The value of the electrocardiogram and chest X-ray for confirming or refuting a suspected diagnosis of heart failure in the community

Transcription

1 The European Journal of Heart Failure 6 (2004) The value of the electrocardiogram and chest X-ray for confirming or refuting a suspected diagnosis of heart failure in the community Cândida Fonseca a,b, *, Teresa Mota c,d,1, Humberto Morais e,f, Fernando Matias e,f, Catarina Costa e,f, António G. Oliveira g,h,fátima Ceia b,i,2 on behalf of the EPICA Investigators 3 a Department of Internal Medicine, Medical Sciences School, New University of Lisbon, Lisbon, Portugal b Serviço de Medicina, Hospital S. Francisco Xavier, 1400 Lisboa, Portugal c Department of Cardiology, Medical Sciences School, New University of Lisbon, Lisbon, Portugal d Serviço de Cardiologia, Hospital Pulido Valente, 1750 Lisboa, Portugal e EPICA Working Group, Lisbon, Portugal f Grupo de Investigação EPICA, Av. António Augusto de Aguiar 128, 1050 Lisboa, Portugal g Datamedica Ltd., Lisbon, Portugal h Datamedica, R. Garcia de Orta 70, 2 D, 1200 Lisboa, Portugal i Department of Medical Therapeutics, Medical Sciences School, New University of Lisbon, Lisbon, Portugal Received 6 July 2004; accepted 8 September 2004 Abstract Background: There is a common assumption that a normal ECG or a normal heart size on chest X-ray virtually rules out a diagnosis of heart failure. Aims: To assess the value of the electrocardiogram and chest X-ray in identifying patients with chronic heart failure in the community. Methods and results: This study was a secondary analysis of data prospectively collected at the time of patient s enrolment in the EPICA study, an epidemiological study of the prevalence of heart failure in Portugal. A total of 6300 subjects were clinically evaluated. Patients who presented with symptoms or signs of heart failure, and/or were receiving diuretics for chronic heart failure (CHF) had a chest X-ray, ECG, and echocardiogram. The diagnosis of heart failure was confirmed in 551 cases. Patients with right atrial enlargement, atrial flutter, atrial fibrillation, 2nd degree-mobitz I atrioventricular block, 1st degree atrioventricular block, left bundle branch block, lung interstitial oedema, and bilateral pleural effusion were more likely to be diagnosed with heart failure. For the diagnosis of heart failure, in the Portuguese population aged over 25 years, an abnormal electrocardiogram had an estimated sensitivity of 81%, and negative predictive value of 75%; an abnormal chest X-ray had an estimated sensitivity of 57%, and negative predictive value of 83%. Twenty five percent of patients with CHF had a normal ECG or chest X-ray. Conclusion: Our results show that electrocardiographic and roentgenographic features are not sufficient to allow heart failure to be reliably predicted in the community and support the recommendation that all patients with suspected heart failure should undergo echocardiography. D 2004 European Society of Cardiology. Published by Elsevier B.V. All rights reserved. Keywords: Heart failure diagnosis; Electrocardiography; Chest radiography; Primary care * Corresponding author. Current address: R. Salvador Barata Feyo no. 1 r/c D.to Oeiras, Portugal. Tel.: ; fax: address: candidafonseca@netc.pt (C. Fonseca). 1 R. do Loreto no , 1200 Lisboa, Portugal. 2 Av. Grão Vasco Esq, Lisboa, Portugal. 3 EPICA Investigators and Steering Committee Members are listed in the appendix (doi: /j.ejheart ) /$ - see front matter D 2004 European Society of Cardiology. Published by Elsevier B.V. All rights reserved. doi: /j.ejheart

2 808 C. Fonseca et al. / The European Journal of Heart Failure 6 (2004) Introduction Many patients with suspected heart failure present first to their general practitioner [1]. Several studies have emphasised the difficulty of diagnosing heart failure in the community. Almost 50% of patients with a clinical diagnosis of heart failure do not have obvious abnormalities on echocardiogram [2 7]. On the other hand, a survey of primary care physicians across six European countries showed that only 5% (Netherlands) to 37% (United Kingdom) of doctors had direct access to echocardiography [8]. Screening patients before referral for echocardiography could improve the use of limited resources [4]. Chest radiography and the electrocardiogram are noninvasive tests that are widely used in the diagnosis of heart failure. Previous reports have suggested that most patients with heart failure will have an abnormal ECG [9] and that a normal ECG virtually rules out left ventricular systolic dysfunction [10,11]. It is also a common belief that a patient cannot have heart failure if the heart size is normal on the chest X-ray [10]. However, most of these studies were not population-based and referred only to patients with left ventricular systolic dysfunction [12]. Little is known about the accuracy of these non-invasive procedures in heart failure due to other types of cardiac dysfunction in the community. We report, in this paper, the value of the electrocardiogram and the chest X-ray in identifying patients with chronic heart failure in the Portuguese adult population aged over 25 years. Data from primary care attendees enrolled in the EPICA study were analysed [7,13 15]. 