Relationship between Serum Lactate with the Severity of Injury in Patients with Spinal Cord TraumaRelationship between Serum Lactate with the Severity of Injury in Patients with Spinal Cord Trauma

Over 1000 patients with traumatic paraplegia or tetraplegia and many more with fractures and dislocations of the spine without damage to the central nervous system have been observed and treated at the Sheffield Spinal Injuries Centre. The vertebral lesions with or without injury to the spinal cord or nerve roots have been classified on the basis of the clinical and roentgenographic findings into five groups: 1. Pure flexion which causes a wedge fracture which is stable. 2. Flexion-rotation which produces an unstable fracture-dislocation with rupture of tue posterior ligament complex, separation of the spinous processes, a slice fracture near the upper border of the lower vertebra, and dislocation of the lower articular processes of the upper vertebra. 3. Extension which causes rupture of the intervertebral disc and the anterior common ligament along with avulsion of a small bone fragment from the anterior border of the dislocated vertebra. The dislocation almost always reduces spontaneously and is stable in flexion. 4. Vertebral compression which results in a fracture of the end plate as the nucleus of the intervertebral disc is forced into the vertebral body and causes it to burst with outward displacement of fragments of the body. Since the ligaments remain intact this comminuted fracture is stable. 5. Shearing which results in forward displacement of the whole vertebra and an unstable fracture of the articular processes or pedicles. Accurate diagnosis and prognosis of the neurological lesion depend on knowledge of the anatomy of the spinal cord and nerve roots, a careful neurological examination shortly after the original injury and repeated examinations thereafter, comparison of the level of spinal injury with the level of paraplegia or tetraplegia, differentiation between paraplegia and tetraplegia of immediate and delayed onset, and the appropriate therapy of the various types and levels of lesion. Simple wedge fractures were treated by bed rest for two to three weeks, mobilization of the back, and ambulation with a back support. Rotational fracture-dislocations in the cervical, thoracolumbar, or lumbar spine were almost invariably associated with tetraplegia or paraplegia. Cervical fracture-dislocations with or without tetraplegia were treated by skull-caliper traction. Thoracolumbar or lumbar fracture-dislocations without paraplegia were treated on a plaster bed for twelve weeks followed by a back support for a few weeks. The thoracolumbar fracture-dislocations with paraplegia were never treated by the plaster bed method but rather by open reduction of the dislocation, and maintenance of the reduction by internal fixation with double plating of the spinous processes. Spontaneous fusion was sufficiently advanced after eight to twelve weeks to get the patient out of bed. If the plates cut out of the bone after twelve weeks, they were removed. Patients with loss of sensation resulting from paraplegia or tetraplegia were turned every two hours to avoid pressure sores. Extension dislocations in the cervical spine, if they had reduced spontaneously, were fitted with a collar to hold the head and neck in sligh flexion for a period of eight to twelve weeks. For dislocations in this region which had not reduced spontaneously, manual manipulation under endotracheal anesthesia was employed. Reduction was maintained by skull tongs applied prior to manipulation. If after four weeks there was roentgenographic evidence of new bone indicating Spontaneous fusion, traction was continued for four to six weeks more followed by a neck collar for an additional six weeks. If new bone did not appear on the roentgenograms after four weeks, anterior fusion was performed followed by skull traction for an additional eight weeks. Vertical compression burst fractures in the cervical spine were treated by skull traction for six weeks followed by a neck collar. In the lumbar spine, burst fractures without paraplegia were treated by immobilization in a plaster bed for eight to twelve weeks followed by back support. The plaster bed was never used in burst fractures with paraplegia. Shear fractures were always associated with complete paraplegia. These fractures were usually stable and did not require operative reduction except when the displacement was great.

