Using conditional power of network meta-analysis (NMA) to inform the design of future clinical trials

Area of Science:

• Biostatistics
• Clinical Epidemiology
• Health Research Methodology

Background:

• Clinical trials aim to test intervention effectiveness but require context from existing evidence.
• Network meta-analyses (NMA) integrate data from multiple trials with competing interventions.
• Planning future trials can be enhanced by understanding the existing evidence network.

Purpose of the Study:

• To present a methodology for evaluating the impact of new studies on the conditional power (CP) of updated network meta-analyses.
• To extend the concept of CP from pairwise meta-analysis to network meta-analysis.
• To provide a framework for estimating required sample sizes for future trials within a network.

Main Methods:

• Developed a methodology to assess how new studies, their information size, comparisons, and heterogeneity affect CP in NMA.
• Extended conditional power calculations from pairwise meta-analysis to network settings.
• Estimated required sample sizes using direct and indirect evidence for future trials.

Main Results:

• Conditional power for treatment comparisons in NMA is influenced by heterogeneity and the balance of direct/indirect evidence.
• The methodology was applied to two published networks, demonstrating its practical utility.
• New studies' characteristics significantly impact the CP of the updated network meta-analysis.

Conclusions:

• The proposed methodology aids investigators in calculating sample sizes under varying heterogeneity assumptions.
• It supports informed decision-making regarding the number and design of future studies in complex evidence networks.
• This approach enhances the efficiency and relevance of clinical trial planning within the context of existing evidence.