← All projects

Research / 2026

Reusable meta-analysis pipelines

R workflows that turn study data into pooled estimates, sensitivity analyses, tables, and figures.

A reusable R pipeline for meta-analysis of proportions, generalized from a diabetic-retinopathy evidence synthesis.

The workflow includes random-effects pooling, subgroup analysis, prediction intervals, meta-regression, leave-one-out sensitivity analyses, and publication-bias diagnostics. Configuration separates each review’s inputs from the analysis code.

It is research tooling rather than a clinical decision system. Individual studies and estimates are not published here.

More to explore

Related work