Lipid profile analysis in spinal trauma patients shows severe distortion of AA/DHA after injury

Cerebral lactate concentration can remain detectable in neonatal hypoxic-ischemic encephalopathy (HIE) after hemodynamic stability. The temporal resolution of regional cerebral lactate concentration in relation to the severity or area of injury is unclear. Furthermore, the interplay between serum and cerebral lactate in neonatal HIE has not been well defined. The study aims to describe cerebral lactate concentration in neonatal HIE in relation to time, injury, and serum lactate. Fifty-two newborns with HIE undergoing therapeutic hypothermia (TH) were enrolled. Magnetic resonance imaging and spectroscopy (MRI + MR spectroscopy) were performed during and after TH at 54.6 ± 15.0 and 156 ± 57.6 h of life, respectively. Severity and predominant pattern of injury was scored radiographically. Single-voxel 1H MR spectra were acquired using short-echo (35 ms) PRESS sequence localized to the basal ganglia (BG), thalamus (Thal), gray matter (GM), and white matter. Cerebral lactate concentration was quantified by LCModel software. Serum and cerebral lactate concentrations were plotted based on age at time of measurement. Multiple comparisons of regional cerebral lactate concentration based on severity and predominant pattern of injury were performed. Spearman's Rho was computed to determine correlation between serum lactate and cerebral lactate concentration at the respective regions of interest. Overall, serum lactate concentration decreased over time. Cerebral lactate concentration remained low for less severe injury and decreased over time for more severe injury. Cerebral lactate remained detectable even after TH. During TH, there was a significant higher concentration of cerebral lactate at the areas of injury and also when injury was more severe. However, these differences were no longer observed after TH. There was a weak correlation between serum lactate and cerebral lactate concentration at the BG (rs = 0.3, p = 0.04) and Thal (rs = 0.35, p = 0.02). However, in infants with moderate-severe brain injury, a very strong correlation exists between serum lactate and cerebral lactate concentration at the BG (rs = 0.7, p = 0.03), Thal (rs = 0.9 p = 0.001), and GM (rs = 0.6, p = 0.04) regions. Cerebral lactate is most significantly different between regions and severity of injury during TH. There is a moderate correlation between serum and cerebral lactate concentration measured in the deep gray nuclei during TH. Differences in injury and altered regional cerebral metabolism may account for these differences.

Background Sepsis is a life-threatening organ dysfunction caused by a dysregulated host response to infection, if not recognized early and managed promptly, it can lead to septic shock, multiple organ failure and death. Sepsis is associated with high mortality, and the early recognition of the signs of tissue hypo perfusion is crucial in its management. Aim The aim of the study was to compare between PCO2 gap, serum lactate and procalcitonin as predictors of clinical outcomes in ICU septic patients regarding 28 days mortality or developing septic shock or multiorgan failure (MOF). Patients and Method This prospective observational study was carried out in the ICU of Ain Shams university hospitals and Egypt air hospital. The study was done on 80 cases. Inclusion criteria: Adult septic patients ≥ 21 years old. Exclusion criteria: Patients with history of chronic obstructive pulmonary disease and bronchial asthma or died <48 h of admission. Method: All patients were subjected