Heart failure (HF) is a condition where the heart cannot pump enough blood to satisfy the bodys demands, causing symptoms such as breathlessness and tiredness. Atrial Fibrillation (AF) describes a type of irregular heartbeat.
HF and AF are common conditions and many people have both, including around 1 in 200 adults in the United Kingdom (UK). Having just one increases the chances of a stroke up to five-fold; having both makes the stroke risk higher still.
Blood thinning medications are recommended for people with HF and AF and can prevent around two thirds of strokes. However, a third of people with AF who could take these medications are currently not prescribed them.
We will report how many people with AF and HF have a stroke, related death, or a serious bleeding episode, which is the main side-effect of the blood thinning treatment. We also want to see how many people receive the recommended medication and which groups of people dont. To do so, we will analyse anonymous patient records held within The Clinical Practice Research Datalink (CPRD), a database containing the general practice records of millions of people in the UK.
We hope that by identifying the groups at risk of not receiving blood thinning medication, we can help target future improvements in healthcare. We will use this data for further research looking at different approaches to organising stroke prevention treatment, to see if we can recommend affordable changes to delivering care that help more people get the recommended treatment.
Heart failure (HF) and atrial fibrillation (AF) are common conditions that often co-exist - around 0.5% of the UK adult population have both. Each is independently associated with an increased stroke risk. In combination, the risk is higher still, but there is considerable variation between previous estimates of effect. Anticoagulation is recommended for patients with HF and AF and can reduce stroke risk by 65%. However, in AF, anticoagulation is not prescribed in one third of eligible patients.
The aim of this study is to determine the stroke incidence, mortality and rate of major haemorrhage for patients with HF and AF, compared to either condition alone or neither condition. We will compare current rates of anticoagulation against national guidelines within these groups and identify factors associated with low rates of anticoagulation prescribing.
We will conduct a retrospective cohort study using CPRD data from 1st January 2000 to 31st December 2018, using a time-dependent analysis to account for the fact people will change between groups over follow-up. Incident cases will be recorded at entry to the study cohort and with each subsequent new diagnosis of either AF or HF. Data will be linked to Hospital Episode Statistics to validate diagnostic codes and Office for National Statistics for mortality data.
A survival time analysis will describe stroke incidence over time in each of the four groups (HF, AF, HF+AF, no HF/AF) alongside cumulative incidence plots of survival rates. Multivariable Cox regression modelling will assess the association between HF, AF and time to stroke based on key variables, adjusting for anticoagulation, co-morbidities and the individual components of the CHA2DS2VASc stroke risk score.
The data collected will be used to inform a health economic analysis, using Markov modelling to predict the comparative clinical and cost effectiveness of future alternative treatment strategies to improve anticoagulation.
The primary outcome measure:
- Incidence of stroke, including type of stroke classified as either ischaemic or haemorrhagic (including either subdural, extradural or other e.g. subarachnoid).
The secondary outcomes measures include:
- All-cause mortality and cardiovascular mortality, including mortality rates related to HF, AF or stroke.
- Incidence and type of major haemorrhage, including either intracerebral or gastrointestinal
- Incidence of transient ischaemic attack (TIA)
- Proportion of eligible patients prescribed anticoagulation, the type of anticoagulation used, (i.e. whether a vitamin K antagonist (VKA) or a direct oral anticoagulant (DOAC)) and the factors associated with high and low rates of anticoagulation prescribing.
Additional information on further relevant covariates that will be included are given in Section N below.
Clare Taylor - Chief Investigator - University of Oxford
Nicholas Jones - Corresponding Applicant - University of Oxford
Andrea Roalfe - Collaborator - University of Oxford
Chang Ho Yoon - Collaborator - University of Oxford
Clare Bankhead - Collaborator - University of Oxford
Margaret Smith - Collaborator - University of Oxford
Richard Hobbs - Collaborator - University of Oxford
Sarah Lay-Flurrie - Collaborator - University of Oxford
Brian Nicholson - Collaborator - University of Oxford
Cynthia Wright Drakesmith - Collaborator - University of Oxford
Joseph Lee - Collaborator - University of Oxford
Margaret Smith - Collaborator - University of Oxford
Rafael Perera - Collaborator - University of Oxford
Ting Cai - Collaborator - University of Oxford
William Elson - Collaborator - University of Oxford
HES Admitted Patient Care;ONS Death Registration Data;Patient Level Index of Multiple Deprivation