Sung-Eun Kim1, Bong Jun Kim2, Steve S Cho3, Heung Cheol Kim4, Jin Pyeong Jeon5. 1. Department of Emergency Medicine, Seoul Emergency Operations Center, Seoul, Korea. 2. Institute of New Frontier Research, Hallym University College of Medicine, Chuncheon, Korea. 3. Perelman School of Medicine at the University of Pennsylvania, Philadelphia, Pennsylvania, USA. 4. Department of Radiology, Hallym University College of Medicine, Chuncheon, Korea. 5. Institute of New Frontier Research, Hallym University College of Medicine, Chuncheon, Korea; Department of Neurosurgery, Hallym University College of Medicine, Chuncheon, Korea. Electronic address: jjs6553@daum.net.
Abstract
OBJECTIVE: To evaluate the incidence of hydrocephalus and implanted shunts in angiogram-negative subarachnoid hemorrhage (SAH) according to hemorrhage patterns: perimesencephalic SAH (PMH) versus non-PMH. METHODS: The online database literature from January 1990 to November 2017 was systematically reviewed. A fixed-effect model was used when heterogeneity was <50%. A Begg funnel plot was used to assess publication bias. An additional trim and fill method was used to estimate the number of missing studies. A subgroup analysis with studies, which defined angiogram-negative SAH using repeated angiography or computed tomography angiography, was further performed. RESULTS: A total of 28 articles including 2577 patients were enrolled. Patients with PMH showed a significantly decreased incidence of hydrocephalus (odds ratio [OR], 0.269; 95% confidence interval [CI], 0.208-0.348) and shunts (OR, 0.263; 95% CI, 0.169-0.411) than did patients without PMH. A subgroup analysis of 7 studies with 675 patients showed less hydrocephalus in patients with PMH than in patients without PMH (OR, 0.358; 95% CI, 0.161-0.793), with possible publication bias. Shunt procedures were marginally less common in patients with PMH compared with those patients without PMH (OR, 0.490; 95% CI, 0.236-1.018) with possible publication bias. After correction of the forest plot, the adjusted OR was 0.617 (95% CI, 0.251-1.513) for hydrocephalus and 0.618 (95% CI, 0.310-1.232) for shunts, suggesting no significant relationships between PMH and the risk of hydrocephalus or shunting. CONCLUSIONS: Hydrocephalus and shunts were more evident in non-PMH than PMH. However, subgroup analyses did not show significant associations between PMH and lower risks of these events after correction for possible publication bias. Further meta-analyses based on individual patient data are necessary.
OBJECTIVE: To evaluate the incidence of hydrocephalus and implanted shunts in angiogram-negative subarachnoid hemorrhage (SAH) according to hemorrhage patterns: perimesencephalic SAH (PMH) versus non-PMH. METHODS: The online database literature from January 1990 to November 2017 was systematically reviewed. A fixed-effect model was used when heterogeneity was <50%. A Begg funnel plot was used to assess publication bias. An additional trim and fill method was used to estimate the number of missing studies. A subgroup analysis with studies, which defined angiogram-negative SAH using repeated angiography or computed tomography angiography, was further performed. RESULTS: A total of 28 articles including 2577 patients were enrolled. Patients with PMH showed a significantly decreased incidence of hydrocephalus (odds ratio [OR], 0.269; 95% confidence interval [CI], 0.208-0.348) and shunts (OR, 0.263; 95% CI, 0.169-0.411) than did patients without PMH. A subgroup analysis of 7 studies with 675 patients showed less hydrocephalus in patients with PMH than in patients without PMH (OR, 0.358; 95% CI, 0.161-0.793), with possible publication bias. Shunt procedures were marginally less common in patients with PMH compared with those patients without PMH (OR, 0.490; 95% CI, 0.236-1.018) with possible publication bias. After correction of the forest plot, the adjusted OR was 0.617 (95% CI, 0.251-1.513) for hydrocephalus and 0.618 (95% CI, 0.310-1.232) for shunts, suggesting no significant relationships between PMH and the risk of hydrocephalus or shunting. CONCLUSIONS:Hydrocephalus and shunts were more evident in non-PMH than PMH. However, subgroup analyses did not show significant associations between PMH and lower risks of these events after correction for possible publication bias. Further meta-analyses based on individual patient data are necessary.