to complete history taking and Physical examinations to exclude systemic diseases. Investigational Studies: Routine laboratory investigations: complete blood count (CBC), erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), liver and kidney functions, PT, PTT and INR. Serum lactate: was done on admission then after 48 hours. PCO2 arterial and central venous (PCo2gap): was done on admission then after 48 hours. Procalcitonin: on admission then after 48 hours. APACHE ΙΙ and SOFA scores were recorded on admission. Results patients were classified in to 2 groups according to mortality: survivors group & non survivors group, survivors were n = 60(75%) and non survivors were n = 20(25%). Regarding serum lactate on 1st day of admission and after 48 h, it revealed that survivors group showed significant reduction in serum lactate level at 48 h compared to baseline (p < 0.001). while in non-survivors group there was statistically significant increase in serum lactate after 48 h than at 1st day of admission (P < 0.001), the cutoff point of s. lactate on 1st day was >3.97 mmo/l showed sensitivity of 90% and specificity of 93.33%, while after 48 hours the cutoff point of s.lactate was >2.68mmo/l showed sensitivity of 100% and specificity of 100%. Regarding PCO2 gap on 1st day of admission and after 48 h, it revealed that survivors group showed significant reduction in PCO2 gap after 48 h compared to baseline (p < 0.001). However, non-survivors group showed an increase in PCO2 gap after 48 h compared to baseline with statistical significance (p < 0.001), the cutoff point of PCO2 gap on 1st day was >7.48 mmHg gave sensitivity of 90% and specificity of 90%. While the cutoff point of PCO2 gap after 48 h. was >6.2 mmHg gave sensitivity of 95% and specificity of 95%. Regarding procalcitonin on 1st day of admission and after 48 h, it revealed that survivors group showed significant reduction in procalcitonin after 48 h compared to baseline (p < 0.001). Non-survivors group showed significant increase in procalcitonin after 48 h compared to baseline (p < 0.001), the cutoff point of procalcitonin on 1st day was >2.336 ng/ml gave sensitivity of 90% and specificity of 81.67%. While the cutoff point of procalcitonin after 48h. was >2.14 ng/ml gave sensitivity of 90% and specificity of 85%. Regarding APACHE II severity score on 1st day it revealed that survivors group ranged between 5-17, while in non survivors group ranged between 15-26, revealed that higher APACHE II score associated significantly with increased mortality in non-survivors compared to survivors group (P < 0.0001). Regarding SOFA severity score on 1st day of admission in survivors group ranged between 1-13,while in non survivors group ranged between 10-17. SOFA score was significantly higher in non-survivors compared to survivors group (P < 0.0001). MOF were 25(41.7%) in survivors group and 20(100%) in non-survivors group, there was statistical significant higher frequency of MOF in non-survivors group than survivors group (P < 0.001). Patients who developed septic shock and required vasopressors were 43(71.7%) in survivors group and 20(100%) in non-survivors group, there was statistical significant higher frequency of vasopressors requirement in non-survivors group than survivors group (P = 0.007). Conclusion PCO2 gap, serum lactate and procalcitonin in addition to severity scores (APACHE II & SOFA scores) were identified as predictors of clinical outcome in ICU septic patients. Serum lactate followed by PCO2 gap then procalcitonin had comparable prognostic accuracy with severity scores (on admission and after 48 h).

