RESUMO
PURPOSE: To describe a pooled estimated incidence of cerebral arterial vasospasm (aVSP) following aneurysmal subarachnoid haemorrhage (aSAH) and to describe sources of variation in the reported incidence. METHODS: We performed a systematic review and meta-analysis of randomised clinical trials (RCTs) and cohort studies. The primary outcome was the proportion of study participants diagnosed with aVSP. We assessed for heterogeneity based on mode of imaging, indication for imaging, study design and clinical characteristics at a study level. RESULTS: We identified 120 studies, including 19,171 participants. More than 40 different criteria were used to diagnose aVSP. The pooled estimate of the proportion of patients diagnosed with aVSP was 0.42 (95% CI 0.39 to 0.46, I2 = 96.5%). There was no evidence that the incidence aVSP was different, nor that heterogeneity was reduced, when the estimate was assessed by study type, imaging modalities, the proportion of participants with high grade CT scores or poor grade clinical scores. The pooled estimate of the proportion of study participants diagnosed with aVSP was higher in studies with routine imaging (0.47, 95% CI 0.43 to 0.52, I2 = 96.5%) compared to those when imaging was performed when indicated (0.30, 95% CI 0.25 to 0.36, I2 = 94.0%, p for between-group difference < 0.0005). CONCLUSION: The incidence of cerebral arterial vasospasm following aSAH varies widely from 9 to 93% of study participants. Heterogeneity in the reported incidence may be due to variation in the criteria used to diagnose aVSP. A standard set of diagnostic criteria is necessary to resolve the role that aVSP plays in delayed neurological deterioration following aSAH. PROSPERO REGISTRATION: CRD42020191895.