Implementation of a Clinical Decision Support Platform for the Efficient Liberation and Assessment of Feeding and Nutritional Data in a Tertiary Neonatal Intensive Care Unit

Abstract

Hospitals collect vast amounts of discrete patient data points, such as diagnoses, demographic characteristics, growth measures, and laboratory findings, all of which are stored in the electronic medical record (EMR). The early promise of using the EMR to support research and provide clinical decision support often falls short. Clinical practice groups generate a wealth of collected data but have no way to assess the quality of this data let alone mine it for useful insights. The purpose of this work was to provide the neonatal intensive care unit (NICU) clinical team with a digital tool to access feeding and nutrition data for preterm infants. The tool (1) extracts data directly from the EMR; (2) analyzes the data to identify missing, incorrect, and inconsistent entries; (3) structures the data for feeding models targeting nutrition-related outcomes and quality improvement initiatives; and (4) evaluates adherence to the hospital’s consensus-based feeding protocols. The EMR-integrated solution, NICUtrition, was implemented at Arkansas Children’s Hospital and provided longitudinal, high-resolution data related to feeding and nutrition for 1992 patients over a 5-year period. This included over 1.2 million feeding events, each with extensive detail. Enteral feeding protocols used during this time were digitized, and historical protocol adherence was evaluated. Visualizations of this data through NICUtrition provided detailed insights on feeding, growth, milestones, protocol adherence, and effectiveness, along with insights related to charting inconsistencies, missing data points, and inaccurate information in the EMR.