Evaluation of ISS, RTS, CASS and TRISS scoring systems for predicting outcomes of blunt trauma abdomen

Evaluation of management of CT scan proved solid organ injury in blunt injury abdomen-a prospective study

BACKGROUND

Trauma especially road traffic injury is one of the major health-related issues throughout the world, especially in developing countries like India (Mattox 2022). Solid organ injury is the most common cause of morbidity and mortality in patients with blunt abdominal trauma. The non-operative management (NOM) is being consistently followed for hemodynamically stable patients with respect to solid organ injuries. This study aims to provide an evidence base for management modalities of solid organ injuries in blunt abdominal trauma.

AIM

The aim of this study is to evaluate the effectiveness of various treatment modalities for solid organ injury in blunt abdominal trauma.

OBJECTIVES

Evaluating the characteristics of blunt abdominal injury with respect to age and gender; distribution, mode of injury, most common organ injured, and severity of injury; effect of delay in getting treatment on the management outcome for patients with solid organ injury; evaluating the various modalities of treatment of CT-proven solid organ injury; incidence of complications in different modes of treatment.

METHODS

All patients aged more than 18 years and suffering from CT-proven solid organ injury secondary to blunt abdominal trauma between February 2021 and September 2022 were included in this prospective observational study. Sixty-five patients were enrolled in the study after meeting the inclusion criteria. Details such as age, gender, mechanism of injury, the time between injury to first hospital contact, presenting complaints, organ and grade of injury, Revised Trauma Score (RTS), Trauma Score and Injury Severity Score (TRISS), management, and outcomes were collected using self-designed pro forma and analyzed. Different modalities of treatment were evaluated and patients undergoing operative and non-operative management were compared. Patients in whom non-operative management failed were compared with patients with successful non-operative management.

RESULTS

The mean age of patients involved were 36.8 years with a male:female ratio of 7.125:1 and the most common age group affected being between 21 and 30 years. The most common mode of injury was noted to be road traffic accidents (72.3%). The most common presenting complaints were abdominal pain (64.6%) followed by chest pain (29.2%) and vomiting (13.8%). There was no significant relationship between latent period and type of intervention or failure of non-operative management. FAST positivity rate was noted to be 92.3%. Chronic alcoholism and bronchial asthma were significant predictors for patients undergoing upfront surgery (p = 0.003 and 0.006 respectively). The presence of pelvic and spine injury was statistically significant for predicting mortality in polytrauma patients (p = 0.003). Concurrent adrenal injury was found in 24.6% of patients but was not related to failure of non-operative management or mortality. RTS significantly predicts the multitude of organ involvement (p = 0.015). The liver was the most common organ injured (60%) followed by the spleen (52.3%) and the kidney (20%). The liver and the spleen (9.2%) were noted to be the most common organ combination involved. No specific organ or organ injury combination was noted to predict failure of non-operative management or mortality. But the multitude of organ involvement was statistically significant for predicting patients undergoing upfront surgery (p = 0.011). Out of 65 patients enrolled in the study, 7 patients (10.8%) underwent immediate surgery, and 58 patients (89.2%) underwent non-operative management. Among the 68 chosen for non-operative management, 6 patients (9.2%) failed non-operative management and 52 patients (80%) had success of non-operative management. A significant drop in hemoglobin (83.3%) on day 1 (66.6%) was seen to be the commonest reason for failure of non-operative management. The spleen was noted to be the most commonly involved organ intra-operatively (61.5%) followed by the liver (30.8%). Concordance between pre-operative and intra-operative grading of organ injuries was highest for liver and kidney injuries (100%) and lowest for pancreatic injuries (0%). Requirement of blood transfusion and liver injuries were significant factors for failure of non-operative management (p = 0.012 and 0.045 respectively). The presence of pancreatic leak was significant between the non-operated patients and patients operated upfront (p = 0.003). Mortality was noted to be 10.8% (7 patients) in our study.

CONCLUSION

Solid organ injury in blunt abdominal trauma is an important cause of morbidity and mortality. RTS was noted to be a good predictor for solid organ injury in blunt abdominal trauma. Pancreatic injuries are notorious for being under-staged on CT findings; hence, the need arises for multimodality imaging for suspected pancreatic injuries. Non-operative management is a successful modality of treatment for majority of patients suffering from multiple solid organ injuries in blunt abdominal trauma provided serial close monitoring of patient's clinical signs and hemoglobin is instituted along with the presence of an emergency surgery team.

