Background: We aimed to build a model using machine learning for the prediction of survival in trauma patients and compared these model predictions to those predicted by the most commonly used algorithm, the Trauma and Injury Severity Score (TRISS). Methods: Enrolled hospitalized trauma patients from 2009 to 2016 were divided into a training dataset (70% of the original data set) for generation of a plausible model under supervised classification, and a test dataset (30% of the original data set) to test the performance of the model. The training and test datasets comprised 13,208 (12,871 survival and 337 mortality) and 5603 (5473 survival and 130 mortality) patients, respectively. With the provision of additional information such as pre-existing comorbidity status or laboratory data, logistic regression (LR), support vector machine (SVM), and neural network (NN) (with the Stuttgart Neural Network Simulator (RSNNS)) were used to build models of survival prediction and compared to the predictive performance of TRISS. Predictive performance was evaluated by accuracy, sensitivity, and specificity, as well as by area under the curve (AUC) measures of receiver operating characteristic curves. Results: In the validation dataset, NN and the TRISS presented the highest score (82.0%) for balanced accuracy, followed by SVM (75.2%) and LR (71.8%) models. In the test dataset, NN had the highest balanced accuracy (75.1%), followed by the TRISS (70.2%), SVM (70.6%), and LR (68.9%) models. All four models (LR, SVM, NN, and TRISS) exhibited a high accuracy of more than 97.5% and a sensitivity of more than 98.6%. However, NN exhibited the highest specificity (51.5%), followed by the TRISS (41.5%), SVM (40.8%), and LR (38.5%) models. Conclusions: These four models (LR, SVM, NN, and TRISS) exhibited a similar high accuracy and sensitivity in predicting the survival of the trauma patients. In the test dataset, the NN model had the highest balanced accuracy and predictive specificity.

Background: This study was designed to measure the effect of stress-induced hyperglycemia (SIH) and diabetic hyperglycemia (DH) versus non-diabetic normoglycemia (NDN) on the outcomes of trauma patients in the intensive care unit (ICU). Methods: Diabetes mellitus (DM) was determined based on patient history and/or a hemoglobin A1c (HbA1c) level of ≥6.5% at admission. The patients who had serum glucose levels of ≥200 mg/dL in the absence or presence of DM were assigned into the groups SIH and DH, respectively. Diabetic normoglycemia (DN) and NDN were determined based on serum glucose levels of <200 mg/dL in patients with and without DM, respectively. Patients with burn injury or incomplete data were excluded. Detailed data of trauma patients in the ICU of a Level-I trauma center from 1 January 2009 to 31 December 2016 were retrieved from the database of the Trauma Registry System. These patients were classified into four exclusive groups, including NDN (n = 1745), DN (n = 306), SIH (n = 225) and DH (n = 206). The Pearson chi-square test was used to compare categorical data between groups. Continuous variables were compared using one-way analysis of variance along with the Games–Howell post hoc test. To decrease the confounding effect of the differences in sex and age, preexisting comorbidities and injury severity score (ISS) among different groups of patients, 1:1 ratio propensity score-matched cohorts were assigned using the NCSS software. The effect of hyperglycemia on the outcomes of patients with and without DM was assessed with a logistic regression analysis. Results: Among those selected propensity score-matched patient cohorts, the patients with SIH and DH had a 3.88-fold (95% CI, 2.13–7.06; p < 0.001) and 1.83-fold (95% CI, 1.00–3.34; p = 0.048) higher mortality, respectively, than those with NDN. Moreover, the patients in the SIH group (10.0 vs. 7.4 days; p = 0.005) and those in the DH group (10.1 vs. 7.4 days; p = 0.006) who were admitted to the ICU had a significantly longer length of stay than those in the NDN group. In addition, the SIH group had a 2.13-fold (95% CI, 1.04–4.36; p = 0.038) higher adjusted odds ratio for mortality than the DH group. Conclusions: This study revealed significantly worse outcomes in terms of mortality among patients with SIH and DH who were admitted to the ICU after controlling for sex and age, preexisting comorbidities and ISS. In addition, patients who had SIH presented significantly higher adjusted odds for mortality than those DH patients. These results suggest that hyperglycemia is detrimental in patients with or without DM who were admitted to the ICU, and there is a different pathophysiological mechanisms behind the SIH and DH.

