• Magnus Olivecrona and Lars-Owe D Koskinen.
• Department of Pharmacology and Clinical Neuroscience, Neurosurgery, Umeå University, Umeå, Sweden. magnus.olivecrona@neuro.umu.se
• Acta Neurochir (Wien). 2012 Sep 1;154(9):1567-73.

BackgroundThe prognosis of severe traumatic brain injury (sTBI) is important. The International Mission on Prognosis in Traumatic Brain Injury (IMPACT) study group has developed a prediction calculator for the outcome of patients with sTBI, and this has been made available on the World Wide Web. We have studied the use of the IMPACT calculator on sTBI patients treated with an ICP-targeted therapy based on the Lund concept.MethodThe individual clinical data of patients in a prospective sTBI protocol-driven trial of the treatment of sTBI using the Lund concept were entered into the prognosis calculator, and the individual prognosis for each patient was calculated and compared with the actual outcome at 6 months.FindingsThe use of the IMPACT calculator led to an overestimation of mortality and of an unfavourable outcome. Compared with the IMPACT database, the absolute risk reduction (ARR) for mortality was 13.6 %. There is a statistically significant probability for the prediction of mortality and unfavourable outcome. A ROC curve analysis shows an area under the curve (AUC) in the Core model for mortality of 0.744 and of unfavourable outcome of 0.731, in the Extended model of 0.751 and 0.721 respectively, and in the Lab model of 0.779 and 0.810 respectively.ConclusionsThe IMPACT prognosis calculator should be used with caution for the prediction of outcome for an individual patient with sTBI treated with an ICP-targeted therapy based on the Lund concept. We conclude that we have to initiate treatment in all patients with blunt sTBI and an initial cerebral perfusion pressure (CPP)≥10 mmHg [corrected]. It seems that the outcome in sTBI patients treated in this fashion is better than would have been expected from the IMPACT prognosis.

48 keywords...

hide…