Unique perspectives sharpen our thinking
Improving Health, Improving Lives
Improving Health, Improving Lives
October 27, 2021
Leveraging technology to mimic the clinician's mindset
by Susan Donahue, IT Director, Clinical and Digital Solutions, Labcorp
I began my career in a clinical laboratory that provided automated interpretations as part of our specialty test results based on clinical guidelines. We learned that many physicians' questions could be answered preemptively through these laboratory insights. For example, test results could include clinical information to help manage patients with kidney stones, offering insights on causes and potential treatments options. After Labcorp acquired that laboratory, we began looking for ways to expand and enhance this program.
Recognizing the customer benefit of adding additional clinical context to laboratory testing to enhance patient care led to the birth of our first clinical decision support program.
When clinicians first hear about clinical decision support, they may think of alerts—a patient didn't get a mammogram or is overdue for a hemoglobin A1c test. Clinicians today are suffering from "alert fatigue." Instead of more alerts, they need patient-specific decision support.
The key to developing effective decision support is a rules engine that can marry technology and science, delivering an interpretation of complex test results in the clinician's workflow based on existing treatment guidelines. Let's explore five considerations for offering real-time insights to help clinicians provide better patient care:
The first key to creating a rules engine is distilling a clinician's process for patient care. Clinicians are trained to think in terms of steps. Therefore, for each program, we must ask, "When you work with a patient, what is your first (or next) step?" because almost every step can become a rule. Then, we must accommodate the clinician's process of simultaneously evaluating all critical factors.
Clinical decision support systems cannot use a simple "if, then" process; a matrix approach is required. The clinician must be able to set parameters for evaluation plus priorities based on the patient's condition to view all possible permutations. The resulting information is especially valuable for chronic or complex diseases such as cardiovascular or kidney disease and diabetes.
Solutions should use more than current test results and demographics. Treatment plans can be better informed by patient history, even if an abnormal test result is buried deep within the patient's records. The system should quickly populate this information—whether 1 or 15 historical results. How much history is relevant? Rules authors must consider the disease and which patient-specific information might impact decision making.
Clinicians are always in control of patient care decisions, so decision support tools must build in capabilities for clinical intervention, interpretation, and customization. Clinicians want the ability to customize reports to meet their needs—for the patient and for the disease. Allowing for unique steps reinforces the concept that the clinician does not work for the system, but rather, the system works for the clinician.
Large health systems have special needs for customized rules, as many have unique guidelines for patient care. Rural versus urban systems, for example, may have different triggers for specialist referrals because of access issues, and this factor will affect guidelines for a given disease.
A decision support system should also set rules for which clinicians receive which interpretations of test results. For example, a kidney specialist may not need test interpretations, but the patient's primary care physician may find them very helpful.
With any evolution of technology, usability is key to adoption. An effective user interface requires that a non-technical person can alter rules quickly and easily. This interface requires a content management system—driven by clinicians, not technicians—that allows updates to the decision matrix, with comments in one field applied to other relevant fields. For example, if a clinician always recommends a given medication, we should apply that information across the system.
In essence, we need to transform current and historical data into insights by applying customizable rules and guidelines and then deliver a holistic view of the patient to inform treatment plans and provide patient-centric care.
Delivering 7 million CDS reports per year, Labcorp serves as a trusted expert to clinicians nationwide. Additionally, Labcorp Drug Development offers expertise in the drug development and clinical trial arena by collecting, processing and delivering massive amounts of actionable data. It's a natural extension to provide more useful insights for clinicians. Developing and delivering this level of industry innovation requires that we seek out and listen to the voice of our customers.
Labcorp Diagnostic Assistant provides a more complete view of a patient’s lab history while delivering actionable guidelines and clinical design support based on the patient’s EHR and Labcorp data. Learn more.
Labcorp Diagnostic Assistant provides a more complete view of a patient’s lab history while delivering actionable guidelines and clinical design support based on the patient’s EHR and Labcorp data.
Susan Donahue has over 15 years of experience in developing clinical decision support for Labcorp. With a BS from the University of Illinois, Urbana-Champaign, she launched her career working in the lab and clinical research. She soon began working with a cross-functional team to deliver expert-driven interpretations as a companion to laboratory results. Applying a passion for leveraging technology to bring Labcorp scientific expertise to clinicians, she today manages a clinical content portfolio across Labcorp Diagnostics, including Labcorp Diagnostic Assistant and Pixel by Labcorp.