On week 4 of Phenotype Phebruary we had a debate on customized versus standardized approach for Phenotype development and evaluation (debate can be found here: Phenotype Phebruary 2023 - Week 3 - Debate – customized versus standardized approach for Phenotype development and evaluation)
For week 4 - we are going to discuss probabilistic phenotyping and its role in phenotype development and evaluation. In phenotype Phebruary of 2022, multiple OHDSI collaborators showed us how probabilistic phenotyping can help us define and capture clinical events of interest with accuracy and precision. Atlas is not the only OHDSI tool to develop phenotypes. Prior work from @Juan_Banda nd others demonstrated the utility of APHRODITE by comparing its performance versus traditional rule-based phenotyping approaches. They also showed how the resulting phenotype models and model construction workflows built with APHRODITE can be shared between multiple OHDSI sites.
In this week discussion we would like to learn from each other’s on how can we couple probabilistic and rule-based phenotyping in a systematic process to improve the development and evaluation of phenotypes? Please share your experience, work from others or even new innovative ideas that you would like to propose to the community!