scrape_meta: Convert forest plot or summary table to meta-analytic dataset

In MetaUtility: Utility Functions for Conducting and Interpreting Meta-Analyses

Description

Given relative risks (RR) and upper bounds of 95% confidence intervals (CI) from a forest plot or summary table, returns a dataframe ready for meta-analysis (e.g., via the metafor package) with the log-RRs and their variances. Optionally, the user may indicate studies for which the point estimate is to be interpreted as an odds ratios of a common outcome rather than a relative risk; for such studies, the function applies VanderWeele (2017)'s square-root transformation to convert the odds ratio to an approximate risk ratio.

Usage

scrape_meta(type = "RR", est, hi, sqrt = FALSE)

Arguments

type	RR if point estimates are RRs or ORs (to be handled on log scale); raw if point estimates are raw differences, standardized mean differences, etc. (such that they can be handled with no transformations)
est	Vector of study point estimates on RR or OR scale
hi	Vector of upper bounds of 95% CIs on RRs
sqrt	Vector of booleans (TRUE/FALSE) for whether each study measured an odds ratio of a common outcome that should be approximated as a risk ratio via the square-root transformation

References

VanderWeele TJ (2017). On a square-root transformation of the odds ratio for a common outcome. Epidemiology.

MetaUtility documentation built on Oct. 30, 2021, 5:07 p.m.