Development of a combined database for meta-epidemiological research

Area of Science:

• Meta-epidemiology
• Biostatistics
• Clinical Trial Methodology

Background:

• Meta-epidemiological studies examine trial characteristics' impact on intervention effects.
• Inconsistent methods and findings across studies hinder reliable conclusions.
• A harmonized database is needed to consolidate and analyze overlapping meta-analyses.

Purpose of the Study:

• To combine data from 10 meta-epidemiological studies into a single, harmonized database.
• To derive a unique dataset by removing overlapping meta-analyses and duplicate trial results.
• To establish a foundation for examining combined evidence on sources of bias in randomized controlled trials.

Main Methods:

• Developed a database design to manage overlapping trials, meta-analyses, and systematic reviews.
• Assigned unique identifiers to references for duplicate trial identification.
• Identified and removed sets of meta-analyses with overlapping trials and duplicated trial results.

Main Results:

• The initial combined database included 427 reviews, 454 meta-analyses, and 4874 trial results.
• After deduplication, 258 meta-analyses were unique, with 196 showing trial overlap.
• Reliability assessments showed median kappa statistics of 0.60 for sequence generation, 0.58 for allocation concealment, and 0.87 for blinding.
• The final database contains 363 meta-analyses and 3477 unique trial results.

Conclusions:

• A strategy for removing overlap between meta-analyses was successfully implemented.
• The harmonized database provides a robust resource for future empirical research on bias in randomized controlled trials.
• The developed methodology for data harmonization may be valuable for future meta-epidemiological research.