Randomized controlled trials (RCTs) remain the gold standard for establishing the causal relationship between medications and health outcomes. However, for some clinical questions RCTs may be infeasible, unethical, costly, or generalizable to only a very narrow population. In these cases, observational studies from routinely collected ?real-world? health data (RWD) are crucial for supplementing the evidence from RCTs. RWD is particularly useful in the context of cardiovascular outcomes, which are typically well-captured in healthcare databases, but concern about the validity of RWD studies mitigate their impact. The purpose of this project is to empirically evaluate the validity of RWD studies by conducting a systematic replication of a large sample of randomized trials of cardiovascular outcomes, including both completed trials with results already available and ongoing trials where results are not yet released, in order to ensure that validity is not influenced by the availability of trial results. RWD studies are designed to match the corresponding RCT as closely as possible, including exposures, outcomes, and inclusion/exclusion criteria, and RWD studies utilize a variety of design and analysis approaches in order to provide guidance on which questions can be answered in RWD and with which methods. As part of this project, a new meta-analysis model will be developed to model the variation in treatment effect estimates between RCTs and RWD, combining data across RWD methodologic approaches and across clinical questions. All treatment effect estimates for all studies, including both RCT and RWD results, will be reported online, and the results of ongoing RCTs will be added as they become available. This project is the first to evaluate the validity of RWD studies of cardiovascular outcomes based on empirical performance in a large sample of clinical questions across a wide variety of epidemiologic and analytical methods. Findings will guide investigators planning RWD studies of cardiovascular outcomes and increase confidence in the ability of RWD studies to answer questions of importance to patients, providers, and other stakeholders.
The proposed project provides fundamental knowledge on what types of clinical questions can be answered with real world data and with which methods. Findings will impact public health by improving the conduct of real world data studies and increasing confidence in their results, allowing for improved treatment decision- making even in the absence of randomized trial evidence.
