Safe and effective management of stroke prevention in atrial fibrillation

The aim of AF treatment is to prevent complications, principally stroke, and alleviate symptoms. Current guidelines from NICE  and also the European Society of Cardiology (ESC) address several clinical areas in which new evidence has become available for stroke and bleeding risk stratification.

Pharmacological therapy recommended to reduce the risk of stroke in AF now only comprises of anticoagulants, with clear evidence to support the fact that anticoagulation with vitamin K antagonists (VKAs) or direct oral anticoagulants (DOACs) reduce stroke and mortality in patients with AF. The recommendations set out by NICE for stroke prevention apply equally, to adults with atrial fibrillation including paroxysmal, persistent and permanent, and atrial flutter

The core aims of this programme included:

• To identify and project actual disease burden (England average prevalence for atrial fibrillation is 1.6% with wide variations amongst practice populations).
• To stratify AF patients according to their level of stroke risk whilst ensuring that patient coding is reflective of actual levels of risk. Stroke risk profiles were based on CHA2DS2-VASc scoring, in line with current guidelines.
• Support the therapeutic management of stroke risk in AF patients in accordance with risk profile, current guidelines and practice preferred anticoagulation strategy.
• Provide patients with educational material relating to stroke prevention in AF in the form of treatment compliance support and lifestyle advice.
• Assist GP practices in the achievement of Quality and Outcome Clinical Indicator targets for AF including assistance with improving the accuracy of prevalence and disease registers.
• Provide an educational legacy that will support ongoing stroke risk profiling and management of antithrombotic therapy.

An initial prevalence of 1.84% demonstrated a higher than average burden of illness for AF (national prevalence 1.7% 2016). Baseline analysis identified 2,990 patients included within practice AF clinical registers and a further 228 patients for consideration of register inclusion and/or investigation of diagnosis. Of the patients included within the AF register, 707 patients with a CHA₂DS₂-VASc ≥ 1 (high risk) where found to be without anticoagulation.

Of those patients currently anticoagulated, 1647 were currently treated with Warfarin and 701 with a direct oral anticoagulant (DOAC). Data published by INR star in 2016 demonstrated that 23.8% of warfarin patients were achieving less than the NICE recommended 65% time in therapeutic range (TTR). On this basis it was estimated that up to 392 patients may not be achieving safe and effective stroke prevention from their prescribed warfarin therapy. No baseline assessment was made with regards to patients receiving DOAC therapy however in the absence of a structured review it was expected that some patients may require adjustments or intervention in this regard.

This programme was intended to assist primary care sites within Sentinel Healthcare GP Federation in taking a lead role in preventing stroke incidence within the AF population. Initial stakeholder engagement between Sentinel Healthcare, New Devon CCG, South West Academic Health Science Network and Interface Clinical Services ensured the approach did not operate in isolation of other strategic objectives.

Working within a defined protocol, a team of pharmacists from Interface Clinical Services worked with clinician(s) in each GP practice to agree the scope and intervention specification for the project. By choosing the criteria to be included, the GP determined the patient cohorts to be clinically assessed. The pharmacist then used the GP electronic patient records to collate data for presentation. Collation involved using the clinical system search functionality to create searches that extract data, the data extracted was then stratified using the ‘Attend2 AF’ toolkit and a baseline report provided to clinicians with opportunities to improve treatment highlighted.

The pharmacist team then undertook a detailed medical record review for individual patients overlaying NICE guidelines to ensure patients were optimally managed. This included considering current and past medication history, compliance, stroke risk, bleed risk (with and without modifiable risk factors), contra-indications to treatment, time in therapeutic range for patients receiving warfarin, relevant comorbidities, lifestyle factors and other documented relevant information.

Patients where then divided into 3 core cohorts:

1. Patients with sub-optimal stroke prevention therapy.
2. Patients receiving warfarin therapy but with sub-optimal time in therapeutic range (65% as recommended by NICE) or with labile INR’s.
3. Patients receiving non-VKA anticoagulants with clinical issues identified.

Individual patients were discussed with clinicians in practice and interventions implemented in accordance with the practice’s usual treatment pathway.

2,990 AF patients were stratified from the AF clinical domain register. From this initial group of patients 3 cohorts were identified for clinical review:

• 707 patients with a CH A₂DS₂-VASc ≥ 1 currently without anticoagulation.
• 1647 patients receiving a VKA anticoagulant required calculation of TTR and assessment of control.
• 701 patients receiving a non-VKA anticoagulant requiring assessment for compliance, dosing etc.

Upon clinical assessment of patients with a CHA₂DS₂-VASc ≥ 1 currently without anticoagulation, 236 patients were appropriately excluded from treatment, 275 patients were recommended for anticoagulation and 220 patients required further investigation.

Of those patients already anticoagulated 78% were found to be receiving optimal therapy. 16% of patients were found to have a clinical issue with their current therapy whilst 5% of patients required GP review or further assessment.

283 patients were initiated onto an anticoagulant (VKA or non-VKA) whilst 241 patients receiving a VKA were transitioned to a VKA anticoagulant. In addition, 125 patients receiving a non-VKA anticoagulant (DOAC – direct oral anticoagulant) required dose optimisation – this represented 17.8% of patients receiving a DOAC prescription.

At the point of reassessment, patients included within AF clinical register had increased by 10.8% from 2,990 to 3,313 demonstrating an element of coding quality issues within the practices and also a potential undiagnosed patient cohort. The number of patients with a CHA₂DS₂-VASc ≥ 1 currently without anticoagulation had fallen by 219 reducing the predicted 12-month stroke incidence from 65.59 to 43.31 and delivering a net saving of £134,255 (£24,855 per stroke³). These predictions do not take into consideration the optimisation of treatment for patients currently receiving VKA anticoagulants but not achieving TTR or indeed for those patients requiring dose optimisation of their non-VKA; as such this figure could be significantly greater.  

3). Progress in improving stroke care: Report on the findings from our modelling of stroke care provision. February 2010. National Audit Office

• Over the past 5 years, there have been significant revisions to NICE guidelines and indeed QOF parameters for Atrial Fibrillation. Whilst primary care has been taking steps to adopt best practice in this domain, there remains a significant gap which can be closed with effective support.
• At the outset of the programme, AF prevalence in the target cohort was above national average at 1.84%. At the point of reassessment, this prevalence had increased to 1.97%. On this basis it could be assumed that the national prevalence is under-reported.
• Structured clinical audit by experienced clinicians can support primary care to identify, review and where appropriate intervene to ensure quality care for all
• Adoption of NICE guidelines requires collaborative working between stakeholders to deliver optimal outcomes.
• The 17.8% of patients receiving DOAC’s with clinical issues identified was greater than expected and illustrates a knowledge gap in relation to prescribing in this area.