BACKGROUND AND OBJECTIVESIn Saudi Arabia (SA), injuries are the second leading cause of death; however, little is known about their frequencies and outcomes. Trauma registries play a major role in measuring the burden on population health. This study aims to describe the population of the only hospital-based trauma registry in the country and highlight challenges and potential opportunities to improve trauma data collection and research in SA.DESIGN AND SETTINGSUsing data between 2001 and 2010, this retrospective study included patients from a large trauma center in Riyadh, SA.PATIENTS AND METHODSA staff nurse utilized a structured checklist to gather information on patients’ demographic, physiologic, anatomic, and outcome variables. Basic descriptive statistics by age group (≤14 vs >14 years) were calculated, and differences were assessed using student t and chi-square tests. In addition, the mechanism of injury and the frequency of missing data were evaluated.RESULTS10 847 patients from the trauma registry were included. Over 9% of all patients died either before or after being treated at the hospital. Patients who were older than 14 years of age (more likely to be male) sustained traffic-related injuries and died in the hospital as compared to patients who were younger than or equal to years of age. Deceased patients were severely injured as measured by injury severity score and Glasgow Coma Scale (P<.001). Overall, the most frequent type of injury was related to traffic (52.0%), followed by falls (23.4%). Missing values were mostly prevalent in traffic-related variables, such as seatbelt use (70.2%).CONCLUSIONThis registry is a key step toward addressing the burden of injuries in SA. Improved injury classification using the International Classification of Disease-external cause codes may improve the quality of the registry and allow comparison with other populations. Most importantly, injury prevention in SA requires further investment in data collection and research to improve outcomes.