Machine learning models predict triage levels, massive transfusion protocol activation, and mortality in trauma utilizing patients hemodynamics on admission

BACKGROUND

The effective management of trauma patients necessitates efficient triaging, timely activation of Massive Blood Transfusion Protocols (MTP), and accurate prediction of in-hospital outcomes. Machine learning (ML) algorithms have emerged as up-and-coming tools in the domains of optimizing triage decisions, improving intervention strategies, and predicting clinical outcomes, consistently outperforming traditional methodologies. This study aimed to develop, assess, and compare several ML models for the triaging processes, activation of MTP, and mortality prediction.

METHODS

In a 10-year retrospective study, the predictive capabilities of seven ML models for trauma patients were systematically assessed using on-admission patients' hemodynamic data. All patient's data were randomly divided into training (80 %) and test (20 %) sets. Employing Python for data preprocessing, feature scaling, and model development, we evaluated K-Nearest Neighbors (KNN), Logistic Regression (LR), Decision Tree (DT), Support Vector Machines (SVM) with RBF kernels, Random Forest (RF), Extreme Gradient Boosting (XGBoost), and Artificial Neural Network (ANN). We employed various imputation techniques and addressed data imbalance through down-sampling, up-sampling, and synthetic minority for the over-sampling technique (SMOTE). Hyperparameter tuning, coupled with 5-fold cross-validation, was performed. The evaluation included essential metrics like sensitivity, specificity, F1 score, accuracy, Area Under the Receiver Operating Curve (AUC ROC), and Area Under the Precision recall Curve (AUC PR), ensuring robust predictive capability.

RESULT

This study included 17,390 adult trauma patients; of them, 19.5 % (3385) were triaged at a critical level, 3.8 % (664) required MTP, and 7.7 % (1335) died in the hospital. The model's performance improved using imputation and balancing techniques. The overall models demonstrated notable performance metrics for predicting triage, MTP activation, and mortality with F1 scores of 0.75, 0.42, and 0.79, sensitivities of 0.73, 0.82, and 0.9, and AUC ROC values of 0.89, 0.95 and 0.99 respectively.

CONCLUSION

Machine learning, especially RF models, effectively predicted trauma triage, MTP activation, and mortality. Featured critical hemodynamic variables include shock indices, systolic blood pressure, and mean arterial pressure. Therefore, models can do better than individual parameters for the early management and disposition of patients in the ED. Future research should focus on creating sensitive and interpretable models to enhance trauma care.

Prediction of mortality among severely injured trauma patients A comparison between TRISS and machine learning-based predictive models

BACKGROUND

Given the huge impact of trauma on hospital systems around the world, several attempts have been made to develop predictive models for the outcomes of trauma victims. The most used, and in many studies most accurate predictive model, is the "Trauma Score and Injury Severity Score" (TRISS). Although it has proven to be fairly accurate and is widely used, it has faced criticism for its inability to classify more complex cases. In this study, we aimed to develop machine learning models that better than TRISS could predict mortality among severely injured trauma patients, something that has not been studied using data from a nationwide register before.

METHODS

Patient data was collected from the national trauma register in Sweden, SweTrau. The studied period was from the 1st of January 2015 to 31st of December 2019. After feature selection and multiple imputation of missing data three machine learning (ML) methods (Random Forest, eXtreme Gradient Boosting, and a Generalized Linear Model) were used to create predictive models. The ML models and TRISS were then tested on predictive ability for 30-day mortality.

RESULTS

The ML models were well-calibrated and outperformed TRISS in all the tested measurements. Among the ML models, the eXtreme Gradient Boosting model performed best with an AUC of 0.91 (0.88-0.93).

CONCLUSION

This study showed that all the developed ML-based prediction models were superior to TRISS for the prediction of trauma mortality.

Use of the Trauma and Injury Severity Score (TRISS) as a Predictor of Patient Outcome in Cases of Trauma Presenting in the Trauma and Emergency Department of a Tertiary Care Institute

BACKGROUND

In this study, we used the anatomic scoring system Abbreviated Injury Scale (AIS) to calculate the Injury Severity Score (ISS) and the physiological scoring system for the Revised Trauma Score (RTS) on the arrival of patients. Both scores were used to calculate the Trauma and Injury Severity Score (TRISS) for predicting the patient outcome in a case of trauma.

METHODS

This prospective, cross-sectional, observational study was carried out at the trauma centre of a tertiary care institute and included patients of either sex, age ≥18 years, and ISS ≥15. A total of 2084 cases of trauma over a period of 18 months were assessed, and 96 cases of blunt trauma meeting the inclusion criteria were studied.

