European Journal of Heart Failure

European Journal of Heart Failure 2003 5(3):349-354; doi:10.1016/S1388-9842(03)00046-1

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Disclaimer Google Scholar Articles by Sparrow, N. Articles by Hampton, J. R. Search for Related Content PubMed PubMed Citation Articles by Sparrow, N. Articles by Hampton, J. R. Social Bookmarking          What's this?

© 2003 European Society of Cardiology

The diagnosis of heart failure in general practice: implications for the UK National Service Framework

Nigel Sparrow^a,*, David Adlam^b, Alan Cowley^b and John R. Hampton^b

^a Newthorpe Medical Practice Chewton Street, Eastwood, Nottingham NG16 3HB, UK
^b Department of Cardiovascular Medicine, Queen's Medical Centre Nottingham NG7 2UH, UK

^* Corresponding author. Tel.: +44-1773-760202. E-mamil address: nigelsparrow{at}aol.com

	Abstract

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 
Background: The UK National Service Framework recommends patients with suspected heart failure undergo echocardiography. Selection of patients for this investigation in primary care is difficult. It is not clear which clinical features best identify patients with left ventricular systolic dysfunction.

Aim: Using echocardiography, to establish the accuracy of primary care diagnosis of left ventricular systolic dysfunction. To investigate the sensitivity, specificity and predictive values of clinical features in the diagnosis of left ventricular systolic dysfunction.

Study: A cross-sectional study of 621 patients from a population prescribed loop diuretics in 7 general practices.

Method: Clinical diagnoses were extracted from general practice records. Symptoms, clinical signs, ECG features, brain natriuretic peptide levels and echocardiographic findings were studied in a research clinic.

Results: Left ventricular systolic dysfunction (ejection fraction <40%) was present in 50% of 621 patients prescribed loop diuretics in primary care. General practice diagnoses showed high false positive rates. Individual or combinations of clinical features did not accurately predict left ventricular systolic dysfunction.

Conclusion: These results suggest the clinical diagnosis of left ventricular systolic dysfunction is inaccurate in this population. General practitioners should have a low threshold for referring patients prescribed loop diuretics for echocardiography. Increased open access echocardiography facilities will be needed.

Key Words: General practice • Echocardiography • Heart failure

Received August 5, 2002; Revised January 8, 2003; Accepted January 21, 2003

	1. Introduction

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 
Chronic heart failure is a common condition with an estimated population prevalence of between 3 and 20 per 1000. The prevalence rises markedly in the elderly population to at least 80 cases per 1000 in those over 75 [1]. The accurate diagnosis of heart failure is becoming increasingly important with the identification of pharmacological interventions that can improve both morbidity and mortality. These include ACE inhibitors [2,3] and more recently Beta blockers [4,5] and Spironolactone [6].

The appropriate investigation and treatment of suspected heart failure forms a core part of the National Service Framework (NSF) for Coronary Heart Disease [1]. The NSF is produced by the Department of Health in the UK and now provides the basis for recommended health care provision across the country. Most patients with heart failure are diagnosed and managed in primary care [7,8]. This is recognised by the NSF, which provides an algorithm for the assessment of suspected heart failure in the community. It recommends that all patients with suspected heart failure on clinical assessment be referred for echocardiography.

Clinical assessment and diagnosis of patients with symptoms and signs suggestive of heart failure is difficult and inaccurate. Studies of patients referred for open access echocardiography show that general practice diagnosis has a high false positive rate [7,9]. Furthermore, community based studies show a substantial population of patients with undiagnosed left ventricular dysfunction some of which are asymptomatic [10,11]. It is likely that these patients would also benefit from early diagnosis and treatment [3,12]. The increasing availability of open access echocardiography will facilitate accurate diagnosis in those referred. However, it remains unclear what patient features or combination of features should trigger a primary care referral for echocardiography.

The aim of this study was to investigate the diagnosis of heart failure in general practice using a population of patients prescribed loop diuretics. The reasons for prescribing these drugs have been described elsewhere [13]. Echocardiography was used as the gold standard investigation to assess the relative importance of a spectrum of clinical features in the identification of patients with left ventricular systolic dysfunction.

