Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2007 August 25; 335(7616): 355–356.
PMCID: PMC1952519

Diagnosing atrial fibrillation in general practice

Henk C P M van Weert, assistant professor of general practice

The combination of a clinical history, clinical signs, and an ECG will pick up most cases

In this week's BMJ, Mant and colleagues and Fitzmaurice and colleagues present the results of the SAFE (screening for atrial fibrillation in the elderly) study. They assess how accurately general practitioners, practice nurses, and an interpretive computer program can diagnose atrial fibrillation on an electrocardiogram (ECG), and they report on the effectiveness of screening patients aged 65 and over for atrial fibrillation in British general practice.1 2 The prevalence of atrial fibrillation rises with age from 1.5% in people in their 60s to more than 10% in those over 90. People with atrial fibrillation have double the mortality and a four to fivefold higher risk of stroke than those without fibrillation. About a quarter of all strokes in elderly people are caused by atrial fibrillation. Strokes caused by atrial fibrillation are often severe and lead to high mortality and a low quality of life.3

Even if normal rhythm cannot be restored, antiplatelet agents reduce the risk of stroke by around 22% and vitamin K antagonists, such as warfarin, reduce the risk by 64% (number needed to treat for one year 37, for patients who have already had a transient ischaemic attack or stroke 12) 4 Thus, diagnosing atrial fibrillation is worthwhile, because effective interventions are available. However, intervention is not without risk and often requires lifelong drug treatment. Therefore, a diagnosis must be made on objective criteria. The 12 lead ECG is the reference standard, but interpretation can be difficult and misinterpretations often occur. 5

The study by Mant and colleagues assesses the accuracy of 49 general practitioners, 49 practice nurses, and interpretative software in diagnosing atrial fibrillation on ECG, without any clinical information.1 Sensitivity was around 80% in all three groups, but specificity was lower in nurses (85%) and general practitioners (92%) than with the software (99%). Because all three methods failed to diagnose about 20% of patients, none seems appropriate for screening purposes. A further disappointment was that training had little effect on the ability to interpret the ECGs correctly. Fortunately, one lead ECGs were as sensitive as 12 lead ECGs, and agreement between the two cardiologists who were the reference standard was very high. The logical conclusion is that a one lead ECG (which saves time and permits the use of loop recorders in daily practice) is sufficient for diagnosis, and that an experienced cardiologist should interpret the ECG.

Another small study in general practice has shown that an experienced nurse and general practitioner could diagnose atrial fibrillation on ECGs with a sensitivity of 96% and 100% and a specificity of 93% and 98%, respectively. 6 However, only one experienced nurse and one specially trained general practitioner interpreted the ECGs. The results of both of these studies are interesting, but the implications for general practice are limited. In daily practice general practitioners do not use ECGs to screen for atrial fibrillation. They use an ECG when disease is suspected, so probability of disease is higher than in the study by Mant and colleagues. Skills in interpreting ECGs improve when useful clinical information is available, especially when interpreters are less experienced. 7 When clinical information points to a rhythm disorder, a general practitioner will scrutinise the ECG for indicative signs, which probably increases sensitivity.

What diagnostic instruments does the general practitioner have to hand? The first is medical history and presenting symptoms. Symptoms of atrial fibrillation are palpitations, breathlessness, dizziness, chest discomfort, and stroke. About 10% of patients presenting with palpitations might have (paroxysmal) atrial fibrillation, but history and symptoms do not discriminate sufficiently between those with and without a serious rhythm disorder. 8

The first diagnostic test a general practitioner would use is to palpate the pulse for any irregularity, which has a sensitivity of 94% for detecting atrial fibrillation (determined in cohorts of elderly patients). However, because of the low specificity (72%) further diagnostic tests are needed. 9 In patients with an irregular pulse or high clinical suspicion the next test would be an ECG. If this shows atrial fibrillation, the diagnosis is clear. However, around a third of patients with atrial fibrillation will have paroxysmal atrial fibrillation. In these cases, a diagnosis is unlikely to be picked up on an ECG measured in the practice, and a patient activated loop recorder may be needed. 8 10

