General Practice

Population based study of use of anticoagulants among patients with atrial fibrillation in the community

BMJ 1997; 314 doi: (Published 24 May 1997) Cite this as: BMJ 1997;314:1529
  1. Mark Sudlowa, MRC training fellow in health services research,
  2. Helen Rodgers, senior lecturer in stroke medicine and servicesa,
  3. Anne Kenny Rose, professor of geriatric medicineb,
  4. Richard Thomson, senior lecturer in epidemiology and public healthc
  1. a Departments of Medicine and Epidemiology and Public Health, The Medical School, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH
  2. b Department of Medicine, The Medical School
  3. c Department of Epidemiology and Public Health, The Medical School
  1. Correspondence to: Dr Sudlow
  • Accepted 12 November 1996


Several randomised controlled trials have shown that warfarin treatment for patients with atrial fibrillation substantially reduces their risk of stroke.1 Studies have found low treatment rates among patients with atrial fibrillation in hospital2 and in primary care,3 but these have been limited by their reliance on identifying patients with atrial fibrillation from coding of records and prescription of antiarrhythmic drugs. We report the use of warfarin among patients with atrial fibrillation in a community survey.

Concern has been expressed about the high rates of exclusion of subjects from randomised trials of warfarin treatment. It has been suggested that the use of similar exclusions in clinical practice would greatly reduce the number of patients eligible for treatment,4 which might explain the low treatment rates. We therefore applied exclusion criteria similar to one of these trials5 to our subjects.

Subjects, methods, and results

As part of a study of atrial fibrillation in the community, we took an age and sex stratified random sample of patients aged 65 and over registered with 10 contiguous general practices in Northumberland, which covered one market town, one industrial town, a dormitory town, and the mining villages and farming communities around them. We invited subjects to attend for electrocardiography, measurement of blood pressure, and completion of a questionnaire including information on contraindications to anticoagulation. We also recorded their current medication.

We identified subjects with atrial fibrillation or flutter from their electrocardiograms, took blood samples, and reviewed patients' medical notes. We sent questionnaires to the subjects' general practitioners asking about their patient's ability to comply with treatment. We derived contraindications to warfarin treatment from the exclusion criteria for the stroke prevention in atrial fibrillation trial5:

  • Self reported history of vomiting blood, rectal blood loss, or haematuria in the six months before the study; alcohol consumption over 28 units in the past week; three or more falls in the past year; and daily use of non-steroidal anti-inflammatory drugs other than aspirin (when questionnaire data were missing we scrutinised medical records for a history of recent bleeding, excessive alcohol consumption, or falls)

  • General practitioner's report of inability to comply with warfarin treatment (when general practitioners did not respond we used a test of ability–subjects failed if they could not read the label on a bottle of warfarin, extract a single tablet, or pick a specified dose after the strength and colour coding of warfarin tablets had been explained)

  • Haemoglobin concentration <100 g/l, platelet count <100x109/l, prothrombin time >15.1 s, and creatinine concentration >300 μmol/l (when subjects refused to give a blood sample we scrutinised medical records for a current diagnosis of liver disease or anaemia)

  • Uncontrolled hypertension at the initial visit–blood pressure >180/100.

The response rate to the survey was 77% (1530/1990), and 100 subjects had atrial fibrillation. Of the subjects for whom notes were available, atrial fibrillation was recorded before the study in the notes of 76% (71/93). We excluded nine subjects–notes could not be traced for seven and for three there was inadequate information to exclude contraindications. Table 1) summarises the results.

Table 1

Details of warfarin treatment for 100 subjects aged ≥65 with atrial fibrillation (values are numbers (percentages (95% confidence interval of percentage)) unless stated otherwise)

View this table:

In our study about half of the patients aged 65-74 with atrial fibrillation were treated with warfarin. A much lower proportion of those aged over 74 were treated (see 1). These low rates can be explained only partially by the presence of contraindications or because subjects were not previously identified by their general practitioner as having atrial fibrillation. It is possible that factors beyond those we considered as contraindications deterred doctors from using warfarin, but the criteria we used cover accepted medical contraindications to warfarin, including poor compliance and falls. If adequate services were available then it should be possible to safely give anticoagulant drugs to most patients without such contraindications. Since treatment is of such benefit and need so widespread, there is an imperative to improve and expand the current use of warfarin.


Funding: The study was funded by the Stroke Association. MS is funded by an MRC Training Fellowship in Health Services Research.

Conflict of interest: None.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract