TCCC Decision Support With Machine Learning Prediction of Hemorrhage Risk, Shock Probability.

INTRODUCTION: Expected future delays in evacuation during near-peer conflicts in remote locales are expected to require extended care including prolonged field care over hours to days. Such delays can increase potential complications, such as insufficient blood flow (shock), bloodstream infection (sepsis), internal bleeding (hemorrhage), and require more complex treatment beyond stabilization. The Trauma Triage Treatment and Training Decision Support (4TDS) system is a real-time decision support system to monitor casualty health and identify such complications. The 4TDS software prototype operates on an Android smart phone or tablet configured for use in the DoD Nett Warrior program. It includes machine learning models to evaluate trends in six vital signs streamed from a sensor placed on a casualty to identify shock probability, internal hemorrhage risk, and need for a massive transfusion. MATERIALS AND METHODS: The project team used a mixed methods approach to create and evaluate the system including literature review, rapid prototyping, design requirements review, agile development, an algorithm "silent test," and usability assessments with novice to expert medics from all three services. RESULTS: Both models, shock (showing an accuracy of 0.83) and hemorrhage/massive transfusion protocol, were successfully validated using externally collected data. All usability assessment participants completed refresher training scenarios and were able to accurately assess a simulated casualty's condition using the phone prototype. Mean responses to statements on evaluation criteria [e.g., fit with Tactical Combat Casualty Care (TCCC), ease of use, and decision confidence] fell at five or above on a 7-point scale, indicating strong support. CONCLUSIONS: Participatory design ensured 4TDS and machine learning models reflect medic and clinician mental models and work processes and built support among potential users should the system transition to operational use. Validation results can support 4TDS readiness for FDA 510k clearance as a Class II medical device.

Validation of a Machine Learning Model for Early Shock Detection.

OBJECTIVES: The objectives of this study were to test in real time a Trauma Triage, Treatment, and Training Decision Support (4TDS) machine learning (ML) model of shock detection in a prospective silent trial, and to evaluate specificity, sensitivity, and other estimates of diagnostic performance compared to the gold standard of electronic medical records (EMRs) review. DESIGN: We performed a single-center diagnostic performance study. PATIENTS AND SETTING: A prospective cohort consisted of consecutive patients aged 18 years and older who were admitted from May 1 through September 30, 2020 to six Mayo Clinic intensive care units (ICUs) and five progressive care units. MEASUREMENTS AND MAIN RESULTS: During the study time, 5,384 out of 6,630 hospital admissions were eligible. During the same period, the 4TDS shock model sent 825 alerts and 632 were eligible. Among 632 hospital admissions with alerts, 287 were screened positive and 345 were negative. Among 4,752 hospital admissions without alerts, 78 were screened positive and 4,674 were negative. The area under the receiver operating characteristics curve for the 4TDS shock model was 0.86 (95% CI 0.85-0.87%). The 4TDS shock model demonstrated a sensitivity of 78.6% (95% CI 74.1-82.7%) and a specificity of 93.1% (95% CI 92.4-93.8%). The model showed a positive predictive value of 45.4% (95% CI 42.6-48.3%) and a negative predictive value of 98.4% (95% CI 98-98.6%). CONCLUSIONS: We successfully validated an ML model to detect circulatory shock in a prospective observational study. The model used only vital signs and showed moderate performance compared to the gold standard of clinician EMR review when applied to an ICU patient cohort.

Decision Support for Tactical Combat Casualty Care Using Machine Learning to Detect Shock.

INTRODUCTION: The emergence of more complex Prolonged Field Care in austere settings and the need to assist inexperienced providers' ability to treat patients create an urgent need for effective tools to support care. We report on a project to develop a phone-/tablet-based decision support system for prehospital tactical combat casualty care that collects physiologic and other clinical data and uses machine learning to detect and differentiate shock manifestation. MATERIALS AND METHODS: Software interface development methods included literature review, rapid prototyping, and subject matter expert design requirements reviews. Machine learning algorithm methods included development of a model trained on publicly available Medical Information Mart for Intensive Care data, then on de-identified data from Mayo Clinic Intensive Care Unit. RESULTS: The project team interviewed 17 Army, Air Force, and Navy medical subject matter experts during design requirements review sessions. They had an average of 17 years of service in military medicine and an average of 4 deployments apiece and all had performed tactical combat casualty care on live patients during deployment. Comments provided requirements for shock identification and management in prehospital settings, including support for indication of shock probability and shock differentiation. The machine learning algorithm based on logistic regression performed best among other algorithms we tested and was able to predict shock onset 90 minutes before it occurred with better than 75% accuracy in the test dataset. CONCLUSIONS: We expect the Trauma Triage, Treatment, and Training Decision Support system will augment a medic's ability to make informed decisions based on salient patient data and to diagnose multiple types of shock through remotely trained, field deployed ML models.

