Risk Management
Clinical Decision Support Systems:
How have they evolved
by James B. Couch, M.D., J.D., FACPE
Managing Partner & Chief Medical Officer
Patient Safety Solutions, LLC
Printable Version of this Article
In continuing our theme of exploring the medical, legal and risk management benefits (and pitfalls) of electronic health records, this article discusses one of the hottest areas surrounding clinical decision support (or CDS) systems.
Clinical decision support (CDS) provides clinicians, staff, patients, or other individuals with knowledge and person-specific information, intelligently filtered or presented at appropriate times, to enhance health and health care. It encompasses a variety of tools and interventions such as computerized alerts and reminders, clinical guidelines, order sets, patient data reports and dashboards, documentation templates, diagnostic support and clinical workflow tools.
-Osheroff, JA, Teich, JM, Middleton, B, Steen, EB, Wright, A, Detmer, DE, A Roadmap for National Action on Clinical Decision Support; J Am Med Inform Assoc. 2007; 14:141-5
In this article, we will attempt to define better CDS systems (what they are and what they are not), how these systems started and evolved, what is required of them to be used to improve the quality and safety of care while reducing its costs, and some of the ways they could be ignored or misused to actually increase the risk of medical liability.
Fact or Fiction?
Fact: CDS systems prompt physicians and other caregivers to consider alternative courses of action in managing the care of patients and their conditions.
Fiction: CDS systems provide computer-generated answers for physicians, caregivers or patients to follow in managing care.
Fact: Certain types of CDS systems (known as Diagnostic Decision Support Systems) may help physicians to diagnose patients with especially complex and confusing patterns of symptoms.
Fiction: Physicians may rely upon the diagnoses suggested by some Diagnostic Decision Support without corroborating evidence from lab, pathology or imaging tests.
Fact: CDS systems do not require physicians to follow their precautions or suggestions so long as they can justify the patient-specific reasons for not doing so.
Fiction: To avoid their disrupting clinical workflow, physicians can just turn off clinical alerts.
Fact: Patients (especially aging Baby Boomers) will likely use CDS systems embedded in their own personal health records (PHR) in the future (including, if not especially, PHR sponsored by large information-based companies like Microsoft and Google).
Fiction: Physicians will not be expected to know how their patients are updating their own PHRs and making clinical decisions of their own based on advice from CDS systems embedded in these PHRs.
Fact: Physicians may be able to use adherence to best practices suggested by a CDS system as at least some evidence that they adhered to the standard of care in a medical liability action.
Fiction: If a patient has an adverse reaction or event following his physician’s adherence to the best practice suggested by the CDS system, the physician will always be absolved from potential liability.
The Early Days
CDS systems have been around since the 1960s at least. Octo Barnett, M.D., the Director of the Laboratory of Computer Sciences at the Massachusetts General Hospital, was one of the early pioneers. In between coming up with the MUMPS software which served as the primary electronic programming language in hospitals for decades, Dr. Barnett began exploring ways by which computers could actually augment the ability of physicians to think and retrieve necessary information at or near the point of care to facilitate better clinical decisions. Out of this work emerged the well-known DxPlain diagnostic decision support system, which is still operational after all these years.
About the same time, while he was developing the well-known Problem Oriented Medical Record (POMR) at Case Western Reserve University School of Medicine, Lawrence Weed, M.D. also began figuring out how physicians link (or couple together) unique sets of symptoms, syndromes, signs and diagnostic test findings with diagnoses. Out of this work, he developed the Knowledge Coupling approach to Diagnostic Decision Support. Both Drs. Weed and Barnett (now octogenarians) are still hard at work on improving these various types of diagnostic decision support systems (both forms of CDS systems) they started developing four decades ago.