Data analysis
Although extracted data regarding the first clinical question, predictors of CT findings, ICI and neurosurgery, could theoretically be summarized in a meta-analysis, the task force decided a priorinot to perform such an analysis for the purpose of development of the guidelines, independent of heterogeneity between studies. We felt combining the data in this way could mislead the working group in the consensus process and opted to instead present uncombined data for studies including their quality assessment. We therefore calculated individual positive likelihood ratios (PLR) and negative likelihood ratios (NLR) for each risk factor with respect to the corresponding reference test