A Mouse Model of Repetitive Blast Traumatic Brain Injury Reveals Post-Trauma Seizures and Increased Neuronal Excitability.

Repetitive blast traumatic brain injury (TBI) affects numerous soldiers on the battlefield. Mild TBI has been shown to have long-lasting effects with repeated injury. We have investigated effects on neuronal excitability after repetitive, mild TBI in a mouse model of blast-induced brain injury. We exposed mice to mild blast trauma of an average peak overpressure of 14.6 psi, repeated across three consecutive days. While a single exposure did not reveal trauma as indicated by the glial fibrillary acidic protein indicator, three repetitive blasts did show significant increases. As well, mice had an increased indicator of inflammation (Iba-1) and increased tau, tau phosphorylation, and altered cytokine levels in the spleen. Video-electroencephalographic monitoring 48 h after the final blast exposure demonstrated seizures in 50% (12/24) of the mice, most of which were non-convulsive seizures. Long-term monitoring revealed that spontaneous seizures developed in at least 46% (6/13) of the mice. Patch clamp recording of dentate gyrus hippocampus neurons 48 h post-blast TBI demonstrated a shortened latency to the first spike and hyperpolarization of action potential threshold. We also found that evoked excitatory postsynaptic current amplitudes were significantly increased. These findings indicate that mild, repetitive blast exposures cause increases in neuronal excitability and seizures and eventual epilepsy development in some animals. The non-convulsive nature of the seizures suggests that subclinical seizures may occur in individuals experiencing even mild blast events, if repeated.

Improving Clinician Decisions and Communication in Critical Care Using Novel Information Technology.

INTRODUCTION: The electronic medical record (EMR) is presumed to support clinician decisions by documenting and retrieving patient information. Research shows that the EMR variably affects patient care and clinical decision making. The way information is presented likely has a significant impact on this variability. Well-designed representations of salient information can make a task easier by integrating information in useful patterns that clinicians use to make improved clinical judgments and decisions. Using Cognitive Systems Engineering methods, our research team developed a novel health information technology (NHIT) that interfaces with the EMR to display salient clinical information and enabled communication with a dedicated text-messaging feature. The software allows clinicians to customize displays according to their role and information needs. Here we present results of usability and validation assessments of the NHIT. MATERIALS AND METHODS: Our subjects were physicians, nurses, respiratory therapists, and physician trainees. Two arms of this study were conducted, a usability assessment and then a validation assessment. The usability assessment was a computer-based simulation using deceased patient data. After a brief five-minute orientation, the usability assessment measured individual clinician performance of typical tasks in two clinical scenarios using the NHIT. The clinical scenarios included patient admission to the unit and patient readiness for surgery. We evaluated clinician perspective about the NHIT after completing tasks using 7-point Likert scale surveys. In the usability assessment, the primary outcome was participant perceptions about the system's ease of use compared to the legacy system.A subsequent cross-over, validation assessment compared performance of two clinical teams during simulated care scenarios: one using only the legacy IT system and one using the NHIT in addition to the legacy IT system. We oriented both teams to the NHIT during a 1-hour session on the night before the first scenario. Scenarios were conducted using high-fidelity simulation in a real burn intensive care unit room. We used observations, task completion times, semi-structured interviews, and surveys to compare user decisions and perceptions about their performance. The primary outcome for the validation assessment was time to reach accurate (correct) decision points. RESULTS: During the usability assessment, clinicians were able to complete all tasks requested. Clinicians reported the NHIT was easier to use and the novel information display allowed for easier data interpretation compared to subject recollection of the legacy EMR.In the validation assessment, a more junior team of clinicians using the NHIT arrived at accurate diagnoses and decision points at similar times as a more experienced team. Both teams noted improved communication between team members when using the NHIT and overall rated the NHIT as easier to use than the legacy EMR, especially with respect to finding information. CONCLUSIONS: The primary findings of these assessments are that clinicians found the NHIT easy to use despite minimal training and experience and that it did not degrade clinician efficiency or decision-making accuracy. These findings are in contrast to common user experiences when introduced to new EMRs in clinical practice.

