NIHR Signal Screen reminders for GPs did not improve anticoagulant prescribing in atrial fibrillation

Published on 18 July 2017

General practice software that generated screen reminders for patients with atrial fibrillation did not increase the proportion taking oral anticoagulants appropriately by six months.

This NIHR-funded trial included GPs in 47 surgeries in England and found that at the start only 63% of eligible patients with atrial fibrillation were being prescribed anticoagulants. Six months later the rate had increased to 66% in intervention practices and 64% in those following usual practice, a non-significant difference between the groups.

Use of the software was associated with increased diagnosis of transient ischemic attack, which could be due to improved detection or over-diagnosis. A reduction in strokes of any type was of borderline statistical significance after 12 months.

Longer term studies, collecting more detail on anticoagulant prescribing and the reasons behind patient and doctor choices, may support practice improvements in the future.

Screen reminders for GPs did not improve anticoagulant prescribing in atrial fibrillation

Share your views on the research.

Why was this study needed?

Atrial fibrillation is an irregular heart rhythm that in 2015 was estimated to affect almost 1.4 million people in England, although over 30% were not diagnosed. Atrial fibrillation contributes to about 12,500 strokes annually in the UK, and each stroke is estimated to cost around £12,000. Effective management of atrial fibrillation could reduce the number of strokes and healthcare costs.

In 2015 more than one in five patients admitted to a hospital with stroke were known to have atrial fibrillation, but less than half were taking anticoagulants.  It is recognised that uptake is not optimal.

In order to improve management of atrial fibrillation it is important to understand why some patients are not taking preventative medicines. This study addressed this challenge by testing software that generated screen reminders when atrial fibrillation patients were seen in general practice.

What did this study do?

This cluster-randomised controlled trial assigned 47 general practices in England to either usual practice or to a new proprietary software tool – Automated Risk Assessment for Stroke in Atrial Fibrillation. The software generated screen reminders when patients eligible for treatment with oral anticoagulants were seen by a clinician, who then had to indicate a reason if treatment was not initiated.

The practices had a combined population of 5,339 atrial fibrillation patients eligible for oral anticoagulants at the beginning of the study, 62.6% of whom were being treated.

The main outcome was the proportion of these patients receiving oral anticoagulants after the six-month intervention period. Secondary outcomes included cardiovascular events and any adverse reports from using the software, at six and 12 months.

What did it find?

  • By six months the proportion of eligible patients receiving oral anticoagulants was 66.3% in the intervention group compared with 63.9% in the control group. This difference was not statistically significant (adjusted mean difference 1.21%, 95% confidence interval [CI] -0.72 to 3.13).
  • After 12 months, the rate of transient ischemic attack among patients with atrial fibrillation was higher in the intervention group: diagnosed in 10 per 1,000 (95% CI 4.2 to 18.2) compared with 2.3 patients per 1,000 (95% CI 0 to 9.0) in the control group.
  • There was a lower rate of haemorrhage in the intervention group at 12 months was of borderline statistical significance: event rate 34.7 patients per 1,000 (95% CI 27.4 to 43.6) at 12 months vs 50.3 per 1,000 (95% CI 33.4 to 57.3) in the control group (p=0.054).  This included haemorrhagic stroke, though there was no difference in the rate of haemorrhagic stroke specifically, or of stroke due to a blood clot. Differences in the rate of all-cause stroke fell short of statistical significance. Incidence of 15 per 1,000 in the intervention group compared with 25 per 1,000 in the controls (p=0.06).
  • There were no reports of adverse effects of the software, such as inappropriate clinical or prescribing decisions.

What does current guidance say on this issue?

The NICE 2014 guideline on atrial fibrillation recommends using the validated CHA2DS2-VASc tool to assess stroke risk and prescribing anticoagulation to people with score two or above, taking bleeding risk into account. Anticoagulation may be considered for men with a stroke risk score of one or above. NICE recommends anticoagulant options are discussed with the patient and that choice is based on their clinical features and preferences.

The use of aspirin as a single treatment (monotherapy) for stroke prevention in people with atrial fibrillation is not recommended.

What are the implications?

It would be useful to explore the reason for these findings and better understand why the software made little difference to anticoagulation prescribing.

The researchers suggest this could be due to practice improving in the control clinics as a consequence of clinicians being aware of being part of a trial. It is possible that a longer intervention period is needed to see the benefits of screen reminders in stroke prevention, supported by the borderline significant reduction of cardiovascular events at 12 months.

