Mixed Treatment Comparison Meta-Analysis of Complex Interventions (Welton component model)
This method helps compare different psychological treatments that have multiple parts (like education, behavior, or support) by analyzing which specific parts make the treatment work better, especially for heart disease patients.
At a glance
Use when
Comparing complex interventions with multiple components; identifying active ingredients of psychological or behavioral programs; conducting network meta-analysis at the component level
Avoid when
Components are highly interdependent or synergistic; insufficient data on component-level breakdown; high heterogeneity in component definitions or delivery
Inputs
Systematic review data on complex interventions with component-level coding; outcome data from randomized controlled trials; network of interventions with shared components
Outputs
Component-level effect estimates; relative efficacy of interventions based on presence/absence of components; rankings of component combinations by effectiveness
How it works
The method uses mixed treatment comparison (MTC) meta-analysis to evaluate complex interventions with multiple components. It enables network meta-analysis across interventions with different combinations of components (e.g., behavioral, cognitive, relaxation), estimating the relative effectiveness of individual components or their combinations by leveraging direct and indirect evidence. Illustrated in a coronary heart disease context, it models component-level effects across multiple outcomes.
- HTA domains
- Clinical Effectiveness, Organisational aspects, Patient and Social Aspects
- Assumptions
- Intervention components act independently or with additive effects; component definitions are consistent across studies; transitivity assumption holds across the network of comparisons
- Strengths
- Enables disentangling of effective components within complex interventions; allows comparison of multi-component interventions even if not directly compared in trials; uses both direct and indirect evidence efficiently
- Limitations
- Relies on accurate coding of intervention components; assumes component additivity which may not reflect real-world interactions; limited by heterogeneity in how components are delivered across studies
- Also known as
- Welton component model, Component-based network meta-analysis
Questions this answers
- › Which specific components of psychological interventions are effective?
- › Are interventions with behavioral components more effective in reducing mortality?
- › Do cognitive components contribute to reducing depression scores?
- › Is the combination of behavioral and cognitive components more effective than single components?
- › How do different intervention components compare in reducing anxiety or cholesterol?
- › What is the relative efficacy of multi-component interventions versus simpler ones?
References & sources
Similar by meaning
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.