New method emulates clinical trial to compare blood thinners for atrial fibrillation

Researchers used real-world clinical data to attempt to emulate a randomized controlled trial testing the effectiveness of two blood thinners, apixaban and warfarin, to prevent stroke in patients with non-valvular atrial fibrillation. The study, led by Emma Maud Powell at the London School of Hygiene and Tropical Medicine, UK, and publishing August 29^th in the open-access journal PLOS Medicine, provides a method to explore the effects of treatments in patients who are underrepresented or excluded from clinical trials.

Patients experiencing atrial fibrillation – a potentially dangerous medical condition in which the upper chambers of the heart beat irregularly – will often be prescribed blood thinners such as apixaban or warfarin to prevent a stroke. However, these treatment recommendations are based on results from randomized controlled trials, and it is unknown if they are applicable to populations of patients who were not included in the trial or present only in very low numbers.

In the new study, researchers used routinely collected health data from patients in the United Kingdom to attempt to emulate a previous randomized controlled trial that compared the effectiveness of apixaban and warfarin. They attempted to emulate the patient eligibility, selection and analysis approaches as the previous trial. They found that patients prescribed apixaban had similar outcomes to patients prescribed warfarin, but unlike the previous trial, they did not find that apixaban was superior. The researchers observed the differences in results may have been linked to higher quality of warfarin control, sub-optimal dosing of apixaban, and differences in the ethnicity of patients and use of concomitant medications compared with the clinical trial population.

Overall, the study established that using an existing randomized controlled trial (the reference trial) as a guide for the design of observational analysis of real patient data is an effective and valid way to estimate the treatment effects and risks of blood thinners given to patients with atrial fibrillation. The methods developed in this study can be used to investigate the effects of these medications in patient groups that are excluded from or underrepresented in these clinical trials, such as the elderly, those with multiple conditions and people with a higher risk of bleeding. This method can also help medical researchers to understand whether results from randomized controlled trials are transferable to "real-world" practices, and provides a framework that can be adapted to investigate treatment effects for other conditions.

The authors add, "Our study aimed to emulate a reference trial in oral anticoagulants in patients with atrial fibrillation using routinely collected UK healthcare data. Reference-trial informed design provides a framework for the study of treatment effects in patient groups excluded from or under-represented in trials."