Training and Assessing Critical Airway, Breathing, and Hemorrhage Control Procedures for Trauma Care: Live Tissue Versus Synthetic Models.

INTRODUCTION: Optimal teaching and assessment methods and models for emergency airway, breathing, and hemorrhage interventions are not currently known. The University of Minnesota Combat Casualty Training consortium (UMN CCTC) was formed to explore the strengths and weaknesses of synthetic training models (STMs) versus live tissue (LT) models. In this study, we compare the effectiveness of best in class STMs versus an anesthetized caprine (goat) model for training and assessing seven procedures: junctional hemorrhage control, tourniquet (TQ) placement, chest seal, needle thoracostomy (NCD), nasopharyngeal airway (NPA), tube thoracostomy, and cricothyrotomy (Cric). METHODS: Army combat medics were randomized to one of four groups: 1) LT trained-LT tested (LT-LT), 2) LT trained-STM tested (LT-STM), 3) STM trained-LT tested (STM-LT), and 4) STM trained-STM tested (STM-STM). Participants trained in small groups for 3 to 4 hours and were evaluated individually. LT-LT was the "control" to which other groups were compared, as this is the current military predeployment standard. The mean procedural scores (PSs) were compared using a pairwise t-test with a Dunnett's correction. Logistic regression was used to compare critical fails (CFs) and skipped tasks. RESULTS: There were 559 subjects included. Junctional hemorrhage control revealed no difference in CFs, but LT-tested subjects (LT-LT and STM-LT) skipped this task more than STM-tested subjects (LT-STM and STM-STM; p < 0.05), and STM-STM had higher PSs than LT-LT (p < 0.001). For TQ, both STM-tested groups (LT-STM and STM-STM) had more CFs than LT-LT (p < 0.001) and LT-STM had lower PSs than LT-LT (p < 0.05). No differences were seen for chest seal. For NCD, LT-STM had more CFs than LT-LT (p = 0.001) and lower PSs (p = 0.001). There was no difference in CFs for NPA, but all groups had worse PSs versus LT-LT (p < 0.05). For Cric, we were underpowered; STM-LT trended toward more CFs (p = 0.08), and STM-STM had higher PSs than LT-LT (p < 0.01). Tube thoracostomy revealed that STM-LT had higher CFs than LT-LT (p < 0.05), but LT-STM had lower PSs (p < 0.05). An interaction effect (making the subjects who trained and tested on different models more likely to CF) was only found for TQ, chest seal, and Cric; however, of these three procedures, only TQ demonstrated any significant difference in CF rates. CONCLUSION: Training on STM or LT did not demonstrate a difference in subsequent performance for five of seven procedures (junctional hemorrhage, TQ, chest seal, NPA, and NCD). Until STMs are developed with improved anthropomorphic and tissue fidelity, there may still be a role for LT for training tube thoracostomy and potentially Cric. For assessment, our STM appears more challenging for TQ and potentially for NCD than LT. For junctional hemorrhage, the increased "skips" with LT may be explained by the differences in anatomic fidelity. While these results begin to uncover the effects of training and assessing these procedures on various models, further study is needed to ascertain how well performance on an STM or LT model translates to the human model.

Validation of an Assessment Tool for Field Endotracheal Intubation.

