Average population attributable fractions are calculated for a set of risk factors (either binary or ordinal valued) for both prospective and case- control designs. Confidence intervals are found by Monte Carlo simulation. The method can be applied to either prospective or case control designs, provided an estimate of disease prevalence is provided. In addition to an exact calculation of AF, an approximate calculation, based on randomly sampling permutations has been implemented to ensure the calculation is computationally tractable when the number of risk factors is large.
|Imports:||MASS (≥ 7.3.0)|
|Author:||John Ferguson [aut, cre]|
|Maintainer:||John Ferguson <john.ferguson at nuigalway.ie>|
|CRAN checks:||averisk results|
|Windows binaries:||r-devel: averisk_1.0.3.zip, r-release: averisk_1.0.3.zip, r-oldrel: averisk_1.0.3.zip|
|OS X El Capitan binaries:||r-release: averisk_1.0.3.tgz|
|OS X Mavericks binaries:||r-oldrel: averisk_1.0.3.tgz|
|Old sources:||averisk archive|
Please use the canonical form https://CRAN.R-project.org/package=averisk to link to this page.