2. Methods The design of the EPICA study has been extensively documented elsewhere [7]. A random sample of 6300 patients, aged over 25 years, attending their general practitioner for a variety of medical compliants was obtained during 1998 in Portugal. Patients with symptoms and/or signs of heart failure, who scored 2 or more points in the Boston questionnaire (sum of parts I and II) [16], and/or who were receiving diuretics for the management of heart failure were investigated further by chest radiography, ECG, haematology, biochemistry, and an echocardiogram. The 12 lead electrocardiograms were classified as either normal or abnormal. In the abnormal electrocardiograms, 19 variables were analysed including rhythm (atrial flutter, atrial fibrillation, atrial rhythm), atrial abnormalities (left atrial enlargement, right atrial enlargement), conduction disturbances (1st degree atrioventricular block, 2nd degree- Mobitz I atrioventricular block, 2nd degree-mobitz II atrioventricular block, 3rd degree atrioventricular block, left bundle branch block, right bundle branch block, left anterior fascicular block, intraventricular conduction defect), presence of abnormal Q waves, poor R wave progression in precordial leads, left ventricular hypertrophy, and abnormal ST-segment/T-wave changes (left ventricular strain, ischemic ST-segment/T-wave changes, non-specific ST-segment/T-wave changes). In the posterioanterior upright chest X-ray, the presence of six signs was evaluated: upper zone flow redistribution, lung interstitial oedema, alveolar pulmonary oedema, bilateral pleural effusion, and cardiac enlargement (cardiomegaly and cardiothoracic ratio above 0.5). Electrocardiograms, chest X-rays and echocardiographic examinations were performed in the community and interpreted by cardiology and radiology specialists, who were unaware of the specific clinical findings of the patients, and reported in standardized forms provided by the EPICA Investigators. In accordance with the Guidelines on Diagnosis of Heart Failure of the European Society of Cardiology, the diagnosis of heart failure was considered in the presence of clinical syndrome and/or anti-congestive medical therapy for heart failure, together with echocardiographic evidence of cardiac dysfunction at rest [12]. The sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio for the diagnosis of heart failure in the Portuguese population aged over 25 years were estimated for each electrocardiogram and chest radiograph variable. 3. Results The final population for analysis included 2025 men and 3409 women with an average age of 68F15 years (S.D.) (range 25 to 99 years). The diagnosis of heart failure was excluded on clinical grounds in 4375 (80.6%) individuals because they scored b2 points on the Boston questionnaire and were not being treated for heart failure. A total of 1058 patients with possible or probable heart failure were referred for further examination; 351 patients (6.5%) scored b2 points on the Boston questionnaire, but were taking diuretics for heart failure and 707 patients (13%) scored 3 to 8 points on the questionnaire, whether they were treated for CHF or not. Among the 1022 individuals who had an echocardiogram, complete measurements could not be taken in 174 subjects (17%) because of a poor echocardiographic window. These patients were excluded from the analysis. A definite diagnosis of heart failure was established in 551 patients (208 men and 343 women). The investigators classified patients into different types of cardiac failure, according to pre-defined echocardiographic criteria [7]: 23% of the patients had ventricular systolic dysfunction, 44% had heart failure with normal systolic function, 15% had valvular heart disease, 1% pericardial disease, 11% right sided heart failure, 2% of the patients had multifactorial causes of heart failure and in 4%, it was impossible to classify the predominant type of cardiac dysfunction.

