Project

Back to overview

What works best? Methods for ranking competing treatments in network meta-analysis

English title What works best? Methods for ranking competing treatments in network meta-analysis
Applicant Salanti Georgia
Number 179158
Funding scheme Project funding (Div. I-III)
Research institution Institut für Sozial- und Präventivmedizin Universität Bern
Institution of higher education University of Berne - BE
Main discipline Medical Statistics
Start/End 01.06.2018 - 31.05.2022
Approved amount 671'999.00
Show all

All Disciplines (3)

Discipline
Medical Statistics
Public Health and Health Services
Methods of Epidemiology and Preventive Medicine

Keywords (4)

network meta-analysis; ranking; benefit-risk ; credibility of evidence

Lay Summary (German)

Lead
Mit welchem Medikament soll man diese Erkrankung behandeln?
Lay summary

Die wichtigste Frage, die Patienten und ihre Ärzte haben, lautet

‘Mit welchem Medikament soll man diese Erkrankung behandeln?’

Um solche Fragen zu beantworten, berücksichtigen Forschende die Ergebnisse von klinischen Studien. Die Erkenntnisse aus relevanten Studien wurden zusammengetragen und (dann) analysiert, um eine Medikamentenhierarchie gemäß ihrer Wirksamkeit zu liefern.

Ärzte und Patienten sind sich darin einig, das ein solcher Ueberblick über die Wirksamkeit der Medikamente zu ihrer Therapieentscheidung und zur Entwicklung von Richtlinien beitragen kann. Forschende debattieren jedoch noch über die dafuer eingesetzte, statistische Methode.

Bei diesem Projekt werden wir die statistischen Methoden zur Medikamentenhierarchie studieren.  Wir werden sie erst beschreiben und bewerten und dann werden wir sie verbessern. Wir werden neue Methoden entwickeln, wobei nicht nur die Wirksamkeit, sondern auch die Verträglichkeit der Medikamente in den Berechnungen der Hierarchie berücksichtigt werden. Unsere Resultate sollen massgeblich die Therapieentscheidung für die optimale medicamentöse Therapie für Patienten mit akuter Depression oder Schizophrenie beitragen. 

 

Direct link to Lay Summary Last update: 03.04.2018

Responsible applicant and co-applicants

Employees

Project partner

Publications

Publication
Agreement between ranking metrics in network meta-analysis: an empirical study
Chiocchia Virginia, Nikolakopoulou Adriani, Papakonstantinou Theodoros, Egger Matthias, Salanti Georgia (2020), Agreement between ranking metrics in network meta-analysis: an empirical study, medRxiv, Online.
Synthesizing existing evidence to design future trials: survey of methodologists from European institutions
Nikolakopoulou Adriani, Trelle Sven, Sutton Alex J., Egger Matthias, Salanti Georgia (2019), Synthesizing existing evidence to design future trials: survey of methodologists from European institutions, in Trials, 20(1), 334-334.
A comparison of arm‐based and contrast‐based models for network meta‐analysis
White Ian R., Turner Rebecca M., Karahalios Amalia, Salanti Georgia (2019), A comparison of arm‐based and contrast‐based models for network meta‐analysis, in Statistics in Medicine, 38(27), 5197-5213.
Optimal dose of selective serotonin reuptake inhibitors, venlafaxine, and mirtazapine in major depression: a systematic review and dose-response meta-analysis
Furukawa Toshi A, Cipriani Andrea, Cowen Philip J, Leucht Stefan, Egger Matthias, Salanti Georgia (2019), Optimal dose of selective serotonin reuptake inhibitors, venlafaxine, and mirtazapine in major depression: a systematic review and dose-response meta-analysis, in The Lancet Psychiatry, 6(7), 601-609.
Extensions of the probabilistic ranking metrics of competing treatments in network meta-analysis to reflect clinically important relative differences on many outcomes
MavridisDimitris, ProcherRaffaele, NikolakopoulouAdriani, SalantiGeorgia, RaveaudPhillipe (2019), Extensions of the probabilistic ranking metrics of competing treatments in network meta-analysis to reflect clinically important relative differences on many outcomes, in Biometrical Journal.
ASSESSING CONFIDENCE IN THE RESULTS OF NETWORK META- ANALYSIS (CINEMA)
Nikolakopoulou Adriani, Higgins Julian, Papakonstantinou Theodore, Chaimani Anna, Del Giovane Cinzia, Egger Matthias, Salanti Georgia, ASSESSING CONFIDENCE IN THE RESULTS OF NETWORK META- ANALYSIS (CINEMA), in PLoS Medicine, 1.

