凭什么人家可以将网状meta分析发在柳叶刀上？ / 开普饭

今天，给大家介绍一篇文章，看看人家是怎样能发在柳叶刀上的？

题目

Comparativeefficacy and tolerability of antidepressants for major depressive disorder inchildren and adolescents: a network meta-analysis

摘要

Background： Major depressive disorder is one of the most common mental disorders inchildren and adolescents. However, whether to use pharmacological interventionsin this population and which drug should be preferred are still matters ofcontroversy. Consequently, we aimed to compare and rank antidepressants andplacebo for major depressive disorder in young people.

背景：在儿童和青少年人群中，重度抑郁症是最常见的精神疾病之一。然而，在这一人群中是否应该使用药物干预措施和那一种药物更加适合治疗仍然存在争议。因此，我们的目的是比较并且排列抗抑郁药与安慰剂治疗青年人重度抑郁症的疗效。

Methods：We did a network meta-analysis to identify both direct and indirect evidence fromrelevant trials. We searched PubMed, the Cochrane Library, Web of Science,Embase, CINAHL, PsycINFO, LiLACS, regulatory agencies’ websites, andinternational registers for published and unpublished, double-blind randomizedcontrolled trials up to May 31, 2015, for the acute treatment of majordepressive disorder in children and adolescents. We included trials of amitriptyline,citalopram, clomipramine, desipramine, duloxetine, escitalopram, fluoxetine,imipramine, mirtazapine, nefazodone, nortriptyline, paroxetine, sertraline, andvenlafaxine. Trials recruiting participants with treatment resistant depression,treatment duration of less than 4 weeks, or an overall sample size of less thanten patients were excluded. We extracted the relevant information from the publishedreports with a predefined data extraction sheet, and assessed the risk of biaswith the Cochrane risk of bias tool. The primary outcomes were efficacy (changein depressive symptoms) and tolerability (discontinuations due to adverseevents). We did pair-wise meta-analyses using the random-effects model and thendid a random-effects network meta-analysis within a Bayesian framework. We assessedthe quality of evidence contributing to each network estimate using the GRADEframework. This study is registered with PROSPERO, number CRD42015016023.

方法：为了确定来自相关试验的直接证据和间接证据，我们做了一项网状meta分析。我们检索了PubMed, theCochrane Library, Web of Science, Embase, CINAHL, PsycINFO, LiLACS, regulatoryagencies’ websites以及国际性注册的发表或者没有发表关于紧急治疗儿童和青少年重度抑郁症的双盲随机对照试验，截止时间为2015年5月31号。我们纳入了amitriptyline,citalopram, clomipramine, desipramine, duloxetine, escitalopram, fluoxetine,imipramine, mirtazapine, nefazodone, nortriptyline, paroxetine, sertraline,venlafaxine的试验。试验招募具有难治性抑郁症的参与者，治疗持续时间少于4周，或者排除总的样本量大小少于10名患者的试验。我们用一个预先确定的数据提取电子表格从已经发表的报告提取相关的信息，并且用the Cochranerisk of bias tool评价偏倚风险。主要的结局指标是疗效（抑郁症状的变化）和耐药性（由于不良事件中止治疗）。我们用随机效应模型做直接比较的meta分析，并且做一个基于贝叶斯框架的随机效应模型的网状meta分析。我们使用GRADE框架评价每个网络贡献的证据的质量。本研究在PROSPERO进行注册，number CRD42015016023。

Findings：We deemed 34 trials eligible, including 5260 participants and 14 antidepressanttreatments. The quality of evidence was rated as very low in most comparisons.For efficacy, only fluoxetine was statistically significantly more effectivethan placebo (standardized mean difference –0·51, 95% credible interval [CrI]–0·99 to –0·03). In terms of tolerability, fluoxetine was also better thanduloxetine (odds ratio [OR] 0·31, 95% CrI 0·13 to 0·95) and imipramine (0·23,0·04 to 0·78). Patients given imipramine, venlafaxine, and duloxetine had morediscontinuations due to adverse events than did those given placebo (5·49, 1·96to 20·86; 3·19, 1·01 to 18·70; and 2·80, 1·20 to 9·42, respectively). In termsof heterogeneity, the global I² values were 33·21% for efficacy and 0% fortolerability.

结果: 我们确定34项合格的试验，包含5260名参与者和14种抗抑郁药物。在大多数比较重中，其证据的质量被评为很低。在疗效这个指标，只有fluoxetine在统计学上显著比安慰剂有效 (standardizedmean difference –0.51, 95% credible interval [CrI] –0.99 to –0.03)。在耐药性方面，fluoxetine 也比duloxetine (odds ratio[OR] 0.31, 95% CrI 0.13 to 0.95) and imipramine (0.23, 0.04 to 0.78)好。与给予安慰剂相比，患者给予imipramine,venlafaxine, duloxetine更加容易出现由不良事件导致的治疗中止(5.49, 1.96 to 20.86; 3.19, 1.01 to 18.70; 2.80, 1.20 to 9.42)

Interpretation： When considering therisk–benefit profile of antidepressants in the acute treatment of majordepressive disorder, these drugs do not seem to offer a clear advantage forchildren and adolescents. Fluoxetine is probably the best option to considerwhen a pharmacological treatment is indicated.

结果：在考虑抗抑郁药在紧急治疗重度抑郁症的风险效益概况时，这些药物似乎没有为儿童和青少年提供一个清楚的优势。在考虑药物治疗时，fluoxetine可能是最佳的治疗方案。

凭什么发在柳叶刀上？

第一：创新强，一下子比较了十四种药，解决了医生和患者的面对多种药物选择的痛苦，可以说是前无古人，但是后无来者就不敢说了。

第二：样本量非常大，纳入了34项研究，5260名参与者，得到了一般人拿不到的数据

第三：方法学上采用贝叶斯统计模型，要是用频率学的模型肯定发不了柳叶刀，并且进行了meta分析的注册，体现了做meta分析的规范性。

第四：文章语言不存在中国式英语，因为文章的第一作者是老外，共同作者是中国人，语言是发现好文章的门槛。即使你的研究内容非常好，语言逻辑不行，表达不好，中国式英语成分多，绝对发不了高分文章。