An important step before clinical intervention selection is the diagnosis of the condition of a patient. Diagnostic tests are commonly used to confirm or exclude a target condition (e.g. a disease). Decision-making rarely relies on a single diagnostic test accuracy (DTA) study; instead, evidence from multiple DTA studies addressing the same question is used.
DTA meta-analysis focuses on evaluating individual tests across separate studies. However, multiple index tests may be available for a target condition, making their comparative accuracy is important for decision-making. Network meta-analysis (NMA) allows for a more integrated and comprehensive evaluation of several diagnostic tests simultaneously in a single model.
Over the past few years, several NMA models have been developed to evaluate the comparative accuracy of multiple diagnostic tests. In this Cochrane Learning Live webinar, the presenters discussed why traditional NMA methods for interventions are unsuitable for DTA studies and highlighted basic principles and assumptions of DTA-NMA models, such as handling multiple thresholds. The session also explored challenges and opportunities in conducting a DTA-NMA.
The session was of particular interest to review authors who would like to incorporate results from DTA studies comparing multiple diagnostic tests in a review. It was delivered in November 2024 and below you will find the videos from the webinar, together with the accompanying slides to download [PDF].
Part 1: Evidence synthesis & geometric representation of DTA studies
Part 2: Scoping review of DTA-NMA methods & a practical example
Presenter Bios
Dr. Areti-Angeliki Veroniki is a Scientist at the Knowledge Translation Program of St. Michael’s Hospital, Unity Health Toronto, and an Assistant Professor at the University of Toronto in the Institute of Health Policy, Management, and Evaluation. She is a co-Convenor of the Cochrane Statistical Methods Group and co-chair of the Cochrane Methods Executive. Her interests are to optimize the processes of evidence-based medicine and statistical modelling for knowledge synthesis of study findings. Specifically, her research focuses on methods for meta-analysis and network meta-analysis.
Dr. Sofia Tsokani is Statistics Editor in the Methods Support Unit of the Cochrane Central Editorial Team. She is also a post-doctoral researcher at the Laboratory of Hygiene, Social & Preventive Medicine and Medical Statistics of Aristotle University of Thessaloniki in Greece. Her research interests encompass evidence synthesis methods and more specifically statistical modelling for meta-analysis and network meta-analysis of both intervention and diagnostic test accuracy studies.