Meta-research in rehabilitation: Reporting guidelines, risk of bias, clinical trials outcomes and interpretability

In the context of healthcare decision-making, high-quality evidence is essential to ensure high value to healthcare. However, concerns persist about the limitations of primary research studies, even within systematic reviews, due to issues related to the way they are planned, conducted, analysed, and reported. To address these challenges and promote research quality, the discipline of meta-research has gained relevance. Meta-research, the study of research itself, not only aids researchers but also guides clinicians in evidence-based decision-making. As the world’s population ages, rehabilitation research plays a crucial role in addressing non-communicable diseases and injuries. Rehabilitation research is complex, with challenges arising from personalised interventions, diverse variables, and evolving disease states. In this context, many key aspects contribute to the research quality in rehabilitation. Comprehensive reporting, for instance, is essential to ensure research replicability and applicability. For this purpose, reporting guidelines and initiatives like the EQUATOR Network are essential for improving reporting standards in the rehabilitation field. The risk of bias and the way in which it is assessed is another critical aspect of research validity. High risk of bias decreases confidence in study results, and it is important to assess whether it influences effect sizes. Moreover, the risk of bias and research reporting are strongly connected. The clear and complete reporting of methods is essential for risk of bias assessment; therefore, investigating the association between risk of bias and completeness of reporting in rehabilitation research is crucial. To take a step further and help clinicians in their daily clinical decision-making, we decided to focus some of the studies included in this thesis on one of the most common and disabling conditions: low back pain. It is a prevalent condition with a significant impact on healthcare systems. Rehabilitation plays a crucial role in managing low back pain patients, and the large amount of research in this field requires rigorous investigations. Randomised controlled trials are essential in evaluating healthcare interventions, but they face unique challenges in the context of low back pain. Methodological flaws, including selective outcome reporting and overlooking clinical relevance, need careful examination. Outcome inconsistency across trials complicates comparisons among the results and makes clinical decision- making difficult. If researchers do not focus their trials on clinically relevant outcomes or discuss clearly the clinical relevance of their results, clinicians cannot decide whether or not an intervention is worth applying. This thesis aimed to explore these key issues in rehabilitation research by evaluating 1) the completeness of reporting, its relationship with risk of bias, and what journals actually do to improve the research reporting, 2) the relationship between risk of bias and effect size, the uptake of the core outcome set and how authors interpret clinical relevance taking low back pain research as a field of investigation.