	2. Method

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 
Seven general practices in Nottingham with a total list size of 60 728 were surveyed between January 1995 and December 1998. The study area included both urban and rural practices with list sizes ranging from 4 to 12 000. This practice population was under the care of an estimated 27 full-time equivalent general practitioners. One thousand three hundred and sixty-six of these patients were found to be taking loop diuretics. The practice records of 1301 of these patients were obtained as part of a separate investigation into loop diuretic prescribing in primary care [13]. A research nurse examined these records to establish the diagnosis or symptom recorded by the general practitioner that lead to prescription of a loop diuretic. For analysis, these data were then retrospectively sorted by one of the authors^a into seven clinical indication categories; ‘definite heart failure’, ‘probable heart failure’, ‘shortness of breath’, ‘ankle swelling’, ‘hypertension’, ‘other’ and ‘no diagnosis recorded’.

These patients were then invited by letter to attend a research clinic at the local hospital. Seven hundred and thirty-seven patients agreed to attend and underwent a full clinical assessment, electrocardiography, measurement of brain natriuretic peptide (BNP) and echocardiography. Each of these elements was carried out blinded to the results of the other parts of the assessment. The presence or absence of clinical symptoms and signs associated with heart failure were recorded for each patient. ECGs were analysed by one of the authors (N. Sparrow) and classified as normal or abnormal. Any other specific ECG features including previous myocardial infarction and ischaemia were also noted. Echocardiography was performed by an experienced technician. Left ventricular ejection fractions were measured using a phased array sector scanner (Vingmed CFM 700) and recorded for each of 621 patients independently of the results of their clinical assessment. Data from the remaining clinic attendees (13%) were excluded from subsequent analysis due to inadequate echocardiographic images. For the purposes of this study an ejection fraction of less than 40% was used as the gold standard definition of left ventricular systolic dysfunction. The BNP assay used has been described elsewhere [18]. BNP data were available for 571 of the 621 patients from which ejection fractions were measured.

Ethical approval for this study was obtained from the Queen's Medical Centre Ethics Committee.

	3. Results

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 
The characteristics of patients prescribed loop diuretics in general practice are shown in Table 1. This compares patients who attended the research clinic with those who, for whatever reason, did not attend. It was the patients who attended this clinic who subsequently underwent clinical assessment, electrocardiography, BNP measurement and echocardiography in this study. These populations were significantly different in a number of ways. There were proportionally more male patients in the clinic population. The age distribution was also different with fewer patients from the over 80s age group able to attend clinic and a correspondingly greater representation of patients from the age groups 60–79 years. A significantly higher recorded prevalent history of angina pectoris but not of other previous illnesses was noted in the general practice records of patients investigated in the research clinic. This was reflected in significantly greater use of long and short acting nitrates amongst these patients. Other cardiac medications notably ACE inhibitors, aspirin and warfarin were also prescribed more frequently in the clinic group.

View this table: [in this window] [in a new window]   Table 1 Patient characteristics

 
Of 621 clinic patients who underwent echocardiography, 314 (50.6%) had ejection fractions less than 40%, with 142 (22.9%) of these having ejection fractions less than 30%. Fifty-seven patients (9.2%) had echocardiographic valve abnormalities. Twenty-six patients had an abnormal mitral valve, 9 patients an abnormal aortic valve and 22 patients had both mitral and aortic valvular disease.

The clinical indication recorded by general practitioners at the time of initiation of loop diuretic therapy was assigned to one of the seven clinical indication categories shown in Table 2. The proportion of patients with left ventricular systolic dysfunction (ejection fraction <40%) and severe left ventricular systolic dysfunction (ejection fractions <30%) are shown for each category. Of patients attributed a primary care diagnosis of definite heart failure, 42.2% did not have impaired left ventricular systolic function on echocardiography. Conversely, of those patients prescribed loop diuretics by their general practitioner who were not given a diagnosis of definite or probable heart failure, 47.9% had left ventricular systolic dysfunction on echocardiography and 20.5% had ejection fractions of less than 30%.

View this table: [in this window] [in a new window]   Table 2 Ejection fraction by primary care clinical indication precipitating loop diuretic prescription

 
The sensitivity, specificity, and predictive values for each of the symptoms, clinical signs, past diagnoses, BNP values and ECG features assessed in the research clinic are shown in Table 3. These are calculated using an ejection fraction as assessed by echo of less than 40% as the gold standard measure of left ventricular systolic dysfunction. Table 4 shows the values calculated for various combinations of these clinical features. The best positive predictive values are derived from the identification of clinical signs with high degrees of specificity for left ventricular systolic dysfunction. The best of these is the presence of a raised JVP. This yields a specificity of 94.1% and a positive predictive value of 73.1%. A previous history of myocardial infarction taken in combination with the presence of other symptoms or signs also yields positive predictive values of greater than 60%. The best negative predictive value 60.8% is derived from interpretation of the ECG as being normal or abnormal.

