What We Learned About Power Calculations for Regression Discontinuity Designs: The eMERGE study

• In Event: AERA Poster Session Wednesday 11:45am
  In Poster Session: General Linear Modeling in Regression Discontinuity Designs

Abstract

The eMERGE study deployed Genome Informed Risk Assessments (GIRAs) to propose altered screening schedules, additional risk-stratification, and/or preventive treatment for those at highest genetic susceptibility. Over N=20,000 primary care patients were recruited and expected to be retained. Their providers received GIRAs for 11 chronic diseases integrating genomic (monogenic, polygenic) risk, clinical and family health history (FHHx). For the main analyses, patients were classified as either at High Polygenic Risk (HR) or non-High Risk (nHR) based on a Polygenic Risk Score (PRS) for each disease. Comparison of these two groups are confounded by the fact the classification is based on the PRS. Regression Discontinuity (RD) analyses are the most appropriate.

Author

• T. Mark Beasley, University of Alabama - Birmingham