Statistical physics approaches to network meta-analysis

Network meta-analysis (NMA) is a statistical technique for the comparison of treatment options in medicine. The nodes of the network graph are the competing treatments and the edges represent comparisons made between treatments in the trials. In this talk I present two projects under the theme of ‘Statistical physics approaches to NMA’. The first is a simulation study that investigates how network topology affects the accuracy and precision of NMA outcomes. We find that disparity in the number of trials involving different treatments leads to a systematic bias in estimated rank probabilities. We define a measure of `degree irregularity' to quantify asymmetry in the number of studies involving each treatment. Our simulations indicate that more regular networks have more accurate and precise parameter estimates. In the second project we present a novel analogy between NMA and random walks (RW). Using the established analogies of both NMA and RW to electrical networks, we are able to provide a RW interpretation of the flow of evidence in NMA. We demonstrate that this approach leads to a closed form derivation of `evidence streams’ that removes the ambiguity of existing algorithms.

 

Zoom: https://us02web.zoom.us/j/83829318876?pwd=Z2pqbUtIMEV3NUQvU0hpakp0NGtsUT09

Meeting ID: 838 2931 8876

Passcode: 797728