Seal and logotype of Brandeis University The Heller School logo

Recent Conferences

June 4, 2008

"Road Testing" Electronic Health Records for Effectiveness Research

Healthcare systems that have adopted electronic medical records (EMRs) are slowly amassing a treasure-trove of data on clinical treatments and patient outcomes. Although EMRs are primarily designed to assist clinicians in treating individual patients, health services researchers and policymakers are interested in aggregating these records to study the effects of treatments on patient populations. To date, much of this work has focused on the management practices of clinicians or the safety/side-effect profiles of new treatments. There is growing interest by policy-makers to determine whether EMRs can generate the evidence needed to evaluate the true effectiveness of a new technology or procedure. 

Compared to expensive and time-consuming clinical trials, EMRs can encompass large numbers of patients, followed over multiple visits in actual clinical-practice settings, and provide researchers with access to important health care endpoints and outcomes, such as glucose levels, blood pressure readings, or tumor size and staging.  But at the core, EMRs are merely a collection of observational data without randomization or carefully-defined control groups, raising concern that any analysis based on EMR data is susceptible to confounding and selection bias. 

This meeting of The Health Industry Forum’s Evidence Workgroup will focus on when and under what conditions observational EMR data can be used to develop credible evidence on the effectiveness of new and existing medical therapies. Through case studies of leading EMR systems, we will review the current capabilities of EMRs to generate new evidence of effectiveness and discuss the strengths and weaknesses of utilizing such analyses for decision-making.  Finally, the group will examine what policymakers can do to better create an infrastructure for systematic effectiveness research using EMRs, from retrospective analyses through assisting clinical trials.

Presentations: