Evidence Synthesis Methods

Evidence Synthesis Methods

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

The aim of this project is to provide a methodological framework to obtain statistical ranking metrics from network meta-analysis and appraise the robustness of the resulting treatment hierarchy. To achieve this aim we have set the following objectives:

  1. Study (theoretically and empirically) the properties of the existing statistical ranking metrics and match them to treatment hierarchy problems (that is, a clear definition about what a ‘preferable’ treatment means)
  2. Extend the existing ranking metrics to address complex hierarchy evaluation problems where multiple health outcomes are of interest, differentiate between clinically important and unimportant treatment effects and account for patient preferences
  3. Develop methods to estimate the impact of within-study biases and selection bias (publication and selective outcome reporting bias) on treatment hierarchy
  4. Develop a measure of the precision of a given treatment hierarchy
  5. Test the extended metrics in ranking antidepressants for acute depression and antipsychotics for schizophrenia.

This research project addresses 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 extends the decision-making arsenal of evidence-based health care and public health with tools that support clinicians, policy makers and patients to make better decisions about the best treatments for a given condition.

 

This project has received funding from the Swiss National Science foundation (SNF) under grant agreement 179158 (Principal Investigator Georgia Salanti).

Group members

Chiocchia V., Nikolakopoulou A., Papakonstantinou T., Egger M., Salanti G. Agreement between ranking metrics in network meta-analysis: an empirical study, BMJ Open http://dx.doi.org/10.1136/bmjopen-2020-037744

Papakonstantinou T., Nikolakopoulou A., Egger M., Salanti G. In network meta-analysis most of the information comes from indirect evidence: empirical study J Clin Epidemiol 2020 Apr 14;124:42-49

Mavridis, D.; Porcher, R.; Nikolakopoulou, A.; Salanti, G.; Ravaud, P. Extensions of the probabilistic ranking metrics of competing treatments in network meta-analysis to reflect clinically important relative differences on many outcomes. Biom J 2020 Mar;62(2):375-385

Nikolakopoulou A., Mavridis D., Chiocchia V., Papakonstantinou T., Furukawa T., Salanti G. PreTA: A network meta-analysis ranking metric measuring the probability of being preferable than the average treatment. Res Synth Methods (submitted, under review)

Nikolakopoulou A., Higgins J.P.T., Papakonstantinou T., Chaimani A., Del Giovane C., Egger M., Salanti G. CINeMA: An approach for assessing confidence in the results of network meta-analysis PLoS Med 2020 Apr; 17(4)

Salanti G. On ranking multiple health interventions. Deutsche Arbeitsgemeinschaft Statistic (DAGStat) Conference 2019, March 2019, Munich, Germany

Nikolakopoulou A. What works best? Methods for ranking competing treatments in network meta-analysis. Joint International Society for Clinical Biostatistics (ISCB) and Australian Statistical Conference, August 2018, Melbourne, Australia

Chiocchia V., Nikolakopoulou A., Papakonstantinou T., Egger M., Salanti G. Empirical evaluation of ranking metrics in network meta-analysis. XXXl Conference of the Austro-Swiss Region [ROeS] of the lnternational Biometric Society (IBS), 9-12th September 2019, Lausanne, Switzerland

Chiocchia V. Methods for ranking competing treatments in network meta-analysis. 11th Symposium Graduate School for Health Sciences 2019, 19-20th November 2019, Thun, Switzerland (poster)

Nikolakopoulou A, Mavridis D, Salanti G. Relative treatment effects against the ‘average treatment’ using an alternative parameterisation of the network meta-analysis model. Deutsche Arbeitsgemeinschaft Statistic (DAGStat) Conference 2019, March 2019, Munich, Germany (poster)

Papakonstantinou T, Nikolakopoulou A, Rücker G, Schwarzer G, Chaimani A, Egger M, Salanti G. Using flow decomposition to estimate the contribution of studies in network meta-analysis. Deutsche Arbeitsgemeinschaft Statistic (DAGStat) Conference, March 2019, Munich

Salanti G, Nikolakopoulou A, Mavridis D. On ranking multiple health interventions. Deutsche Arbeitsgemeinschaft Statistic (DAGStat) Conference, March 2019, Munich

Nikolakopoulou A, Trelle S, Egger M, Salanti G. The emerging evidence synthesis tools: Actively Living Network Meta-Analysis. 25th Cochrane Colloquium, September 2018, Edinburgh, UK.

Papakonstantinou T, Nikolakopoulou A, Rücker G, Schwarzer G, Chaimani A, Egger M, Salanti G. Using flow to estimate the percentage contribution of studies in network meta-analysis. 25th Cochrane Colloquium, September 2018, Edinburgh, UK (poster)

Nikolakopoulou A, Mavridis D, Salanti G. What works best? Methods for ranking competing treatments in network meta-analysis. Joint International Society for Clinical Biostatistics (ISCB) and Australian Statistical Conference, August 2018, Melbourne

Papakonstantinou T, Nikolakopoulou A, Rücker G, Chaimani A, Schwarzer G, Egger M, Salanti G. Using flow to estimate the percentage contribution of studies in network meta-analysis. Joint International Society for Clinical Biostatistics (ISCB) and Australian Statistical Conference, August 2018, Melbourne

Chiocchia V. “Methods for ranking competing treatments in network meta-analysis.”: Prize for best poster at the 11th Symposium Graduate School for Health Sciences 2019, 19-20th November 2019, Thun, Switzerland