Background: Hyperglycemia at the time of hospital admission is associated with higher morbidity and mortality rates in patients with traumatic brain injury (TBI). Using data from the Chang Gung Research Database (CGRD), this study aimed to compare mortality outcomes between patients with stress-induced hyperglycemia (SIH), diabetic hyperglycemia (DH), and nondiabetic normoglycemia (NDN). The study occurred at Keelung, Linkou, Chiayi, and Kaohsiung Chang Gung Memorial Hospitals (CGMHs). Methods: A total of 1166, 6318, 3622, and 5599 health records from Keelung, Linkou, Chiayi, and Kaohsiung CGMHs, respectively, were retrieved from the CGRD for hospitalized patients with TBI between January 2001 and December 2015. After propensity score matching for sex, age, and Glasgow Coma Scale (GCS) score, the matched cohorts were compared to evaluate differences in the primary outcome between patients with SIH, DH, and NDN. In-hospital mortality was the primary outcome. Results: The analysis of matched patient populations revealed that at the Kaohsiung CGMH, patients with SIH had 1.63-fold (95% CI: 1.09–2.44; p = 0.017) and 1.91-fold (95% CI: 1.12–3.23; p = 0.017) higher odds of mortality than patients with NDN and DH, respectively. Similar patterns were found at the Linkou CGMH; patients with SIH had higher odds of mortality than patients with NDN and DH. In contrast, at the Keelung CGMH, patients with SIH had significantly higher odds of mortality than those with NDN (OR: 3.25; 95% CI: 1.06–9.97; p = 0.039). At the Chiayi CGMH, there were no significant differences in mortality rates among all groups. Conclusions: This study’s results suggest that SIH and DH differ in their effect on the outcomes of patients with TBI. The results were similar between medical centers but not nonmedical centers; in the medical centers, patients with SIH had significantly higher odds of mortality than patients with either NDN or DH.

Background: In recent years, several versions of the Abbreviated Injury Scale (AIS) were updated and published. It was reported that the codeset in the dictionary of AIS-2005 had significant change from that of AIS-1998. This study was designed to evaluate the potential impact of adapting the AIS-2005 codeset from the AIS-1998 in an established trauma system of a single level I trauma center. The patients’ outcome was measured in different Injury Severity Score (ISS) strata according to the double-coded injuries in a three-year period. Methods: The double-coded injuries sustained by 7520 trauma patients between 1 January, 2016, and 31 December, 2018, in a level I trauma center were used to compare the patient injury characteristics and outcomes between AIS-1998 and AIS-2005 and under different ISS strata, defined as 24 (critical injury). Results: The mean ISS was significantly lower using AIS-2005 than using AIS-1998 (7.5 ± 6.3 vs. 8.3 ± 7.1, respectively, p < 0.001). AIS-2005 scores in the body regions of the head/neck (2.94 ± 1.08 vs. 3.40 ± 1.15, respectively, p < 0.001) and extremity (2.19 ± 0.56 vs. 2.24 ± 0.58, respectively, p < 0.001), but not in other body regions, were significantly lower than AIS-1998 scores. The critically injured patients (ISS >24), but not severely injured patients or patients with mild-to-moderate injury, coded by AIS-2005 had a significantly higher mortality rate (34.2% vs. 26.2%, respectively, p = 0.031) than did patients coded by AIS-1998. The rate of intensive care unit admission was significantly higher for patients in all ISS strata after adapting AIS-2005 as the scoring system than after adapting AIS-1998. Regarding patients with major trauma, which was defined as ISS > 15, the number of patients with major trauma in this study was 17.0% (n = 1276) for AIS-1998 and 9.7% (n = 733) for AIS-2005. As a consequence, the mortality rate of patients with major trauma was significantly higher in AIS-2005 than in AIS-1998 (15.4% vs. 9.1%, respectively, p < 000.1). Conclusions: In this study, we revealed that the adaptation of AIS-2005 from AIS-1998 had resulted in a significant decrease of severity scores in the measurement of the same injuries. The number of head/neck injuries classified as 16–24 was the key difference between AIS-1998 and AIS-2005. Furthermore, critically injured patients who had ISS > 24 coded by AIS-2005 had significantly higher mortality rates than did the patients coded by AIS-1998. This study also indicated that a direct comparison of the measurements that are generated from these two AIS versions can produce misleading results.

