Traumatic brain injury (TBI) is responsible for up to 45% of in-hospital trauma mortality. Computed tomography (CT) is central to acute TBI diagnostics, and millions of brain CT scans are conducted yearly worldwide. Though many studies have addressed individual predictors of outcome from findings on CT scans, few have done so from a multivariate perspective. As these parameters are interrelated in a complex manner, there is a need for a better understanding of them in this context. CT scans from 861 TBI patients were reviewed according to an extensive protocol. An extended analysis of CT parameters with respect to outcome was performed using linear and non-linear methods. We identified complex interactions and mutual information in many of the parameters. Variables predicting death differ from those predicting unfavorable versus favorable outcomes (Glasgow Outcome Scale scores of 1-3 versus 4-5 [GOS]). The most important parameter for prediction of unfavorable outcome is the magnitude of midline shift. In fact, this parameter, as a continuous variable, is by itself a better predictor and is better calibrated than the Marshall CT score, even for predicting death. In addition, hematoma volumes are nearly co-linear with midline shift and can be substituted for it. A score of traumatic subarachnoid/intraventricular blood components adds substantially to model calibration. A CT scoring system geared toward dichotomous GOS scores is suggested. CT parameters were found to add 6-10% additional estimated explained variance in the presence of the important clinical variables of age, Glasgow Coma Scale score, and pupillary response. Finally we present a practical clinical "rule of thumb" to help predict the probability of unfavorable outcome using clinical and CT variables.

No abstract

High prevalence of thrombotic events in severely ill COVID-19 patients have been reported. Pulmonary embolism as well as microembolization of vital organs may in these individuals be direct causes of death. The identification of patients at high risk of developing thrombosis may lead to targeted, more effective prophylactic treatment. The primary aim of this study was to test whether rotational thromboelastometry (ROTEM) at admission indicates hypercoagulopathy and predicts the disease severity, assessed as care level, in COVID-19 patients. The study was designed as a prospective, observational study where COVID-19 patients over 18 years admitted to hospital were eligible for inclusion. Patients were divided into two groups depending on care level: (1) regular wards or (2) wards with specialized ventilation support. Conventional coagulation tests, blood type and ROTEM were taken at admission. 60 patients were included; age 61 (median), 67% men, many with comorbidities (e.g. hypertension, diabetes). The ROTEM variables Maximum Clot Firmness (EXTEM-/FIBTEM-MCF) were higher in COVID-19 patients compared with in healthy controls (p < 0.001) and higher in severely ill patients compared with in patients at regular wards (p < 0.05). Our results suggest that hypercoagulopathy is present early in patients with mild to moderate disease, and more pronounced in severe COVID-19 pneumonia. Non-O blood types were not overrepresented in COVID-19 positive patients. ROTEM variables showed hypercoagulopathy at admission and this pattern was more pronounced in patients with increased disease severity. If this feature is to be used to predict the risk of thromboembolic complications further studies are warranted.

Pretreatment with levosimendan in pigs subjected to endotoxin shock improved cardiac output and systemic and gut oxygen delivery. In addition, pulmonary hypertension largely was attenuated without any adverse effects on gas exchange. These results are promising in several aspects, but the role of levosimendan in the treating circulatory failure in sepsis remains to be established.