3 C. Fonseca et al. / The European Journal of Heart Failure 6 (2004) Table 1a Absolute numbers of patients with ECG abnormalities in the enrolled population with suspected heart failure based on clinical findings Characteristics Number of Patients Overall No HF HF HF and LVSD Abnormal ECG Atrial flutter Atrial fibrillation Abnormal atrial rhythm, other Left atrial enlargement Right atrial enlargement st degree AV block nd degree AV block, Mobitz I nd degree AV block, Mobitz II rd degree AV block Left ventricular hypertrophy Poor R wave progression Abnormal Q wave Left bundle branch block Right bundle branch block Left anterior fascicular block IV conduction defect Nonspecific ST-T abnormalities Left ventricular strain Ischemic ST-T changes HF: heart failure, LVSD: left ventricular systolic dysfunction, IV: intraventricular. Electrocardiographic findings are presented in Tables 1a and b. From the 1058 enrolled patients with a suspected diagnosis of CHF based on clinical features, 1034 had an ECG performed. Of these, 778 had an abnormal ECG, including 460 of the 539 with heart failure, and 111 of 127 with left ventricular systolic dysfunction. Of the 256 patients who had a normal ECG, 79 had heart failure and 16 had left ventricular systolic dysfunction. Of the YY patients who had a normal ECG, XX had heart failure and ZZ had left ventricular systolic dysfunction. Overall, an abnormal electrocardiogram had an estimated sensitivity of 81%, a specificity of 51%, a positive predictive value of 59%, a negative predictive value of 75% and a likelihood ratio of 1.7, in identifying patients with heart failure, in the Portuguese adult population with a suspected diagnosis. For the diagnosis only of left ventricular systolic dysfunction, an abnormal electrocardiogram had a sensitivity of 80%, a specificity of 40%, a positive predictive value of 17%, a negative predictive value of 93% and a likelihood ratio of 1.3. Individual ECG variables had high specificity (74 99%) but low sensitivity (1 42%) for the overall diagnosis of heart failure. The most sensitive electrocardiographic abnormalities were non-specific ST-segment/ T-wave changes (43%), left ventricular hypertrophy (30%), left ventricular strain (26%) and left atrial enlargement (18%). However, these abnormalities were not specific. The estimated prevalence of atrial arrhythmia was 12.8% (S.E.: 1.8) for atrial fibrillation, 4.1% (S.E.: 1.9) for atrial flutter and 1.67 (S.E.: 0.5) for chaotic atrial rhythm. The presence of any one of the following electrocardiographic abnormalities significantly increased the likelihood of a patient having heart failure: right atrial enlargement (likelihood ratio 8.6), atrial flutter (likelihood ratio 5.3), 2nd degree Mobitz I atrioventricular block (likelihood ratio Table 1b Estimated sensitivity, specificity, predictive values and likelihood ratio of ECG variables for the diagnosis of heart failure in the primary care setting (Portuguese population aged over 25 years) Characteristics Sensitivity Specificity PPV NPV LR % S.E. % S.E. % S.E. % S.E. Abnormal ECG Atrial flutter Atrial fibrillation Abnormal atrial rhythm, other Left atrial enlargement Right atrial enlargement st degree AV block nd degree AV block, Mobitz I nd degree AV block, Mobitz II 3rd degree AV block Left ventricular hypertrophy Poor R wave progression Abnormal Q wave Left bundle branch block Right bundle branch block Left anterior fascicular block IV conduction defect Nonspecific ST-T abnormalities Left ventricular strain Ischemic ST-T changes PPV: positive predictive value; NVP: negative predictive value; LR: likelihood ratio; S.E.: standard error; AV: atrioventricular; IV: intraventricular; ST-T: ST segment-t wave.

