Abstract
INTRODUCTION: The leading cause of preventable traumatic death is hemorrhage. Early detection of hemorrhagic shock remains a critical challenge. For the early prediction of hemorrhagic shock-related cardiovascular decompensation, our team has developed the compensatory reserve measurement (CRM) algorithm. CRM uses a photoplethysmography waveform to quantify the body's capacity to compensate during hypovolemia. This study focuses on the development and use of an application that can predict CRM in real-time (CRM(RT)) during simulated hypovolemia experiments. METHODS: The CRM(RT) application was developed in Python to generate CRM predictions and highlight trend trajectories in real-time (RT). Data were collected during a human research protocol that was reviewed and approved by the Institutional Review Board. Participants (n = 20) meeting the inclusion criteria underwent a simulated hypovolemia procedure in a lower-body negative pressure chamber while wearing a Masimo® MightySat® Rx pulse oximeter. Data were streamed in RT via a Bluetooth® connection to a computer running the CRM(RT) application. RESULTS: CRM was successfully implemented for RT data capture during the research study. The CRM(RT) application achieved a median performance error of -0.95%, while the median absolute performance error was higher at 19.00%. CRM(RT) resulted in an average early prediction time of 18.3 min by tracking the slope trend changes in RT. DISCUSSION: The CRM(RT) application effectively tracked CRM during simulated hypovolemia using a wearable non-invasive sensor. Predictions served as an earlier indicator of hemorrhage compared to traditional vital signs, addressing a limitation of current triage practices. Overall, the CRM(RT) application represents a promising advancement toward RT prediction of hypovolemic decompensation.