ABSTRACT
Cerebral vasospasm is a life-threatening complication following aneurysmal subarachnoid hemorrhage (aSAH). While digital subtraction angiography (DSA) is the current gold standard for detection, the diagnostic performance of computed tomography angiography (CTA) and transcranial Doppler (TCD) remains controversial. We aimed to summarize the available evidence and provide recommendations for their use based on GRADE criteria. A literature search was conducted for studies comparing CTA or TCD to DSA for adults ≥ 18 years with aSAH for radiographic vasospasm detection. The DerSimonian-Laird random-effects model was used to pool sensitivity and specificity and their 95% confidence intervals (CI) and derive positive and negative pooled likelihood ratios (LR + /LR -). Out of 2070 studies, seven studies (1646 arterial segments) met inclusion criteria and were meta-analyzed. Compared to the gold standard (DSA), CTA had a pooled sensitivity of 82% (95%CI, 68-91%) and a specificity of 97% (95%CI, 93-98%), while TCD had lower sensitivity 38% (95%CI, 19-62%) and specificity of 91% (95%CI, 87-94%). Only the LR + for CTA (27.3) reached clinical significance to rule in diagnosis. LR - for CTA (0.19) and TCD (0.68) approached clinical significance (< 0.1) to rule out diagnosis. CTA showed higher LR + and lower LR - than TCD for diagnosing radiographic vasospasm, thereby achieving a strong recommendation for its use in ruling in or out vasospasm, based on the high quality of evidence. TCDs had very low LR + and a reasonably low LR - , thereby achieving a weak recommendation against its use in ruling in vasospasm and weak recommendation for its use in ruling out vasospasm.
