Lipman Paula Darby, Loudon Kirsty, Dluzak Leanora, Moloney Rachael, Messner Donna, Stoney Catherine M
Westat, 1600 Research Boulevard, Rockville, MD, 20850, USA.
NMAHP Research Unit, Unit 13 Scion House, Stirling University Innovation Park, Stirling, FK9 4NF, UK.
Trials. 2017 Nov 10;18(1):532. doi: 10.1186/s13063-017-2267-y.
There continues to be debate about what constitutes a pragmatic trial and how it is distinguished from more traditional explanatory trials. The NIH Pragmatic Trials Collaborative Project, which includes five trials and a coordinating unit, has adopted the Pragmatic-Explanatory Continuum Indicator Summary (PRECIS-2) instrument. The purpose of the study was to collect PRECIS-2 ratings at two points in time to assess whether the tool was sensitive to change in trial design, and to explore with investigators the rationale for rating shifts.
A mixed-methods design included sequential collection and analysis of quantitative data (PRECIS-2 ratings) and qualitative data. Ratings were collected at two annual, in-person project meetings, and subsequent interviews conducted with investigators were recorded, transcribed, and coded using NVivo 11 Pro for Windows. Rating shifts were coded as either (1) actual change (reflects a change in procedure or protocol), (2) primarily a rating shift reflecting rater variability, or (3) themes that reflect important concepts about the tool and/or pragmatic trial design.
Based on PRECIS-2 ratings, each trial was highly pragmatic at the planning phase and remained so 1 year later in the early phases of trial implementation. Over half of the 45 paired ratings for the nine PRECIS-2 domains indicated a rating change from Time 1 to Time 2 (N = 24, 53%). Of the 24 rating changes, only three represented a true change in the design of the trial. Analysis of rationales for rating shifts identified critical themes associated with the tool or pragmatic trial design more generally. Each trial contributed one or more relevant comments, with Eligibility, Flexibility of Adherence, and Follow-up each accounting for more than one.
PRECIS-2 has proved useful for "framing the conversation" about trial design among members of the Pragmatic Trials Collaborative Project. Our findings suggest that design elements assessed by the PRECIS-2 tool may represent mostly stable decisions. Overall, there has been a positive response to using PRECIS-2 to guide conversations around trial design, and the project's focus on the use of the tool by this group of early adopters has provided valuable feedback to inform future trainings on the tool.
关于什么构成实用性试验以及它如何与更传统的解释性试验区分开来，一直存在争议。美国国立卫生研究院实用性试验协作项目包括五项试验和一个协调单位，采用了实用性-解释性连续统指标总结（PRECIS-2）工具。本研究的目的是在两个时间点收集PRECIS-2评分，以评估该工具对试验设计变化是否敏感，并与研究者探讨评分变化的理由。
采用混合方法设计，包括对定量数据（PRECIS-2评分）和定性数据进行顺序收集和分析。在两次年度面对面项目会议上收集评分，并对随后与研究者进行的访谈进行录音、转录，并使用适用于Windows的NVivo 11 Pro进行编码。评分变化被编码为以下三种情况之一：（1）实际变化（反映程序或方案的改变），（2）主要是反映评分者变异性的评分变化，或（3）反映关于该工具和/或实用性试验设计的重要概念的主题。
根据PRECIS-2评分，每项试验在规划阶段都高度实用，并且在试验实施的早期阶段一年后仍保持如此。九个PRECIS-2领域的45对评分中，超过一半表明从时间1到时间2有评分变化（N = 24，53%）。在这24个评分变化中，只有三个代表试验设计的真正变化。对评分变化理由的分析更广泛地确定了与该工具或实用性试验设计相关的关键主题。每项试验都提供了一条或多条相关评论，其中入选标准、依从性灵活性和随访各自占比超过一条。
事实证明，PRECIS-2有助于在实用性试验协作项目成员之间“构建关于试验设计的对话”。我们的研究结果表明，PRECIS-2工具评估的设计要素可能大多代表稳定的决策。总体而言，对于使用PRECIS-2来指导围绕试验设计的对话有积极反馈，并且该项目对这组早期采用者使用该工具的关注提供了有价值的反馈，为该工具未来的培训提供参考。
