You are here

Article Text

Article menu
Download PDFPDF
EBM learning
Combining and converting groups when extracting data for meta-analysis
  1. Kathryn S Taylor1,
  2. Kamal R Mahtani1,
  3. Jeffrey K Aronson2
  1. 1 Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, UK
  2. 2 Centre for Evidence Based Medicine, University of Oxford, Oxford, UK
  1. Correspondence to Dr Kathryn S Taylor, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford OX1 2JD, UK; kathryn.taylor{at}phc.ox.ac.uk

https://doi.org/10.1136/bmjebm-2020-111647

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    It is important that as much evidence as possible is included in the pooled analyses of systematic reviews. Therefore, systematic reviewers may want to analyse outcomes that are not completely specified in the reports of studies that they would like to include. However, using equations and other methods it is sometimes possible to fill in missing data and also check the validity of the reported data. Equations that are particularly useful are those for changing the grouping of patients from that which is reported. The desire to change groups often arises in systematic reviews and meta-analyses of interventions. For example, you may wish to combine the baseline data for two or more groups. Consider a clinical question as an example:

    What is the effect of inhibiting the renin-angiotensin aldosterone system, compared with placebo, on urinary albumin concentrations in patients with type 2 diabetes?

    Patients with type 2 diabetes may be subgrouped into those with diabetic nephropathy (microalbuminuric) and those without diabetic nephropathy (normoalbuminuric), such as in a study by Sawaki et al,1 which we shall use as an example. They reported the effects of losartan on urinary albumin, measured by the urinary albumin:creatinine ratio (UACR), expressed as milligram of albumin per gram of creatinine (mg/g). Baseline characteristics were reported by treatment group, and some of the data needed for meta-analysis were reported for the subgroups within each treatment group.

    When compiling a table of baseline characteristics of the study populations of the included studies in a systematic review, you may find that some studies report baseline data for one or more groups, split by treatment, or by …

    View Full Text

    Footnotes

    • Twitter @dataextips, @JKAronson

    • Contributors KST and KRM conceived the idea of the series of which this is one part. KST wrote the first draft of the manuscript. All authors revised the manuscript and agreed the final version.

    • Funding This research was supported by the National Institute for Health Research Applied Research Collaboration Oxford and Thames Valley at Oxford Health NHS Foundation Trust.

    • Disclaimer The views expressed in this publication are those of the author(s) and not necessarily those of the National Institute for Health Research

      or the Department of Health and Social Care.

    • Competing interests KRM and JKA were Associate Editors of BMJ Evidence Medicine at the time of submission.

    • Provenance and peer review Commissioned; internally peer reviewed.