EValue: Sensitivity Analyses for Unmeasured Confounding in Observational Studies and Meta-Analyses

Conducts sensitivity analyses for unmeasured confounding for either an observational study or a meta-analysis of observational studies. For a single observational study, the package reports E-values, defined as the minimum strength of association on the risk ratio scale that an unmeasured confounder would need to have with both the treatment and the outcome to fully explain away a specific treatment-outcome association, conditional on the measured covariates. One can use one of the evalues.XX() functions to compute E-values for the relevant outcome types. Outcome types include risk ratios, odds ratio with common or rare outcomes, hazard ratios with common or rare outcomes, and standardized differences in outcomes. Optionally, one can use the bias_plot() function to plot the bias factor as a function of two sensitivity parameters. (See VanderWeele & Ding, 2017 [<http://annals.org/aim/article/2643434>] for details.) For a meta-analysis, use the function confounded_meta to compute point estimates and inference for: (1) the proportion of studies with true causal effect sizes more extreme than a specified threshold of scientific importance; and (2) the minimum bias factor and confounding strength required to reduce to less than a specified threshold the proportion of studies with true effect sizes of scientifically significant size. The functions sens_plot() and sens_table() create plots and tables for visualizing these meta-analysis metrics across a range of bias values, and scrape_meta helps scrape study-level data from a published forest plot or summary table to obtain the needed estimates when these are not reported. (See Mathur & VanderWeele [<https://osf.io/jkhfg/>] for details.) Most of the analyses available in this package can also be conducted using web-based graphical interfaces (for a single observational study: <https://mmathur.shinyapps.io/evalue/>; for a meta-analysis: <https://mmathur.shinyapps.io/meta_gui_2/>).

Downloads:

Linking:

Please use the canonical form https://CRAN.R-project.org/package=EValue to link to this page.