To assess the clinical application of magnetic resonance imaging (MRI) in patients with acute spinal cord trauma (SCT) according to the type, extension, and severity of injury and the clinical-radiological correlation. Diagnostic imaging [computed tomography (CT) and MRI] tests of 98 patients with acute SCT were analyzed to assess their clinical diagnostic value. The following radiological findings of SCT were investigated: vertebral compression fractures, bursts and dislocations, posterior element fractures, C1 and C2 lesions, vertebral listhesis, bone swelling, spinal canal compression, disk herniation, extradural hematoma, spinal cord contusions, spinal cord swelling, and posterior ligamentous complex (PLC) injuries. The radiological findings were better visualized using MRI, except for the posterior elements (p = 0.001), which were better identified with CT. A total of 271 lesions were diagnosed as follows: 217 using MRI, 154 using CT, and 100 (36.9 %) using both MRI and CT. MRI detected 117 more lesions than CT. MRI was significantly superior to CT in the diagnosis of bone swelling, PLC injury, disk herniation, spinal canal compression, spinal cord contusion and swelling present in SCT. MRI detected a larger number of lesions than CT and is highly useful for the diagnosis of soft tissue and intrathecal injuries.

Introduction: Spinal cord trauma (SCT) is an important cause of morbidity and mortality around the world. It affects different age groups, especially young adults who are victims of high-energy trauma. The most effective way to reduce the incidence of spinal cord trauma and its consequences is through preventive campaigns and control and surveillance measures through public agencies. The objective of this study is to outline the epidemiological profile of patients with spinal cord trauma attended at a tertiary care center in the city of São Paulo. Methods: Retrospective, cross-sectional study performed at a reference center for the care of patients with spinal cord injury in the State of São Paulo. Data were collected from the medical records of patients with spinal cord trauma between 2012 and 2016. Results: Of the 515 patients with spinal trauma, 153 (29.7%) had spinal cord injury of which 131 (85.62%) were male, and 22 (14.37%) were female, in a ratio of approximately 6:1. The mean age was 39.45 years. The main cause of spinal cord trauma observed was the fall from heights, with 72 cases (47.05%), and 52.94% were classified as Frankel A. Conclusions: The results showed that the majority of the patients were young, economically active, with low educational level, exposed to accidents that could be largely avoided. Most of these patients also had severe disabling injuries, which usually bring considerable psychological sequelae and economic consequences to the individual and to society. Level of evidence: IV. Type of study: Case series.

Head injury.

Background: Renal trauma is among the most important trauma challenges. Better management of traumatic renal patients is necessary to improve the patients’ clinical outcome. Objectives: This study aimed to analyze the pattern, severity, mechanism of injury, and outcome in renal trauma. Methods: This cross-sectional study was conducted in the largest trauma center in southern Iran from March 2018 to June 2022. Adult patients with diagnosed renal trauma based on abbreviated injury scale guideline were included. Variables of age, gender, anatomy of injury, mechanism of injury, lenghth of hospitalization, level of blood pressure, severity of injury, and the outcome of patients have been collected. Results: In total, 4,416 traumatic patients were admitted to the Emergency Department of the Hospital during the study time, of which 46 cases had traumatic renal injury. The rate of renal injury in the level one trauma center in southern Iran was 0.96%, and the death rate in renal injury victims was 12(26.1%). Most of the injured with renal injury were men 39(84.8%). Blunt trauma was the dominant type of trauma in most victims with renal injury 34(73.9%). Thirty-seven percent (n=17) of traumatic renal injury victim was in mild level (grade 1 and 2), and 16(34.8%) in severe level (grade 4 and 5). Renal injury was mostly associated with thorax injury (n=46) as extra-abdominal organ injury, and liver injury (n=16) and spleen (n=15) as intra-abdominal organ injury. Conclusion: The results of the present study showed that the severity of injury in patients with renal injury is high, and it is usually associated by injury to other body organs. It is suggested that the traumatic patients should be immediately examined for renal injury in the resuscitation department due to the high mortality rate and severity of these patients.

