Hemoperitoneum semiquantitative analysis on admission of blunt trauma patients improves the prediction of massive transfusion

Is a Positive Prehospital FAST Associated with Severe Bleeding? A Multicenter Retrospective Study

INTRODUCTION

Severe hemorrhage is the leading cause of early preventable death in severe trauma patients. Delayed diagnosis is a poor prognostic factor, and severe hemorrhage prediction is essential. The aim of our study was to investigate if there was an association between the detection of peritoneal or pleural fluid on prehospital sonography for trauma and posttraumatic severe hemorrhage.

METHODS

We retrospectively studied data from records of thoracic or abdominal trauma patients managed in mobile intensive care units from January 2017 to December 2021 in four centers in France. Severe hemorrhage was defined as a condition necessitating transfusion of at least four packed red blood cells or surgical intervention/radioembolization for hemostasis within the first 24 h. Using a multivariate analysis, we investigated the predictive performance of focused assessment with sonography for trauma (FAST) alone or in combination with the five Red Flags criteria validated by Hamada et al.

RESULTS

Among the 527 patients analyzed, 371 (71%) were men, the mean age was 41 ± 19 years, and the Injury Severity Score was 11 (Interquartile range = [5; 22]). Seventy-three (14%) patients had severe hemorrhage - of whom 28 (38%) had a positive FAST, compared to 61 (13%) without severe hemorrhage (p < 0.01). For severe hemorrhage prediction, FAST had a sensitivity of 38% (95%CI = [27%; 50%]) and a specificity of 87% (95%CI = [83%; 90%]) (AUC = 0.62, 95%CI = [0.57; 0.68]). The comparison of the other outcomes between positive and negative FAST was: hemostatic procedure, 22 (25%) vs 28 (6%), p < 0.01; intensive care unit admission 71 (80%) vs 190 (43%), p < 0.01; mean length of hospital stay 11 [4; 27] vs 4 [0; 14] days, p = 0.02; 30-day mortality 13 (15%) vs 22 (5%), p < 0.01.

CONCLUSION

A positive FAST performed in the prehospital setting is associated with severe hemorrhage and all prognostic criteria we studied.

A Multiscale Deep Learning Method for Quantitative Visualization of Traumatic Hemoperitoneum at CT: Assessment of Feasibility and Comparison with Subjective Categorical Estimation

PURPOSE

To evaluate the feasibility of a multiscale deep learning algorithm for quantitative visualization and measurement of traumatic hemoperitoneum and to compare diagnostic performance for relevant outcomes with categorical estimation.

MATERIALS AND METHODS

This retrospective, single-institution study included 130 patients (mean age, 38 years; interquartile range, 25-50 years; 79 men) with traumatic hemoperitoneum who underwent CT of the abdomen and pelvis at trauma admission between January 2016 and April 2019. Labeled cases were separated into five combinations of training (80%) and test (20%) sets, and fivefold cross-validation was performed. Dice similarity coefficients (DSCs) were compared with those from a three-dimensional (3D) U-Net and a coarse-to-fine deep learning method. Areas under the receiver operating characteristic curve (AUCs) for a composite outcome, including hemostatic intervention, transfusion, and in-hospital mortality, were compared with consensus categorical assessment by two radiologists. An optimal cutoff was derived by using a radial basis function-based support vector machine.

RESULTS

Mean DSC for the multiscale algorithm was 0.61 ± 0.15 (standard deviation) compared with 0.32 ± 0.16 for the 3D U-Net method and 0.52 ± 0.17 for the coarse-to-fine method (P < .0001). Correlation and agreement between automated and manual volumes were excellent (Pearson r = 0.97, intraclass correlation coefficient = 0.93). The algorithm produced intuitive and explainable visual results. AUCs for automated volume measurement and categorical estimation were 0.86 and 0.77, respectively (P = .004). An optimal cutoff of 278.9 mL yielded accuracy of 84%, sensitivity of 82%, specificity of 93%, positive predictive value of 86%, and negative predictive value of 83%.

CONCLUSION

A multiscale deep learning method for traumatic hemoperitoneum quantitative visualization had improved diagnostic performance for predicting hemorrhage-control interventions and mortality compared with subjective volume estimation. Supplemental material is available for this article. © RSNA, 2020.

