Overview of the project (extendNMA)
The project titled “Extending Network Meta Analysis to include Non-Randomized Evidence and Single-Arm Trials (extendNMA)” is funded by the Hellenic Foundation for Research & Innovation (project number 15467) under the call “Sub-action 2. Funding Projects in Leading-Edge Sectors – RRFQ: Basic Research Financing(Horizontal support for all Sciences)”. The research team consists of the following members
Dimitris Mavridis, Professor of Statistics in the Department of Primary Education, University of Ioannina
Stavros Nikolakopoulos, Assistant Professor of Statistics in the Department of Psychology, University of Ioannina
Ourania Koutsiouroumpa, PhD student in the Department of Primary Education, University of Ioannina
Katerina-Maria Kontouli, PhD student in the Department of Primary Education, University of Ioannina
Christos Christogiannis, Post-doctoral research associate in the Department of Psychology, University of Southampton, UK
Georgios Seitidis, Post-doctoral research associate in the Department of Psychology, University of Ioannina
Sofia Tsokani, Post-doctoral research associate in the School of Medicine, Aristotle University of Thessaloniki
Iro Ntaga, PhD student in the Department of Primary Education, University of Ioannina
Afroditi Tsatsi, PhD student in the Department of Psychology, University of Ioannina
Meta-analysis is a statistical method used to synthesize quantitative data from multiple independent studies that address a common research question to compare the effectiveness of two interventions. Network Meta-Analysis (NMA) is an established method that extends meta-analysis for comparative effectiveness research across a network of studies and interventions. This project has several aims including evaluating the main assumptions of NMA and developing tests for comparing direct and indirect evidence, conducting two systematic reviews for identifying methodological papers for synthesizing randomized evidence (RE) and non-randomized evidence (NRE) and for including single-arm trials (SAT) in NMAs. We aim to compare some of the retrieved methodologies both theoretically and through a simulation study. We aim to develop new methods that will use studies proximities to get indirect effect estimates using studies with, as similar as possible, distribution of covariates.
We conducted a systematic review to identify all methods available in the literature for including NRE and SATs in meta-analyses and NMAs. The protocol was drafted by six research team members and the principal investigator, who also organized search strategies and conducted the searches, screening and data extraction. More analytically, searches were performed across three databases, and two levels of screening by a pool of four independent reviewers yielded 38 and 39 publications for NRE and SATs, respectively. Then three independent extractors worked on the data extraction from the included publication and are currently summarizing the extracted information. Two technical reports were submitted to the Hellenic Foundation for Research and Innovation as planned. Results of this work will be submitted for presentation at the upcoming 2nd Conference in Biostatistics & Health Analytics in Aegina, The conference is co-organised by our group (the Evidence Synthesis Methods Statistics Team of the University Ioannina), the Department of Statistics of Athens University of Economics and Business, the MSc of Biostatistics & Health Data Science and the Frontier Science Foundation-Hellas. We will give several presentations and posters regarding the progress of extendNMA.
Bayesian methods to evaluate consistency have been developed and included in an R package, available both in github (https://github.com/georgiosseitidis/ssifs/blob/master/README.md) and CRAN repositories.
Methods for incorporating non-randomized evidence and single-arm trials in network meta-analysis of randomized clinical trials
Extending NMA to include non randomized studies results of a scoping review
Oral presentation and poster prepared by Koutsiouroumpa Ourania for the 11th Panhellenic conference of public health and epidemiology. Athens, Greece, 21-23 November 2025 and poster presented for the 2nd Conference in Biostatistics & Health Analytics, Aegina, Greece, 7-9 July 2025.
Bridging Single Arm Studies with Individual Participant Data
Poster prepared by Katerina-Maria Kontouli for the 46th conference of the International Society for Clinical Biostatistics in Basel, Switzerland, 24th-28th August, 2025 and for the 2nd Conference in Biostatistics & Health Analytics, Aegina, Greece, 7-9 July 2025 .
Synthesing randomized and non-randomized evidence in meta-analysis models
Workshop prepared by Ourania Koutsiouroumpa and presented in the 19th Panhellenic Conference of the Hellenic Society of Psychology, Ioannina, Greece, 7th-11th May, 2025