AN APPRAISAL OF REAL-WORLD EVIDENCE (RWE) IN PATIENTS WITH NON-VALVULAR ATRIAL FIBRILLATION (NVAF) TREATED WITH NON-VITAMIN K ANTAGONIST ORAL ANTICOAGULANTS (NOACS).
Author(s)
Deitelzweig S1, Cichewicz A2, DiFusco M3, Kang A4, Russ C3, Snook K2, Bergrath E2, Earley A2, Cappelleri JC5
1Ochsner Health System, New Orleans, LA, USA, 2Evidera, Waltham, MA, USA, 3Pfizer, Inc, New York, NY, USA, 4Bristol-Myers Squibb, Lawrence Township, NJ, USA, 5Pfizer, Groton, CT, USA
OBJECTIVES A systematic literature review (SLR) was undertaken to summarize reviews assessing real-world clinical outcomes of NOACs in NVAF. METHODS MEDLINE, MEDLINE In-Process, Embase, and Cochrane Database of Systematic Reviews were systematically searched from 01/ 2013–09/2018 for SLRs or network meta-analyses (NMAs) of RWE reporting effectiveness and/or safety of individual NOACs or Vitamin K Antagonists (VKAs). Outcomes of interest were major bleeding (MB) and stroke/systemic embolism (S/SE). RESULTS Thirteen unique SLRs/NMAs were identified. Eleven/13 compared at least one NOAC with VKA; 4/11 also compared NOACs. Two/13 compared only NOACs. Nine/13 reported treatment effects for MB; four also reported S/SE. In most reviews, apixaban (n=2/2), dabigatran (n=2/3), and rivaroxaban (n=2/3) were associated with a similar risk of S/SE versus VKA; one reported reduced S/SE risk with dabigatran and rivaroxaban versus VKA. Risk of MB varied among NOACs; apixaban was consistently associated with lower risk versus VKA (n=6/6), dabigatran (n=3/4) and rivaroxaban (n=4/4). Dabigatran was associated with lower (n=3/6) or similar (n=3/6) MB risk versus VKA. Rivaroxaban was associated with similar (n=6/6) risk for MB versus VKA, and higher risk versus dabigatran (n=3/3). All reviews followed SLR/NMA guidelines for preferred reporting items. Eleven/13 reported methods to address clinical/methodological heterogeneity: subgroup analyses (n=9) included evaluation by study design (n=2), dose (n=5), follow-up time (n=2); sensitivity analyses were less frequent (n=5). Twelve reviews did not control for varying outcome definitions. All reviews assessed statistical heterogeneity. Three quality assessment tools were used across 10 reviews; three reviews did not formally assess study quality. CONCLUSIONS Despite differences in methodologies, reviews indicated similar risk of S/SE and varying risk of MB among most comparisons, with apixaban consistently associated with lower MB risk than rivaroxaban and dabigatran, and dabigatran associated with lower MB risk versus rivaroxaban.
Conference/Value in Health Info
2019-05, ISPOR 2019, New Orleans, LA, USA
Value in Health, Volume 22, Issue S1 (2019 May)
Code
PCV14
Topic
Clinical Outcomes
Topic Subcategory
Comparative Effectiveness or Efficacy
Disease
Cardiovascular Disorders