Marshall and Rotterdam Computed Tomography scores in predicting early deaths after brain trauma

Abstract

Trauma is one of the most important issues of most healthcare systems accompanying with head trauma in the most cases. We sought to determine the scoring system and initial Computed Tomography (CT) findings predicting the death at hospital discharge (early death) in patients with traumatic brain injury based on Marshall and Rotterdam CT scores. This is a cross sectional study on traumatic neurosurgical patients with mild-to-severe traumatic brain injury admitted to the emergency department of Emam Reza Hospital, Birjand University of Medical Sciences. Patients≥18 years old with TBI during last 24 hours with GCS≤13 were included and exclusion criteria were multiple trauma, penetrating injuries, previous history of anticoagulant therapy, pregnancy, not willingness for participation. Their initial CT and status at hospital discharge, one and three months (dead or alive) were reviewed, and both CT scores were calculated. We examined whether each score is related to death using SPSS11 by The Mann–Whitney U at the level of p≤0.05. Overall, 98 patients were included. Mean age was 43.52±21.29. Most patients were male (63.3%). Mean Marshall and Rotterdam CT scores were 3.2±1.3 and 2.5±1. The mortality at two weeks, one moth and three months were 19.4%, 20.4%, and 20.4%. Rotterdam CT score was significantly different based on type of hematoma. Median GCS score in alive and dead patients on 2 weeks were 10 and 4 (p=0.0001), at one month were 10 and 4 (p=0.0001), and at three months were 10 and 4 (p=0.0001). The median Marshall CT score on 2 weeks were 2 and 4 (p=0.0001), at one month were 2 and 4 (p=0.0001), and at three months were 2 and 4 (p=0.0001). The median Rotterdam CT score on 2 weeks were 2 and 4 (p=0.0001), at one month were 2 and 3 (p=0.001), and at three months were 2 and 3 (p=0.001). The Rotterdam CT score was significantly correlated with mortality at two weeks, one month and three months (p=0.004, p=0.001, and p=0.001, respectively). The Marshall CT score was not significantly correlated with mortality at any time. The Rotterdam CT score was more accurate for prediction of mortality on 2 weeks (ROC80.9), at one month (ROC80.7), and at three months were (ROC80.7) than The Rotterdam CT score (ROC 76, 74.1, and 74.1, respectively). This study concluded that The Marshall CT score was more accurate for prediction of mortality on 2 weeks, at one month, and at three months were than The Marshall CT score with higher ROC. The correlation of the Rotterdam CT score with mortality was significant.

Predictive Value of Rotterdam Score and Marshall Score in Traumatic Brain Injury: A Contemporary Review

“…15 Sensitivity and specificity for mortality prediction at 2 weeks for Rotterdam and Marshall score were 56, 94.11%, and 87.34, 52.63%, respectively, and 57.69, 94.44%, and 87.34, 52.63%, at 1 month, respectively. 15 Sensitivity and specificity for mortality prediction at 2 weeks for Rotterdam and Marshall score were 57.69, 94.44%, and 87.17, 50%, respectively. 15 Receiver operating characteristic (ROC) area under the curve (AUC) showed higher accuracy in predicting mortality at 2 weeks, 1 month, and 3 months for Rotterdam score as compared with the Marshall score.…”

“…Mean age was 43.36 ± 21.65 years, mean GCS was 8.7 ± 3, and mean Marshall and Rotterdam CT score was 3.2 ± 1.3 and 2.5 ± 1.0, respectively, in this study. 15 They found a significant correlation between the Rotterdam score and mortality at 2 weeks, 1 month, and after 3 months; however, no such correlation was found with Marshall score. 15 Sensitivity and specificity for mortality prediction at 2 weeks for Rotterdam and Marshall score were 56, 94.11%, and 87.34, 52.63%, respectively, and 57.69, 94.44%, and 87.34, 52.63%, at 1 month, respectively.…”

“…15 They found a significant correlation between the Rotterdam score and mortality at 2 weeks, 1 month, and after 3 months; however, no such correlation was found with Marshall score. 15 Sensitivity and specificity for mortality prediction at 2 weeks for Rotterdam and Marshall score were 56, 94.11%, and 87.34, 52.63%, respectively, and 57.69, 94.44%, and 87.34, 52.63%, at 1 month, respectively. 15 Sensitivity and specificity for mortality prediction at 2 weeks for Rotterdam and Marshall score were 57.69, 94.44%, and 87.17, 50%, respectively.…”

Rationalising neurosurgical head injury referrals: development and validation of the Liverpool Head Injury Tomography Score (Liverpool HITS) for mild TBI

“…This is the first scoring system of its kind to the authors' knowledge. There are CT based scoring systems that accurately predict TBI mortality (such as Marshall and Rotterdam scores) [34][35][36] however none of these are specific to mild TBI, nor do they reference referral criteria for being accepted to a neurosurgical unit 37 . This means the potential impact on practice is high, as there is no current way of deciphering a surgically significant mild TBI at the level of national guidance.…”

Novel management of glioma by molecular therapies, a review article

“…3 Approximately 70% of cancers of brain and 20% of cancers of spinal cord are glial cell specific. 4 The factors that would be considered as hazard of production neuroglia tumorigenesis are include viral infection, ionizing radiation, contact with chemicals, genetic modification (histone, epidermal growth factor receptor (EGFR), cyclin dependent kinase inhibitor 2A (CDKN2A), tumor protein 53 (TP53), TACC, IDH, phosphatase and tensin homolog (PTEN), etc.). 5 The main characterizations of glioma are tumor recurrence, potential in proliferative growth, behavior of infiltrative growth and high degree of intratumor genetic heterogeneity.…”