Gravar-mail: An Interdisciplinary Approach to Reducing Errors in Extracted Electronic Health Record Data for Research