Precision Medicine Beginning to Identify Important Clinical Uses

Regenstrief Institute software engineers have begun to design a second-generation genomic clinical decision support (CDS) system.  We are taking the opportunity to improve upon our prior genomic CDS implementation, which required replacement after migrating to a new EHR system.  We are impressed by the tight link between precision medicine and CDS.

Precision medicine is at an early stage, with the underlying science beginning to identify important clinical uses, and EHR support following.  It is completely possible today to practice effective medical care without robust genomic CDS, with genetic counselors or external resources providing clinical interpretation as the need arises.  For purposes of creating IT infrastructure, it will be useful to build increased support of genomic standards into existing EHR systems, improve methods for sharing such data across institutions, supplement genomic data with other personalized data extracted from EHRs or captured directly from patients, and create repositories of shareable genomic CDS logic.

The truly exciting CDS aspects related to precision medicine lie in the future.  When its value is fully realized, precision medicine will require the automated synthesis of medical knowledge with the patient's own genetic, proteomic, metabolomic, electronic health record, fitness, and other big data.  In return, precision medicine will improve the identification of conditions that an individual patient is at risk for or has already acquired, as well as identify optimal preventive and treatment measures specific to that patient.  

Such synthesis will greatly exceed the abilities of the human brain.  At such time precision medicine proves useful for a majority of medical decisions, this synthesis could be required hundreds of times daily for a practice, precluding the possibility of relying solely on external resources.  At that time, physicians and computers will need to work in a harmony currently unseen.  There will be no choice but to add a robust layer of computer logic that binds an increasingly complex medical literature, the patient's genomic and other personalized data, clinicians, and patients' symptoms and signs. 

This will represent an opportunity not just for genomic CDS, but CDS generally.  CDS has largely acted on the periphery of medical care, reminding clinicians of allergies, drug interactions, and preventive care.  CDS only rarely provides recommendations critical to identifying and solving the patient's central issues.  Insofar as the overwhelming complexity of precision medicine recommendations forces adoption of CDS that addresses the patient's central issues, all forms of CDS recommendations and resultant medical decision-making will benefit. 

About the Contributors

Paul Dexter, MDResearch Scientist, Center for Biomedical Informatics, Regenstrief Institute, Inc. and Associate Professor of Clinical Medicine, Indiana University School of Medicine. Dr. Dexter has been a research scientist at Regenstrief Institute for the last twenty years with a focus on clinical decision support and electronic health records.

Shaun Grannis, MD, MSInterim Director and Research Scientist, Center for Biomedical Informatics, Regenstrief Institute, Inc. and Associate Professor of Family Medicine, Indiana University School of Medicine. Dr. Grannis collaborates closely with national and international public health stakeholders to advance technical infrastructure and data-sharing capabilities. He is a member of World Health Organization (WHO) Collaborating Center for the Design, Application, and Research of Medical Information Systems, where he provides consultancy on issues related to health information system identity management and implementing automated patient record matching strategies.