IMPORTANCE Spontaneous intracranial hypotension (SIH) is a highly disabling but often misdiagnosed disorder. The best management options for patients with SIH are still uncertain.OBJECTIVE To provide an objective summary of the available evidence on the clinical presentation, investigations findings, and treatment outcomes for SIH.DATA SOURCES Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) reporting guideline-compliant systematic review and meta-analysis of the literature on SIH. Three databases were searched from inception to April 30, 2020: PubMed/MEDLINE, Embase, and Cochrane. The following search terms were used in each database: spontaneous intracranial hypotension, low CSF syndrome, low CSF pressure syndrome, low CSF volume syndrome, intracranial hypotension, low CSF pressure, low CSF volume, CSF hypovolemia, CSF hypovolaemia, spontaneous spinal CSF leak, spinal CSF leak, and CSF leak syndrome. STUDY SELECTION Original studies in English language reporting 10 or more patients with SIH were selected by consensus.DATA EXTRACTION AND SYNTHESIS Data on clinical presentation, investigations findings, and treatment outcomes were collected and summarized by multiple observers. Random-effect meta-analyses were used to calculate pooled estimates of means and proportions.
MAIN OUTCOMES AND MEASURESThe predetermined main outcomes were the pooled estimate proportions of symptoms of SIH, imaging findings (brain and spinal imaging), and treatment outcomes (conservative, epidural blood patches, and surgical).
RESULTSOf 6878 articles, 144 met the selection criteria and reported on average 53 patients with SIH each (range, 10-568 patients). The most common symptoms were orthostatic headache (92% [95% CI, 87%-96%]), nausea (54% [95% CI, 46%-62%]), and neck pain/stiffness (43% [95% CI, 32%-53%]). Brain magnetic resonance imaging was the most sensitive investigation, with diffuse pachymeningeal enhancement identified in 73% (95% CI, 67%-80%) of patients. Brain magnetic resonance imaging findings were normal in 19% (95% CI, 13%-24%) of patients. Spinal neuroimaging identified extradural cerebrospinal fluid in 48% to 76% of patients. Digital subtraction myelography and magnetic resonance myelography with intrathecal gadolinium had high sensitivity in identifying the exact leak site. Lumbar puncture opening pressures were low, normal (60-200 mm H 2 O), and high in 67% (95% CI, 54%-80%), 32% (95% CI, 20%-44%), and 3% (95% CI, 1%-6%), respectively. Conservative treatment was effective in 28% (95% CI, 18%-37%) of patients and a single epidural blood patch was successful in 64% (95% CI, 56%-72%). Large epidural blood patches (>20 mL) had better success rates than small epidural blood patches (77% [95% CI, 63%-91%] and 66% [95% CI, 55%-77%], respectively).CONCLUSIONS AND RELEVANCE Spontaneous intracranial hypotension should not be excluded on the basis of a nonorthostatic headache, normal neuroimaging findings, or normal lumbar puncture opening pressure. Despite the heterogeneous nature of the studies...