There has been an increase in prescription of newer oral anticoagulants since this trial started and it is not clear if this may have altered the discussion of risks with patients in general practice.

Citation and Funding

Holt TA, Dalton A, Marshall T, et al. Automated Software System to Promote Anticoagulation and Reduce Stroke Risk: Cluster-Randomized Controlled Trial. Stroke. 2017r;48(3):787-90.

This project was funded by the National Institute for Health Research [School for Primary Care Research] (grant FR6/173).

Bibliography

ABPI. One year on: Why are patients still having unnecessary AD-related strokes? London: The Association of the British Pharmaceutical Industry; 2015.

NHS Choices. Atrial fibrillation. London: Department of Health; 2015.

NHS. Cost impact and benefits. Manchester; 2014.

NICE. Atrial fibrillation: management. CG180. London: National Institute for Health and Clinical Excellence; 2014.

NICE. Thousands of strokes and deaths preventable from “silent killer”. London: National Institute for Health and Clinical Excellence; 2014.

Public Health England. Atrial fibrillation prevalence estimates in England: Application of recent population estimates of AF in Sweden. London: Pubic Health England; 2015.

RCP Sentinel Stroke National Audit Programme (SSNAP). Clinical audit April-June 2015: Public report. London: Royal College of Physicians; 2015.

Why was this study needed?

Atrial fibrillation is an irregular heart rhythm that in 2015 was estimated to affect almost 1.4 million people in England, although over 30% were not diagnosed. Atrial fibrillation contributes to about 12,500 strokes annually in the UK, and each stroke is estimated to cost around £12,000. Effective management of atrial fibrillation could reduce the number of strokes and healthcare costs.

In 2015 more than one in five patients admitted to a hospital with stroke were known to have atrial fibrillation, but less than half were taking anticoagulants.  It is recognised that uptake is not optimal.

In order to improve management of atrial fibrillation it is important to understand why some patients are not taking preventative medicines. This study addressed this challenge by testing software that generated screen reminders when atrial fibrillation patients were seen in general practice.

What did this study do?

This cluster-randomised controlled trial assigned 47 general practices in England to either usual practice or to a new proprietary software tool – Automated Risk Assessment for Stroke in Atrial Fibrillation. The software generated screen reminders when patients eligible for treatment with oral anticoagulants were seen by a clinician, who then had to indicate a reason if treatment was not initiated.

The practices had a combined population of 5,339 atrial fibrillation patients eligible for oral anticoagulants at the beginning of the study, 62.6% of whom were being treated.

The main outcome was the proportion of these patients receiving oral anticoagulants after the six-month intervention period. Secondary outcomes included cardiovascular events and any adverse reports from using the software, at six and 12 months.

What did it find?

  • By six months the proportion of eligible patients receiving oral anticoagulants was 66.3% in the intervention group compared with 63.9% in the control group. This difference was not statistically significant (adjusted mean difference 1.21%, 95% confidence interval [CI] -0.72 to 3.13).
  • After 12 months, the rate of transient ischemic attack among patients with atrial fibrillation was higher in the intervention group: diagnosed in 10 per 1,000 (95% CI 4.2 to 18.2) compared with 2.3 patients per 1,000 (95% CI 0 to 9.0) in the control group.
  • There was a lower rate of haemorrhage in the intervention group at 12 months was of borderline statistical significance: event rate 34.7 patients per 1,000 (95% CI 27.4 to 43.6) at 12 months vs 50.3 per 1,000 (95% CI 33.4 to 57.3) in the control group (p=0.054).  This included haemorrhagic stroke, though there was no difference in the rate of haemorrhagic stroke specifically, or of stroke due to a blood clot. Differences in the rate of all-cause stroke fell short of statistical significance. Incidence of 15 per 1,000 in the intervention group compared with 25 per 1,000 in the controls (p=0.06).
  • There were no reports of adverse effects of the software, such as inappropriate clinical or prescribing decisions.

What does current guidance say on this issue?

The NICE 2014 guideline on atrial fibrillation recommends using the validated CHA2DS2-VASc tool to assess stroke risk and prescribing anticoagulation to people with score two or above, taking bleeding risk into account. Anticoagulation may be considered for men with a stroke risk score of one or above. NICE recommends anticoagulant options are discussed with the patient and that choice is based on their clinical features and preferences.

