动物模型肝体积CT、MR测量准确性的荟萃分析

发布时间：2018-08-25 12:12

【摘要】：目的采用循证医学方法,对动物模型研究进行荟萃(Meta)分析,评价肝体积影像学测量的准确性,探讨影响肝体积测量准确性的因素。方法检索Cochrane图书馆、文摘数据库(英文:Medline、Embase,中文:中国生物医学文献数据库)和全文数据库(英文:OVID、Springer、Sciencedirect等,中文:CNKI、万方、维普等),获得动物模型肝体积影像学研究的中英文文献,文献筛选采用Cochrane协作网推荐的诊断试验纳入标准,对纳入文献进行质量评价、信息提取和异质性检验,根据异质性分析结果选择效应量合并模型,并根据实验动物肝体积大小、成像设备、肝状态、测量方法和有无对比剂增强进行亚组分析,合并效应量为加权均数差值,最后进行敏感性分析和发表性偏倚分析。结果符合纳入标准的14篇文献报道了27项研究,包括189只实验动物,敏感性分析去除对汇总结果影响较大的1篇文献报道的4项研究(包括15只实验动物),最终13篇文献报道的23项研究进行了Meta分析,纳入研究之间无异质性(I2=0%,Q=17.65,P=0.7265),根据固定效应模型合并效应量,动物模型影像学测量肝体积和实际肝体积的加权均数差值为-0.0453(95%可信区间:-0.198~0.1074),两者之间无显著性差别(Z=-0.5810,P=0.5612),根据不同实验动物肝体积大小、成像设备、肝状态、测量方法和有无对比剂增强进行分层后,各亚组间汇总结果无显著性差异,敏感性分析显示最终汇总结果稳定,漏斗图及发表性偏倚检验提示无发表性偏倚(t=-0.6843,P=0.5013)。结论肝体积的影像学测量结果准确可靠,不受肝体积大小、成像设备、肝状态、测量方法和有无对比剂增强的影响,能为临床提供重要参考信息。
[Abstract]:Objective to evaluate the accuracy of liver volume imaging measurement and explore the factors influencing the accuracy of liver volume measurement by meta-( Meta) analysis on animal model by evidence-based medicine. Methods Cochrane Library, Abstract Database (Chinese: Chinese Biomedical Literature Database) and Full-text Database (Cochrane Library) were searched, and the Chinese and English literatures on liver volume imaging of animal models were obtained. The quality evaluation, information extraction and heterogeneity test were carried out. According to the results of heterogeneity analysis, the combined model of effect quantity was selected, and the liver volume of experimental animals was selected according to the size of the liver. Imaging equipment, liver state, measurement method and contrast enhancement were used for subgroup analysis. The combined effect was weighted mean difference. Finally, sensitivity analysis and published bias analysis were performed. Results Twenty-seven studies, including 189 experimental animals, were reported in 14 articles that met the inclusion criteria. Sensitivity analysis removed 4 studies (including 15 experimental animals) that had a significant impact on the summary results, and Meta analysis was performed on 23 of the 13 papers. There was no heterogeneity among the inclusion studies (I _ 2Q _ (17. 65) P ~ (0.7265), combining the effects according to the fixed effect model, The weighted mean difference of liver volume and actual liver volume in animal model was -0.0453 (95% confidence interval: -0.198 / 0.1074). There was no significant difference between the two groups (Z _ (-0.5810) P _ (0.5612). After stratification with or without contrast enhancement, there was no significant difference among the subgroups. Sensitivity analysis showed that the final aggregate results were stable, and funnel map and published bias test showed no published bias (t _ (-0.6843) P _ (0.5013). Conclusion the results of imaging measurement of liver volume are accurate and reliable. It is not affected by the size of the liver, the imaging equipment, the state of the liver, the method of measurement and the enhancement of contrast agent, which can provide important reference information for clinical practice.
【作者单位】：天津市第一中心医院放射科;日照市人民医院;
【基金】：天津市卫生局科技基金面上项目(编号:2012KY07)
【分类号】：R816.5;R445.2

【参考文献】