Meta-analysis in clinical trials revisited

Area of Science:

• Biostatistics
• Clinical Trials
• Epidemiology

Background:

• Revisiting the seminal 1986 paper introducing the random-effects model for meta-analysis.
• The DerSimonian and Laird method has become a standard approach in medical research due to its simplicity and ease of implementation.
• The method is crucial for summarizing treatment efficacy and characterizing heterogeneity across clinical trials.

Purpose of the Study:

• To review the historical context and evolution of the random-effects meta-analysis model.
• To explore the application and trends of the DerSimonian and Laird method over time.
• To propose a refinement using a robust variance estimator and discuss its repurposing for Big Data and genetic studies.

Main Methods:

• Review of the original random-effects model for meta-analysis.
• Exploration of the method's usage in diverse research settings.
• Introduction of a robust variance estimator for enhanced effect testing.

Main Results:

• The DerSimonian and Laird method is highly cited and widely adopted in clinical research.
• The method effectively provides overall effect estimates and quantifies study heterogeneity.
• A refined approach with robust variance estimation is recommended for improved accuracy.

Conclusions:

• The random-effects meta-analysis model remains a valuable tool in biostatistics and clinical research.
• Refinements can enhance the method's performance, particularly in large-scale data analyses.
• The model shows promise for applications in Big Data meta-analysis and Genome Wide Association Studies.