Family Practice Vol. 20, No. 1, 16-21
© Oxford University Press 2003
Clinical Research |
Is treatment of atrial fibrillation in primary care based on thromboembolic risk assessment?
Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, The Netherlands.
a Present address Institute for Research in Extramural Medicine, Free University Medical Centre, Amsterdam, The Netherlands.
Correspondence to Frans H Rutten, MD; E-mail: F.H.Rutten{at}med.uu.nl
Rutten FH, Hak E, Stalman WAB, Verheij TJM and Hoes AW. Is treatment of atrial fibrillation in primary care based on thromboembolic risk assessment? Family Practice 2003; 20: 16–21.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background. Antithrombotic treatment in atrial fibrillation should be guided by the risk of thromboembolic events. Although practice studies have shown underutilization of antithrombotics, it is not clear whether physicians make use of thromboembolic risk stratification in their treatment decisions, as recommended in current guidelines.
Objective. The aim of the present study was to assess which clinical determinants influence coumarin and aspirin prescription in patients with atrial fibrillation in primary care.
Methods. In a cross-sectional study of 15 computerized general practices covering 38 000 patients in The Netherlands, we identified patients with atrial fibrillation using several search algorithms. Determinants of antithrombotic therapy were assessed using polytomous logistic regression analysis.
Results. A total of 247 patients with atrial fibrillation were identified among 10 335 patients aged 55 years or over (prevalence 2.4%). Mean age was 77 years (range 55–95); 51% were male. Overall, 52% of the patients used coumarin and 27% used aspirin. About 50% of patients were also treated by a cardiologist. Referred patients were prescribed coumarin more often (63% versus 35%, P < 0.001). Prior stroke/transient ischaemic attack or systemic embolism [adjusted odds ratio (OR) 5.3, 95% confidence interval (CI) 1.1–24.8], and chronic heart failure (adjusted OR 2.1, 95% CI 0.8–5.1) were independent determinants for coumarin prescription. These associations were less strong for aspirin prescription. However, other established risk factors for future thromboembolic events, such as hypertension and diabetes mellitus, did not lead to higher prescription rates. In contrast, several factors not associated with thromboembolic events, such as chronic (versus paroxysmal) atrial fibrillation (adjusted OR 3.7, 95% CI 1.7–8.2) and longstanding (versus recent) diagnosis (adjusted OR 2.2, 95% CI 1.1–4.5), were also associated with higher coumarin prescription.
Conclusion. These results illustrate that physicians’ decisions about initiating antithrombotic therapy in atrial fibrillation are not evidence based. They are guided partly by thromboembolic risk stratification, but also by factors insignificant for thromboembolic risk.
Keywords. Antithrombotics, atrial fibrillation, general practice, thromboembolic risk stratification.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Atrial fibrillation is the most common cardiac arrhythmia, with a prevalence of ~5% in subjects aged 60 years or over.1–3 The significance of atrial fibrillation lies in the increased risk of thromboembolic events, notably ischaemic strokes. The presence of non-valvular atrial fibrillation increases the risk of ischaemic stroke 5-fold,4 and the risk of mortality by 50 and 90% in men and women, respectively.5 Of all strokes, ~15% are attributable to atrial fibrillation.6
In the past decade, several randomized controlled trials provided evidence for the effectiveness of antithrombotic drugs in reducing the occurrence of ischaemic strokes and death in patients with non-rheumatic atrial fibrillation. The effect size of warfarin (68% relative risk reduction) seems more impressive than that of aspirin treatment (21% relative risk reduction),7,8 although others reported comparable effectiveness in a direct head-to-head comparison.9
Importantly, there is a marked clinical heterogeneity in risk of stroke among atrial fibrillation patients.10 Older patients with atrial fibrillation (aged >75 years), and atrial fibrillation patients with rheumatic valve disease, mitral valve stenosis, prior stroke/transient ischaemic attack or systemic embolism, hypertension, diabetes mellitus and possibly those with a history of chronic heart failure, are at an increased risk for future stroke,7,11–13 and are therefore more likely to benefit from warfarin treatment, despite the risk of iatrogenic bleeding. Patients without these risk factors are considered ‘low risk patients’, and possibly can best be treated with aspirin.10,14 Available guidelines recommend antithrombotic treatment in most patients with atrial fibrillation, but it must be individualized, taking into account the patient’s risk profile, bleeding risk and the impact of treatment on the quality of life.15 Available practice studies demonstrate efficacy of anticoagulation,16 but also underutilization of warfarin and aspirin treatment.2,17,18 It is known that patient’s health beliefs and attitudes influence the acceptance of treatment with warfarin.18 Little, however, is known about whether treatment of patients in primary care is guided by assessment of the absolute risk of thromboembolic events and bleeding risks.13,19 This knowledge, however, is indispensable to guide strategies to improve treatment uptake.19
