Meta-research to improve the planning and reporting of randomized clinical trials

The ongoing challenge of poor research persists in the scientific community, highlighting concerns about unreliable findings, misguiding decision-making processes, and losing public trust. A key solution lies in addressing the burden of inadequate methods by making adjustments. Better planning and reporting are essential to tackle this issue effectively. Meta-research, involving interdisciplinary research on research methods, offers insights into existing challenges. In this PhD, we proposed two main meta-research projects to identify problems and suggest solutions, ultimately improving the planning, and reporting of randomized controlled trials (RCTs): i) The pattern of RECRUITment In randomized clinical trials (RECRUIT-IT) study, and ii) Subprojects of the ASPIRE (Adherence to SPIRIT Recommendations) Study on different methodological topics.
i) RECRUIT-IT study:
We conducted an empirical study using individual participant recruitment data from RCTs, gathered from convenient national, and international networks, to empirically identify common trial recruitment patterns. Our findings indicate that approximately two-thirds of RCTs had an overall linear recruitment trajectory, facilitating straightforward predictions of future recruitment. Principal investigator (PI) sites generally contributed more, longer, and faster in participant recruitment, underscoring their crucial role in ensuring successful trial conduct.
ii) Subprojects of the ASPIRE study:
Our team has built a database of 760 RCT protocols already approved by research ethics committees in Switzerland (Basel, Bellinzona, Bern, Geneva, Lausanne, St. Gallen, Thurgau, Zurich), Germany (Freiburg), Canada (Hamilton), and UK; 360 RCT protocols from 2012 and 400 from 2016. We used this database and complemented it with additional information from RCT protocols and corresponding publications to provide empirical evidence on A) non-registration, discontinuation, and non-publication of RCTs, B) the prevalence and reporting of patient reported outcomes (PROs) in RCT protocols and publications, and C) reporting of prespecified subgroup analyses in RCT protocols. The results of subproject A showed that non-registration (6%), and non-availability of results (20%) remain significant issues, with non-industry trials being more affected than industry trials. Additionally, approximately one third of all RCTs were prematurely discontinued, mainly due to poor participant recruitment. The results of subproject B showed that 70% of RCTs specified PROs as either primary or secondary outcomes, with significant variability among medical disciplines and interventions. The reporting standard for PROs in both protocols and publications was suboptimal, with a considerable proportion failing to adhere to protocol specifications when reporting PRO results. Similarly, the results of subproject C showed that planned subgroup analyses in the majority of RCT protocols were remained persistently inadequately in addressing fundamental scientific principles such as prior research considerations, limiting the number of subgroup analyses, and applying appropriate statistical methods.
The insights from the RECRUIT-IT study provide investigators with an overview of common trial recruitment patterns on the trial-level and the site-level, facilitating prediction and monitoring of participant recruitment in RCTs. Consequently, they can intervene to improve recruitment if it is needed, reducing the risk of unsuccessful recruitment and trial discontinuation. Assuring recruitment preserves research integrity by allowing the study to progress as intended, thereby minimizing the chance of discontinuation and research waste.
Insights from ASPIRE subprojects have raised awareness and highlighted the importance of various methodological challenges in trial registration, results publication, and planning and reporting of PROs and subgroup analyses. Enhancing registration and publication practices can reduce duplication, publication bias, and enhance transparency, thus reducing research waste. Simultaneously, effective planning and adherence to PRO protocols and improving methodological quality in subgroup analysis have a crucial role in ensuring the credibility of RCT results, aiding in waste reduction, and promoting more robust and transparent clinical research.