The Marshall and Rotterdam scoring systems can be used to predict early death in patients with traumatic brain injury (TBI), according to a study published in the May issue of Academic Radiology.
Because outcome prediction at patient admission for TBIs is vital for clinical decision making, resource allocation and family counseling, two scoring systems have been developed to improve patient assessment: the Marshall and Rotterdam scores. Though previous studies have evaluated the performance of these systems, they have not been assessed for prediction of death at hospital discharge.
Lead author Daddy Mata-Mbemba, MD, PhD, of the Tohoku University Graduate School of Medicine in Sendai, Japan, and colleagues compared the scoring systems’ performances in predicting early death and also determined specific CT findings that independently predict early death in patients with TBI.
Comprised of 245 adult patients with mild-to-severe TBI, the study reviewed the patients’ initial CT, status at hospital discharge and both CT scores. At the time of discharge, 220 of the 245 patients were alive and 25 were dead. They examined scores for basal cistern status, presence of midline shift, presence of epidural hematoma (EHD), presence of intraventricular hemorrhage and/or subarachnoid hemorrhage (IVH/SAH) and volume of hemorrhagic mass. The researchers determined if each score was related to early death and compared the two systems’ abilities to predict death at hospital discharge.
More deaths were reported among patients with higher Marshall and Rotterdam scores. The areas under the receiver operating characteristic curve (AUCs) revealed that both scoring systems had similar discriminative power in predicting early death, with an AUC of 0.85 for Marshall and 0.85 for Rotterdam.
Mata-Mbema identified basal cistern absence, positive midline shift, hemorrhagic mass volume greater than or equal to 25 milliliters, and intraventricular or subarachnoid hemorrhage were independent predictors of early death.
“The performance of Marshall scoring system is equal or possibly slightly better than that of Rotterdam scoring system. This statement is supported by our specific results showing that the two strongest predictors of early death—basal cistern absence and positive midline shift—are included in both CT scoring systems, but voluminous hemorrhagic mass, the next strongest predictor, is included only in Marshall score,” wrote the authors. “Thus, although older than Rotterdam score, Marshall score remains useful for the purpose of outcome prediction after TBI.”