.2
Ng, LL; et al 2003 UK	Random selection of male patients (45-80y) and female patients (55-80y) invited to attend screening for heart failure from 21 General Practices in the Leicester area. 2393 patients invited, 1360 patients responded.	Patients attended hospital and underwent echocardiography, ECG and blood testing. Left ventricular function assessed by LVEF and left ventricular wall motion index. ECGs examined independently.	Prediction of decreased LVEF by abnormal ECG	Sensitivity 88%, specificity 60.7%, PPV 0.031.	Only 17 patients out of the total actually had documented left ventricular systolic dysfunction.
Kruger, S; et al 2004 Germany	128 consecutive patients referred for an elective echocardiogram for suspected heart failure. All patients subsequently had an ECG, echocardiography and cardiac catheterisation carried out.	Left ventricular systolic dysfuntion defined as an ejection fraction <50%. QRS complexes measured from ECGs.	Prediction of decreased LVEF by QRS >0.12s	Sensitivity 75.7%, specificity 90.5%, PPV 0.94, NPV 0.79, LR 15.65.
			Prediction of decreased LVEF by QRS>0.10s	Sensitivity 84.8%, specificity 61.3%, PPV 0.7, NPV 0.79, positiive LR 2.19.
Fonseca, C; et al 2004 Portugal	6300 patients over the age of 25y who were attending their GP for an unrelated problem were enrolled at random into the EPICA study which was seeking to determine the prevalence of heart failure. Patients who scored over 2 on the Boston questionnaire or who were receiving treatment for heart failure underwent further investigation. This consisted of a chest x-ray, an ECG and an echocardiogram.	Secondary analysis of prospectively collected data assessing the value of the electrocardiogram and chest radiograph in identifying patients in the community with heart failure. The majority of patients did not score over 2 in the questionnaire nor received any treatment for heart failure. Other patients did not receive echocardiography or the investigation was not technically possible. Patients who did go on to have an echocardiogram had the results compared with the ECG finding.	Prediction of heart failure by abnormal ECG	Senstivity 81%, specificity 51%, PPV of 0.59, NPV of 0.75 and a LR of 1.7.	Boston questionnaire not defined. Various forms of heart failure specified by echocardiography but quantative results not given in this paper.
			Prediction of left ventricular SD by abnormal ECG	Sensitivity 80%, specificity 40%, PPV 0.17, NPV 0.93 and LR 1.3.
Wang, CS; et al 2005 Various	Meta-analysis incorporating data from 7 studies of dyspnoeic patients presenting to the emergency department.	Presence of various ECG findings compared with the likelihood of heart failure.	Prediction of heart failure by the presence of AF	LR, 3.8; 95% CI 1.7-8.8
			Prediction of heart failure by normal ECG	LR 0.64; 95% CI 0.47-0.88.

Comment(s)

These papers represent a variety of patients who present with symptoms suggestive of heart failure. A variety of different criteria are used to classify the ECG in an attempt to maximise its utility in predicting heart failure or left ventricular systolic dysfunction. The patients generally have stable symptoms and are selected on clinical grounds with various exclusions. The results of these studies are quite variable but overall it is reasonable to surmise that a normal ECG makes the diagnosis of heart failure or LVSD unlikely but by no means rules it out.

Clinical Bottom Line

A normal ECG makes heart failure less likely but further investigation of the patient is required in order to rule out this diagnosis. There is no consistent ECG abnormality that is specfic enough to make the diagnosis of heart failure purely from this investigation.

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.

References

1. Davie, AP; Francis, CM; Love MP; et al Value of the electrocardiogram in identifying heart failure due to left ventricular systolic dysfunction BMJ 1996; 222
2. Gillespie, ND; McNeill, G; Pringle, T; et al Cross sectional study of contribution of clinical assessment and simple cardiac investigations to diagnosis of left ventricular systolic dysfunction in patients admitted with acute dyspnoea BMJ 1997 (29 March);936
3. Houghton, AR; Sparrow, NJ; Toms, E; Cowley, AJ Should general practitioners use the electrocardiogram to select patients with suspected heart failure for echocardiography? International Journal of Cardiology 1997 pp31-36
4. Murkofsky, RL; Dangas, G; Diamond, JA; Mehta, D; Schaffer, A; Ambrose, JA A Prolonged QRS Duration on Surface Electrocardiogram Is a Specific Indicator of Left Ventricular Dysfunction Journal of the American College of Cardiology 1998; pp476-82
5. Ng, LL; Loke, I; Davies, JE; Kunti, K; Stone, M; Abrams, KR; Chin, DT; Squire, IB Identification of previously undiagnosed left ventricular systolic dysfunction: community screening using natriuretic peptides and electrocardiography The European Journal of Heart Failure 2003; pp775-782
6. Kruger, S; Filzmaier, K; Graf, J; Kunz, D; Stickel, T; Hoffmann, R; Hanrath, P; Janssens, U QRS prolongation on surface ECG and brain natriuretic peptide as indicators of left ventricular systolic dysfunction Journal of Internal Medicine 2004; pp206-212
7. Fonseca, C; Mota, T; Morais, H; Matias, F; Costa, C; Oliveira, AG; Ceia, F The value of the electrocardiogram and chest X-ray for confirming or refuting a suspected diagnosis of heart failure in the community The European Journal of Heart Failure 2004; pp807-812
8. Wang, CS; Fitzgerald, JM; Schulzer, M; Mak, E; Ayas, NT Does This Dyspneic Patient in the Emergency Department Have Congestive Heart Failure? JAMA 2005; pp1944-1956