We assessed which clinical factors, notably those related to thromboembolic risk, influence the physician’s choice for oral anticoagulation, antiplatelet treatment or no antithrombotic treatment in patients with atrial fibrillation.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In this cross-sectional study, the study population was recruited in 1997, from 15 general practices covering 38 000 patients in The Netherlands. These practices are part of a computerized general practice network, co-ordinated by the Julius Center for General Practice and Patient Oriented Research, University Medical Center Utrecht. Age and sex ratios in these practices are comparable with those of the Dutch population.20 Since 1992, GPs enter all data related to daily practice in a standardized manner into a general practice information system, classifying each patient contact and diagnosis according to the International Classification of Health Problems in Primary Care, second edition (ICHPPC-2).21,22
Data collection and definitions
Patients with atrial fibrillation were identified using several search algorithms. First, all patients with the International Classification of Primary Care code K78 (atrial fibrillation or atrial flutter) were tracked. Secondly, all computerized patient records covering the preceding 5 years (1992–1997) were searched for the Dutch (abbreviated) words referring to the diagnosis of atrial fibrillation; ‘fibril’, ‘atriumf’, ‘boezemf’ and ‘irr’. Finally, the patient files were searched for electrocardiographic-confirmed atrial fibrillation, since diagnosis of atrial fibrillation based on physical examination alone is not reliable enough.23
For all patients with atrial fibrillation, the GP was asked to fill out a questionnaire, including information on: (i) diagnosis of atrial fibrillation; (ii) treatment by a cardiologist; (iii) drug prescriptions; (iv) medical history, including co-morbidity and (possible) causes of atrial fibrillation; (v) prior thromboembolic events (i.e. stroke/transient ischaemic attack or systemic embolism elsewhere, such as mesenteric artery thrombosis); (vi) bleeding complications; and (vii) possible contraindications for coumarin or aspirin therapy, during the previous 5-year period.
Information on determinants of antithrombotic therapy that were considered clinically relevant were obtained: these included age, gender, time since diagnosis of atrial fibrillation, treatment by a cardiologist, prior thromboembolic events and co-morbidities. Co-morbidities included hypertension, diabetes mellitus, hypercholesterolaemia, coronary heart disease and possible causes of atrial fibrillation such as hyperthyroidism, chronic heart failure, chronic obstructive pulmonary disease, rheumatic heart valve disease and mitral valve stenosis. A prior thromboembolic events was defined as prior stroke/transient ischaemic attack or systemic embolism. In accordance with the Stroke Prevention in Atrial Fibrillation investigators, we considered patients with non-valvular atrial fibrillation as ‘low risk patients’ when they were without a history of hypertension, chronic heart failure, diabetes mellitus or thromboembolic event, and age <75 years.12 Iatrogenic complications included minor and major bleedings. Major bleeding was defined as intracranial haemorrhage, fatal bleeding or bleeding leading to admission to hospital, emergency procedures or urgent transfusion. All other bleedings were classified as minor. Contraindications for anticoagulation were increased bleeding risks (gastrointestinal or genitourinary bleeding or haemorrhagic stroke in the past), risk for falls, chronic alcohol abuse, renal failure and dementia. Both the GPs’ questionnaire and patient files were scrutinized to collect the data mentioned above. Patient files were anonymized before data collection. This study was approved by the medical ethical committee of the University Medical Center Utrecht, The Netherlands.
Statistical analysis
First, baseline characteristics of non-users and users of coumarin and aspirin were compared, using chi-square tests for categorical variables, and Students t-test for continuous variables. Of all potentially clinically relevant determinants of antithrombotic therapy (see above), univariable logistic regression analysis was performed to assess the association between clinical characteristics and the use of antithrombotics. Determinants associated with antithrombotic use (indicated by a P-value of <0.10) subsequently were included in a polytomous logistic regression analysis with the dependent variable comprising three categories: coumarin use, aspirin use and no use of antithrombotics (reference category). Thus, the independent influence of potential determinants on the prescription of aspirin, coumarin or non-use of antithrombotics was assessed. All analyses were performed using SPSS for Windows version 9.0 (SPSS, Chicago, IL, USA).
