Accession Number : AD1003900

Title :   Automated Analysis of Vital Signs Identified Patients with Substantial Bleeding Prior to Hospital Arrival

Descriptive Note : Conference Paper

Corporate Author : U.S. Army Medical Research and Materiel Command Fort Detrick United States

Personal Author(s) : Reifman,Jaques ; Liu,Jianbo ; Khitrov,Maxim Y ; Edla,Shwetha ; Reisner,Andrew T

Full Text :

Report Date : 01 Oct 2015

Pagination or Media Count : 16

Abstract : Uncontrolled bleeding is the leading cause of preventable death on the battlefield. For the recent conflicts in Iraq and Afghanistan, it has been reported that as many as 22 of such casualties could potentially survive. Protocols for substantial bleeding, typically activated after the patients arrival in a hospital, are known to improve trauma outcomes. Early identification of patients with substantial bleeding could facilitate faster implementation of these protocols, thereby improving patient outcomes. Over the last decade, our interdisciplinary research team has been developing technologies to automatically diagnose hemorrhage in trauma casualties, culminating with the first and only deployment of an automated emergency care decision system on board active air ambulances: the APPRAISE system, a hardware/software platform for automated, real-time analysis of vital-sign data. After developing the APPRAISE system using data from trauma patients transported by Memorial Hermann Life Flight (MHLF), we field-tested it on two active Boston MedFlight (BMF) helicopters during emergency transport of adult trauma patients to three Level 1 trauma centers between February 2010 and December 2012. Between the MHLF and BMF populations, we observed that there were significant differences in terms of vital signs as a function of 24-hr blood transfusion requirements. Despite these differences, the APPRAISE system provided consistent determination of whether or not patients were bleeding. We found that the automated APPRAISE system using a multivariate classifier could automatically diagnose casualties in need of massive blood transfusion with 78 sensitivity and 90 specificity within 6-10 min (median time) after the start of transport to a trauma center. In addition to casualty triage and evacuation decision-making, this capability could be useful to expedite preparedness at medical treatment facilities for receiving patients with substantial blood loss.

Descriptors :   HEMORRHAGE , automation , triage , decision making , decision aids , decision support systems , computer programs , CASUALTIES , MILITARY PERSONNEL , prevention , MORTALITY RATE , COMPUTER AIDED DIAGNOSIS , MULTIVARIATE ANALYSIS , BLOOD TRANSFUSION , sensitivity , accuracy , vital signs , determination , detection

Distribution Statement : APPROVED FOR PUBLIC RELEASE