Battlefield Medic and Clinician Decision Support
Event Type
Oral Presentations
TimeWednesday, April 142:20pm - 2:40pm EDT
LocationDigital Health
Battlefield Medic and Clinician Decision Support

Our Trauma Triage Treatment and Training Decision Support (4TDS) project team developed a decision support system (DSS) for casualty care in austere battlefield environments. As an application on a Android smart phone and tablet, 4TDS includes training scenarios in skills such as shock identification and management. 4TDS pairs with various vital signs sensors that can stream data for a machine learning algorithm to detect the probability of life-threatening shock in a casualty. A “silent test” compared algorithm performance with actual clinical diagnoses at Mayo Clinic, Rochester, MN. Usability assessments enabled us to determine medic and clinician acceptance of 4TDS and verify how well it aligns with their work and their decision making.

In Tactical Combat Casualty Care (TCCC), medics stabilize battlefield casualties at point of injury and transport them to field care facilities such as a Battalion Aid Station or Field Hospital where clinicians provide critical care. Care provider experience and ability vary, and training in the field can help to improve recall and performance of infrequently used critical care skills. This becomes more necessary during Prolonged Field Care (PFC) when evacuation is not immediately available, and medics may need to perform more complex treatment.

Our presentation will describe the 2-year Trauma Triage Treatment and Training Decision Support (4TDS) project, including:
Research—How the team used Department of Defense field medical procedures to develop use cases, work flows, and rough prototypes of user interfaces for use in Design Requirements Reviews with qualified TCCC subject matter experts.
Development—How the team created an operating prototype for use on a Samsung Galaxy smart phone configured for use in the DoD Nett Warrior program. They integrated a durable, inexpensive sensor developed for Department of Homeland Security first responders to stream casualty vital signs. They further created refresher training scenarios on two life-critical skills that are available for use on the phone. They also programmed a Nett Warrior tablet that Battalion Aid Station clinicians can use as they care for multiple casualties. The machine learning team developed and evaluated multiple models to detect and predict shock, settling on a logistic regression model they tested on publicly available MIMIC data, then Mayo Clinic ICU patient data.
Evaluation—How the team created four evaluations to assess the Nett Warrior smart phone and tablet applications. Medic TCCC was designed for use with the smart phone and vital signs monitor. Three were designed for use with the tablet: Medic Battalion Aid Station, Clinician Battalion Aid Station, and Training Instructor. We will describe results of field assessments in December 2020 and January 2021 with a convenience sample of 27 Army, Air Force, and Navy medics. We will also describe results from the Silent Test of algorithm performance, which ran at Mayo Clinic in Rochester, MN from May to October 2020.

Battlefield care is complex, emergent, and hazardous. Medics who provide such care require support to enable them to diagnose and treat life-threatening afflictions such as shock. Applications outside of the military from first responders to humanitarian relief workers can benefit from these results.

Take Away Points
a. Decision support systems can minimize time to act and reduce likelihood of error
b. Training in life critical skills can sustain skills that would otherwise erode
c. Effective decision support relies on rigorous research and evaluation
d. Machine learning can identify clinically meaningful trends in patient data