Title:EHRs Data Harmonization Platform, an easy-to-use shiny app based on recodeflow for harmonizing and deriving clinical features

Abstract:Electronic health records (EHRs) contain important longitudinal information on individuals who have received medical care. Traditionally, EHRs have been used to support a wide range of administrative activities such as billing and clinical workflow, but, given the depth and breadth of clinical and demographic data they contain, they are increasingly being used to provide real-world data for research. Although EHR data have enormous research potential, the full realization of that potential requires a data management strategy that extracts from large EHR databases, that are collected from a range of care settings and time periods, well-documented research-relevant data that can be used by different researchers. Having a common well-documented data management strategy for EHR will support reproducible research and sharing documentation on research variables that are derived from EHR variables is important to open science. In this short paper, we describe the EHRs Data Harmonization Platform. The platform is based on an easy to use web app a publicly available at this https URL and as a standalone software package at this https URL Harmonization-Platform, that is linked to an existing R library for data harmonization called recodeflow. The platform can be used to extract, document, and harmonize variables from EHR and it can also be used to document and share research variables that have been derived from those EHR data.