4 810 C. Fonseca et al. / The European Journal of Heart Failure 6 (2004) Table 2a Absolute number of patients with X-ray abnormalities in the enrolled population with suspected heart failure based on clinical findings Characteristics Number of Patients Overall No HF HF HF and LVSD Abnormal X-ray Pulmonary vessel cephalisation Lung interstitial oedema Alveolar oedema Bilateral pleural effusion Cardiomegaly Cardiothoracic ratio N HF: heart failure, LVSD: left ventricular systolic dysfunction. 5.2), 1st degree atrioventricular block (likelihood ratio 3.5), and left bundle branch block (likelihood ratio 3.4). The estimated prevalence of left bundle branch block was 6.2% (S.E.: 1.2). Chest roentgenographic findings are presented in Tables 2a and b. From the 1058 enrolled patients with a suspected diagnosis of CHF based on clinical features, 1039 had a chest X-ray performed. Of these, 602 had an abnormal chest X-ray, including 370 of the 543 with heart failure. Of the 437 patients who had a normal chest X- ray, 173 had heart failure. The overall estimated prevalence of cardiac enlargement was 53% (S.E.: 4), 63% (S.E.: 7) in CHF due to LV systolic dysfunction and 26% (S.E.: 4) in CHF with preserved systolic function. The estimated prevalence of upper zone flow redistribution and interstitial oedema was 18% and that of pleural effusion and of alveolar oedema was minimal (1% and 2%, respectively). An abnormal chest X-ray had an estimated sensitivity of 57%, specificity of 78%, positive predictive value of 50%, negative predictive value of 83%, and a likelihood ratio of 3.0 in identifying patients with heart failure. Radiographic variables had high specificity (79 99%) but modest sensitivity (1 54%) for the diagnosis of heart failure in the Portuguese adult population. The most sensitive chest radiographic abnormalities were cardiomegaly (54%) and a cardiothoracic ration0.50 (43%). Pulmonary vessel cephalisation and lung interstitial oedema had a sensitivity of 18%. 4. Discussion Previous studies comparing diagnostic findings of ECG and chest X-ray in patients with heart failure have reached different conclusions, due in part to different clinical definitions of heart failure and lack of echocardiographic criteria [17 19]. Most studies have focused on left ventricular systolic dysfunction [20] or included only hospitalised patients [21], patients referred for heart transplantation [22] or patients with coronary heart disease [10]. In contrast, the EPICA study included all types of heart failure and reflects actual everyday practice [7]. Cardiac enlargement was the most informative radiological measurement, but was present in only about half of the patients with heart failure, regardless of the type of cardiac dysfunction. A qualitative impression of heart size (cardiomegaly) was a little more predictive of heart failure than cardiac measurement (cardiothoracic ratio (CTR)N50%). Other studies have also suggested that less than half of patients with left ventricular systolic dysfunction have unequivocal cardiac enlargement (CTRN0.55) [23,24]. Clark et al. [25] found only weak correlations between CTR and left ventricular ejection fraction measured by radionuclide ventriculography and left ventricular dimensions and function evaluated by echocardiography in patients with CHF. Comparing patients with CHF with depressed and preserved systolic function, except for pulmonary venous flow cephalisation, Thomas et al. [21] reported no significant difference in the prevalence of a variety of X-ray abnormalities. Some large studies [4,6,10] support the view that a normal electrocardiogram virtually excludes chronic heart failure due to left ventricular systolic dysfunction. It is suggested that the use of the electrocardiogram as the initial investigation could be the most cost-effective approach to the diagnosis. In the EPICA study, which Table 2b Estimated sensitivity, specificity, predictive values and likelihood ratio of X-ray variables for the diagnosis of heart failure in the primary care setting (Portuguese population aged over 25 years) Characteristics Sensitivity Specificity PPV NPV LR % S.E. % S.E. % S.E. % S.E. Abnormal X-ray Pulmonary vessels cephalisation Lung interstitial oedema Alveolar oedema Bilateral pleural effusion Cardiomegaly Cardiothoracic ratio N PPV: positive predictive value; NPV: negative predictive value; LR: likelihood ratio; S.E.: standard error.

