It’s been a lazy summer on our part, so it’s time to get back to business. And we’ll start off with this article by Moustafa et al, which was recently published in the Scandinavian Journal of Trauma, Resuscitation and Emergency Medicine. The authors deal with the issue of predicting risk of intracranial haemorrhage in TBI patients taking anti-platelet agents. As is pointed out in the article, 90% of CT-scans carried out in this patient group turn out negative, with the radiation burden and cost that this entails.
In a registry-study they reviewed the records of almost 1000 patients on anti-platelet drugs who were admitted with a history of head trauma. Among those subjected to a CT-scan around 9% were found to have an intracranial bleeding. These patients were more likely to have a history of severe bleeding, to be on dual anti-platelet therapy and to present with headache or vomiting and with an abnormal GCS. The authors concluded that the absence of any of these factors predicts a normal CT-scan with 98.9% specificity and a negative predictive value of 92%.
This is an interesting finding, which may be a start in the development of a more robust prediction model for a rather large, and potentially at-risk, group of patients. Such a model would be useful for informing future guidelines, so as to both limit unnecessary use of radiation and make sure we don’t miss any significant injury.