View this table: [in this window] [in a new window]   Table 3 The value of clinical features from patients prescribed loop diuretics in primary care, for predicting left ventricular dysfunction at echocardiography (ejection fraction <40%)

 

View this table: [in this window] [in a new window]   Table 4 The value of combinations of patient clinical features from patients prescribed loop diuretics in primary care, for predicting left ventricular dysfunction at echocardiography (ejection fraction <40%)

 

	4. Discussion

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 
This study examines the prevalence of left ventricular systolic dysfunction in patients prescribed loop diuretics in a primary care setting. Previous studies have looked at the overall population prevalence of left ventricular systolic dysfunction [10,11], or examined patients referred to open access echocardiography services [14,15] and for radionucleotide ventriculography [16]. One previous study examined the echocardiographic features of 78 patients on loop diuretics for heart failure [7]. In that study 41% of patients had evidence of left ventricular systolic dysfunction.

Many studies of this nature endeavour to use data derived from a population of patients who attend hospital for investigation, to draw conclusions about the community population as a whole. In this study the population were unselected beyond those taking loop diuretics. Patients attended clinic voluntarily rather than at the behest of their general practitioner. However, even in this relatively unselected group there remain significant differences between those who attend and do not attend hospital clinic. Those attending clinic were more likely to be younger, male and have a past history of angina. The clinic patients were also taking significantly more cardiac acting drugs, notably nitrates, ACE inhibitors, digoxin, aspirin and warfarin, than those who did not attend. These population differences are important as most clinical evidence on investigation and management of patients with left ventricular systolic dysfunction is derived from hospital-based research. This study confirms the difficulty in obtaining a study population that genuinely reflects the community population as a whole.

The results of this study demonstrate that left ventricular systolic dysfunction, as defined by an ejection fraction of less than 40%, is common in this population. In total about half of all patients studied fulfilled these criteria. Furthermore, these results show that symptoms and signs are not an accurate predictor of left ventricular systolic dysfunction. In general practice this leads to difficulties in accurate patient diagnosis.

The NSF for ischaemic heart disease sets out an algorithm for the assessment of suspected heart failure in primary care. This recommends that if clinical assessment, patient history or hospital records suggest heart failure, a patient should be referred for echocardiography. The results of this study however suggest that specific clinical features or even combinations of features have poor predictive value for left ventricular systolic dysfunction in the population studied. Even the raised JVP with the best positive predictive value is not so strongly associated with left ventricular systolic dysfunction that patients could be treated in confidence without recourse to echocardiography. There is also no feature that has a sufficiently high negative predictive value to identify patients in whom left ventricular systolic dysfunction can confidently be excluded without referral for echocardiography. It is therefore likely that referring patients for echocardiography in the manner recommended by the NSF will leave a significant population of patients in the community with unidentified and under-treated left ventricular systolic dysfunction.

The results of this study suggest that general practitioners should have a low threshold for referring patients prescribed loop diuretics for echocardiography. The pre-test probability for left ventricular systolic dysfunction in the population studied is approximately 50%. This prevalence lies in the optimal range for carrying out an investigation, as this is the range within which a potential investigation can have the greatest impact on post-test probability [17]. Hence the clinical decision-making involved in loop diuretic prescription in primary care may of itself identify patients ‘ripe’ for echocardiography. Clearly this has implications for the provision of open access echocardiography services. From this study the prevalence of loop diuretic prescription is 2249/100 000 of which approximately 50% will have left ventricular systolic dysfunction. However, given that a number of these patients will be taking diuretics over a long period and a further substantial proportion of this elderly population may not attend hospital for echocardiography (47.7% in this study), the increase in service provision required may still be attainable.

For the remaining patients, a practical and cost effective test for diagnosing left ventricular systolic dysfunction is needed. It has been suggested that BNP levels may have a potential role in this area. However, we have recently found in our population that the correlation between BNP and left ventricular systolic dysfunction is less accurate than that reported in previous studies [18]. There is a significant difference (Wilcoxon Rank Sum Test; P<0.001) between the BNP values measured for patients with ejection fractions greater than 39% vs. those less than 40% (Table 3). However, the area under a receiver operator curve constructed from the BNP data derived from this study was 0.587. This indicates poor overall diagnostic accuracy of this test in this population. This is reflected by the predictive values for BNP shown in Table 4 where the median value for BNP in this population is used as a cut off. Research into the potential role of peptide hormones in the diagnosis of heart failure is ongoing.