5 C. Fonseca et al. / The European Journal of Heart Failure 6 (2004) included patients with a broad range of causes of heart failure, an abnormal electrocardiogram had a lower negative predictive value and, if used alone, could have missed 25% of the patients with heart failure. The high prevalence of non-specific ST-segment/T-wave-changes, left ventricular hypertrophy, left ventricular strain and left atrial enlargement reflects the high prevalence of hypertension in this population (66% of the heart failure patients) [7]. Other studies [6] have found a higher prevalence of abnormal Q-waves, probably reflecting the inclusion of more patients with coronary heart disease and left ventricular systolic dysfunction Limitations of the study One limitation of this study is that we may have missed patients with cardiac dysfunction and few symptoms. Because the EPICA study [7] estimated a higher prevalence of heart failure in the adult general population 4.36% (95% CI: 3.96 to 5.02) as compared to other published epidemiological studies [6], we feel that our Diagnostoic threshold was sufficiently low for this possibility to be unlikely. Another potential limitation to the generalisation of our results is the initial selection performed on clinical grounds, resulting in a sample with a high prevalence of heart failure. This fact may affect the predictive values, and decreases the negative predictive value of the variables in our study. However, it is closer to clinical practice than screening the whole population. Another limitation of this study is that electrocardiograms, chest radiographs and echocardiograms were not performed in a central laboratory and the skills of the reporting physicians were not assessed. This could have increased technical variability and reduced diagnostic accuracy [26 28]. Patients in whom the echocardiographic examination was not complete because of poor ultrasonic window, which could have introduced bias, were excluded. The final echocardiographic diagnosis depended on the interpretation of data by a cardiologist and it is likely that diagnosis in some cases was difficult and prone to inter-observer (or even intra-observer) variation. In summary, a normal ECG or chest X-ray does not exclude a diagnosis of heart failure. These results support the recommendations of the European Society of Cardiology that all patients, with suspected heart failure, should undergo echocardiography. Acknowledgements The EPICA Project is supported by Servier Research Group, and has the scientific sponsorship of the Portuguese Society of Cardiology and of the Working Group on Heart Failure of the European Society of Cardiology. We are especially grateful to Prof. John G. Cleland for his advice. The authors are indebted to the EPICA investigators, without whom this work would not have been possible. We are grateful to Dr. Micaela Seeman Monteiro for revising the manuscript. References [1] Johansson S, Wallander M-W, Ruigomez A, Rodriguez LAC. Incidence of newly diagnosed heart failure in UK general practice. Eur J Heart Fail 2001;2: [2] Wheeldom NM, MacDonald TM, Flucker CJ, Mckendrick AD, Mcdevitt DG, Struthers AD. Echocardiography in chronic heart failure in the community. Q J Med 1993;86: [3] Francis CM, Caruana L, Kearney P, et al. Open access echocardiography in management of heart failure in the community. Br Med J 1995;310: [4] Davie AP, Francis CM, Caruana L, Sutherland GR, McMurray JJV. Assessing diagnosis in heart failure: which features are any use? Q J Med 1997;90: [5] Cowie MR, Wood DA, Coats AJS, et al. Incidence and aetiology of heart failure. A population based study. Eur Heart J 1999;20: [6] Davies MK, Hobbs FDR, Davis RC, et al. Prevalence of leftventricular systolic dysfunction and heart failure in the Echocardiographic Heart of England Screening study: a population based study. Lancet 2001;358: [7] Ceia F, Fonseca C, Mota T, et