通里攻下法治疗重症急性胰腺炎的meta分析-meta - analysis of treating severe acute pancreatitis with purgative therapy.docxVIP

成都中医药大学硕士学位论文 成都中医药大学硕士学位论文 PAGE PAGE 1 中文摘要 目的：通过 Meta 分析比较通里攻下法结合西医治疗（简称通里攻下法治疗）与 单纯西医治疗对于急性重症胰腺炎（severe acute pancreatitis，SAP）患者的治疗 效果。 方法：检索中国生物医学文献光盘数据库、中国期刊网全文专题数据库、维普数 据库，收集具有通里攻下作用的中药复方、中成药、单方治疗急性重症胰腺炎随 机对照试验。所有检索均截至 2008 年 12 月。根据纳入标准，摘录 8 篇随机对照 研究，对纳入研究进行文献质量评价，并用 RevMan 5.0 软件进行 Meta 分析。 结果：共检索到 218 篇文献，其中符合纳入标准的文献共有 8 篇，均为中文发表， 质量评价显示大部分研究的质量较低，1 篇文献 Jadad 积分为 3 分，7 篇文献 Jadad 积分为 2 分。Met a 分析结果显示：通里攻下法治疗与单纯西医治疗 SAP 效果有 显著差异：8 篇纳入研究中 8 篇报道了治疗结束时的病死率、5 篇报道了中转手 术率、6 篇报道了并发症发生率、5 篇报道了住院时间等指标。结果显示通里攻 下法治疗急性重症胰腺炎病死率（RR=0.43，95％CI 0.26～0.71）、中转手术人 数（RR=0.42，95％CI 0.27～0.65）及并发症发生率（RR=0. 47，95％CI 0.38～ 0.57）较单纯西医治疗组低，住院时间较单纯西医治疗组缩短（SMD= -1.69， 95%CI [-2.35，-1.03]，P0.00001）。 结论：通里攻下法治疗重症急性胰腺炎与单纯西医治疗相比，可以有助于降低病 死率、中转手术率、并发症率，缩短住院时间。但由于研究文献质量较低，仍需 要进行大规模、高质量、随访时间足够的随机对照研究为临床提供证据。 关键词:急性重症胰腺炎 通里攻下 系统评价 ABSTRACT Objective: To analyze and compare the effect of combination Purgation treatment and western treatment for patients with severe acute pancreatitis． Methods: Review Chinese Biomedical Literature Database CD-ROM, the full text of the topic CNKI database, VIP database, collect Chinese randomized controlled trials which treat severe acute pancreatitis with herbal compound medicine or unilateral Chinese Purgation herbs. As all searches end in 2008. According to the including criteria，eight randomized controlled trials of this topic were enrolled into the analysis．The detail about the trial design characters of the subjects， results of the studies were reviewed and analyzed by Revman5 software． Results: Retrieved a total of 218 researches, eight randomized controlled trials meting the inclusion criteria were selected and reviewed. Retrieved a total of 314 researches, fifteen randomized controlled trials meting the inclusion criteria were selected and reviewed. One researches’ Jadad score is 3,seven researches’ Jadad score is 2.Compared with western treatment，Purgation treatment was associated with a significantly lower incidence of m