Polypharmacy and the impact of drug interactions with direct oral anticoagulants (DOACs)

Date of ISAC Approval: 
03/10/2019
Lay Summary: 
Warfarin is commonly used to prevent blood clots in people with heart diseases. Another alternative class of drugs known as direct oral anticoagulants (DOACs) including dabigatran, rivaroxaban, apixaban, edoxaban are increasingly being used instead of the older alternative, warfarin to thin the blood. Many individuals requiring blood thinning treatment may have several clinical conditions and take many different medications, potentially increasing the risk of interactions between drugs. Some studies suggested how drugs interact with DOACs but the evidence was conflicting. Little is known about the consequences of these interactions. Moreover, proton pump inhibitors (PPIs), such as omeprazole and pantoprazole, a group of drugs used to treat digestive disorders, have been recently recommended to prevent DOAC-related bleeding in the stomach and intestines. However, whether the treatment effects of DOACs would be compromised by concurrently taking PPIs and DOACs remains largely unknown. This project will examine the effects of interaction between DOAC and other drugs in the UK. Using anonymised electronic health records, we will find out which medications are being used at the same time as DOACs and warfarin respectively from 2011 to 2019. We will study whether there are interactions between these medications and DOACs, leading to adverse outcomes, for example, bleeding. We will also study how these interactions compare with those seen with warfarin. We will also estimate the number of individuals concurrently receiving PPIs and DOACs and study whether PPIs can reduce DOAC-related bleeding in the stomach and intestines, whilst not adversely affecting the benefits of DOACs.
Technical Summary: 
The aim of this study is to explore patterns of polypharmacy and to quantify the health consequences of drug interactions with direct oral anticoagulants (DOACs). We will focus on drugs which potentially interact with DOACs but currently have conflicting mechanistic and epidemiological evidence including amiodarone, dronedarone, verapamil, diltiazem, digoxin, cyclosporine, simvastatin, lovastatin, atorvastatin, erythromycin, clarithromycin, ketoconazole, itraconazole, voriconazole, posaconazole. We will also investigate the benefits and risks of using proton pump inhibitors (PPIs) to prevent DOAC-related gastrointestinal bleeding. We will identify new DOACs and warfarin users, aged 18 or above during 1/1/2011-31/07/2019 using the data from the Clinical Practice Research Datalink. We will describe the patterns of co-prescribed medications and polypharmacy among oral anticoagulant users. We will then use a cohort study with propensity score and case-crossover study design to investigate if 1)DOACs are associated with interactions with the drugs of interest); 2)whether DOACs have a lower risk of drug interactions compared with warfarin; 3)whether concurrently receiving PPIs and DOACs reduces risk of bleeding without affecting the risk of cardiovascular and mortality outcomes. Outcomes will be cardiovascular events, bleeding and mortality. Using cohort study design, we will identify the exposed group (DOAC with the concomitant drug/PPIs) and three comparison groups (DOAC alone; warfarin with the concomitant drug/PPIs; warfarin alone). Propensity score methods will be used to control for confounding. We will compare the hazards of each outcome among the exposed group with each comparison group using Cox proportional-hazard regression. Using modified case-crossover study design, we will identify patients with each outcome who were exposed to at least one of the two interacting drugs (oral anticoagulants and the concomitant drug/PPIs). We will compare each patient's exposure in a time period prior to the outcome (hazard window) to that during a control period (referent window) within an individual using conditional logistic regression.
Health Outcomes to be Measured: 
Primary outcomes (cardiovascular effectiveness endpoints): ischaemic stroke, myocardial infarction, venous thromboembolism and cardiovascular death. Secondary outcomes (safety endpoints): intracranial and/or gastrointestinal bleeding, other-bleeding and all-cause mortality.
Collaborators: 

Yun "Angel" Wong - Chief Investigator - London School of Hygiene & Tropical Medicine (LSHTM)
Dr Amitava Banerjee - Collaborator - University College London (UCL)
Esther Chan - Collaborator - University of Hong Kong
Ian Douglas - Collaborator - London School of Hygiene & Tropical Medicine (LSHTM)
Professor Ian Wong - Collaborator - UCL School Of Pharmacy
Dr Krishnan Bhaskaran - Collaborator - London School of Hygiene & Tropical Medicine (LSHTM)
Professor Liam Smeeth - Collaborator - London School of Hygiene & Tropical Medicine (LSHTM)
Yun "Angel" Wong - Corresponding Applicant - London School of Hygiene & Tropical Medicine (LSHTM)

Linkages: 
HES Admitted;ONS;Patient IMD