Many patients with atrial fibrillation do not have symptoms, so screening has been advocated because it is such a life threatening condition. 11 The SAFE study found that in general practice, screening leads to an increase of newly detected atrial fibrillation of 6/1000 patients aged 65 and over and provides evidence that simple opportunistic screening of elderly people is just as effective as a much more labour intensive systematic strategy and seems quite acceptable to the patients. Opportunistic screening involves feeling the pulse of elderly patients who visit their general practitioner for any reason, and carrying out electrocardiography if the pulse is irregular. 2 As most patients with atrial fibrillation have serious comorbidity they will visit a doctor regularly. When the ECG does not provide a diagnosis and doubt remains, an automatically triggered loop recorder could be used.12 This strategy will identify about one new case of atrial fibrillation for each 70 pulses taken. Five ECGs will have to be measured to find this one such patient. A general practitioner who is experienced in interpreting ECGs could do this, but general practitioners vary greatly in this respect. A sensible strategy would be for an experienced second reader, such as a cardiologist, to interpret the ECGs. Modern technology should make this feasible and not too expensive.


Competing interests: None declared.

Provenance and peer review: Commissioned; not externally peer reviewed.


1. Mant J, Fitzmaurice DA, Hobbs FDR, Jowettt S, Murray ET, Holder R, et al. Accuracy of diagnosing atrial fibrillation on electrocardiogram by primary care practitioners and interpretative diagnostic software: analysis of data from screening for atrial fibrillation in the elderly (SAFE) trial. BMJ 2007;335:380-2. [PMC free article] [PubMed]
2. Fitzmaurice DA, Hobbs FDR, Jowett S, Mant R, Murray ET, Holder R, et al. Screening versus routine practice in detection of atrial fibrillation in patients aged 65 or over: cluster randomised controlled trial. BMJ 2007;335:383-6. [PMC free article] [PubMed]
3. Lin HJ, Wolf PA, Kelly-Hayes M, Beiser AS, Kase CS, Benjamin EJ, et al. Stroke severity in atrial fibrillation: the Framingham study. Stroke 1996;27:1760-4. [PubMed]
4. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007;146:857-67. [PubMed]
5. Salerno SM, Alguire PC, Waxman HS. Competency in interpretation of 12-lead electrocardiograms: a summary and appraisal of published evidence. Ann Intern Med 2003;138:751-60. [PubMed]
6. Somerville S, Somerville J, Croft P, Lewis M. Atrial fibrillation: a comparison of methods to identify cases in general practice. Br J Gen Pract 2000;50:727-9. [PMC free article] [PubMed]
7. Hatala R, Norman GR, Brooks LR. Impact of a clinical scenario on accuracy of electrocardiogram interpretation. J Gen Intern Med 1999;14:126-9. [PubMed]
8. Hoefman E, van Weert HC, Reitsma JB, Koster RW, Bindels PJ. Diagnostic yield of patient-activated loop recorders for detecting heart rhythm abnormalities in general practice: a randomised clinical trial. Fam Pract 2005;22:478-84. [PubMed]
9. Cooke G, Doust J, Sanders S. Is pulse palpation helpful in detecting atrial fibrillation? A systematic review. J Fam Pract 2006;55:130-4. [PubMed]
10. Kinlay S, Leitch JW. Cardiac event recorders yield more diagnoses and are more cost-effective than 48-hour Holter monitoring in patients with palpitations. Ann Intern Med 1996;124:16-20. [PubMed]
11. Dewar RI, Lip GYH; on behalf of the guidelines development group for the NICE clinical guideline for the management of atrial fibrillation. Identification, diagnosis and assessment of atrial fibrillation. Heart 2007;93:25-8. [PMC free article] [PubMed]
12. Reiffel JA, Schwarzberg R, Murry M. Comparison of autotriggered memory loop recorders versus standard loop recorders versus 24-hour Holter monitors for arrhythmia detection. Am J Cardiol 2005;95:1055-9. [PubMed]

Articles from The BMJ are provided here courtesy of BMJ Publishing Group