VERMONT OXFORD NETWORK DATA INGESTION AUTOMATION

What is the Vermont Oxford Network

The Vermont Oxford Network (VON) supports NICUs worldwide by forming a community focused on improving neonatal care. J P Systems helped VON improve care for very low birth weight infants through data standardization and education. Here we standardized VON’s patient database using a coordinated program for quality improvement, research, and training. Then Clinicians clearly enter detailed data manually using a questionnaire that affects infants in under-resourced global settings. Our project aims to automate and standardize data entry internationally for faster ingestion and better usability. Once standardized, we can design FHIR™ apps to automate workflows by ingesting data from EHRs like Cerner or Epic.

Our standardization work helps neonatal providers achieve data-driven quality improvement across the global community. J P Systems analyzed data collected from 1,200 neonatal units using the completed questionnaires. Our Teams broke down each question into parts to match them with international reference terminologies. Also, we mapped simple answers to individual SNOMED concepts when possible. We mapped compound answers to multiple SNOMED concepts as needed. We aim to locate the same data fields in member hospitals’ EHR systems. A future FHIR app will pull data directly from EHRs, eliminating the need for manual questionnaires. The goal of our project is to enable automated and internationally standardized data entry so data can be ingested faster.

Some questions are too complex to map directly to EHR fields. These answers require analyzing multiple fields and treatment records to draw accurate conclusions. For example, complex questions include: “Temperature within the first NICU hour” and “Sepsis/Meningitis by Day 3.” These questions cannot map to single EHR fields and need further study for automation.