BACKGROUND Studies on several preclinical models of type 2 diabetes mellitus have been conducted to establish the hypoglycemic activity of Momordica charantia L. Concerned with appropriateness of these models, we designed a systematic review to establish the efficacy and safety of M. charantia L. in preclinical models of type 2 diabetes mellitus. METHODS Review authors will search without language restriction in MEDLINE/PubMed, Web of Science, Embase, Scopus, and CINAHL databases through April 2019. Search filters will be applied to enhance search efficiency. The authors will search for gray literature in Google and Google Scholar, OpenGrey, and ProQuest Dissertations & Theses. Two authors will evaluate full texts, extract data, and asses risk of bias independently. The review will include randomized or non-randomized studies that assessed the efficacy or safety of M. charantia L. with vehicle control group. The primary endpoint will be fasting blood glucose level. We will use Egger's test to assess publication biases. Chi-square test and I2 will be used to assess heterogeneity in effect size of the primary outcome. Using RevMan software version 5.3, the authors will perform a meta-analysis of quantitative data. DISCUSSION The strength of evidence will be rated as high, moderate, low, or very low using GRADE framework for animal studies. This systematic review will potentially improve research practice by identifying risks of bias and design features that compromise translatability and contribute to evidence-based clinical trial design. SYSTEMATIC REVIEW REGISTRATION PROSPERO CRD42019119181.

中文翻译：

---

苦瓜对 2 型糖尿病动物模型的功效和安全性的系统评价和荟萃分析方案。


背景 对几种 2 型糖尿病的临床前模型进行了研究，以确定苦瓜的降血糖活性。考虑到这些模型的适当性，我们设计了一项系统评价，以确定苦瓜在临床前的有效性和安全性。 2型糖尿病模型。方法 综述作者将在 2019 年 4 月之前在 MEDLINE/PubMed、Web of Science、Embase、Scopus 和 CINAHL 数据库中不受语言限制地进行检索。将应用检索过滤器来提高检索效率。作者将在 Google 和 Google Scholar、OpenGrey 以及 ProQuest Dissertations &Theses 中搜索灰色文献。两位作者将独立评估全文、提取数据并评估偏倚风险。审查将包括评估苦瓜与载体对照组的功效或安全性的随机或非随机研究。主要终点是空腹血糖水平。我们将使用艾格检验来评估发表偏倚。卡方检验和 I2 将用于评估主要结果效应大小的异质性。作者将使用 RevMan 软件 5.3 版对定量数据进行荟萃分析。讨论 使用动物研究的 GRADE 框架将证据强度分为高、中、低或极低。这项系统评价将通过识别偏倚风险和损害可翻译性并有助于循证临床试验设计的设计特征来潜在地改善研究实践。系统审查注册 PROSPERO CRD42019119181。