Collaboration

Group / person Country
Types of collaboration
MRC Clinical Trials Unit at UCL Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Department of Psychiatry University of Oxford Great Britain and Northern Ireland (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Exchange of personnel
Department of Health Promotion and Human Behavior, Kyoto University Japan (Asia)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
- Exchange of personnel
Department of Psychiatry, Technical University of U Munich Germany (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication
Institut für Medizinische Biometrie und Statistik (IMBI) Albert-Ludwigs-Universität Freiburg Germany (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Exchange of personnel
University of Ioannina, Department of Education Greece (Europe)
- in-depth/constructive exchanges on approaches, methods or results
- Publication

Awards

Title Year
Best poster for the PhD project of Virginia Chioccia at 11th Symposium Graduate School of Health Sciences 2019 (19-20th November 2019) 2019

Use-inspired outputs

Associated projects

Number Title Start Funding scheme
166656 Enhancing methods for evaluating the comparative safety of medical interventions 01.09.2016 Project funding (Div. I-III)

Abstract

Background: The most critical question raised by patients and clinicians at the point of care is “what is the drug of choice for the given condition?”. Comparative effectiveness research and its quantitative component, network meta-analysis (NMA), have been used to answer this question for over a decade now. NMA estimates all relative effects between competing treatments and can produce statistical ranking metrics (such as the , the surface under the cumulative ranking curve) that lead to a treatment hierarchy from the least desirable to the most desirable option. While clinicians and guideline developers unanimously agree that a treatment hierarchy is essential, methodologists debate several issues underpinning the ranking metrics obtained from NMA. Although about half of the published NMAs report a treatment hierarchy for their primary outcome, methods for simultaneously ranking treatments for multiple outcomes, for example efficacy and safety, are underdeveloped. A coherent approach to critically appraise a treatment hierarchy is also lacking; it is not clear how to estimate the impact of within-study biases and publication bias. Finally, although uncertainty around ranks could be calculated, a single measure of the precision of the most likely treatment hierarchy is still needed. Aim: This project aims a) to extend the existing ranking metrics to produce treatment hierarchies that account for multiple health outcomes and reflect clinically important treatment effects b) to provide methods to evaluate the robustness of the treatment hierarchy in the presence of bias and estimate its precision. Methods: We will start by evaluating the existing ranking metrics (using both a statistical-theoretical approach and an empirical re-analysis of a large collection of networks) and match them with treatment hierarchy problems (that is, a clear definition about what a ‘preferable’ treatment means). We will then develop ranking metrics for multiple health outcomes by considering emerging statistical models, a re-parametrization of the standard NMA model and different probabilistic summary statistics that account for clinically important differences between treatments and patient values. We will develop methods that characterize the robustness of the treatment hierarchy in the presence of study-level bias and selection bias. Finally, we will suggest a measure of the precision of treatment hierarchy. We will develop user-friendly open-source tools (software routines, packages and web applications) to implement our methods and we will apply them in two clinical settings; in the hierarchy of 21 antidepressants in acute depression and 34 antipsychotics for schizophrenia. Significance: This project will address an urgent need to develop fit-for-purpose ranking metrics to generate treatment hierarchies for a variety of research and clinical questions and evaluate their robustness. Our project will extend the decision-making arsenal of evidence-based health care and public health with tools that support clinicians, policy makers, and health-care decision makers to make better decisions about the best treatments for a given condition.
-