Search

The deployment of predictive analytic algorithms that can safely and seamlessly integrate into existing healthcare workflows remains a significant challenge. Here, we present a scalable, cloud-based, fault-tolerant platform that is capable of extracting and processing electronic health record (EHR) data for any patient at any time following admission and transferring results back into the EHR. This platform has been successfully deployed within the UC San Diego Health system and utilizes interoperable data standards to enable portability.Clinical relevanceThis platform is currently hosting a deep learning model for the early prediction of sepsis that is operational in two emergency departments.

ObjectiveOpioid Overdose Network is an effort to generalize and adapt an existing research data network, the Accrual to Clinical Trials (ACT) Network, to support design of trials for survivors of opioid overdoses presenting to emergency departments (ED). Four institutions (Medical University of South Carolina (MUSC), Dartmouth Medical School (DMS), University of Kentucky (UK), and University of California San Diego (UCSD)) worked to adapt ACT network. This paper reports their progress.Materials and MethodsThe approach taken to enhancing ACT network focused on four activities: cloning and extending the ACT infrastructure, developing an e-phenotype and corresponding registry, developing portable natural language processing (NLP) tools to enhance data capture, and developing automated documentation templates to enhance extended data capture.ResultsAll four institutions were able to replicate their i2b2 and Shared Health Research Information Network (SHRINE) infrastructure. A five category e-phenotype model based on ICD-10 coding was developed from prior published work. Ongoing work is refining this via machine learning and artificial intelligence methods. Portable NLP tools, focused at the sentence level, were also developed to identify uncoded opioid overdose related concepts in provider notes.Optimal performance was seen in NLP tools that combined rule-based with deep learning methods (F score, 0.94). A template for ED overdose documentation was developed to improve primary data capture. Interactive prompts to physicians inside ED progress notes were effective in promoting use of the template. The template had good system usability and net promoter scores (0.72 and 0.75, respectively, n=13). Work to design ED trials based on the network’s data is underway.Discussion and ConclusionsOverall, initial results suggest that tailoring of existing multipurpose research networks to specific tasks is feasible; however, substantial efforts are required for coordination of the subnetwork and development of new tools for extension of available data.