Decompressive Craniectomy Evidence For Association With Post-Traumatic Hydrocephalus:A Meta-Analysis

Objectives There is increasing evidence from observational studies that Decompressive craniectomy(DC) and subdural hygroma(SDG) are associated with post-traumatic hydrocephalus(PTH).A number of these studies have been published with conflicting results.We conduct a meta-analysis based on case-control studies,to determine whether the DC and SDG are risk factors for the PTH.The purpose of the meta-analysis is twofold:(1) to estimate the association between DC and PTH;(2) to estimate the association between SDG and PTH.Methods(1) By combining Medical Subject Headings(Me SH) terms and keyword,We performed a literature search in following databases:Chinese Biomedical literature database(CBM), China National Knowledge Internet(CNKI), VIP database, wanfang database,Pub Med,EMBASE, MEDLINE, Web of Knowledge(WOK) and the Cochrane Database of Systematic Reviews.We also Use the services platform of academic searching and system of full text delivery to retrieve open data resources, patents, conference papers and letters.The published time was from 2000 to 2015.(2) Two commentators selected the literature and extracted the targeted data according to the inclusion and exclusion standard. The quality of study was evaluated on the basis of the NOS.The software of Comprehensive Meta- Analysis v3 was applied to tested the heterogeneity.It could use the fixed effect model to calculate the pooled OR when the homogeneity is moderate, and the factor of high heterogeneity should be dealt firstly.The subgroup analysis was variable on occasion.In the process of sensitivity analysis,it assessed the proportion of single study in the whole effect size by eliminating a single research successively.We carried out a cumulative meta-analysis according to time sequence.The publication bias was shown by the funnel figure and the fill-vacancy graph of funnel figure; We made a regression analysis to test the association between the heterogeneity and the quality of study by defining the scores of NOS as explanatory variables, and OR value as outcome variable.Outcomes There were two independent meta-analysis in the paper,one about association between DC and PTH, the other about association between SDG and PTH.(1)The Meta-analysis of the correlation between DC and PTH contained 18 published papers from 2006 to 2006, and 3 studies were English-language publications,other 15 studies Chinese-language publications. There were 4665 cases of the research in all,including exposed group and unexposed group.The areas of the research covered China, South Korea and Australia?(1)The exposed factor included unilateral and bilateral DC.In the Meta-analysis, the result of heterogeneity test,I2, is 74.955%.It used the random effects model to show the pooled meta-analysis results(OR 5.531;95%CI 3.889,7.868;p<0.05).Results of publication bias indicated bias was obvious.(2)We carried out a subgroup analysis by defined unilateral and bilateral DC as grouping variable.The pooled meta-analysis results of the unilateral DC group(OR 3.503;95%CI 2.498,4.912;p<0.05),bilateral DC group(OR 21.481;95%CI 10.599,43.537;p<0.05).(3)Unilateral DC was set as independent exposed factor.The result of heterogeneity test,I2, was 45.155%.By using the fixed effects model combined effect quantity(OR 3.809;95%CI 3.074,4.721;p<0.05).In the process of sensitivity analysis,the OR value was around 3.8,and 95% CI coincided evidently.OR value of the Cumulative meta-analysis fluctuated up to 4.0 and down to 3.8.There did not exist publication bias in the test.(4) Bilateral DC was set as independent exposed factor.There was no heterogeneity in the test.The pooled results of the bilateral DC(OR 21.481;95%CI 10.599,43.537;p<0.05);Sensitivity analysis and the cumulative meta-analysis indicated that pooled effect quantity is relatively steady.No publication bias existed there.Regression analysis showed no linear relation.(2)The Meta-analysis of the correlation between SDG and PTH contained 8 published papers from 2012 to 2015,and 5 studies were Chinese-language publications,other 3 studies English-language publications.The areas of the research also covered China, South Korea and Australia.(1)The exposed factors included the position of SDG in unilateral and interhemispheric subdural.In the Meta-analysis, the result of heterogeneity test,I2, was 38.894%, so it adopted the fixed effects model.The pooled meta-analysis results(OR 3.267;95%CI 2.325,4.589;p<0.05).(2)Unilateral SDG as independent exposed factor in the Meta-analysis,the result of heterogeneity test,I2, was 47.583%,and pooled effect size(OR 3.259;95%CI 2.312,4.594;p<0.05).Sensitivity analysis and cumulative meta-analysis were relatively stable.There was no publication bias.(3) Interhemispheric SDG as independent exposed factor.No heterogeneity existed.The pooled results of the fixed effects model(OR 14.895; 95%CI 6.296, 35.239; p<0.05).The differences existed in sensitivity analysis.The cumulative meta-analysis showed the tendency of pooled effect size from 2014 to 2015 was moderate.Publication bias was not perceived.According to the Regression analysis, we did not confirm the association between the quality evaluation of studies and heterogeneity.Conclusion Our meta-analysis of case-control studies finds a significant association between these two exposed factors(DC and SDG) and PTH. Especially,bilateral DC and interhemispheric SDG increased the risk of PTH observably and independently. Before executing the DC,it should give mature deliberation.If the SDG appears after TBI,active intervention is essential.