Dr David Archer explains how his practice improved standards of care for their patients on warfarin using local guidelines


The demand for anticoagulation services has increased with the recognition of atrial fibrillation as a major risk factor for strokes 1. As fundholders, we felt we could offer a better service to patients at lower cost by running an anticoagulant clinic, for patients on warfarin therapy, at the surgery. We wanted to demonstrate that such clinics can be managed safely and effectively in primary care 2.


Over the past five years, increasing emphasis has been placed on patient-centred care. The change of service delivery closer to the patient's home, in their home, or at the local doctor's surgery have been brought about because of increasing emphasis on quality.

Rapid response teams comprising social services, home care and the district nursing service are an example of this. They enable the patient to receive exemplary care in their own home rather than in hospital thus avoiding the risk of institutionalisation.

Our practice has long been interested in delivering care as close to the patient's home as possible. The partners were concerned that a patient who needed transport to a hospital-run anticoagulant clinic might be picked up by ambulance as early as eight in the morning. A ride around the town and other areas of Hemel Hempstead may see the patient in clinic by ten, bloods taken, an hour wait for results and then back in the ambulance by two. Frequently, patients, many of them elderly and/or infirm returned home by four, completely exhausted, having forgotten to inform meals on wheels of their clinic time and missed lunch. We believed that there must be a way to improve the quality of this service for these people.

Most patients needing anticoagulation only need warfarin for three months. These patients have usually had deep vein thrombosis, requiring heparinisation either in hospital or at home. Some patients, such as those diagnosed with a pulmonary embolus, need six months' treatment. However, there is an increasing number requiring life-long warfarin for certain conditions, the most common being lone atrial fibrillation. These patients often have a comorbidity that makes hospital transport difficult, but warfarin treatment essential. The clinic was originally set up for these patients, but once the local haematologist was satisfied that the clinic was running smoothly, the practice saw a gradual shift of patients from secondary care provision to their local surgery.


The first requirement was to identify those patients on warfarin. A number of different approaches had to be made in order to ensure that all patients were captured into the one datafile. Obviously, the easiest search was on repeat prescriptions, but this usually only identified long-term warfarin usage. The hospital anticoagulant clinic was asked for a list of current patients from the surgery, and the practice computer was interrogated for patients with atrial fibrillation, deep vein thrombosis or pulmonary emboli. However, those patients requiring INR targets of 3.5 (see below) would be managed more appropriately in hospital settings, because of the increased risk of haemorrhage if control was less than ideal.


Patients were then categorised by their previous control in order to ascertain frequency of monitoring. Some, mainly those on long-term treatment were relatively well controlled and would need testing relatively infrequently, often only monthly. Others, whose control was more erratic, needed almost perpetual testing in order to ensure maximum benefit from treatment. We decided initially to familiarise ourselves with the INR results from the hospital, rather than investing in near-patient testing.

The results of the INRs are faxed to the surgery. Similarly we have not yet purchased computer decision support software systems to manage individual INR changes as the patient numbers in general practice were too small to warrant the expense. This has meant that in practice, the INRs have been checked more frequently.

The patients were then booked into the anticoagulation clinic. They were taken through a questionnaire, which was required for optimum control. The patients requiring treatment, the clinician initiating treatment, and the clinical indication for treatment were all recorded. The target INR was noted, and the duration of treatment. Previous personal or family history of venous thromboembolism was ascertained together with concurrent medical conditions and drug therapies. The amount of alcohol consumed weekly was discussed, and where relevant, mode of contraception.

The locally agreed guidelines for various conditions requiring anticoagulation with the therapeutic ranges, INR targets and duration are shown in table 1.

Table 1: Therapeutic ranges, INR targets and duration of conditions requiring anticoagulation
Deep vein thrombosis
2.0 – 3.0
3 months
Pulmonary embolus
2.0 – 3.0
6 months
Tissue heart valves
2.0 – 3.0
Lifelong AF with valvular disease
3.0 – 4.5
Lifelong mechanical heart valves
3.0 – 4.5
Lifelong recurrent venous embolism
3.0 – 4.5
Lifelong lone AF
2.0 – 3.0


Lifelong dilated cardiomyopathy
2.0 – 3.0



During the first appointment, patients are counselled about the mode of action of warfarin, complications of poor control, (bleeding or bruising), the appropriate action to take if this should occur. Self-prescribing with aspirin or health shop remedies is also discussed. The patient is consulted on alcohol, diet, immunisations, surgical procedures (especially dentistry) as well as hobbies and leisure activities. Patients are given a warfarin card, a patient information booklet (figure 1), and shown the different tablet colours. Blood samples are taken in the morning.

Figure 1: Patient information from the anticoagulant therapy record
patients information

Warfarin is usually started on Day 1 and 2 at 10mg and then depending on the INR on Day 3 the dosage modified.3 Testing is obviously more frequent when the patient has only just started therapy, (weekly or twice a week), but as the dosage requirements become clearer, the interval can be increased to as much as monthly. However even in the best of control, testing should never be less frequent than 12 weeks. Following a dose adjustment, (with INR outside the therapeutic range), testing should be brought forward to assess the effect of the change.


The practice is following local and regional audits of anticoagulant control. These audits have shown that at any time:

  • Fewer than 5% of patients should be over-anticoagulated
  • Fewer than 5% should be under-anticoagulated
  • Each patient should spend more than 60% of the therapy time within the therapeutic range, the practice setting the standard nearer to 80%.

Initially, the results were faxed back to the practice in the evening at 5pm. It was one of the on-call doctors tasks to comment on all the results, write instructions in the daybook for the receptionists to give to the patients that evening and enter into the patient's notes. Unfortunately, there was a problem with this. The practice is run from two sites and the notes of the 1,200 patients at the branch surgery are not immediately to hand.

Sometimes, best practice could not be achieved, as the INRs were kept at the branch surgery, and the range was unknown. As soon as this problem was discovered, all target INRs were placed on the practice computer, as are previous results, therapeutic range, in order to aid the on-call doctor in their decision making. The computer database now has target INR, results ±0.5 from target limits, and outside ±0.5 target limits. The last ten results are recorded and this is used for audit control.


Despite the siren's call that anticoagulation belongs in secondary care, the practice has tried to improve services to their patients. Most patients prefer the quick turnaround in the surgery, with a telephone call in the evening. The long ambulance journey belongs to a prehistoric era, which in retrospect was not patient-centred.

Without doubt, the service could not run if the doctors and nurses of the practice stopped at the primary/secondary care interface. Care in the community is not rhetoric, it can really happen, the patients want it: as health care professionals, we should listen to our patients.


  1. Atrial Fibrillation Investigators. Risk factors for stroke and efficacy of antithrombolytic therapy in atrial fibrillation. Ann Intern Med. 1994:154: 1449-57.
  2. Pell J P, McIver B, Stuart P et al. Comparison of anticoagulant control among patients attending general practice and hospital anticoagulant clinic. Br J Gen Practice 1993:43: 152-54
  3. Fennerty A, Dolben J, Thomas P et al. Flexible induction dose regimen for warfarin and prediction of maintenance dose. Br Med J 1984:277: 1268-70

Guidelines in Practice, October 1998, Volume 1
© 1998 MGP Ltd
further information | subscribe