Determining the Diagnostic Utility of Lumbar Punctures in Computed Tomography Negative Suspected Subarachnoid Hemorrhage: A Systematic Review and Meta-analysis

BACKGROUND: While headache is a common neurologic symptom, subarachnoid hemorrhage (SAH) is a rare and potentially catastrophic cause of sudden-onset severe headache. The utility of the imaging modalities and interventional procedures are central to the investigation of the causes of headache; however, they are not without their limitations, risks, and complications. METHODS: A meta-analysis in accordance with the Preferred Reporting for Systematic Reviews and Meta-analysis guidelines was conducted searching PubMed, EMBASE, and Google Scholar. Patients investigated for suspected subarachnoid hemorrhage (SAH) with a negative computed tomography (CT) and positive lumbar puncture (LP) and final diagnosis of SAH were included. The sensitivity of LP in the context of a negative CT and vsubsequent imaging confirming the cause of SAH (computed tomography angiography, magnetic resonance angiography, digital subtraction angiography [DSA]) was quantified. The pooled data were analyzed using the DerSimonian-Laid random effects model. RESULTS: Four studies with 2782 patients who presented with headache suspicious for SAH were included with an initial negative CT report and a subsequent LP to rule out SAH. All included studies had an observational prospective cohort design. A combined pooled proportion of 0.383 (0.077, 0.756); 0.086 (0.007, 0.238); and 0.22 (0.04, 0.49) for LP+, DSA+, and DSA/computed tomography angiography+ investigations were estimated with a 95% confidence interval. CONCLUSIONS: The current clinical workflow of an LP after a negative CT head for a patient presenting with a sudden-onset severe headache is observed to have a high enough proportion to warrant its continued use despite the sensitivity of modern CT scanners of ≥97%.