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
In all, 247 patients with electrocardiographic-confirmed atrial fibrillation were identified among the 10 334 patients aged 55 years or over (prevalence 2.4%). The prevalence clearly increased with age (Table 1
). Mean age of the patients with atrial fibrillation was 77 years (SD 9, range 55–94), and 51% were male. Sixty-two (25%) patients had paroxysmal, and 185 (75%) patients had chronic atrial fibrillation. Hypertension was the most prevalent co-morbidity (37%), and chronic heart failure (28%) the most common (possible) cause. Overall, 79% of the patients received antithrombotics, 127 (52%) coumarin and 65 (27%) aspirin (Table 2). Of the 14 atrial fibrillation patients with mitral valve stenosis or rheumatic valve disease, 12 (86%) patients received coumarin and two (14%) received no antithrombotic therapy (Table 2). Those patients (53%) treated by a cardiologist received coumarin more often (63% versus 38%, P < 0.001) and aspirin less often (21% versus 33%, P = 0.036) than those treated by the GP only, while 15% did not receive antithrombotic therapy (Table 2). In 44 patients, there were reasons for not prescribing coumarin. The most prominent reasons were non-compliance (13 patients, 30%) and increased bleeding risk (27%). Other reasons were dementia or risk of falls (20%), patient’s refusal (16%) and chronic alcohol abuse (7%). In five patients, there was more than one reason.
|
|
Polytomous logistic regression analysis showed that prior stroke/transient ischaemic attack or systemic embolism was a strong independent determinant of prescribing antithrombotics [adjusted odds ratio (OR) 3.6, 95% confidence interval (CI) 0.7–18.2 for aspirin and OR 5.3, 95% CI 1.1–24.8 for coumarin] (Table 3
). In addition, chronic (versus paroxysmal) atrial fibrillation, treatment by a cardiologist and longstanding diagnosis of atrial fibrillation (diagnosis >2 years versus 
2 years ago) were independent determinants for the use of coumarin (Table 3). The presence of diabetes mellitus or hypertension did not influence prescription rates of antithrombotics.
|
Atrial fibrillation patients with a ‘low risk profile’ (i.e. non-valvular atrial fibrillation without established risk factors and age <75 years) were treated less often with antithrombotics (adjusted OR 0.5, 95% CI 0.2–1.2 for aspirin and 0.6, 95% CI 0.3–1.2 for coumarin). There was no evidence for a shift from coumarin to aspirin treatment in this patient group.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Our study showed that antithrombotic treatment of atrial fibrillation in primary care was not guided by thromboembolic risk stratification, since it was not based on available scientific evidence. In particular, several known risk factors for future thromboembolism, such as hypertension and diabetes mellitus,7,11,12 did not influence antithrombotic prescriptions, while some factors not associated with future thromboembolic events, such as duration of atrial fibrillation or chronic (versus paroxysmal) atrial fibrillation, led to increased coumarin prescriptions.
In total, ~80% of the patients with atrial fibrillation received antithrombotic treatment. This proportion is higher than in other practice studies in primary care.2,17,18,24 Indeed, the number of patients that refused antithrombotic therapy (3%) or had contraindications (14%) for antithrombotics was rather low compared with other studies.2,25
Atrial fibrillation patients with a ‘low risk profile’ were treated less often with coumarin, but also aspirin. This is partly evidence based, since guidelines advise at least some kind of thromboembolic prevention in this category of patients.
The prevalence of atrial fibrillation found in our ‘practice’ study (2.4%) is about half of that observed in community-based studies on atrial fibrillation in a comparable age group. Interestingly, other ‘practice’ studies (i.e. based on available patient records) in primary care also reported a prevalence of atrial fibrillation in patients aged 
60 years remarkably similar (2.5%) to our estimate.17,26 This implies that a substantial number of patients with atrial fibrillation in the population at large are not known to the GP,27 probably because they do not experience or report symptoms.
Our study has several limitations. The (contra) indications of antithrombotic therapy were assessed retrospectively by using a GP questionnaire and patient files. Any misclassification resulting from the lack of prospective data, however, is likely to be small and non-differential. Secondly, our polytomous logistic regression analysis included several empty cells because of a limited number of patients in some of the exposure categories of the independent variables. However, we preferred the polytomous regression analysis because it allowed us to distinguish both coumarin and aspirin as separate outcome categories, in contrast to conventional multivariate logistic regression analysis. Furthermore, our study lacked statistical power to detect some clinical determinants, in particular those with a rather low prevalence, reflected by the wide 95% confidence interval.
