Strategies to improve retention in randomised trials: a Cochrane systematic review and meta-analysis
Strategies to improve retention in randomised trials: a Cochrane systematic review and meta-analysis
复制标题
DOI：
10.1136/bmjopen-2013-003821
复制
发表时间：
2014-01-01
期刊：
影响因子：
2.9
通讯作者：
Rait, G.
中科院分区：
文献类型：
--
作者：
Brueton, V. C.;Tierney, J. F.;Rait, G.
目的 量化提高随机试验保留率的策略的效果。设计 系统评价和荟萃分析。数据来源 检索来源：MEDLINE、EMBASE、PsycINFO、DARE、CENTRAL、CINAHL、C2-SPECTR、ERIC、PreMEDLINE、Cochrane Methodology Register、Current Controlled Trials metaRegister、WHO 试验平台、临床试验协会 (SCT) 会议记录以及对所有英国临床试验研究的调查审查方法 纳入的试验是对宿主随机试验中提高保留率的策略进行随机评估。主要结果是试验参与者的保留。使用固定效应模型汇总试验数据。亚组分析用于探索异质性，并确定不同策略类型的效果是否存在差异。结果 确定了 38 项保留试验。评估了六大类策略。增加邮寄问卷答复的策略是：添加，即给予金钱奖励（RR 1.18；95% CI 1.09至1.28）和更高价值的奖励（RR 1.12；95% CI 1.04至1.22）。提供金钱奖励，即在收到完整问卷后给予奖励，也增加了电子问卷的答复（RR 1.25；95% CI 1.14 至 1.38）。较短的问卷（RR 1.04；95% CI 1.00 至 1.08）和与疾病/状况相关的问卷（RR 1.07；95% CI 1.01 至 1.14）的证据不太清楚。根据单项试验的结果，以下策略似乎可以有效提高问卷答复：记录问卷的发送（RR 2.08；95% CI 1.11至3.87）；一揽子邮政通信策略（RR 1.43；95% CI 1.22 至 1.67）和开放试验设计（RR 1.37；95% CI 1.16 至 1.63）。没有充分的证据表明以下策略会影响试验反应/保留： 增加非金钱激励（RR=1.00；95% CI 0.98 至 1.02）；提供非金钱激励（RR=0.99；95% CI 0.95 至 1.03）；增强字母（RR=1.01；95% CI 0.97 至 1.05）；与提供抽奖机会相比的金钱激励（RR=1.04；95% CI 0.91 至 1.19）；优先邮寄（RR=1.02；95% CI 0.95 至 1.09）；行为激励策略（RR=1.08；95% CI 0.93 至 1.24）；对参与者的额外提醒（RR=1.03；95% CI 0.99 至 1.06）和问卷问题顺序（RR=1.00、0.97 至 1.02）。同样基于单次试验，这些策略似乎并不有效：电话调查与金钱激励加问卷调查相比（RR=1.08；95% CI 0.94 至 1.24）；提供慈善捐款（RR=1.02，95% CI 0.78 至 1.32）；发送站点提醒（RR=0.96；95% CI 0.83 至 1.11）；尽早发送调查问卷（RR=1.10；95% CI 0.96 至 1.26）；更长、更清晰的调查问卷（RR=1.01、0.95至1.07）以及试验助理对参与者进行案例管理（RR=1.00；95% CI 0.97至1.04）。 结论 大多数试验评估的是问卷反应，而不是改善参与者返回现场进行随访的方法。货币激励和货币激励的提供增加了邮寄和电子问卷的答复。一些策略需要进一步评估。这些结果的应用将取决于试验背景和后续程序。
Objective To quantify the effect of strategies to improve retention in randomised trials.Design Systematic review and meta-analysis.Data sources Sources searched: MEDLINE, EMBASE, PsycINFO, DARE, CENTRAL, CINAHL, C2-SPECTR, ERIC, PreMEDLINE, Cochrane Methodology Register, Current Controlled Trials metaRegister, WHO trials platform, Society for Clinical Trials (SCT) conference proceedings and a survey of all UK clinical trial research units.Review methods Included trials were randomised evaluations of strategies to improve retention embedded within host randomised trials. The primary outcome was retention of trial participants. Data from trials were pooled using the fixed-effect model. Subgroup analyses were used to explore the heterogeneity and to determine whether there were any differences in effect by the type of strategy.Results 38 retention trials were identified. Six broad types of strategies were evaluated. Strategies that increased postal questionnaire responses were: adding, that is, giving a monetary incentive (RR 1.18; 95% CI 1.09 to 1.28) and higher valued incentives (RR 1.12; 95% CI 1.04 to 1.22). Offering a monetary incentive, that is, an incentive given on receipt of a completed questionnaire, also increased electronic questionnaire response (RR 1.25; 95% CI 1.14 to 1.38). The evidence for shorter questionnaires (RR 1.04; 95% CI 1.00 to 1.08) and questionnaires relevant to the disease/condition (RR 1.07; 95% CI 1.01 to 1.14) is less clear. On the basis of the results of single trials, the following strategies appeared effective at increasing questionnaire response: recorded delivery of questionnaires (RR 2.08; 95% CI 1.11 to 3.87); a package' of postal communication strategies (RR 1.43; 95% CI 1.22 to 1.67) and an open trial design (RR 1.37; 95% CI 1.16 to 1.63). There is no good evidence that the following strategies impact on trial response/retention: adding a non-monetary incentive (RR=1.00; 95% CI 0.98 to 1.02); offering a non-monetary incentive (RR=0.99; 95% CI 0.95 to 1.03); enhanced' letters (RR=1.01; 95% CI 0.97 to 1.05); monetary incentives compared with offering prize draw entry (RR=1.04; 95% CI 0.91 to 1.19); priority postal delivery (RR=1.02; 95% CI 0.95 to 1.09); behavioural motivational strategies (RR=1.08; 95% CI 0.93 to 1.24); additional reminders to participants (RR=1.03; 95% CI 0.99 to 1.06) and questionnaire question order (RR=1.00, 0.97 to 1.02). Also based on single trials, these strategies do not appear effective: a telephone survey compared with a monetary incentive plus questionnaire (RR=1.08; 95% CI 0.94 to 1.24); offering a charity donation (RR=1.02, 95% CI 0.78 to 1.32); sending sites reminders (RR=0.96; 95% CI 0.83 to 1.11); sending questionnaires early (RR=1.10; 95% CI 0.96 to 1.26); longer and clearer questionnaires (RR=1.01, 0.95 to 1.07) and participant case management by trial assistants (RR=1.00; 95% CI 0.97 to 1.04).Conclusions Most of the trials evaluated questionnaire response rather than ways to improve participants return to site for follow-up. Monetary incentives and offers of monetary incentives increase postal and electronic questionnaire response. Some strategies need further evaluation. Application of these results would depend on trial context and follow-up procedures.