The predictive values from this study are generally lower than those in a study of patients referred for open access echocardiography by Davie et al. [14]. Our results support the finding that a past history of myocardial infarction in combination with other clinical symptoms or signs, is one of the best predictors of left ventricular systolic dysfunction (Table 4). However, the relationship is not strong enough to recommend treating such patients without prior echo. Our results do not confirm the high negative predictive value of a normal ECG shown by Davie et al. [15]. These differences are likely to result from differences in the patient populations studied. Our patients were unselected beyond those taking loop diuretics that were willing and able to attend a research clinic. Those of Davie et al. were derived from patients selected by general practitioners for referral to an open access echocardiography service. There were also methodological differences between these studies. A cardiologist performed all the clinical assessments and echoes by Davie et al. Our study was designed to reflect the circumstances in community practice. Hence a primary care physician performed most of the clinical assessments and a trained technician performed the echocardiography.

One of the limitations of this study is the use of a population of patients already treated with diuretics. This may reduce the number of clinical symptoms and signs displayed by patients. Hence the predictive value of these clinical features may be better at the time of acute presentation. However, there is a poor correlation between the recorded general practice diagnosis at the time of first presentation and left ventricular systolic function. This suggests that even immediate clinical assessment may not be a sufficiently accurate predictor of left ventricular systolic function. A further problem is the assumption that most patients with left ventricular systolic dysfunction will be taking loop diuretics. Patients prescribed thiazide diuretics for symptoms and signs of heart failure and others with left ventricular systolic dysfunction not prescribed loop diuretics were not included in our study population. This study also focused on patients with left ventricular systolic dysfunction. Patients with diastolic dysfunction are not included in this analysis.

Finally, for practical reasons an individual investigator performed clinical assessment and ECG analysis. Echocardiography was carried out by a single technician. There is therefore no measure in this study of inter-observer variability.

The results of this study confirm the difficulty of accurate diagnosis of left ventricular systolic dysfunction in primary care. If the aim of the NSF to ‘help people with heart failure to live longer and achieve a better quality of life’ is to be achieved, greater diagnostic accuracy will be needed to identify patients who may benefit from specific treatments. This study suggests clinical features cannot be relied upon to achieve accurate diagnosis and general practitioners should therefore have a low threshold for referring patients requiring loop diuretics for echocardiography.

	Acknowledgements

 
The study was funded by a grant from Trent Regional Health Authority and the Nottinghamshire Multi-disciplinary Audit Advisory Group. Part of the funding for the echocardiography was supported by a grant from Merck, Sharp and Dohme.

	References

Top Abstract 1. Introduction 2. Method 3. Results 4. Discussion References

 

1. National Service Framework—Coronary Heart Disease. Department of Health, 2000.
2. CONSENSUS Trial Study Group. Effects of enalapril on mortality in severe congestive heart failure. Results of the Co-operative North Scandinavian Enalapril Survival Study (CONSENSUS). N Engl J Med (1987) 316:1429–1435.[Abstract]
3. SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N Engl J Med (1991) 325:293–302.[Abstract]
4. CIBIS-II Investigators and Committees. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet (1999) 353:9–13.[CrossRef][Web of Science][Medline]
5. MERIT-HF Study Group. Effect of metoprolol CR/XL in chronic heart failure: metoprolol CR/XL randomised intervention trial in congestive heart failure (MERIT-HF). Lancet (1999) 353:2001–2007.[CrossRef][Web of Science][Medline]
6. Pitt B., Zannad F., Remmie W.J., et al. The effect of spironolactone on morbidity and mortality in patients with severe heart failure. N Engl J Med (1999) 341:709–717.[Abstract/Free Full Text]
7. Wheeldon N.M., MacDonald T.M., Flucker C.J., et al. Echocardiography in chronic heart failure in the community. QJM (1993) 86:17–23.[Abstract/Free Full Text]
8. Clarke K.W., Gray D., Hampton J.R. Evidence of inadequate investigation and treatment of patients with heart failure. Br Heart J (1994) 71:584–587.[Abstract/Free Full Text]
9. Francis C.M., Caruana L., Kearney P., et al. Open access echocardiography in management of heart failure in the community. Br Med J (1995) 310:634–636.[Abstract/Free Full Text]
10. Mosterd A., Hoes A.W., de Bruyne M.C., et al. Prevalence of heart failure and left ventricular dysfunction in the general population. Eur Heart J (1999) 20:447–455.[Abstract/Free Full Text]
11. McDonagh T.A., Morrison C.E., Lawrence A., et al. Symptomatic and asymptomatic left-ventricular systolic dysfunction in an urban population. Lancet (1997) 350:829–833.[CrossRef][Web of Science][Medline]
12. Pfeffer M.A., Brunwald E., Moye L.A., et al. Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions. N Engl J Med (1992) 327:669–677.[Abstract]
13. Sparrow N, Adlam D, Cowley A, Hampton JR. Difficulties of introducing the national service framework for heart failure into general practice in the UK. Eur J Heart Fail 2003;5:355–360.
14. Davie A.P., Francis C.M., Caruana L., et al. Assessing diagnosis in heart failure: which features are any use? QJM (1997) 90:335–339.[Abstract/Free Full Text]
15. Davie A.P., Francis C.M., Caruana L., et al. Value of the electrocardiogram in identifying heart failure due to left ventricular systolic dysfunction. Br Med J (1996) 312:222.[Free Full Text]
16. Marantz P.R., Tobin J.N., Wassertheil-Smoller S., et al. The relationship between left ventricular systolic function and congestive heart failure diagnosed by clinical criteria. Circulation (1998) 77:607.
17. Sackett D.L., Haynes R.B., Guyatt G.H., Tubwell P. Clinical epidemiology a basic science for clinical medicine (1991) 2nd ed. London: Little Brown and Company.
18. Hetmanski D.J., Sparrow N.J., Curtis S., et al. Failure of plasma brain natriuretic peptide to identify left ventricular systolic dysfunction in the community. Heart (2000) 84:440–441.[Free Full Text]

