In the era of widespread endoscopy use, randomized trials may,strongly underestimate the effects of colorectal cancer screening
In the era of widespread endoscopy use, randomized trials may,strongly underestimate the effects of colorectal cancer screening
复制标题
DOI：
10.1016/j.jclinepi.2013.05.008
复制
发表时间：
2013-10-01
影响因子：
7.2
通讯作者：
Hoffmeister, Michael
中科院分区：
文献类型：
--
作者：
Brenner, Hermann;Stock, Christian;Hoffmeister, Michael
目的：尽管随机对照试验（RCT）被认为是评估筛查效果的最有效方法，但它们也面临着许多方法学挑战，包括不依从性和污染。我们的目的是量化它们在评价内镜筛查结直肠癌（CRC）的随机对照试验中的潜在影响。研究设计和设置：我们使用合理的非依从性和污染水平进行模型计算。假设依从性（70%）和污染（即使用下消化道内镜检查而不是筛选（30%））的中等值，在意向筛查分析中，预期CRC发病率或死亡率的70%、50%或30%的真实降低将分别减弱至43%、29%和16%的降低。在低水平的依从性（50%）和高但现实的污染水平（50%）的情况下，预计筛选效果甚至会更严重地衰减到估计的27%、17%和9%的降低。在敏感性分析的估计只有轻微的修改，另外允许根据CRC risk.Conclusion的差异坚持和污染：在广泛使用内窥镜检查的时代，即使在筛选程序之外，随机对照试验可能会大大低估CRC筛选的影响。解释不依从和污染的额外分析对于揭示真实的筛查效果至关重要。(C)2013 Elsevier Inc. All rights reserved.
Objectives: Although randomized controlled trials (RCTs) are considered the most valid approach to evaluate screening effects, they also face a number of methodological challenges, including nonadherence and contamination. We aimed to quantify their potential impact in RCTs evaluating endoscopic screening for colorectal cancer (CRC).Study Design and Setting: We carried out model calculations using plausible levels of nonadherence and contamination.Results: Assuming medium values of adherence (70%) and contamination, that is, use of lower gastrointestinal endoscopy other than the one offered for screening (30%), true reductions in CRC incidence or mortality of 70%, 50%, or 30% would be expected to be attenuated to reductions of 43%, 29%, and 16%, respectively, in intention-to-screen analyses. In case of low levels of adherence (50%) and high but realistic levels of contamination (50%), even more severe attenuation of screening effects to estimated reductions of 27%, 17%, and 9% would be expected. The estimates are only slightly modified in sensitivity analyses, additionally allowing for differential adherence and contamination according to CRC risk.Conclusion: In the era of widespread endoscopy use even outside the screening programs, RCTs may strongly underestimate the effects of CRC screening. Additional analyses accounting for nonadherence and contamination are crucial for disclosing the true screening effects. (C) 2013 Elsevier Inc. All rights reserved.