Apart from clinical determinants, behavioural aspects (both physician and patient related) play a role in the decision to initiate antithrombotic therapy.13,15,18,28–30 Earlier studies have shown that the patients’ and physicians’ health beliefs and attitudes towards change and death are of pivotal importance for whether they would accept treatment with warfarin.18,28 In addition, practising clinicians are more likely to be influenced by treatment risks than by treatment benefits, leading to underutilization of antithrombotic treatment.31
Our results further illustrate that current evidence concerning thromboembolic risks in subgroups of patients with atrial fibrillation is not applied optimally in clinical practice. Initiation of antithrombotic therapy in atrial fibrillation is not always rational. Implementation of evidence-based guidelines to improve antithrombotic treatment uptake in patients with atrial fibrillation should be encouraged further.
|
Acknowledgments |
|---|
We thank the participating GPs: JA Andriesse, HAM Asbreuk, JG Blankestijn, JG Blommestein, JF Bolderink, MW van den Broek, GJA Daggelders, MEL van Dillen, WH Eizenga, EFHM Hendrickx, MM de Groot, R Hirsch, HHG de Jong, JMA Juffermans, L Truijens, HJ Nederhorst, N Nicolay, ME Numans, HM Pieters, CD Rijkens, BE van der Snoek, Y Stoutenbeek, NJ de Wit and WLG van Zijl.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
1 Kannel WB, Abbott RD, Savage DD, McNamara PM. Coronary heart disease and atrial fibrillation: the Framingham Study. Am Heart J 1983; 106: 389–396.[CrossRef][Web of Science][Medline]
2 Sudlow M, Thomson R, Thwaites B, Rodgers H, Kenny RA. Prevalence of atrial fibrillation and eligibility for anticoagulants in the community. Lancet 1998; 352: 1167–1171.[CrossRef][Web of Science][Medline]
3 Langenberg M, Hellemons BS, van Ree JW et al. Atrial fibrillation in elderly patients: prevalence and comorbidity in general practice. Br Med J 1996; 313: 1534.
4 Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke 1991; 22: 983–988.
5 Benjamin EJ, Wolf PA, D’Agostino RB, Silbershatz H, Kannel WB, Levy D. Impact of atrial fibrillation on the risk of death: the Framingham Heart Study. Circulation 1998; 98: 946–952.
6 Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation: a major contributor to stroke in the elderly. The Framingham Study. Arch Intern Med 1987; 147: 1561–1564.
7 Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation. Analysis of pooled data from five randomized controlled trials. Arch Intern Med 1994; 154: 1449–1457.
8 Atrial Fibrillation Investigators. The efficacy of aspirin in patients with atrial fibrillation. Analysis of pooled data from 3 randomized trials. Arch Intern Med 1997; 157: 1237–1240.
9 Taylor F, Cohen H, Ebrahim S. Systematic review of long term anticoagulation or antiplatelet treatment in patients with non-rheumatic atrial fibrillation. Br Med J 2001; 322: 321–326.
10 Stroke Prevention in Atrial Fibrillation Investigators. Patients with nonvalvular atrial fibrillation at low risk of stroke during treatment with aspirin: Stroke Prevention in Atrial Fibrillation III Study. J Am Med Assoc 1998; 279: 1273–1277.
11 Stroke Prevention in Atrial Fibrillation Investigators. Predictors of thromboembolism in atrial fibrillation: I. Clinical features of patients at risk. Ann Intern Med 1992; 116: 1–5.
12 Hart RG, Pearce LA, McBride R, Rothbart RM, Asinger RW. Factors associated with ischemic stroke during aspirin therapy in atrial fibrillation: analysis of 2012 participants in the SPAF I–III clinical trials. Stroke 1999; 30: 1223–1229.
13 Connolly SJ. Anticoagulation for patients with atrial fibrillation and risk factors for stroke. Warfarin reduces the risk by two thirds, but doctors still aren’t prescribing it enough [editorial]. Br Med J 2000; 320: 1219–1220.
14 Hellemons B, Langenberg M, Lodder J et al. Primary prevention of arterial thromboembolism in non-rheumatic atrial fibrillation in primary care: randomised controlled trial comparing two intensities of coumarin with aspirin. Br Med J 1999; 319: 958–964.
15 Lancaster TR, Singer DE, Sheehan MA et al. The impact of long-term warfarin therapy on quality of life. Evidence from a randomized trial. Arch Intern Med 1991; 151: 1944–1949.