The use of aspirin as a single treatment (monotherapy) for stroke prevention in people with atrial fibrillation is not recommended.

What are the implications?

It would be useful to explore the reason for these findings and better understand why the software made little difference to anticoagulation prescribing.

The researchers suggest this could be due to practice improving in the control clinics as a consequence of clinicians being aware of being part of a trial. It is possible that a longer intervention period is needed to see the benefits of screen reminders in stroke prevention, supported by the borderline significant reduction of cardiovascular events at 12 months.

There has been an increase in prescription of newer oral anticoagulants since this trial started and it is not clear if this may have altered the discussion of risks with patients in general practice.

Citation and Funding

Holt TA, Dalton A, Marshall T, et al. Automated Software System to Promote Anticoagulation and Reduce Stroke Risk: Cluster-Randomized Controlled Trial. Stroke. 2017r;48(3):787-90.

This project was funded by the National Institute for Health Research [School for Primary Care Research] (grant FR6/173).

Bibliography

ABPI. One year on: Why are patients still having unnecessary AD-related strokes? London: The Association of the British Pharmaceutical Industry; 2015.

NHS Choices. Atrial fibrillation. London: Department of Health; 2015.

NHS. Cost impact and benefits. Manchester; 2014.

NICE. Atrial fibrillation: management. CG180. London: National Institute for Health and Clinical Excellence; 2014.

NICE. Thousands of strokes and deaths preventable from “silent killer”. London: National Institute for Health and Clinical Excellence; 2014.

Public Health England. Atrial fibrillation prevalence estimates in England: Application of recent population estimates of AF in Sweden. London: Pubic Health England; 2015.

RCP Sentinel Stroke National Audit Programme (SSNAP). Clinical audit April-June 2015: Public report. London: Royal College of Physicians; 2015.

Automated Software System to Promote Anticoagulation and Reduce Stroke Risk

Published on 24 January 2017

T Holt, A Dalton, T Marshall, M Fay, N Qureshi, S Kirkpatrick, J Hislop, D Lasserson, K Kearley, J Mollison, L Yu. R Hobbs, D Fitzmaurice

Stroke , 2017

Background and Purpose—Oral anticoagulants (OAC) substantially reduce risk of stroke in atrial fibrillation, but uptake is suboptimal. Electronic health records enable automated identification of people at risk but not receiving treatment. We investigated the effectiveness of a software tool (AURAS-AF [Automated Risk Assessment for Stroke in Atrial Fibrillation]) designed to identify such individuals during routine care through a cluster-randomized trial. Methods—Screen reminders appeared each time the electronic health records of an eligible patient was accessed until a decision had been taken over OAC treatment. Where OAC was not started, clinicians were prompted to indicate a reason. Control practices continued usual care. The primary outcome was the proportion of eligible individuals receiving OAC at 6 months. Secondary outcomes included rates of cardiovascular events and reports of adverse effects of the software on clinical decision-making. Results—Forty-seven practices were randomized. The mean proportion–prescribed OAC at 6 months was 66.3% (SD=9.3) in the intervention arm and 63.9% (9.5) in the control arm (adjusted difference 1.21% [95% confidence interval −0.72 to 3.13]). Incidence of recorded transient ischemic attack was higher in the intervention practices (median 10.0 versus 2.3 per 1000 patients with atrial fibrillation; P=0.027), but at 12 months, we found a lower incidence of both all cause stroke (P=0.06) and hemorrhage (P=0.054). No adverse effects of the software were reported. Conclusions—No significant change in OAC prescribing occurred. A greater rate of diagnosis of transient ischemic attack (possibly because of improved detection or overdiagnosis) was associated with a reduction (of borderline significance) in stroke and hemorrhage over 12 months.

Expert commentary

The use of on-screen reminders in general practice has escalated in the past decade, despite the modest benefits found by the Cochrane review cited in this paper. Perhaps this proliferation explains the negative result found in this trial. By necessity GPs are becoming more selective about which reminders they respond to.

Consultations with patients having atrial fibrillation will often include considerations of co-morbidity as well as the patients’ own agendas. Complex shared decisions such as starting anticoagulants need more time than GPs can currently offer.

Andrew Wilson, Professor of Primary Care Research, University of Leicester