RESULTS

Patients injured in road traffic accidents constituted the maximum caseload. Out of a sample size of 96 patients with ISS ≥15, 77 died during the treatment and 19 survived. The ISS ranged from 15 to 66, with a mean ± SD score of 27.48 ± 8.79. Non-survivors had a statistically higher significant ISS than survivors (p<0.001). The RTS ranged from <1 to 7.84, with a mean ± SD score of 4.52 ± 2.08. Non-survivors had low RTS (RTS <5, n=52) compared to survivors, and the difference was statistically significant (p<0.001). The mean ± SD TRISS (Ps) score was 0.69 ± 2.288. In the non-survivor (NS) group, 15 patients had TRISS (Ps) between 0.26-0.50, followed by 0.51-0.75 (n=18), 0.76-0.90 (n=12), and 0.90-0.95 (n=11). While 16 survivors had TRISS (Ps) between 0.96 and 1, a statistically significant association was found between TRISS and patient outcome (p-value <0.001). On the receiver operating characteristic (ROC) curve analysis, the sensitivity of TRISS (94.7%) and RTS was found to be comparable (94.7%), whereas ISS was less sensitive (36.8%) in predicting the patient outcome. RTS (79.2%) and TRISS (76.6%) scores were more specific than ISS (5.2%) for outcome analysis.

CONCLUSION

The TRISS score is useful in the management of trauma patients as it can satisfactorily predict mortality in a case of trauma. The trauma scores are of immense help in determining the nature of injury in medicolegal cases.

Development and Asian-wide validation of the Grade for Interpretable Field Triage (GIFT) for predicting mortality in pre-hospital patients using the Pan-Asian Trauma Outcomes Study (PATOS)

Background

Field triage is critical in injury patients as the appropriate transport of patients to trauma centers is directly associated with clinical outcomes. Several prehospital triage scores have been developed in Western and European cohorts; however, their validity and applicability in Asia remains unclear. Therefore, we aimed to develop and validate an interpretable field triage scoring systems based on a multinational trauma registry in Asia.

Methods

This retrospective and multinational cohort study included all adult transferred injury patients from Korea, Malaysia, Vietnam, and Taiwan between 2016 and 2018. The outcome of interest was a death in the emergency department (ED) after the patients' ED visit. Using these results, we developed the interpretable field triage score with the Korea registry using an interpretable machine learning framework and validated the score externally. The performance of each country's score was assessed using the area under the receiver operating characteristic curve (AUROC). Furthermore, a website for real-world application was developed using R Shiny.

Findings

The study population included 26,294, 9404, 673 and 826 transferred injury patients between 2016 and 2018 from Korea, Malaysia, Vietnam, and Taiwan, respectively. The corresponding rates of a death in the ED were 0.30%, 0.60%, 4.0%, and 4.6% respectively. Age and vital sign were found to be the significant variables for predicting mortality. External validation showed the accuracy of the model with an AUROC of 0.756-0.850.

Interpretation

The Grade for Interpretable Field Triage (GIFT) score is an interpretable and practical tool to predict mortality in field triage for trauma.

Funding

This research was supported by a grant of the Korea Health Technology R&D Project through the Korea Health Industry Development Institute (KHIDI), funded by the Ministry of Health & Welfare, Republic of Korea (Grant Number: HI19C1328).

The reverse shock index multiplied by Glasgow coma scale (rSIG) is predictive of mortality in trauma patients according to age

OBJECTIVE

The role of reverse shock index multiplied Glasgow coma scale (rSIG) in patients post-trauma with traumatic brain injury (TBI) has not yet been defined well. Our study aimed to investigate the predictive performance of rSIG according to age group.

METHOD

This is a prospective multi-national and multi-center cohort study using Pan-Asian Trauma Outcome Study registry in Asian-Pacific, conducted on patients post-trauma who visited participating hospitals. The main exposure was low rSIG measured at emergency department. The main outcome was in-hospital mortality. We performed multilevel logistic regression analysis to estimate the association low rSIG and study outcomes. Interaction analysis between rSIG and age group were also conducted.

RESULTS

Low rSIG was significantly associated with an increase in in-hospital mortality in patients post-trauma with and without TBI (aOR (95% CI): 1.49 (1.04-2.13) and 1.71 (1.16-2.53), respectively). The ORs for in-hospital mortality differed according to the age group in patients post-trauma with TBI (1.72 (1.44-1.94) for the young group and 1.13 (1.07-1.52) for the old group; p < 0.05).

CONCLUSION

Low rSIG is associated with an increase in in-hospital mortality in adult patients post-trauma. However, in patients with TBI, the prediction of mortality is significantly better in younger patient group.