16 Kalra L, Yu G, Perez I, Lakhani A, Donaldson N. Prospective cohort study to determine if trial efficacy of anticoagulation for stroke prevention in atrial fibrillation translates into clinical effectiveness. Br Med J 2000; 320: 1236–1239.
17 Lip GY, Golding DJ, Nazir M, Beevers DG, Child DL, Fletcher RI. A survey of atrial fibrillation in general practice: the West Birmingham Atrial Fibrillation Project. Br J Gen Pract 1997; 47: 285–289.[Web of Science][Medline]
18 Howitt A, Armstrong D. Implementing evidence based medicine in general practice: audit and qualitative study of antithrombotic treatment for atrial fibrillation. Br Med J 1999; 318: 1324–1327.
19 Bungard T, Ghali W, Teo K, McAlister F, Tsuyuki R. Why do patients with atrial fibrillation not receive warfarin? Arch Intern Med 2000; 160: 41–46.
20 Stalman, WAB. Management of sinusitis-like complaints in general practice. Thesis Utrecht University, The Netherlands, 1997.
21 Classification Committee of WONCA. ICHPPC-2 Defined. International Classification of Primary Care. Oxford: Oxford University Press, 1983.
22 Metsemakers JF, Höppener P, Knottnerus JA, Kocken RJ, Limonard CB. Computerized health information in The Netherlands: a registration network of family practices. Br J Gen Pract 1992; 42: 102–106.[Web of Science][Medline]
23 Sudlow M, Rodgers H, Kenny RA, Thomson R. Identification of patients with atrial fibrillation in general practice: a study of screening methods. Br Med J 1998; 317: 327–328.
24 Gurwitz JH, Monette J, Rochon PA, Eckler MA, Avorn J. Atrial fibrillation and stroke prevention with warfarin in the long-term care setting. Arch Intern Med 1997; 157: 978–984.
25 Antani MR, Beyth RJ, Covinsky KE et al. Failure to prescribe warfarin to patients with nonrheumatic atrial fibrillation. J Gen Intern Med 1996; 11: 713–720.[Web of Science][Medline]
26 Hill J, Mottram E, Killeen P. Study of the prevalence of atrial fibrillation in general practice patients over 65 years of age. J R Coll Gen Pract 1987; 37: 172–173.[Web of Science][Medline]
27 Rutten FH, Hak E, Hoes AW. Using anticoagulation or aspirin to prevent stroke. Results of the study cannot be generalised to the general practice population [letter]. Br Med J 2000; 320: 1009–1010.[Medline]
28 Devereaux P, Anderson D, Gardner M et al. Differences between perspectives of physicians and patients on anticoagulation in patients with atrial fibrillation: observational study. Br Med J 2001; 323: 1–7.
29 Rodgers H, Sudlow M, Dobson R, Kenny RA, Thomson RG. Warfarin anticoagulation in primary care: a regional survey of present practice and clinicians’ views. Br J Gen Pract 1997; 47: 309–310.[Web of Science][Medline]
30 Protheroe J, Fahey T, Montgomery AA, Peters TJ. The impact of patients’ preferences on the treatment of atrial fibrillation: observational study of patient based decision analysis. Br Med J 2000; 320: 1380–1384.
31 Hobbs R. Identification and treatment of patients with atrial fibrillation in primary care [editorial]. Heart 1999; 81: 333–334.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  L. Friberg, N. Hammar, M. Ringh, H. Pettersson, and M. Rosenqvist Stroke prophylaxis in atrial fibrillation: who gets it and who does not?: Report from the Stockholm Cohort-study on Atrial Fibrillation (SCAF-study) Eur. Heart J., August 2, 2006; 27(16): 1954 - 1964. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. G. Manuel, P. Tanuseputro, C. A. Mustard, S. E. Schultz, G. M. Anderson, S. Ardal, D. A. Alter, and A. Laupacis The 2003 Canadian recommendations for dyslipidemia management: Revisions are needed Can. Med. Assoc. J., April 12, 2005; 172(8): 1027 - 1031. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Glasziou and B. Haynes The paths from research to improved health outcomes Evid. Based Nurs., April 1, 2005; 8(2): 36 - 38. [Full Text] [PDF]
|
|
|
|
 |
|
 P. Glasziou and B. Haynes The paths from research to improved health outcomes Evid. Based Med., February 1, 2005; 10(1): 4 - 7. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||