Findings requiring immediate surgery in blunt abdominal trauma patients with isolated free fluid without solid organ injury on abdominal computed tomography: Retrospective laboratory, clinical and radiologic analysis. A case control study

BACKGROUND

Determining surgical treatment is difficult in blunt abdominal trauma (BAT) patients with isolated free fluid without solid organ injury (IFFWSOI) on abdominal computed tomography (CT). We investigated the laboratory, clinical, and radiologic features of BAT patients with IFFWSOI on abdominal CT requiring surgery.

METHODS

A retrospective medical record review was performed for patients treated at our government-established regional tertiary trauma center from March 2014 to August 2018. A total of 501 patients were identified and reviewed. Patients were divided into Surgery and No Surgery groups for analysis. The Surgery group included patients who underwent surgery during the index admission, while the No Surgery group included patients who did not undergo surgery.

RESULTS

There were significantly more cases of severe fluid collection (61.5% vs. 11.8%; p < 0.001), car accidents (69.2% vs. 35.3%; p = 0.018), and abdominal pain (87.2% vs. 58.8%; p = 0.031) at the emergency department in the Surgery group. Regarding laboratory studies performed at the emergency department, only the median amylase level was significantly higher in the No Surgery group (54.5 U/L vs. 62.5 U/L; p = 0.048). On multivariate logistic regression analysis with adjustments for age and sex, the odds ratio (OR) for severe fluid collection on abdominal CT to predict surgery was 13.52 (p = 0.006), while the OR for abdominal pain was 7.34 (p = 0.036) and the OR for car accident was 2.14 (p = 0.329). In addition, a multivariate logistic regression with adjustment for age, sex, delta neutrophil index, and C-reactive protein, showed the same propensity as the other model, although statistical significance was retained only for severe fluid collection.

CONCLUSION

Surgical treatment should be actively considered in the presence of a large volume of intra-abdominal free fluid, especially when concomitant with abdominal pain or after car accidents in BAT patients without solid organ injury.

Quantity of hemoperitoneum is associated with need for intervention in patients with stable blunt splenic injury

Background

In patients with hemodynamically stable blunt splenic injury (BSI), there is no consensus on whether quantity of hemoperitoneum (HP) is a predictor for intervention with splenic artery embolization (SAE) or failing nonoperative management (fNOM). We sought to analyze whether the quantity of HP was associated with need for intervention.

Methods

This retrospective cohort study included adult trauma patients with hemodynamically stable BSI admitted to six trauma centers between 2014 and 2016. Quantity of HP was defined as small (perisplenic blood or blood in Morrison's pouch), moderate (blood in one or both pericolic gutters), or large (additional finding of free blood in the pelvis). Multivariate logistic regression was performed to identify predictors of intervention with SAE or fNOM versus successful observation.

Results

There were 360 patients: hemoperitoneum was noted in 214 (59%) patients, of which the quantity was small in 92 (43%), moderate in 76 (35.5%), and large in 46 (21.5%). Definitive management was as follows: 272 (76%) were observed and 88 (24%) had intervention (83 SAE, 5 fNOM). The rate of intervention was univariately associated with quantity of HP, even after stratification by American Association for the Surgery of Trauma (AAST) grade. After adjustment, larger quantities of HP significantly increased odds of intervention (p=0.01). Compared with no HP, the odds of intervention were significantly increased for moderate HP (OR=3.51 (1.49 to 8.26)) and large HP (OR=2.89 (1.03 to 8.06)), with similar odds for small HP (OR=1.21 (0.46 to 2.76)). Other independent predictors of intervention were higher AAST grade, older age, and presence of splenic vascular injury.

Conclusion

Greater quantity of HP was associated with increased odds of intervention, with no difference in risk for moderate versus large HP. These findings suggest quantity of HP should be incorporated in the management algorithm of BSI as a consideration for angiography and/or embolization to maximize splenic preservation and reduce the risk of splenic rupture.

Level of evidence

III, retrospective epidemiological study.

Systematic reviews of scores and predictors to trigger activation of massive transfusion protocols

BACKGROUND

The use of massive transfusion protocols (MTPs) in the resuscitation of hemorrhaging trauma patients ensures rapid delivery of blood products to improve outcomes, where the decision to trigger MTPs early is important. Scores and tools to predict the need for MTP activation have been developed for use to aid with clinical judgment. We performed a systematic review to assess (1) the scores and tools available to predict MTP in trauma patients, (2) their clinical value and diagnostic accuracies, and (3) additional predictors of MTP.

