Safety and effectiveness of direct oral anticoagulants versus vitamin K antagonists: results from 3 Italian regions

Background: In Italy, direct oral anticoagulant drugs (DOACs) were authorized for stroke prevention in patients with non-valvular atrial fibrillation (NVAF) in 2013. There is conflicting evidence on their benefit-risk profile under real world conditions. Methods: The Italian Medicines Agency funded this study to investigate effectiveness and safety of DOACs compared to vitamin K antagonists (VKAs) in three Italian regions. An observational study was conducted with a sequential propensity-score-matched new user parallel-cohort design in the period July 2013-December 2015 using administrative health data. DOAC users with NVAF diagnosis were 1:1 matched to VKA users based on a PS which accounted for over 90 potential confounders at baseline. Applying an as-treated approach with a 90-day renewal grace time, patients were followed from the day after the first prescription of the study drug until occurrence of the outcome, death, discontinuation, switch, end of health plan enrolment, or study end. Outcomes were total and cardiovascular mortality, acute myocardial infarction, ischemic and haemorrhagic stroke, and gastrointestinal bleeding. Analyses were performed, using Cox proportional hazard models stratified by matched set. The results of the regional analyses were combined through a random-effects meta-analysis. Results: During the first 30 months of authorisation for NVAF, DOACs were increasingly prescribed. Overall, 72,434 new anticoagulant users were enrolled, 34% of whom received a DOAC. After PS matching, 37,266 patients contributed to the analysis. No differences between the study groups were found for total and cardiovascular mortality, myocardial infarction and ischemic stroke. DOAC users were at higher risk of gastrointestinal bleeding (HR 1.41, 95%CI 1.07-1.86) and at a not significant lower risk of haemorrhagic stroke (HR 0.36, 95%CI 0.10-1.33). Conclusions: The present study confirms findings from previous research regarding bleeding events, whereas we did not find a reduced risk of mortality in DOAC users. Further research on single active agents and specific populations is warranted.