Modelling of ROC curves in meta-analysis of diagnostic test accuracy studies and network meta-analysis

• Applicant: Dr. Gerta Rücker
• Subject Area: Epidemiology and Medical Biometry/Statistics
• Term: from 2012 to 2020
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 214969570

• Final Report Year: 2020

Final Report Abstract

Meta-analysis of diagnostic accuracy (DTA) studies. We introduced and published a new approach to analyse data from meta-analysis of DTA studies reporting data on sensitivity and specificity of arbitrarily many biomarker threshold values reported in the primary studies. It has been implemented in the open source R package diagmeta which is continuously updated. We applied our approach to the diagnosis of asthma and compare it to competing approaches. At present, I am working with a group led by Antonia Zapf to compare our approach to others in an extensive simulation study. Network meta-analysis. Guido Schwarzer and I continued to extend our R package netmeta. The main methodological developments during the second phase of this project were one-stage methods to analyse binary NMA data using one-stage model such as the Mantel-Haenszel method or GLMMs, methods for component network meta-analysis (CNMA), including methods for disconnected networks, and a measure of the importance of each primary study for each network-based treatment effects estimate. We also wrote a new function netbind() to create flexible forest plots for comparing multiple NMAs, for example several CNMA models. All major advances were implemented in the R package netmeta.

Publications

• (2021) Component network meta-analysis compared to a matching method in a disconnected network: A case study. Biometrical journal. Biometrische Zeitschrift 63 (2) 447–461
  Rücker, Gerta; Schmitz, Susanne; Schwarzer, Guido
  (See online at https://doi.org/10.1002/bimj.201900339)
• (2016). Modelling multiple thresholds in meta-analysis of diagnostic test accuracy studies. BMC Medical Research Methodology, 16:97:97
  Steinhauser, S., Schumacher, M., and Rücker, G.
  (See online at https://doi.org/10.1186/s12874-016-0196-1)
• (2017). Resolve conflicting rankings of outcomes in network meta-analysis: Partial ordering of treatments. Research Synthesis Methods, 8(4):526–536
  Rücker, G. and Schwarzer, G.
  (See online at https://doi.org/10.1002/jrsm.1270)
• (2020). Network metaanalysis of multicomponent interventions. Biometrical Journal, 62:808–821
  Rücker, G., Petropoulou, M., and Schwarzer, G.
  (See online at https://doi.org/10.1002/bimj.201800167)