METHODS

MEDLINE, EMBASE, and CENTRAL were searched from inception to June 2017. All studies that utilized scores or predictors of MTP activation in adult (age, ≥18 years) trauma patients were included. Data collection for scores and tools included reported sensitivities and specificities and accuracy as defined by the area under the curve of the receiver operating characteristic.

RESULTS

Forty-five articles were eligible for analysis, with 11 validated and four unvalidated scores and tools assessed. Of four scores using clinical assessment, laboratory values, and ultrasound assessment the modified Traumatic Bleeding Severity Score had the best performance. Of those scores, the Trauma Associated Severe Hemorrhage score is most well validated and has higher area under the curve of the receiver operating characteristic than the Assessment of Blood Consumption and Prince of Wales scores. Without laboratory results, the Assessment of Blood Consumption score balances accuracy with ease of use. Without ultrasound use, the Vandromme and Schreiber scores have the highest accuracy and sensitivity respectively. The Shock Index uses clinical assessment only with fair performance. Other clinical variables, laboratory values, and use of point-of-care testing results were identified predictors of MTP activation.

CONCLUSION

The use of scores or tools to predict MTP need to be individualized to hospital resources and skill set to aid clinical judgment. Future studies for triggering nontrauma MTP activations are needed.

LEVEL OF EVIDENCE

Systematic review, level III.

Relationship between Obesity and Massive Transfusion Needs in Trauma Patients, and Validation of TASH Score in Obese Population: A Retrospective Study on 910 Trauma Patients

Background

Prediction of massive transfusion (MT) is challenging in management of trauma patients. However, MT and its prediction were poorly studied in obese patients. The main objective was to assess the relationship between obesity and MT needs in trauma patients. The secondary objectives were to validate the Trauma Associated Severe Hemorrhage (TASH) score in predicting MT in obese patients and to use a grey zone approach to optimize its ability to predict MT.

Methods and Findings

An observational retrospective study was conducted in a Level I Regional Trauma Center Trauma in obese and non-obese patients. MT was defined as ≥10U of packed red blood cells in the first 24h and obesity as a BMI≥30kg/m². Between January 2008 and December 2012, 119 obese and 791 non-obese trauma patients were included. The rate of MT was 10% (94/910) in the whole population. The MT rate tended to be higher in obese patients than in non-obese patients: 15% (18/119, 95%CI 9‒23%) versus 10% (76/791, 95%CI 8‒12%), OR, 1.68 [95%CI 0.97‒2.92], p = 0.07. After adjusting for Injury Severity Score (ISS), obesity was significantly associated with MT rate (OR, 1.79[95%CI 1.00‒3.21], p = 0.049). The TASH score was higher in the obese group than in the non-obese group: 7(4–11) versus 5(2–10)(p<0.001). The area under the ROC curves of the TASH score in predicting MT was very high and comparable between the obese and non-obese groups: 0.93 (95%CI, 0.89‒0.98) and 0.94 (95%CI, 0.92‒0.96), respectively (p = 0.80). The grey zone ranged respectively from 10 to 13 and from 9 to 12 in obese and non obese patients, and allowed separating patients at low, intermediate or high risk of MT using the TASH score.

Conclusions

Obesity was associated with a higher rate of MT in trauma patients. The predictive performance of the TASH score and the grey zones were robust and comparable between obese and non-obese patients.

Prediction of critical haemorrhage following trauma: A narrative review

Introduction: Traumatic haemorrhagic shock can be difficult to diagnose. Models for predicting critical bleeding and massive transfusion have been developed to aid clinicians. The aim of this review is to outline the various available models and report on their performance and validation. Methods: A review of the English and non-English literature in Medline, PubMed and Google Scholar was conducted from 1990 to September 2015. We combined several terms for i) haemorrhage AND ii) prediction, in the setting of iii) trauma. We included models that had at least two data points. We extracted information about the models, their developments, performance and validation. Results: There were 36 different models identified that diagnose critical bleeding, which included a total of 36 unique variables. All models were developed retrospectively. The models performed with variable predictive abilities–the most superior with an area under the receiver operating characteristics curve of 0.985, but included detailed findings on imaging and was based on a small cohort. The most commonly included variable was systolic blood pressure, featuring in all but five models. Pattern or mechanism of injury were used by 16 models. Pathology results were used by 15 models, of which nine included base deficit and eight models included haemoglobin. Imaging was utilised in eight models. Thirteen models were known to be validated, with only one being prospectively validated. Conclusions: Several models for predicting critical bleeding exist, however none were deemed accurate enough to dictate treatment. Potential areas of improvement identified include measures of variability in vital signs and point of care imaging and pathology testing.