al. Prevalence of chronic heart failure in southwestern Europe: the EPICA study. Eur J Heart Fail 2002;4: [8] Hobbs FDR, Wilson S, Jones MI, Tobias RS, Allen TF, Fletcher M. European survey (EuroHF) of primary care physician perceptions and practice in heart failure diagnosis and management. Eur Heart J 1998;19:154s. [9] The NETWORK Investigators. Clinical outcome with enalapril in symptomatic chronic heart failure; a dose comparicon. Eur Heart J 1998;19: [10] Rihal CS, Davies KB, Kennedy JW, et al. The utility of clinical electrocardiographic, and roentgenographic variables in the prediction of left ventricular function. Am J Cardiol 1995;75: [11] Davie AP, Francis CM, Love MP, et al. Value of the electrocardiogram in identifying heart failure due to left ventricular systolic dysfunction. Br Med J 1996;312:222. [12] Task Force of the Working Group on Heart Failure of the European Society of Cardiology. The treatment of heart failure. Eur Heart J 1997;18: [13] Fonseca C, Morais H, Mota T, et al. The diagnosis of heart failure in primary care: value of symptoms and signs. Eur J Heart Fail 2004;6: (doi: /j.ejheart ). [14] Fonseca C, Oliveria AG, Mota T, et al. Evaluation of the performance and concordance of clinical questionnaires for the diagnosis of heart failure in primary care. Eur J Heart Fail 2004;6: (doi: / j.ejheart ). [15] Ceia F, Fonseca C, Mota T, et al. Aetiology, co-morbidity and drug therapy of chronic heart failure in the real world: the EPICA sub-study. Eur J Heart Fail 2004;6:801 6 (doi: /j.ejheart ). [16] Carlson KJ, Lee DC-S, Goroll AH, Leahy M, Johnson RA. An analysis of physician s reasons for prescribing long-term digitalis therapy in outpatients. J Chronic Dis 1985;38: [17] Marantz PR, Tobin JN, Wassertheil-Smoller S, Steingart RM, Wexler N, Budner N, et al. The relationship between left ventricular systolic function and congestive heart failure diagnosed by clinical criteria. Circulation 1988;77(3): [18] McKee PA, Castelli WP, McNamara PM, et al. The natural history of congestive heart failure: the Framingham study. N Engl J Med 1971;26:

6 812 C. Fonseca et al. / The European Journal of Heart Failure 6 (2004) [19] Parameshwar J, Schachell MM, Richardson A, Poole-Wilson PA, Sutton GC. Prevalence of heart failure in three general practices in north west London. Br J Gen Pract 1992;42: [20] Davies MK, Hobbs FDR, Davis RC, et al. Prevalence of leftventricular systolic dysfunction and heart failure in the Echocardiographic Heart of England Screening study: a population based study. Lancet 2001;358: [21] Thomas JT, Kelly RF, Thomas SJ, Stamos TD, Albasha K, Parrillo JE, et al. Utility of history, physical examination, electrocardiogram, and chest radiograph for differentiating normal from decreased systolic function in patients with heart failure. Am J Med 112:2002; [22] Chakko S, Woska D, Martinez H, Marchena E, Futterman L, Kessler K, et al. Clinical, radiographic, and hemodynamic correlations in chronic congestive heart failure: conflicting results may lead to inappropriate care. Am J Med 1991;90: [23] Petrie MC, McMurray JJV. It cannot be cardiac failure because the heart is not enlarged on the chest X-ray. Eur J Heart Fail 2003;5: [24] The Digitalis Investigation Group: the effect of digoxin on mortality and morbidity in patients with heart failure. N Engl J Med 1997;336: [25] Clark AL, Coats AJ. Unreability of cardiothoracic ratio as a marker of left ventricular impairment: comparison with radionuclide ventriculography and echocardiography. Postgrad Med J 2000;76: [26] Davidson C. Can heart failure be diagnosed in primary care? Chest radiography is still useful. Br Med J 2000;321:1414. [27] Keats TE. Atlas of Roentgenographic Measurement, 6th ed. St Louis Mosby-Year Book. p [28] Chickos PM, Figley MM, Fisher L. Correlation between chest film and angiographic assessment of left ventricular size. J Roentgenol 1977;128:363 7.