Spectrophotometry or visual inspection to most reliably detect xanthochromia in subarachnoid hemorrhage: systematic
Kevin Chu1, Angus Hann2, Jaimi Greenslade1
1School of Medicine, University of Queensland, Brisbane, Queensland, Australia; Department of Emergency Medicine, Royal Brisbane and Women's Hospital, Brisbane, Queensland, Australia.
Study Objective:
We assess the sensitivity and specificity of xanthochromia as adjudicated by visual inspection and spectrophotometry at predicting the presence of cerebral aneurysm in patients with suspected subarachnoid hemorrhage who have a normal computed tomography (CT) head scan result.
Methods:
A systematic review was performed. MEDLINE and EMBASE databases were searched. Relevant studies with clinical data on the diagnostic accuracy of visual inspection or spectrophotometry were considered. Patients who had a normal CT head scan result followed by a lumbar puncture were included in this review. Sensitivities, specificities, and heterogeneity (I(2)) were calculated. Subgroup analyses were performed to explore reasons for the heterogeneity.
Results:
There were major methodological limitations in the studies found. Twenty-two relevant articles were heterogeneous in regard to time to lumbar puncture, spectrophotometry methods, and follow-up of patients not undergoing cerebral angiography. Twelve of the 22 studies selected patients on the basis of a cerebral aneurysm or subarachnoid hemorrhage on imaging, or a positive lumbar puncture result. These studies were excluded from our initial analysis, which included only patients with clinically suspected subarachnoid hemorrhage. In this initial analysis, pooled estimates of sensitivity and specificity for spectrophotometry were 87% (95% confidence interval [CI] 71% to 96%; I(2)=26%) and 86% (95% CI 84% to 88%; I(2)=96%), respectively. For visual inspection, pooled sensitivity and specificity were 83% (95% CI 59% to 96%; I(2)=52%) and 96% (95% CI 93% to 97%; I(2)=76%), respectively. Sensitivity estimates are difficult to interpret without knowing time to lumbar puncture.
Conclusion:
The heterogeneity in the underlying studies, combined with significant overlap in pooled confidence limits, makes it impossible to provide a definite conclusion about the diagnostic accuracy of spectrophotometry versus visual inspection.
