The evidence was initially of high quality but was down- graded due to limitations in study design (mostly selection bias), indirectness (outcomes were rarely reported) and impreciseness (different magnitudes of predictive power of risk factors between studies). However, the strength of the recommendation was view as strong by the working group, considering the seriousness of the complication and health/economic impact of missing a patient with a neurosurgical lesion. The working group also discussed older age (≥60 years and ≥65 years) as well as antiplatelet
medication as risk factors of importance, partly due to the presence of these criteria in other guidelines and decision rules. However, the predictive ability was only moderate and these individual risk factors would lead to an unaccep- table CT increase and so consensus was not to include these in our recommendation.