Trends in the prescription of oral anticoagulants in UK primary care (2009-2015)

Date of Approval
Application Number
Technical Summary

A longitudinal cohort study will be used to analyze trends in OAC prescription rates and patient profiles in the UK, from 2009 to 2015. Incident prescription rates and corresponding 95% confidence intervals of OAC (VKA, NOAC, and individual NOAC) for each calendar year will be estimated using Poisson distribution. Prescription rates will be stratified by age, sex, and indication (where available). We will also estimate the proportion of prescriptions attributable to each OAC in each calendar year, and changes in proportions for VKA and NOAC will be analyzed over time using a chi-squared test for trend. The profile of incident users will be described for each OAC and each calendar year of study, including demographic characteristics, risk factors, comorbidities, medications, and a measure of health utilization. Baseline profiles will be stratified by indication, and analyzed over time and between OAC using one-way ANOVA for means and chi-squared test for proportions. Multivariate logistic regression will be used to model the probability of a NOAC prescription as compared to VKA. The profile of users who switch anticoagulants over the course of the study period will also be analyzed, and stratified by the direction of the switch and by calendar year.

Health Outcomes to be Measured

All OAC in the British National Formulary (BNF) and that are available in the UK for the prevention of VTE and/or the prevention of stroke in NVAF patients will be identified. Warfarin, phenindione, and acenocoumarol are VKA that fall under the coumarins and phenindione category of the BNF. All VKA (warfarin, acenocoumarin, and phenindione) will be grouped and analyzed as a single OAC class. NOAC to be studied include dabigatran, a direct thrombin inhibitor, as well as rivaroxaban and apixaban which are Factor Xa inhibitors. These NOAC will be analyzed both individually, and together as a single OAC class. Patient demographics including age and sex, as well as risk factors, comorbidities, medications and combined risk factor scores will be used to describe the average patient profile at the time of first prescription, for every year of study and for each OAC class (VKA and NOAC) and individual NOAC (i.e. dabigatran, rivaroxaban, and apixaban). The number of physician visits will also be included as a measure of health utilization. Comorbidities and vascular risk factors to be identified include obesity, smoking status, hyperlipidemia, hypertension, diabetes, ischemic stroke and transient ischemic attack, atrial fibrillation, coronary artery disease, congestive heart failure, peripheral vascular disease, chronic obstructive pulmonary disease, chronic kidney disease, cancer, liver disease, history of bleeding, venous thromboembolism and pulmonary embolism. Concomitant medications will include antiplatelets, NSAIDs, lipid lowering agents, and antihypertensive drugs (beta-blockers, thiazide diuretics, calcium-channel blockers, angiotensin receptor blockers, and angiotensin-converting enzyme inhibitors). CHADS2, CHA2DS2-VASc, and HAS-BLED index scores will also be included within patient profiles as combined risk factor scores and global indicators of susceptibility to stroke (CHADS2 and CHA2DS2-VASc) and major bleeding (HAS-BLED).


Samy Suissa - Chief Investigator - McGill University
Christel Renoux - Corresponding Applicant - McGill University
Laetitia Huiart - Collaborator - University Hospital of La Reunion
Simone Loo - Collaborator - McGill University
Sophie Dell'Aniello - Collaborator - McGill University