Background: Malnutrition is frequently underdiagnosed in geriatric patients and is considered to be a contributing factor for worse outcomes during hospitalization. In addition, elderly patients who undergo trauma are often malnourished at the time of incurring fractures. The geriatric nutritional risk index (GNRI), calculated based on the serum albumin level and the ratio of present body weight to ideal body weight, was proposed for the assessment of the nutritional status of elderly patients with various illnesses. This study aimed to investigate whether the GNRI has a prognostic value that links the nutritional status and mortality outcomes of elderly patients who have previously undergone trauma with femoral fractures. Methods: From January 1, 2009 to December 31, 2019, a total of 678 elderly patients with femoral fractures were categorized into four nutritional risk groups: a major-risk group (GNRI < 82; group 1, n = 127), moderate-risk group (GNRI 82–92; group 2, n = 179), low-risk group (GNRI 92–98; group 3, n = 123), and no-risk group (GNRI > 98; group 4, n = 249). To minimize the confounding effects of sex, age, preexisting comorbidities, and injury severity of patients on outcome measurements, propensity score-matched patient cohorts were created to assess the impact of patients being in different nutritional risk groups on the in-hospital mortality outcomes against the no-risk group. Results: The patients in groups 1–3 were significantly older and presented a significantly lower body mass index and lower serum albumin levels than those in group 4. Compared with patients in group 4 (3.6%), a significantly higher mortality rate was found in the patients in group 1 (17.3%, p < 0.001), but not in those in group 2 (6.7%) or group 3 (2.4%). The study of propensity score-matched patient cohorts provided similar results; group 1 patients had significantly higher odds of mortality than group 4 patients (odds ratio, 6.3; 95% confidence interval, 1.34–29.37; p = 0.009), but there were no significant differences in mortality risks among patients in groups 2 and 3 compared with those in group 4. Conclusions: This preliminary study suggested that the GNRI may be used as a screening tool to identify patients with malnutrition at a high risk of mortality among elderly patients with femoral fractures. A prospective study is needed to validate the suggestion.

Exosomes secreted by adipose-derived stem cells (ADSCs) enhance angiogenesis and wound healing. However, in clinical settings, wounds may be infected by various bacteria or pathogens. We investigated whether human ADSCs stimulated with lipopolysaccharide (LPS) secrete exosomes (ADSC-LPS-exo) that augment the angiogenesis of human umbilical vein endothelial cells (HUVECs). ExoQuick-TC exosome precipitation solution was used to purify exosomes from human ADSC culture media in the presence or absence of 1 µg/mL LPS treatment for 24 h. The uptake of ADSC-LPS-exo significantly induced the activation of cAMP response element binding protein (CREB), activating protein 1 (AP-1), and nuclear factor-κB (NF-κB) signaling pathways and increased the migration of and tube formation in HUVECs. RNA interference with CREB, AP-1, or NF-κB1 significantly reduced the migration of and tube formation in HUVECs treated with ADSC-LPS-exo. An experiment with an antibody array for 25 angiogenesis-related proteins revealed that only interleukin-8 expression was significantly upregulated in HUVECs treated with ADSC-LPS-exo. In addition, proteomic analysis revealed that eukaryotic translation initiation factor 4E, amyloid beta A4 protein, integrin beta-1, and ras-related C3 botulinum toxin substrate 1 may be potential candidates involved in ADSC-LPS-exo-mediated enhanced angiogenesis.