Pain Management Guidelines for Blunt Thoracic Trauma Bruce Simon;James Cushman;Robert Barraco;Vivian Lane;Fred Luchette;Maurizio Miglietta;David Roccaforte;Ruth Spector; The Journal of Trauma: Injury, Infection, and Critical Care

The beneficial effects of a thromboxane receptor antagonist on spinal cord perfusion following experimental cord injury

The Role of Therapeutic Hypothermia After Traumatic Spinal Cord Injury—A Systematic Review

&lt;h4&gt;Background&lt;/h4&gt;&lt;p dir="ltr"&gt;Trauma is a leading cause of death and disability worldwide, particularly among younger populations, but increasingly also among the elderly. Reliable data are crucial both for epidemiological understanding and for guiding improvements in trauma care. Trauma registers such as the Swedish Trauma Register (SweTrau) have therefore become vital resources as they can provide standardized information on injury mechanisms, severity, management, and outcomes. This thesis utilizes SweTrau data to address epidemiological questions and to evaluate the potential of machine learning methods for outcome prediction.&lt;/p&gt;&lt;h4&gt;Methods&lt;/h4&gt;&lt;p dir="ltr"&gt;All four studies used data collected by SweTrau. Study 1 was a retrospective cohort study of patients with and without pelvic fractures. Baseline characteristics, injury severity, physiological status, and mortality at 30 days and one year after injury were compared between the two groups. Univariate logistic regression was used to identify crude mortality associations. Potential confounders of these associations were assessed using the change-in-estimate method, and variables identified as confounders were subsequently included together with pelvic fracture in a multivariable logistic regression model. Study 2 was a descriptive epidemiological study examining the distribution of musculoskeletal injuries among trauma patients. Studies 3 and 4 were both based on a machine learning (ML) approach. In Study 3, three different ML models were applied to predict mortality, while in Study 4, they were used to predict admission to intensive care unit (ICU) and hospital length of stay (LOS). Results were evaluated with C-statistics, calibration curves, and decision curve analysis.&lt;/p&gt;&lt;h4&gt;Results&lt;/h4&gt;&lt;p dir="ltr"&gt;Across the four studies, more than 37,000 trauma patients from the Swedish national trauma register were analyzed. In study 1, pelvic fracture was found to be associated with a higher crude mortality compared to patients without pelvic fracture (30-day mortality 9% vs. 4%). However, after adjustment for confounders including age, circulatory shock, severe head injury, and overall injury severity, a pelvic fracture was not a risk factor for mortality, suggesting they reflect injury burden rather than uniquely drive mortality. Study 2 demonstrated that musculoskeletal injuries in trauma were highly prevalent, affecting 41% of all trauma patients, with fractures representing the vast majority of the musculoskeletal injuries. The spine was the most frequently injured region, followed by upper and lower extremities, respectively. Patients with musculoskeletal injuries showed higher Injury Severity Score (ISS), longer hospital stay and slightly increased mortality. Distinct patterns were observed across injury mechanisms: traffic accidents dominated, while penetrating trauma showed clear associations with extremity injuries. In study 3, three ML methods were compared with the Trauma and Injury Severity Score (TRISS) for mortality prediction in 9,208 severely injured trauma patients. All tested ML models, particularly the extreme Gradient Boosting (XGB) model, outperformed TRISS, achieving an Area Under Curve (AUC) of 0.91 (95% CI: 0.88-0.93) versus 0.85 (95% CI: 0.82-0.88) for TRISS. The most important predictors identified for mortality were age, Glasgow Coma Scale (GCS), base excess, New Injury Severity Score (NISS), severity of head and thoracic injuries, systolic blood pressure, and American Society of Anaesthesiologists (ASA) class. The ML models also demonstrated better calibration and higher clinical utility than TRISS. In study 4, the ML approach was extended to the prediction of ICU admission and LOS in 9,056 severely injured trauma patients. The XGB model achieved excellent performance for ICU admission with AUC 0.85 (95% CI: 0.84-0.87), but only moderate accuracy for LOS prediction with AUCs between 0.64 and 0.71 depending on the category. The models were implemented in an online tool for individualized estimation of ICU needs and LOS.&lt;/p&gt;&lt;h4&gt;Conclusion&lt;/h4&gt;&lt;p dir="ltr"&gt;Together, the four papers demonstrated that trauma outcomes are influenced by injury patterns, physiological status, and comorbidities. They further showed how insights into these factors can be leveraged into predictive models that outperform traditional statistical methods for trauma prediction.&lt;/p&gt;&lt;h3&gt;List of scientific papers&lt;/h3&gt;&lt;p dir="ltr"&gt;I. The pelvic Fracture - an indicator of injury severity or a lethal fracture? &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Peyman Bakhshayesh, and Anders Enocson. Injury, Volume 49, Issue 8, August 2018, Pages 1568-1571. &lt;a href="https://doi.org/10.1016/j.injury.2018.06.016" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2018.06.016&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;II. Musculoskeletal injuries in trauma patients: a Swedish nationwide register study including 37,266 patients. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, and Anders Enocson. Acta Orthopaedica, 2023; 94: 171-177. &lt;a href="https://doi.org/10.2340/17453674.2023.11960" rel="noreferrer" target="_blank"&gt;https://doi.org/10.2340/17453674.2023.11960&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;III. Prediction of mortality among severely injured trauma patients: A comparison between TRISS and machine learning-based predictive models. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Mathias Mosfeldt, Anders Enocson, and Hans E Berg. Injury, Volume 55, Issue 8, 111702 August 2024. &lt;a href="https://doi.org/10.1016/j.injury.2024.111702" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2024.111702&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;IV. Development of a new tool for prediction of hospital length of stay and intensive care needs in trauma patients using Machine Learning. Mathias Mosfeldt, &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, Anders Enocson. [Submitted]&lt;/p&gt;

