Abstract

Continuous Risk Analytics System for Tracking the Likelihood for Inadequate Oxygen Delivery in Pediatric Patients

Author: Dimitar Baronov

Background: In critically ill patients, assuring adequate systemic oxygen delivery (DO2) is essential to maintaining or recovering vital organ function, and therefore has an enormous impact on patient outcomes. However, the potential mismatch between DO2 and patient oxygen demand cannot be assessed directly. Instead physicians rely on a subjective interpretation of multiple physiologic proxy measurements. The Inadequate Oxygen Delivery Index (IDO2) algorithm combines all of the available physiologic proxies from continuously monitored vital signs and biomarkers into a single objective measure interpreting the patient’s risk of inadequate oxygen delivery. The IDO2 index has been cleared by the FDA for use with postoperative neonatal patients. The reported work presents new validation studies aimed at demonstrating the performance of the index with an expanded patient population -pediatric patients less than 12 years of age. Methods: The correlation of IDO2 with inadequate oxygen delivery was established through a retrospective study of a patient population that included neonates (0-28 days of age), infants (29 days-2 years of age), and children (2–12 years of age) from multiple institutions. The study included 1,504 patients whose data had been previously collected by the Etiometry T3 Data Aggregation and Visualization system. The purpose of the study was to validate the ability of the IDO2 index to correctly predict mixed venous oxygen blood saturation below 40% as an objective measure of inadequate oxygen delivery. The performance of the index was tested under different data availability: 1) Full Dataset – which included the full set of streamed physiologic measures and laboratory results captured by the T3 system, including venous blood gas measurements; 2) Medium Dataset – which included all data in the Full Dataset except for venous blood gases; and 3) Minimum Dataset – which was intended to demonstrate the performance of the index on minimally monitored patients having a Heart Rate and SpO2 measurement once every 60 seconds, and a Blood Pressure measurement once every 10 minutes. The performance of the algorithm was assessed based on Area Under the Curve (AUC). Results: For all pediatric patients the IDO2 Index had an AUC of 0.83 (0.82 – 0.85) under the full data set, 0.79 (0.77-0.80) under the medium data set, and 0.76 under the minimum data set. A similar trend was observed for each pediatric subgroup, with the neonatal subpopulation having respective AUCs under each data set of 0.84 (0.80 – 0.87), 0.82 (0.79 – 0.86), and 0.74 (0.69 – 0.78), the infant subpopulation AUCs of 0.84 (0.82 0.86), 0.79 (0.77 – 0.81), and 0.77 (0.75 – 0.8), and the child subpopulation AUCs of 0.78 (0.73 – 0.83), 0.78 (0.73 – 0.83), and 0.78 (0.74 – 0.83). Conclusion: The IDO2 Index presents a novel multivariate algorithm to assess the likelihood of physiologic states associated with poor perfusion. The current work demonstrates its ability to reliably predict critical venous blood gas measurements indicating inadequate oxygen delivery under different data availability including under minimal monitoring. There is ongoing work to effectively implement this tool in clinical practice to assist clinicians in achieving a timely and accurate diagnosis of suboptimal tissue DO2, thus affording opportunities for targeted intervention.

Co Author/Co-Investigator Names/Professional Title: Dimitar Baronov PhD, Michael McManus BS, Conor Holland BS, Joshua Salvin MD, Melvin Almodovar MD, and Peter Laussen MBBS
Funding Acknowledgement (If Applicable): Research reported in this publication was supported by the National Heart, Lung, And Blood Institute of the National Institutes of Health under Award Number R44HL117340. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.