Estimating individual absolute risk reductions and increases of DOAC treatment in AF patients

Individual Treatment Effect Estimation of 2 Doses of Dabigatran on Stroke and Major Bleeding in Atrial Fibrillation: Results from the RE-LY Trial.

Literature - Stam-Slob MC, Connolly SJ, van der Graaf Y et al. - Circulation. 2019 May 3. doi: 10.1161/CIRCULATIONAHA.118.035266. [Epub ahead of print]

Introduction and methods

The effectiveness of direct oral anticoagulants (DOACs) in comparison with vitamin K antagonists (VKAs) in the reduction of ischemic stroke and systemic embolism (SE) in patients with nonvalvular atrial fibrillation (AF) has been well established [1-4]. Regarding safety, they appear to induce equal or fewer major bleeds than VKA, depending on DOAC type and dose. International guidelines therefore recommend DOACs for patients with nonvalvular AF.

The RE-LY trial results demonstrated that high-dose dabigatran (150 mg twice daily) was more effective in prevention of ischemic stroke/SE than low-dose dabigatran (110 mg twice daily), at the expense of an increased risk of gastrointestinal bleeds [1]. Similar observations have been done in a general practice setting [5].

The trial results give one estimation of the treatment effect for all patients, although absolute treatment effects for an individual will likely differ between individuals. Relying on a single characteristic is suboptimal, since multiple patient characteristics affect the risk of ischemic stroke/SE and the risk of bleeding differently [6,7]. Estimating the individual treatment effect allows identification of patients who have a favourable balance between absolute benefit and harm of treatment with dabigatran, compared with no treatment. Moreover, the optimal dose can be selected for each individual patient.

This study set out to develop and validate a model to estimate the absolute treatment effect of dabigatran on ischemic stroke/SE and major bleeding in individual patients with AF. Access to anonymized individual patient data of RE-LY was obtained through AF patients in the RE-LY trial (n=18133) had a mean age of 71 years (SD: 8), and most had an indication for oral anticoagulation therapy with a CHA2DS2-VASc score ≥2. Patients were followed for a median duration of 2.0 (IQR: 1.6-2.4) years. For this study, the RE-LY population was randomly divided into a derivation (n=11.955) and validation (n=6.158) cohort. 294 ischemic stroke/SE occurred in the derivation cohort and 161 in the validation cohort. 751 Major bleedings were recorded in the derivation cohort and 431 in the validation cohort.

Main results

  • The relative treatment effect of dabigatran 110 mg for stroke/SE was 1.06 (95%CI: 0.86-1.32), as compared with warfarin, and 0.80 (95%CI: 0.70-0.93) for major bleeding (incl. hemorrhagic stroke).
  • For dabigatran 150 mg, the relative treatment effect was 0.76 (95%CI: 0.60-0.96) for ischemic stroke/SE and 0.94 (95%CI: 0.82-1.08) for major bleeding.
  • The derivation model revealed that predictors associated with a large increase in the risk of ischemic stroke/SE were a previous stroke and peripheral artery disease. Age was associated with increases in the risk of both stroke and major bleeding. Decreasing eGFR, low baseline haemoglobin, current/former smoking and type 2 diabetes strongly increased the risk of major bleeding.
  • The models calibrated well. For stroke prediction, a c-statistic of 0.67 (95%CI: 0.65-0.70) was found in the derivation set, and 0.65 (95%CI: 0.62-0.70) in the validation set. For major bleeding, it was 0.69 (95%CI: 0.67-0.71) in the derivation set and 0.69 (95%CI: 0.66-0.71) in the validation set.
  • Individual treatment effect estimations for 2 years revealed that the 5-year absolute risk reduction (ARR) in ischemic stroke/SE with dabigatran 150 mg twice daily was<10% in 20% of patients, 10-15% in 32%, 15-20% in 22%, 20-25% in 12% and >25% in 7% of patients. The 5-years absolute risk increase (ARI) in major bleeding for this dose was <5% in 53% of patients, 5-10% in 38%, 10-15% in 8% and 15-20% in 1% of patients.
  • When comparing the treatment effects of high-dose and low-dose dabigatran, the 5-year ARR in ischemic stroke/SE with 150 mg was <1% for 16% of patients, 1-2% for 51%, 2-3% for 21%, and >3% for 12% of patients.
  • The study also considered the balance between the individual absolute benefit and harm with dabigatran 150 mg vs. 110 mg and identified patient characteristics that are associated with more benefit (ARR) than harm (ARI). For 60% of the patients younger than 75 years, ARR for high-dose vs. low-dose was higher than the ARI. For patients over 75 years, ARR was higher than ARI in 24% of patients.


This study shows that readily available patient characteristics can be used to assess the absolute treatment benefit and harm from dabigatran therapy in individual patients with AF. In a population of AF patients with additional risk factors for ischemic stroke/SE, the indication for anticoagulation was confirmed for most patients. Individual variation was seen in the 5-year ARR for ischemic stroke/SE and the ARI for major bleeding with the two treatment doses.


1. Connolly SJ, Ezekowitz MD, Yusuf S et al. Dabigatran versus warfarin in patients with atrial fibrillation. N Engl J Med. 2009;361:1139-1151.

2. Giugliano RP, Ruff CT, Braunwald E, et al. Edoxaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2013;369:2093-2104.

3. Granger CB, Alexander JH, McMurray JJ, et al. Apixaban versus warfarin in patients with atrial fibrillation. N Engl J Med. 2011;365:981-992.

4. Patel MR, Mahaffey KW, Garg J, et al. Rivaroxaban versus warfarin in nonvalvular atrial fibrillation. N Engl J Med. 2011;365:883-891.

5. Graham DJ, Reichman ME, Wernecke M, et al. Cardiovascular, bleeding, and mortality risks in elderly medicare patients treated with dabigatran or warfarin for nonvalvular atrial fibrillation. Circulation. 2015;131:157-164.

6. Dorresteijn JA, Kaasenbrood L, Cook NR, et al. How to translate clinical trial results into gain in healthy life expectancy for individual patients. BMJ. 2016;352:i1548.

7. van der Leeuw J, Ridker PM, van der Graaf Y et al. Personalized cardiovascular disease prevention by applying individualized prediction of treatment effects. Eur Heart J. 2014;35:837-843.

Find this article online at Circulation

Facebook Comments


We’re glad to see you’re enjoying PACE-CME…
but how about a more personalized experience?

Register for free