[Multidisciplinary consensus document on the management of massive haemorrhage (HEMOMAS document)]

Massive haemorrhage is common and often associated with high morbidity and mortality. We perform a systematic review of the literature, with extraction of the recommendations from the existing evidences because of the need for its improvement and the management standardization. From the results we found, we wrote a multidisciplinary consensus document. We begin with the agreement in the definitions of massive haemorrhage and massive transfusion, and we do structured recommendations on their general management (clinical assessment of bleeding, hypothermia management, fluid therapy, hypotensive resuscitation and damage control surgery), blood volume monitoring, blood products transfusion (red blood cells, fresh frozen plasma, platelets and their best transfusion ratio), and administration of hemostatic components (prothrombin complex, fibrinogen, factor VIIa, antifibrinolytic agents).

Multidisciplinary consensus document on the management of massive haemorrhage (HEMOMAS document)

Massive haemorrhage is common and often associated with high morbidity and mortality. We perform a systematic review of the literature, with extraction of the recommendations from the existing evidences because of the need for its improvement and the management standardization. From the results we found, we wrote a multidisciplinary consensus document. We begin with the agreement in the definitions of massive haemorrhage and massive transfusion, and we do structured recommendations on their general management (clinical assessment of bleeding, hypothermia management, fluid therapy, hypotensive resuscitation and damage control surgery), blood volume monitoring, blood products transfusion (red blood cells, fresh frozen plasma, platelets and their best transfusion ratio), and administration of hemostatic components (prothrombin complex, fibrinogen, factor VIIa, antifibrinolytic agents).

A simple assessment of haemothoraces thickness predicts abundant transfusion: a series of 525 blunt trauma patients

BACKGROUND

The goal of this study was to evaluate the performance of haemothorax quantification to predict an abundant transfusion in blunt thoracic trauma patients.

METHODS

This study included all severe trauma patients admitted into our trauma centre from January 2005 to January 2010, who presented a blunt thoracic trauma (thoracic AIS ≥1) and had a CT scan within the first hour following admission. For each haemothorax, thickness of dominant side and the cumulated thicknesses of both sides (Dominant-t, Cumulated-t), as well as lengths (Dominant-l, Cumulated-l) and the calculated volumes (Dominant-v, Cumulated-v provided by a previously validated formula) were retrospectively measured by CT scan. A multiple logistic regression was performed to define the independent risk factors for an abundant transfusion (≥5 packed red blood cells in the first 24h). Finally, ROC curves have been drawn on an isolated thoracic trauma subgroup to predict abundant transfusion. The more specific cut-offs were extracted from this analysis.

RESULTS

From the 525 blunt thoracic trauma patients (75% males, mean age 38.2 (SD18.7) years, mean ISS 22.5 (SD16.4)), 31% received an abundant transfusion. In multivariable analysis, Cumulated-t was significantly associated with an abundant transfusion (OR 1.3 [95% CI 1.1-1.4], P=0.002). In isolated thoracic trauma subgroup (n=251), the global abilities of different CT measurements to predict abundant transfusion were significantly comparable (AUCs 0.69-0.70). The more specific cut-offs were established at 28mm for Cumulated-t (specificity 92%, positive predictive value 47%) and at 24mm for Dominant-t (specificity 92%; positive predictive value 43%).

CONCLUSION

The haemothorax quantification upon admission may help to predict transfusion needs. Cumulated-t was found independent risk factor for abundant transfusion in a large population of severe trauma patients. Beyond a Cumulated-t of 28mm or a Dominant-t of 24mm, abundant transfusion will be very frequently necessary.

LEVEL OF EVIDENCE

Retrospective review, level III.