CDS Hooks: The “so what” of FHIR for clinicians


Editor’s Note: This is the second in a series from Rimidi about CDS Hooks, a new specification that builds on FHIR, describing how EHRs can automatically invoke external platforms or apps. Read the first article here

Healthcare providers continue to be pushed to their limits. Many are working long hours, caring for an increasing number of patients, managing through workforce shortages, and juggling growing administrative tasks. While their workload is already full, they are now also tasked with needing to evaluate increasingly untenable quantities of data — from what’s in electronic health records (EHRs) to information generated from consumer and home medical devices as well as continuously evolving evidence and guidelines – all with the goal of delivering the highest quality of care.

To standardize and connect these disparate sources of health data, the industry has adopted Fast Healthcare Interoperability Resources (FHIR) as the protocol of choice for application programming interface (API)-based data exchange. FHIR enables electronic health records, digital health applications, patient reported outcomes measures, healthcare payers and consumer devices to seamlessly exchange structured health information.

Bringing data from different systems together is just the foundational step. The value begins with how those aggregate data drive improvements in patient care and clinical workflows. Clinical decision support (CDS) is a key concept in realizing that value. CDS Hooks defines the clinical decision support standard of HL7 FHIR. It’s the ‘so what’ of FHIR because it allows for information, insights and actions to be driven in the proper patient context at the proper time within a clinician’s workflow. CDS Hooks will help support the continued movement to value-based care as we continue to target cost-effectiveness, provider efficiency, patient outcomes and an overall improvement in the healthcare value equation.

Clinical decision support systems lag in adoption

CDSS has rapidly evolved since it was first debuted in the 1980s. Yet decision support has not been widely adopted for a number of reasons. One of the main challenges is that CDSS can disrupt clinician workflows if they are not integrated with EHRs and require additional documentation outside of traditional workstations. Even when CDSS and EHRs work in concert, decision support tools can be more disruptive than useful if they are not both patient and context-specific enough.

As many as 95% of CDSS alerts are not meaningful. Clinicians will tend to ignore alerts if they become overwhelmed by how many they receive. Additionally, clinicians may become distrustful and decide not to act on alerts if they believe they are not critical enough or just wrong.

A clinician recently gave me the example of a sepsis alert on an 18-year-old athlete being seen in the ER for a broken leg. His elevated heart rate triggers an alert for sepsis, which can also be characterized by elevated heart rate, even though sepsis was highly unlikely in his clinical scenario. Knowing this frequently happens, providers may ignore other warnings, much like adults did when the boy called “wolf” too often.

Gaining greater context with CDS Hooks

FHIR-based APIs are helpful for connecting systems and providing broader context about a patient from a wide assortment of sources like remote patient monitoring (RPM), continuous glucose monitoring (CGM) and patient-reported outcomes (PRO) solutions into one integrated application and user experience within the EHR. This standardization and interoperability allow for decision support tools to be able to ingest data beyond the native EHR data when driving clinical insights and actions. With CDS Hooks, these insights can be delivered directly within EHR workflows.

This tool is finally seeing commercial support from the leading EHR vendors and this can encourage FHIR app developers to build narrower, more specific and contextual rules – breaking it down by clinical site or users – to better help CDSS guide the most appropriate actions to take. CDS Hooks envisions several forms that the CDS prompt, called CDS Hooks Cards, can take — information, suggestion, or smart app link.

For example, a CDS Card could identify a patient with diabetes, who had a recent hospital admission and uses a continuous glucose monitor that shows high blood glucose readings more than 10% of the time. This Card would alert the patient’s primary care practice or endocrinologist about this relevant insight (information card).

Alternatively, the Card could go beyond presenting information to allow the clinician to place an order for a medication or send a text message to the patient directly (suggestion). Should the clinician wish to see more information and trends of the CGM data, the Card could allow them to launch a FHIR app where that data can be explored (smart app link). The key here is using all available data, putting it into context, and alerting the appropriate clinician involved in the patient’s care to proactively avoid additional hospitalizations or other complications, and achieve the best possible outcomes.

CDS Hooks are no longer purely academic and theoretical — leading EHRs are supporting the use of CDS Hooks today. Adding CDS Hooks to the technology mix will streamline clinical workflows, particularly as the sources and quantities of data continue to expand at a rapid pace. By breaking down data silos and enabling more specific, contextual and actional decision support rules, CDS Hooks can deliver better insights, helping clinicians understand the next best steps in patient care, while also enabling efficiency in clinical practice.

Photo: Natali_Mis, Getty Images



Source link

Leave a Reply

Your email address will not be published.