Comparative accuracy of CT perfusion in diagnosing acute ischemic stroke: A systematic review of 27 trials

To systematically evaluate and compare the diagnostic accuracy of CT perfusion (CTP), non-enhanced computed tomography (NCCT) and computed tomography angiography (CTA) in detecting acute ischemic stroke.

We searched seven databases and screened the reference lists of the included studies. The risk of bias in the study quality was assessed using QUADASII. We produced paired forest plots in RevMan to show the variation of the sensitivity and specificity estimates together with their 95% CI. We used a hierarchical summary ROC model to summarize the sensitivity and specificity of CTP in detecting ischemic stroke.

We identified 27 studies with a total of 2168 patients. The pooled sensitivity of CTP for acute ischemic stroke was 82% (95% CI 75-88%), and the specificity was 96% (95% CI 89-99%). CTP was more sensitive than NCCT and had a similar accuracy with CTA. There were no statistically significant differences in the sensitivity and specificity between patients who underwent CTP within 6 hours of symptom onset and beyond 6 hours after symptom onset. No adverse events were reported in the included studies.

CTP is more accurate than NCCT and has similar accuracy to CTA in detecting acute ischemic stroke. However, the evidence is not strong. There is potential benefit of using CTP to select stroke patients for treatment, but more high-quality evidence is needed to confirm this result.