Exact Inference for Random Effects Meta-Analyses for Small, Sparse Data

Area of Science:

• Biostatistics
• Pharmaceutical Research
• Clinical Trials

Background:

• Meta-analysis is crucial for drug safety and efficacy assessment.
• Classical meta-analysis methods struggle with small study numbers and rare events.
• Existing approaches like study removal or continuity corrections can invalidate results.

Purpose of the Study:

• To develop a novel exact inference method for random effects meta-analysis.
• To address challenges posed by rare events and small sample sizes in meta-analysis.
• To provide a statistically valid approach for meta-analysis in challenging scenarios.

Main Methods:

• Introduced XRRmeta, an exact inference method for random effects meta-analysis.
• Designed for two-sample settings with rare events.
• Validated through extensive numerical simulations.

Main Results:

• XRRmeta provides valid statistical inference for meta-analysis.
• The method performs reliably even with between-study heterogeneity.
• It is effective when event rates, study numbers, or sample sizes are small.
• Numerical studies show XRRmeta does not lead to overly conservative inference.

Conclusions:

• XRRmeta offers a robust solution for meta-analysis of rare events.
• The method enhances the reliability of statistical inference in challenging clinical trial settings.
• An open-source R package is available for applying the XRRmeta method.