Health Services Research Unit, University of Aberdeen, Aberdeen, UK.
Syst Rev. 2018 Feb 20;7(1):30. doi: 10.1186/s13643-018-0696-7.
Randomised control trials are regarded as the gold standard for evaluating the effectiveness and efficacy of healthcare interventions with thousands of trials published every year. Despite significant investment in infrastructure, a staggering number of clinical trials continue to face challenges with retention. Dropouts could lead to negative consequences-from lengthy delays to missing data that can undermine the results and integrity of the trial. Summarising evidence from non-randomised evaluations of retention strategies could provide complementary information to randomised evaluations that could guide trialists to the most effective ways of increasing retention of participants in clinical trials.
The following electronic databases will be searched for relevant studies: EMBASE, MEDLINE, the Cochrane Controlled Trials Register, and Cochrane Methodology Register and the search will be limited to English-published studies during the last 10 years to increase relevance to current trials. Non-randomised studies (observational studies) including a comparison of two or more strategies to increase participant retention in randomised trials or comparing one or more strategies with no strategy will be included. The primary outcome will be the proportion of participants remained at the primary analysis as determined in each retention study.
This review aims to gather and evaluate evidence on the effect of retention strategies examined in non-randomised studies. It is imperative to collect evidence from obseravational studies to infer whether or not these studies could be considered a practical way to complement or even replace a broadly favourable randomised design. If we find that non-randomised studies to be included in this review are of high quality with adequate control of biases, we will recommend to trialists and others not to rely exclusively on randomised studies and to give meticulous attention to the plentiful evidence that can be obtained from non-randomised studies. Should the results of this review suggest that evaluating retention strategies in observational studies provides insufficient evidence to trialists planning their retention strategies, we will be able to say that there is little point in conducting non-randomised studies and that they would do better to invest their time and resources in a randomised evaluation if possible. Where a non-randomised study design is chosen, the review authors will offer recommendations to trialists and others regarding how to ensure that these studies are conducted in a way that can minimise the risk of bias and increase confidence in the findings.
PROSPERO 2017: CRD42017072775 .
随机对照试验被认为是评估医疗干预措施有效性和效果的金标准，每年都有数千项试验发表。尽管在基础设施方面投入了大量资金，但仍有大量临床试验在保留方面面临挑战。脱落可能导致从冗长的延迟到数据缺失等负面后果，这可能会破坏试验的结果和完整性。总结保留策略的非随机评估证据可以为随机评估提供补充信息，从而指导试验人员找到最有效的方法来提高临床试验参与者的保留率。
将在以下电子数据库中搜索相关研究：EMBASE、MEDLINE、Cochrane 对照试验登记处和 Cochrane 方法学登记处，并且搜索将限于过去 10 年内发表的英文研究，以增加与当前试验的相关性。将纳入非随机研究（观察性研究），包括比较两种或多种策略以增加随机试验中参与者的保留率，或比较一种或多种策略与无策略的研究。主要结局将是每个保留研究中确定的主要分析中保留的参与者比例。
本综述旨在收集和评估非随机研究中检查的保留策略的效果证据。从观察性研究中收集证据以推断这些研究是否可以被认为是一种补充甚至替代广泛有利的随机设计的实际方法至关重要。如果我们发现纳入本综述的非随机研究质量高，且偏倚得到充分控制，我们将建议试验人员和其他人不要仅依赖随机研究，并对可以从非随机研究中获得的大量证据给予细致关注。如果本综述的结果表明，在观察性研究中评估保留策略为计划保留策略的试验人员提供的证据不足，我们将能够说进行非随机研究意义不大，如果可能的话，他们最好将时间和资源投入到随机评估中。如果选择非随机研究设计，综述作者将向试验人员和其他人提供有关如何确保这些研究以可最大程度降低偏倚风险并增加对研究结果的信心的方式进行的建议。
PROSPERO 2017：CRD42017072775。
