A model for network meta-analysis of rare events

ID: 

346

Session: 

Poster session 3

Date: 

Tuesday 18 September 2018 - 12:30 to 14:00

All authors in correct order:

Efthimiou O1, Rücker G2, Egger M1, Salanti G1
1 Institute of Social and Preventive Medicine (ISPM), University of Bern, Switzerland
2 Institute for Medical Biometry and Statistics, Medical Faculty and Medical Center, University of Freiburg, Germany
Presenting author and contact person

Presenting author:

Orestis Efthimiou

Contact person:

Abstract text
Background:
The Mantel-Haenszel method has been used for decades now to synthesize data obtained from studies that compare two interventions with respect to a binary outcome. It has been shown to perform better than the inverse-variance method or Peto’s odds ratio when data are sparse. Network met-analysis (NMA) is increasingly being used to compare the safety of medical interventions, synthesising, for example, data on mortality or serious adverse events. In this setting sparse data occur often, and yet there is to date no extension of the Mantel-Haenszel method for the case of NMA.

Objectives:
In this work, we present a Mantel-Haenszel NMA model for odds ratios. Similarly to the pairwise Mantel-Haenszel method we provide a model that assumes common treatment effects. We implement our approach in R, and we provide easy-to-use routines. We illustrate our model using data from two previously published networks. We compare our results to those obtained with three other NMA models: an inverse-variance NMA, an NMA model with non-central hypergeometric likelihood, and a Bayesian NMA with a binomial likelihood.

Conclusions:
Our Mantel-Haenszel NMA offers a reliable approach to network meta-analysis of sparse data, especially when there are studies in the dataset with zero events, and when the assumption of homogeneity is justifiable.

Patient or healthcare consumer involvement:
Not applicable.

Relevance to patients and consumers: 

Not applicable