OBJECTIVES: Endotracheal intubation (ETI) is an important skill for all emergency providers; our ability to train and assess our learners is integral to providing optimal patient care. The primary aim of this study was to assess the inter-rater reliability (IRR) and discriminant validity of a novel field ETI assessment tool using a checklist-derived performance score (PS) and critical failure (CF) rate. METHODS: Forty-three participants (18 paramedic students, 11 paramedics, and 14 emergency physicians [EPs]) performed ETI during a simulated trauma scenario on a pseudo-ventilated cadaver. Each participant was assessed by two experienced raters. IRR was calculated using the intraclass correlation coefficient. Regarding discriminant validity, a Kruskal-Wallis test was used to analyze PSs and a χ2 test was used for CFs. Mean global rating scale (GRS) scores were compared using an analysis of variance. RESULTS: The ETI assessment tool had excellent IRR, with an intraclass correlation coefficient of 0.94. There was a significant difference in PSs, CFs, and GRSs (p < 0.05) between cohorts. CONCLUSION: The novel field ETI assessment tool has excellent reliability among trained raters and discriminates between experienced ETI providers (EPs) and less experienced ETI performers using PSs, CFs, and GRSs on a fresh cadaveric model.

When physics meets biology: low and high-velocity penetration, blunt impact, and blast injuries to the brain.

The incidence of traumatic brain injuries (TBI) in the US has reached epidemic proportions with well over 2 million new cases reported each year. TBI can occur in both civilians and warfighters, with head injuries occurring in both combat and non-combat situations from a variety of threats, including ballistic penetration, acceleration, blunt impact, and blast. Most generally, TBI is a condition in which physical loads exceed the capacity of brain tissues to absorb without injury. More specifically, TBI results when sufficient external force is applied to the head and is subsequently converted into stresses that must be absorbed or redirected by protective equipment. If the stresses are not sufficiently absorbed or redirected, they will lead to damage of extracranial soft tissue and the skull. Complex interactions and kinematics of the head, neck and jaw cause strains within the brain tissue, resulting in structural, anatomical damage that is characteristic of the inciting insult. This mechanical trauma then initiates a neuro-chemical cascade that leads to the functional consequences of TBI, such as cognitive impairment. To fully understand the mechanisms by which TBI occurs, it is critically important to understand the effects of the loading environments created by these threats. In the following, a review is made of the pertinent complex loading conditions and how these loads cause injury. Also discussed are injury thresholds and gaps in knowledge, both of which are needed to design improved protective systems.

Blast Testing Issues and TBI: Experimental Models That Lead to Wrong Conclusions.

Over the past several years, we have noticed an increase in the number of blast injury studies published in peer-reviewed biomedical journals that have utilized improperly conceived experiments. Data from these studies will lead to false conclusions and more confusion than advancement in the understanding of blast injury, particularly blast neurotrauma. Computational methods to properly characterize the blast environment have been available for decades. These methods, combined with a basic understanding of blast wave phenomena, enable researchers to extract useful information from well-documented experiments. This basic understanding must include the differences and interrelationships of static pressure, dynamic pressure, reflected pressure, and total or stagnation pressure in transient shockwave flows, how they relate to loading of objects, and how they are properly measured. However, it is critical that the research community effectively overcomes the confusion that has been compounded by a misunderstanding of the differences between the loading produced by a free field explosive blast and loading produced by a conventional shock tube. The principles of blast scaling have been well established for decades and when properly applied will do much to repair these problems. This paper provides guidance regarding proper experimental methods and offers insights into the implications of improperly designed and executed tests. Through application of computational methods, useful data can be extracted from well-documented historical tests, and future work can be conducted in a way to maximize the effectiveness and use of valuable biological test data.

Biophysical mechanisms of traumatic brain injuries.

Despite years of effort to prevent traumatic brain injuries (TBIs), the occurrence of TBI in the United States alone has reached epidemic proportions. When an external force is applied to the head, it is converted into stresses that must be absorbed into the brain or redirected by a helmet or other protective equipment. Complex interactions of the head, neck, and jaw kinematics result in strains in the brain. Even relatively mild mechanical trauma to these tissues can initiate a neurochemical cascade that leads to TBI. Civilians and warfighters can experience head injuries in both combat and noncombat situations from a variety of threats, including ballistic and blunt impact, acceleration, and blast. It is critical to understand the physics created by these threats to develop meaningful improvements to clinical care, injury prevention, and mitigation. Here the authors review the current state of understanding of the complex loading conditions that lead to TBI and characterize how these loads are transmitted through soft tissue, the skull and into the brain, resulting in TBI. In addition, gaps in knowledge and injury thresholds are reviewed, as these must be addressed to better design strategies that reduce TBI incidence and severity.