Identifying Conditions Favouring Multiplicative Heterogeneity Models in Network Meta-Analysis

2026-01-16

AI Summary

This study addresses the susceptibility of between-study heterogeneity in network meta-analysis to extreme or imprecise observations by empirically comparing the fit and robustness of multiplicative effects models against conventional additive random-effects models. Using two-arm trial data from the nmadb database, model performance was evaluated via weighted least squares estimation and assessed using the Akaike Information Criterion. The findings indicate that, in the presence of substantial heterogeneity, the multiplicative effects model achieves comparable or superior model fit. Moreover, by assigning lower weights to outlying studies, this approach demonstrates greater robustness to publication bias, thereby enhancing the reliability of meta-analytic results.

Technology Category

• Search and Optimization: Metareasoning and Metaheuristics
• Reasoning under Uncertainty: Relational Probabilistic Models
• Machine Learning: Ensemble Methods

Application Category

• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Social Networks and Social Media: Fairness and bias in social network and social media analysis
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Explicit modelling of between-study heterogeneity is essential in network meta-analysis (NMA) to ensure valid inference and avoid overstating precision. While the additive random-effects (RE) model is the conventional approach, the multiplicative-effect (ME) model remains underexplored. The ME model inflates within-study variances by a common factor estimated via weighted least squares, yielding identical point estimates to a fixed-effect model while inflating confidence intervals. We empirically compared RE and ME models across NMAs of two-arm studies with significant heterogeneity from the nmadb database, assessing model fit using the Akaike Information Criterion. The ME model often provided comparable or better fit to the RE model. Case studies further revealed that RE models are sensitive to extreme and imprecise observations, whereas ME models assign less weight to such observations and hence exhibit greater robustness to publication bias. Our results suggest that the ME model warrant consideration alongside conventional RE model in NMA practice.

Problem

• network meta-analysis
• heterogeneity
• multiplicative-effect model
• random-effects model
• publication bias

Innovation

• multiplicative-effect model
• network meta-analysis
• heterogeneity modeling
• robustness to publication bias
• model comparison