Statistical Machines for Trauma Hospital Outcomes Research: Application to the PRospective, Observational, Multi-Center Major Trauma Transfusion (PROMMTT) Study

Artificial intelligence assistance in deciding management strategies for polytrauma and trauma patients

<b><br>Introduction:</b> Artificial intelligence (AI) is an emerging technology with vast potential for use in several fields of medicine. However, little is known about the application of AI in treatment decisions for patients with polytrauma. In this systematic review, we investigated the benefits and performance of AI in predicting the management of patients with polytrauma and trauma.</br> <b><br>Methods:</b> This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. Studies were extracted from the PubMed and Google Scholar databases from their inception until November 2022, using the search terms "Artificial intelligence," "polytrauma," and "decision." Seventeen articles were identified and screened for eligibility. Animal studies, review articles, systematic reviews, meta-analyses, and studies that did not involve polytrauma or severe trauma management decisions were excluded. Eight studies were eligible for final review.</br> <b><br>Results:</b> Eight studies focusing on patients with trauma, including two on military trauma, were included. The AI applications were mainly implemented for predictions and/or decisions on shock, bleeding, and blood transfusion. Few studies predicted death/survival. The identification of trauma patients using AI was proposed in a previous study. The overall performance of AI was good (six studies), excellent (one study), and acceptable (one study).</br> <b><br>Discussion:</b> AI demonstrated satisfactory performance in decision-making and management prediction in patients with polytrauma/severe trauma, especially in situations of shock/bleeding.</br> <b><br>Importance:</b> The present study serves as a basis for further research to develop practical AI applications for the management of patients with trauma.</br>.

Prediction of Prognosis in Patients with Trauma by Using Machine Learning

Background and Objectives: We developed a machine learning algorithm to analyze trauma-related data and predict the mortality and chronic care needs of patients with trauma. Materials and Methods: We recruited admitted patients with trauma during 2015 and 2016 and collected their clinical data. Then, we subjected this database to different machine learning techniques and chose the one with the highest accuracy by using cross-validation. The primary endpoint was mortality, and the secondary endpoint was requirement for chronic care. Results: Data of 5871 patients were collected. We then used the eXtreme Gradient Boosting (xGBT) machine learning model to create two algorithms: a complete model and a short-term model. The complete model exhibited an 86% recall for recovery, 30% for chronic care, 67% for mortality, and 80% for complications; the short-term model fitted for ED displayed an 89% recall for recovery, 25% for chronic care, and 41% for mortality. Conclusions: We developed a machine learning algorithm that displayed good recall for the healthy recovery group but unsatisfactory results for those requiring chronic care or having a risk of mortality. The prediction power of this algorithm may be improved by implementing features such as age group classification, severity selection, and score calibration of trauma-related variables.

Whole blood for blood loss: hemostatic resuscitation in damage control

Hemorrhagic shock and its complications are a major cause of death among trauma patients. The management of hemorrhagic shock using a damage control resuscitation strategy has been shown to decrease mortality and improve patient outcomes. One of the components of damage control resuscitation is hemostatic resuscitation, which involves the replacement of lost blood volume with components such as packed red blood cells, fresh frozen plasma, cryoprecipitate, and platelets in a 1:1:1:1 ratio. However, this is a strategy that is not applicable in many parts of Latin America and other low-and-middle-income countries throughout the world, where there is a lack of well-equipped blood banks and an insufficient availability of blood products. To overcome these barriers, we propose the use of cold fresh whole blood for hemostatic resuscitation in exsanguinating patients. Over 6 years of experience in Ecuador has shown that resuscitation with cold fresh whole blood has similar outcomes and a similar safety profile compared to resuscitation with hemocomponents. Whole blood confers many advantages over component therapy including, but not limited to the transfusion of blood with a physiologic ratio of components, ease of transport and transfusion, less volume of anticoagulants and additives transfused to the patient, and exposure to fewer donors. Whole blood is a tool with reemerging potential that can be implemented in civilian trauma centers with optimal results and less technical demand.

Trauma-induced coagulopathy: The past, present, and future

Trauma remains a leading cause of death worldwide, and most early preventable deaths in both the civilian and military settings are due to uncontrolled hemorrhage, despite paradigm advances in modern trauma care. Combined tissue injury and shock result in hemostatic failure, which has been identified as a multidimensional molecular, physiologic and clinical disorder termed trauma-induced coagulopathy (TIC). Understanding the biology of TIC is of utmost importance, as it is often responsible for uncontrolled bleeding, organ failure, thromboembolic complications, and death. Investigations have shown that TIC is characterized by multiple phenotypes of impaired hemostasis due to altered biology in clot formation and breakdown. These coagulopathies are attributable to tissue injury and shock, and encompass underlying endothelial, immune and inflammatory perturbations. Despite the recognition and identification of multiple mechanisms and mediators of TIC, and the development of targeted treatments, the mortality rates and associated morbidities due to hemorrhage after injury remain high. The purpose of this review is to examine the past and present understanding of the multiple distinct but highly integrated pathways implicated in TIC, in order to highlight the current knowledge gaps and future needs in this evolving field, with the aim of reducing morbidity and mortality after injury.

Discrepancies between conventional and viscoelastic assays in identifying trauma-induced coagulopathy

BACKGROUND

Trauma-induced coagulopathy can present as abnormalities in a conventional or viscoelastic coagulation assay or both. We hypothesized that patients with discordant coagulopathies reflect different clinical phenotypes.

METHODS

Blood samples were collected prospectively from critically injured patients upon arrival at two urban Level I trauma centers. International normalized ratio (INR), partial thromboplastin time (PTT), thromboelastography (TEG), and coagulation factors were assayed.

RESULTS

278 patients (median ISS 17, mortality 26%) were coagulopathic: 20% with isolated abnormal INR and/or PTT (CONVENTIONAL), 49% with isolated abnormal TEG (VISCOELASTIC), and 31% with abnormal INR/PTT and TEG (BOTH). Compared with VISCOELASTIC, CONVENTIONAL and BOTH had higher ISS, lower GCS, larger base deficit, and decreased factor activities (all p < 0.017). They received more blood products and had more ICU/ventilation days (all p < 0.017). Mortality was higher in CONVENTIONAL (40%) and BOTH (49%) than VISCOELASTIC (6%, p < 0.017).

CONCLUSIONS

Although TEG-guided resuscitation improves survival after injury, INR and PTT identify coagulopathic patients with highest mortality regardless of TEG and likely represent distinct mechanisms independent of biochemical clot strength.