To the Editor
In the United States, the 2009 Affordable Care Act provided incentives to install EMRs that resulted in a dramatic increase in their adoption.2 Nearly all Americans now have their data stored in some type of EMR. While details of the storage may vary from system to system, in almost all cases the EMR serves as the central repository for patient data. This includes not just vital signs and orders, but also laboratory results and pertinent medical history. It is precisely because of this vast amalgamation of clinical knowledge that we suggest that its integration into safety systems would bring tremendous benefits.
In describing their example of a patient-controlled analgesia pump, the authors state, “The system queries the hospital information system for the patient’s weight, age, and medication list …” Unfortunately at this time there are no existing standards for structuring these data, let alone for querying them. The 2 largest EMR systems—EPIC (Verona, WI) and Cerner (Kansas City, MO)—use significantly different data structures, and, in the case of EPIC, fundamental elements (such as the units for weight) differ from 1 institution to the next.
A potential solution to this problem is the adoption of Fast Healthcare Interoperability Resources (FHIR), an open-source health level-7 standard to promote the transmission of data out of the EMR.3 Unfortunately, EPIC and Cerner only support FHIR to a limited extent, and even in the upcoming EPIC release, FHIR support will still be limited to the core patient-centric elements, such as vital signs, laboratory results, medications, providers, allergies, etc, and overlook data necessary for perioperative integration, such as procedure scheduling and admission/discharge/transfer data. Additionally, the fact that many of the FHIR resources do not provide real-time data significantly diminishes the impact of safe interoperability. Moreover, the specific calls and structure for FHIR differ not only between the EMRs, but also from 1 institution to the next. This forces each institution to create its own custom interface to support the FHIR and integrate with other devices. Collectively, these factors create a costly barrier to entry and greatly inhibit integration, especially in community hospitals that lack technical resources.
While FHIR may not be the perfect solution today, we encourage the broader anesthesia community to work with the EMR vendors and others to leverage this framework to create true interoperability between EMRs and devices.