&lt;h4&gt;Background&lt;/h4&gt;&lt;p dir="ltr"&gt;Trauma is a leading cause of death and disability worldwide, particularly among younger populations, but increasingly also among the elderly. Reliable data are crucial both for epidemiological understanding and for guiding improvements in trauma care. Trauma registers such as the Swedish Trauma Register (SweTrau) have therefore become vital resources as they can provide standardized information on injury mechanisms, severity, management, and outcomes. This thesis utilizes SweTrau data to address epidemiological questions and to evaluate the potential of machine learning methods for outcome prediction.&lt;/p&gt;&lt;h4&gt;Methods&lt;/h4&gt;&lt;p dir="ltr"&gt;All four studies used data collected by SweTrau. Study 1 was a retrospective cohort study of patients with and without pelvic fractures. Baseline characteristics, injury severity, physiological status, and mortality at 30 days and one year after injury were compared between the two groups. Univariate logistic regression was used to identify crude mortality associations. Potential confounders of these associations were assessed using the change-in-estimate method, and variables identified as confounders were subsequently included together with pelvic fracture in a multivariable logistic regression model. Study 2 was a descriptive epidemiological study examining the distribution of musculoskeletal injuries among trauma patients. Studies 3 and 4 were both based on a machine learning (ML) approach. In Study 3, three different ML models were applied to predict mortality, while in Study 4, they were used to predict admission to intensive care unit (ICU) and hospital length of stay (LOS). Results were evaluated with C-statistics, calibration curves, and decision curve analysis.&lt;/p&gt;&lt;h4&gt;Results&lt;/h4&gt;&lt;p dir="ltr"&gt;Across the four studies, more than 37,000 trauma patients from the Swedish national trauma register were analyzed. In study 1, pelvic fracture was found to be associated with a higher crude mortality compared to patients without pelvic fracture (30-day mortality 9% vs. 4%). However, after adjustment for confounders including age, circulatory shock, severe head injury, and overall injury severity, a pelvic fracture was not a risk factor for mortality, suggesting they reflect injury burden rather than uniquely drive mortality. Study 2 demonstrated that musculoskeletal injuries in trauma were highly prevalent, affecting 41% of all trauma patients, with fractures representing the vast majority of the musculoskeletal injuries. The spine was the most frequently injured region, followed by upper and lower extremities, respectively. Patients with musculoskeletal injuries showed higher Injury Severity Score (ISS), longer hospital stay and slightly increased mortality. Distinct patterns were observed across injury mechanisms: traffic accidents dominated, while penetrating trauma showed clear associations with extremity injuries. In study 3, three ML methods were compared with the Trauma and Injury Severity Score (TRISS) for mortality prediction in 9,208 severely injured trauma patients. All tested ML models, particularly the extreme Gradient Boosting (XGB) model, outperformed TRISS, achieving an Area Under Curve (AUC) of 0.91 (95% CI: 0.88-0.93) versus 0.85 (95% CI: 0.82-0.88) for TRISS. The most important predictors identified for mortality were age, Glasgow Coma Scale (GCS), base excess, New Injury Severity Score (NISS), severity of head and thoracic injuries, systolic blood pressure, and American Society of Anaesthesiologists (ASA) class. The ML models also demonstrated better calibration and higher clinical utility than TRISS. In study 4, the ML approach was extended to the prediction of ICU admission and LOS in 9,056 severely injured trauma patients. The XGB model achieved excellent performance for ICU admission with AUC 0.85 (95% CI: 0.84-0.87), but only moderate accuracy for LOS prediction with AUCs between 0.64 and 0.71 depending on the category. The models were implemented in an online tool for individualized estimation of ICU needs and LOS.&lt;/p&gt;&lt;h4&gt;Conclusion&lt;/h4&gt;&lt;p dir="ltr"&gt;Together, the four papers demonstrated that trauma outcomes are influenced by injury patterns, physiological status, and comorbidities. They further showed how insights into these factors can be leveraged into predictive models that outperform traditional statistical methods for trauma prediction.&lt;/p&gt;&lt;h3&gt;List of scientific papers&lt;/h3&gt;&lt;p dir="ltr"&gt;I. The pelvic Fracture - an indicator of injury severity or a lethal fracture? &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Peyman Bakhshayesh, and Anders Enocson. Injury, Volume 49, Issue 8, August 2018, Pages 1568-1571. &lt;a href="https://doi.org/10.1016/j.injury.2018.06.016" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2018.06.016&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;II. Musculoskeletal injuries in trauma patients: a Swedish nationwide register study including 37,266 patients. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, and Anders Enocson. Acta Orthopaedica, 2023; 94: 171-177. &lt;a href="https://doi.org/10.2340/17453674.2023.11960" rel="noreferrer" target="_blank"&gt;https://doi.org/10.2340/17453674.2023.11960&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;III. Prediction of mortality among severely injured trauma patients: A comparison between TRISS and machine learning-based predictive models. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Mathias Mosfeldt, Anders Enocson, and Hans E Berg. Injury, Volume 55, Issue 8, 111702 August 2024. &lt;a href="https://doi.org/10.1016/j.injury.2024.111702" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2024.111702&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;IV. Development of a new tool for prediction of hospital length of stay and intensive care needs in trauma patients using Machine Learning. Mathias Mosfeldt, &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, Anders Enocson. [Submitted]&lt;/p&gt;

