Effectiveness of the tromboembolic venous disease treatments estimated through clinical trials vs. observational studies

Abstract

It is unknown whether the efficacy of treatments for venous thromboembolic disease (VTE) estimated by observational studies with propensity analysis is similar to the one estimated by clinical trials (ECs). A systematic review of PubMed and Web of Science was conducted to identify observational studies with propensity analysis that would have evaluated the effect of VTE treatments on short term all-cause mortality. After identifying the treatment scenarios evaluated by the observational studies, a standardized algorithm was used to match them with at least one CE or a meta-analysis of ECs. The efficacy of the treatment (relative odds ratio [OR]) between the observational studies and the paired Ecs was compared. The global relative OR for all clinical scenarios was also calculated. Results: observational studies and CSs evaluated 7 treatment scenarios for VTE. There was no statistically significant difference in the effect on mortality obtained by observational studies compared to CSs (OR 0.89 [95%CI,0.32-1.46]). However, in 2 scenarios (thrombolysis vs. anticoagulation for lung thromboembolism, treatment with enoxaparin once or twice a day for VTE) there were statistically significant differences in the direction of effect, and there were no significant differences in the magnitude of the effect in 2 other scenarios (rivaroxaban vs. vitamin K antagonists for the treatment of VTE, home treatment vs. hospitalization for deep vein thrombosis). Conclusions: observational studies that use propensity analysis and CSs find similar results regarding the effect of VTE treatments, although there are occasionally differences in the direction/magnitude of this effect.