The interpretation of clinical relevance in randomised clinical trials in patients with chronic low back pain: protocol for a meta-research study

Abstract

ABSTRACTBackgroundChronic low back pain (cLBP) is one of the leading worldwide causes of disability, which accounts for large costs for healthcare systems and work productivity. Many treatment options are available for patients with cLBP and to determine the effectiveness of an intervention, several stakeholders (e.g. researchers, policymakers) should make a judgment about the statistical significance and clinical relevance of the results. Many parameters have been proposed to find a threshold of clinical relevance and the Minimal Important Difference (MID) has been used for a long time. However, the MID has many potential flaws: it is not intervention-specific, and it should be used to interpret a within-group change on an outcome measure rather than as a parameter to interpret a between-group effect. To overcome these issues, the smallest worthwhile effect (SWE) parameter has been proposed. The SWE captures the judgments of recipients of care; allows patients to weigh the benefits of treatment against the risks, costs, and side effects; and potentially provides estimates based on an intervention-control comparison. Specific values of the SWE have been calculated in patients with cLBP for pain intensity, physical functioning and time to recovery for physiotherapy (combining exercise and manual therapy) compared with no intervention. These values are expressed as a percentage of the smallest amount of patient-valued benefit that an intervention would require, compared to an alternative, to justify associated costs, risks, and other inconveniences.Objectives1) To evaluate how authors of original, eligible, and published RCTs have interpreted the clinical relevance of the effect of physiotherapy compared to no-intervention on pain, physical functioning and time to recovery; 2) To reinterpret the clinical relevance of these between-group differences of these original, eligible and published RCTs based on the available SWE estimates for physiotherapy in cLBP compared to no intervention; 3) To evaluate, for descriptive purposes, whether the studies are adequately powered or underpowered considering the published SWE values and a power threshold of 80%.MethodsA systematic search in Medline, PEDro, Embase and Cochrane CENTRAL from inception to November 30th2022 will be conducted. We will search for randomized controlled trials investigating the effectiveness of physiotherapy (combining exercise and manual therapy) as compared to no interventions in people with cLBP. Two authors will perform the data selection and extraction independently; a third reviewer will be involved in case of disagreement. We will compare the authors’ interpretation of results for statistical significance and clinical relevance with their results to determine if they meet their a-priori definitions. Then, we will perform a re-interpretation of the between-group differences for every individual trial based on SWE values published for cLBP. Lastly, we will calculate the power for each included study considering the alpha error used in the study, sample size and a between-group difference of 20% on the primary outcome.Ethics and disseminationA manuscript will be prepared and submitted for publication in an appropriate peer-reviewed journal upon study completion. We believe that the results of this investigation will be relevant to researchers paying more attention to the interpretation of the research results to translate clinical implications to key stakeholders (healthcare providers and patients).VERSION 2.0 – CHANGESWe changed the title replacing “meta-epidemiological” with “meta-research”. This decision is made to reflect the absence of an inferential approach (e.g. regression analyses) in our study that usually characterizes meta-epidemiological studies, as highlighted in a recent publication (see: Puljak L. Caution is needed when describing a study design as meta-epidemiological. J Clin Epidemiol. 2022 Dec;152:326-327. doi: 10.1016/j.jclinepi.2022.10.017. Epub 2022 Oct 26. PMID: 36309145).We replaced meta-epidemiological with “meta-research” in the whole text.