A Markov Chain approach for ranking treatments in network meta-analysis

Chaimani, Anna; Porcher, Raphaël; Sbidian, É.; Mavridis, Dimitris

doi:10.1101/19008722

Cited by 3 publications

(4 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ranking methods have attracted considerable interest in recent years 14,16,17,18,19,20 . It is generally recognised that the accuracy of ranking statistics and the treatment effect estimates are likely to be affected by the geometry of the network of treatments and trials 21,22,17,23,5 .…”

Section: Introductionmentioning

confidence: 99%

Degree irregularity and rank probability bias in network meta-analysis

Galla

2020

Preprint

View full text Add to dashboard Cite

Network meta-analysis (NMA) is a statistical technique for the comparison of treatment options. The nodes of the network are the competing treatments and edges represent comparisons of treatments in trials. Outcomes of Bayesian NMA include estimates of treatment effects, and the probabilities that each treatment is ranked best, second best and so on. How exactly network geometry affects the accuracy and precision of these outcomes is not fully understood. Here we carry out a simulation study and find that disparity in the number of trials involving different treatments leads to a systematic bias in estimated rank probabilities. This bias is associated with an increased variation in the precision of treatment effect estimates. Using ideas from the theory of complex networks, 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 precise treatment effect estimates and smaller bias of rank probabilities. We also find that degree regularity is a better indicator of NMA quality than both the total number of studies in a network and the disparity in the number of trials per comparison. These results have implications for planning future trials. We demonstrate that choosing trials which reduce the network's irregularity can improve the precision and accuracy of NMA outcomes. KEYWORDS:network meta-analysis, simulation study, rank probability, network geometry and degree irregularity, planning future trials arXiv:2003.07662v1 [stat.ME]

show abstract

Section: Introductionmentioning

confidence: 99%

Degree irregularity and rank probability bias in network meta-analysis

Galla

2020

Preprint

View full text Add to dashboard Cite

show abstract

“…Ranking methods have attracted considerable interest in recent years 15,17‐21 . It is generally recognised that the accuracy of ranking statistics and the treatment effect estimates are likely to be affected by the topology of the network of treatments and trials 5,18,22‐24 .…”

Section: Introductionmentioning

confidence: 99%

“…16 Ranking methods have attracted considerable interest in recent years. 15,[17][18][19][20][21] It is generally recognised that the accuracy of ranking statistics and the treatment effect estimates are likely to be affected by the topology of the network of treatments and trials. 5,18,[22][23][24] PRISMA guidelines ("Preferred Reporting Items for Systematic Reviews and Meta-Analyses") therefore recommend that authors provide graphical and qualitative descriptions of network geometry.…”

Section: Introductionmentioning

confidence: 99%

Degree irregularity and rank probability bias in network meta‐analysis

Galla

2020

Research Synthesis Methods

View full text Add to dashboard Cite

Network meta‐analysis (NMA) is a statistical technique for the comparison of treatment options. Outcomes of Bayesian NMA include estimates of treatment effects, and the probabilities that each treatment is ranked best, second best and so on. How exactly network topology affects the accuracy and precision of these outcomes is not fully understood. Here we carry out a simulation study and find that disparity in the number of trials involving different treatments leads to a systematic bias in estimated rank probabilities. This bias is associated with an increased variation in the precision of treatment effect estimates. Using ideas from the theory of complex networks, 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 precise treatment effect estimates and smaller bias of rank probabilities. Conversely, these topological effects are not observed for the accuracy of treatment effect estimates. This reinforces the importance of taking into account multiple measures, rather than making decisions based on a single metric. We also find that degree regularity is a better indicator for the accuracy and precision of parameter estimates in NMA than both the total number of studies in a network and the disparity in the number of trials per comparison. These results have implications for planning future trials. We demonstrate that choosing trials which reduce the network's irregularity can improve the precision and accuracy of parameter estimates from NMA.

show abstract

“… 19 The Probability of Selecting a Treatment to Recommend incorporates important information such as the confidence in the evidence or clinical priors in the ranking algorithm. 20 A first approach to evaluate the confidence in rankings from NMA was described by Salanti et al but it has not yet been implemented into a proper framework like CINeMA. 3 21 The aim to create evidence-based guidelines also inspired the threshold analysis approach, which is not a new ranking method per se, but it informs on the robustness of treatment recommendations by quantifying how much the evidence could change before the ranking of the treatments changes.…”

mentioning

confidence: 99%