Table 1.
Recommendations.
| Recommendations for best practice | Authors |
|---|---|
| Statisticians should be involved in all trial stages, from the beginning (i.e. conceptualisation and planning) to end (i.e. reporting and dissemination). | [25,28,30,33,42,[47], [48], [49], [50]]; Meeks et al., 2018 |
| Statisticians who sit on RECs should have appropriate training. | [26,36] |
| More general, formal training should be made available for statisticians to advance their statistical methodology knowledge | [31,38,39,44] |
| Statisticians should be involved in writing budgets and trial funding grant applications. | [4,29,35] |
| Statisticians should be involved in preparing and designing case reports forms and interacting with data management team members and database developers. | Crewson and Applegate, 2001 [4,24,42,49,51,72]; |
| Statisticians should create clear and comprehensive statistical analysis plans (SAPs), adhere to them while analysing a trial and pre-publish them to reduce bias. | [59]; Snow et al., 2015, Pyke et al., 2010 [59], |
| Statisticians should be involved in reporting trials and ensure honest and accurate reporting of all trial results (regardless of the outcome being positive or negative) and be transparent in the methods sections of manuscripts. | [[60], [61], [62], [63]]. [33,53,64] |
| Statisticians should actively communicate with fellow statisticians and other team members (including clinicians) by understanding the others' perspective | [30,53,71] |
| Statisticians should avoid the use of technical statistical jargon and ensure team members understand the statistical aspects. | [4,33,42,46,71] |
| Non-statisticians involved in trials should have an understanding of the fundamentals of research methods and statistical reasoning. | [4,21,41,47,48,55,65,70,88]. [45]. [66,80] |
| Study statisticians or statisticians involved in trial leadership should be independent or external to statisticians conducting interim analyses and reporting to DMCs. A firewall should separate the sponsor/steering committee/the primary trial statistician and the DMC or independent statistician. |
[25,27,83,84] |