CiteULike    Connotea    Del.icio.us    What's this?

This article has been cited by other articles:

				U. Dahlstrom, J. Hakansson, K. Swedberg, and A. Waldenstrom Adequacy of diagnosis and treatment of chronic heart failure in primary health care in Sweden Eur J Heart Fail, January 1, 2009; 11(1): 92 - 98. [Abstract] [Full Text] [PDF]

				T Fahey, S Jeyaseelan, C McCowan, E Carr, B. Goudie, S. Pringle, P. Donnan, F. Sullivan, and A. Struthers Diagnosis of left ventricular systolic dysfunction (LVSD): development and validation of a clinical prediction rule in primary care Fam. Pract., December 1, 2007; 24(6): 628 - 635. [Abstract] [Full Text] [PDF]

				M. F Dawwas Spironolactone and risk of upper gastrointestinal events: Association is unproved. BMJ, September 2, 2006; 333(7566): 501 - 501. [Full Text]

				N. Aspromonte, V. Ceci, A. Chiera, C. Coletta, A. D'Eri, M. Feola, P. Giovinazzo, L. Milani, F. Noventa, A. B. Scardovi, et al. Rapid Brain Natriuretic Peptide Test and Doppler Echocardiography for Early Diagnosis of Mild Heart Failure Clin. Chem., September 1, 2006; 52(9): 1802 - 1808. [Abstract] [Full Text] [PDF]

				J. Latour-Perez, F. J. Coves-Orts, C. Abad-Terrado, V. Abraira, and J. Zamora Accuracy of B-type natriuretic peptide levels in the diagnosis of left ventricular dysfunction and heart failure: A systematic review Eur J Heart Fail, June 1, 2006; 8(4): 390 - 399. [Abstract] [Full Text] [PDF]

				D. Adlam, P. Silcocks, and N. Sparrow Using BNP to develop a risk score for heart failure in primary care Eur. Heart J., June 1, 2005; 26(11): 1086 - 1093. [Abstract] [Full Text] [PDF]

				S J Leslie, S P McKee, E A Imray, and M A Denvir Management of chronic heart failure: perceived needs of general practitioners in light of the new general medical services contract Postgrad. Med. J., May 1, 2005; 81(955): 321 - 326. [Abstract] [Full Text] [PDF]

This Article Abstract FREE Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to My Personal Archive Download to citation manager Request Permissions Disclaimer Google Scholar Articles by Sparrow, N. Articles by Hampton, J. R. Search for Related Content PubMed PubMed Citation Articles by Sparrow, N. Articles by Hampton, J. R. Social Bookmarking          What's this?

Online ISSN 1879-0844 - Print ISSN 1388-9842

Copyright © 2010 European Society of Cardiology

Oxford Journals Oxford University Press

• Site Map
• Privacy Policy
• Frequently Asked Questions

Other Oxford University Press sites: Oxford University Press Oxford Journals China Oxford Journals Japan American National Biography Booksellers' Information Service Children's Fiction and Poetry Children's Reference Corporate & Special Sales Dictionaries Dictionary of National Biography Digital Reference English Language Teaching Higher Education Textbooks Humanities International Education Unit Law Medicine Music Online Products Oxford English Dictionary Reference Rights and Permissions Science School Books Social Sciences Very Short Introductions World's Classics