|
Abstract:
|
Comparative effectiveness research (CER) seeks to evaluate which treatments or interventions work best for a given patient or subgroup. Although randomized trials are often considered the gold standard for CER, the ever-expanding role of electronic health records, data networks, and other sources of 'big data', emphasize the need for effective modeling strategies in observational CER. While a substantial volume of literature has been published on approaches such as propensity score-based methods, identifying the most effective strategies for a given scenario requires a clear understanding of the causal question, adequacy of the data, underlying treatment assignment mechanism, and underlying assumptions associated with the selected approaches. To address these needs, we developed (through a PCORI contract) a Decision Tool for Causal Inference and Observational Data Analysis Methods in Comparative Effectiveness Research (DECODE CER). DECODE CER is a publicly-available set of Google Slides for guiding researchers through the process of observational CER. Future work focuses on implementing the tool into research and education efforts with partnering medical centers.
|
