Basic sensitivity analysis of the observed relative risks adjusting for unmeasured confounding and misclassification of the exposure/outcome, or both. It follows the bias analysis methods and examples from the book by Lash T.L, Fox M.P, and Fink A.K. "Applying Quantitative Bias Analysis to Epidemiologic Data", ('Springer', 2009).
The R package episensr allows to do basic sensitivity analysis of epidemiological results as described in Applying Quantitative Bias Analysis to Epidemiological Data by Timothy L. Lash, Matthew P. Fox, and Aliza K. Fink (ISBN: 978-0-387-87960-4, bias.analysis). A similar function is available in Stata (episens).
This package is free and open source software, licensed under GPL2.
We will use a case-control study by Stang et al. on the relation between mobile phone use and uveal melanoma. The observed odds ratio for the association between regular mobile phone use vs. no mobile phone use with uveal melanoma incidence is 0.71 [95% CI 0.51-0.97]. But there was a substantial difference in participation rates between cases and controls (94% vs 55%, respectively) and so selection bias could have an impact on the association estimate. The 2X2 table for this study is the following:
|regular use||no use|
We use the function
selection as shown below.
library(episensr)selection(matrix(c(136, 107, 297, 165),dimnames = list(c("UM+", "UM-"), c("Mobile+", "Mobile-")),nrow = 2, byrow = TRUE),bias_parms = c(.94, .85, .64, .25))#> --Observed data--#> Outcome: UM+#> Comparing: Mobile+ vs. Mobile-#>#> Mobile+ Mobile-#> UM+ 136 107#> UM- 297 165#>#> 2.5% 97.5%#> Observed Relative Risk: 0.7984287 0.6518303 0.9779975#> Observed Odds Ratio: 0.7061267 0.5143958 0.9693215#> ---#>#> Selection Bias Corrected Relative Risk: 1.483780#> Selection Bias Corrected Odds Ratio: 1.634608
The 2X2 table is provided as a matrix and selection probabilities given with the
bias_parms, a vector with the 4 probabilities (guided by the participation
rates in cases and controls) in the following order: among cases exposed, among
cases unexposed, among noncases exposed, and among noncases unexposed.
The output shows the observed 2X2 table, the observed odds ratio (and relative
risk) followed by the corrected ones.
You can get the latest release from CRAN:
Or install the development version from GitHub with devtools package:
devtools::install_github('dhaine/episensr', ref = "develop")
probsens.conffunction for prevalence of exposure among the non-exposed (as producing NaNs) (#1).
misclassification_covfor a misclassified covariate (confounder or effect modifier).
misclassificationin the previous version is now removed.
Fix bug when building 2-by-2 table.
Various formatting improvements in output of
Standardize use of
Selection bias factor now available in output of
Add bootstrap option
Harmonization of arguments across functions.
New distributions added to
probsens series of functions: constant,
logit-logistic, logit-normal, log-logistic, and log-normal.
Probabilistic analysis of person-time data added with
exposure misclassification, and
probsens.irr.conf for unmeasured confounder.
Sensitivity analysis to correct for selection bias caused by M bias with
mbias function, including DAG plot and print function.
Fix CI formatting.
NAMESPACE: add imports to
stats functions to avoid new R CMD CHECK warnings