05/31/2023

Time Savings from Integrating Discrete, Genomic Data Into the Electronic Health Record Using the Epic Aura Network

ASCO 2023 Abstract
Authors Leo Posteraro, Patrick Mergler, Dennis Urbanek, Joseff Schulin, Howard Anton, Nancy Kokkinos, and James Lin Chen

Background: The capability to deliver, order, and receive discrete genomics data from external pathology laboratories into the electronic health record (EHR) has improved dramatically. As opposed to unstructured PDF integrations, discrete genomic results allow physicians real-time receipt of results which may trigger rapid clinical actions and expedite patient care. However, these genomic integrations, traditionally point-to-point (P2P), can be time consuming. On the other hand, Epic’s order and results network (Aura) connects pathology labs directly to Epic, which has pre-existing connections to Epic EHR sites. This eliminates the need to create a new connection, and is designed to provide a faster and simpler approach. Tempus was the first vendor to use the Epic genomics module for delivery of discrete, somatic mutational data and now is also the first company to leverage Aura for this purpose. Here, we report on our early Aura integration experiences.

Methods: Using the Tempus Database, we retrospectively analyzed medical systems that were all Epic customers and determined their EHR genomic integration time using Aura. Integration time was defined as the time between initial start date of creating “orderables” to time of site testing. The number of EHR medical center analysts was also quantified. We used our prior P2P EHR deployments as a historical control.

Results: Six bidirectional Tempus genomic integrations using Aura were performed from July 2022 to December 2022. All medical centers were expansive, multisite hospital systems. Mean number of integrated departments per hospital was 249 (range: 19-699). In 17 P2P integrations prior to Aura, mean integration time was 6.7m (range: 4m-18m). In contrast, mean integration time using Aura over the 6 sites was 1.5m (range: 0.75m-2m), which is a 78% improvement. No correlation was noted between the integration time and number of EHR analysts staffed or departments integrated. The two slowest integration sites (2m and 2.5m) had competing EHR deployment priorities.

Conclusions: Using Aura, the speed of Tempus EHR integration was improved by 78% for bidirectional, discrete genomic data. Dedicated EHR integration resourcing reduced deployment times to as low as 3 weeks. This reduction in integration time may further the adoption of genomic EHR integrations and improve patient care by expediting time to appropriate, targeted treatments.

VIEW THE PUBLICATION