&lt;h4&gt;Background&lt;/h4&gt;&lt;p dir="ltr"&gt;Trauma is a leading cause of death and disability worldwide, particularly among younger populations, but increasingly also among the elderly. Reliable data are crucial both for epidemiological understanding and for guiding improvements in trauma care. Trauma registers such as the Swedish Trauma Register (SweTrau) have therefore become vital resources as they can provide standardized information on injury mechanisms, severity, management, and outcomes. This thesis utilizes SweTrau data to address epidemiological questions and to evaluate the potential of machine learning methods for outcome prediction.&lt;/p&gt;&lt;h4&gt;Methods&lt;/h4&gt;&lt;p dir="ltr"&gt;All four studies used data collected by SweTrau. Study 1 was a retrospective cohort study of patients with and without pelvic fractures. Baseline characteristics, injury severity, physiological status, and mortality at 30 days and one year after injury were compared between the two groups. Univariate logistic regression was used to identify crude mortality associations. Potential confounders of these associations were assessed using the change-in-estimate method, and variables identified as confounders were subsequently included together with pelvic fracture in a multivariable logistic regression model. Study 2 was a descriptive epidemiological study examining the distribution of musculoskeletal injuries among trauma patients. Studies 3 and 4 were both based on a machine learning (ML) approach. In Study 3, three different ML models were applied to predict mortality, while in Study 4, they were used to predict admission to intensive care unit (ICU) and hospital length of stay (LOS). Results were evaluated with C-statistics, calibration curves, and decision curve analysis.&lt;/p&gt;&lt;h4&gt;Results&lt;/h4&gt;&lt;p dir="ltr"&gt;Across the four studies, more than 37,000 trauma patients from the Swedish national trauma register were analyzed. In study 1, pelvic fracture was found to be associated with a higher crude mortality compared to patients without pelvic fracture (30-day mortality 9% vs. 4%). However, after adjustment for confounders including age, circulatory shock, severe head injury, and overall injury severity, a pelvic fracture was not a risk factor for mortality, suggesting they reflect injury burden rather than uniquely drive mortality. Study 2 demonstrated that musculoskeletal injuries in trauma were highly prevalent, affecting 41% of all trauma patients, with fractures representing the vast majority of the musculoskeletal injuries. The spine was the most frequently injured region, followed by upper and lower extremities, respectively. Patients with musculoskeletal injuries showed higher Injury Severity Score (ISS), longer hospital stay and slightly increased mortality. Distinct patterns were observed across injury mechanisms: traffic accidents dominated, while penetrating trauma showed clear associations with extremity injuries. In study 3, three ML methods were compared with the Trauma and Injury Severity Score (TRISS) for mortality prediction in 9,208 severely injured trauma patients. All tested ML models, particularly the extreme Gradient Boosting (XGB) model, outperformed TRISS, achieving an Area Under Curve (AUC) of 0.91 (95% CI: 0.88-0.93) versus 0.85 (95% CI: 0.82-0.88) for TRISS. The most important predictors identified for mortality were age, Glasgow Coma Scale (GCS), base excess, New Injury Severity Score (NISS), severity of head and thoracic injuries, systolic blood pressure, and American Society of Anaesthesiologists (ASA) class. The ML models also demonstrated better calibration and higher clinical utility than TRISS. In study 4, the ML approach was extended to the prediction of ICU admission and LOS in 9,056 severely injured trauma patients. The XGB model achieved excellent performance for ICU admission with AUC 0.85 (95% CI: 0.84-0.87), but only moderate accuracy for LOS prediction with AUCs between 0.64 and 0.71 depending on the category. The models were implemented in an online tool for individualized estimation of ICU needs and LOS.&lt;/p&gt;&lt;h4&gt;Conclusion&lt;/h4&gt;&lt;p dir="ltr"&gt;Together, the four papers demonstrated that trauma outcomes are influenced by injury patterns, physiological status, and comorbidities. They further showed how insights into these factors can be leveraged into predictive models that outperform traditional statistical methods for trauma prediction.&lt;/p&gt;&lt;h3&gt;List of scientific papers&lt;/h3&gt;&lt;p dir="ltr"&gt;I. The pelvic Fracture - an indicator of injury severity or a lethal fracture? &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Peyman Bakhshayesh, and Anders Enocson. Injury, Volume 49, Issue 8, August 2018, Pages 1568-1571. &lt;a href="https://doi.org/10.1016/j.injury.2018.06.016" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2018.06.016&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;II. Musculoskeletal injuries in trauma patients: a Swedish nationwide register study including 37,266 patients. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, and Anders Enocson. Acta Orthopaedica, 2023; 94: 171-177. &lt;a href="https://doi.org/10.2340/17453674.2023.11960" rel="noreferrer" target="_blank"&gt;https://doi.org/10.2340/17453674.2023.11960&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;III. Prediction of mortality among severely injured trauma patients: A comparison between TRISS and machine learning-based predictive models. &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Mathias Mosfeldt, Anders Enocson, and Hans E Berg. Injury, Volume 55, Issue 8, 111702 August 2024. &lt;a href="https://doi.org/10.1016/j.injury.2024.111702" rel="noreferrer" target="_blank"&gt;https://doi.org/10.1016/j.injury.2024.111702&lt;/a&gt;&lt;/p&gt;&lt;p dir="ltr"&gt;IV. Development of a new tool for prediction of hospital length of stay and intensive care needs in trauma patients using Machine Learning. Mathias Mosfeldt, &lt;b&gt;Jonas Holtenius&lt;/b&gt;, Hans E Berg, Anders Enocson. [Submitted]&lt;/p&gt;