get_relative_risk

The input <code>data.frame</code>. Ideally this should be the output <code>recurrence_data</code> from the
<code><a rd-options="autoCovariateSelection" href="/link/get_recurrence_covariates?package=autoCovariateSelection&version=1.0.0&to=autoCovariateSelection" data-mini-rdoc="autoCovariateSelection::get_recurrence_covariates">get_recurrence_covariates</a></code> function. The first column should be the patient identifier column and
all other columns should be binary covariates. The values of these binary columns should be 1 indicating occurrence of covariate and 0 indicating
no occurrence of the covariate.

The 1-D outcome vector indicating whether or not the patient experienced the outcome of interest (value = 1) or not (value =0).
The length of this vector should be equal to the number of rows of <code>df</code>. The order of elements in this vector should resonate with
the order of patients in <code>df</code>

outcomeVec

<code>get_relative_risk</code> function is a helper function used within the <code><a rd-options="autoCovariateSelection" href="/link/get_prioritised_covariates?package=autoCovariateSelection&version=1.0.0&to=autoCovariateSelection" data-mini-rdoc="autoCovariateSelection::get_prioritised_covariates">get_prioritised_covariates</a></code> function.
This function computes the prevalence in the exposed and that in the unexposed and simply returns the relative risk for all the covariates in the
input <code>data.frame</code>

Contains functions to implement automated covariate selection using methods described in the
high-dimensional propensity score (HDPS) algorithm by Schneeweiss et.al. Covariate adjustment in real-world-observational-data (RWD) is important for
for estimating adjusted outcomes and this can be done by using methods such as, but not limited to, propensity score
matching, propensity score weighting and regression analysis. While these methods strive to statistically adjust for
confounding, the major challenge is in selecting the potential covariates that can bias the outcomes comparison estimates
in observational RWD (Real-World-Data). This is where the utility of automated covariate selection comes in.
The functions in this package help to implement the three major steps of automated covariate selection as described by
Schneeweiss et. al elsewhere. These three functions, in order of the steps required to execute automated covariate
selection are, get_candidate_covariates(), get_recurrence_covariates() and get_prioritised_covariates().
In addition to these functions, a sample real-world-data from publicly available de-identified medical claims data is
also available for running examples and also for further exploration. The original article where the algorithm is described
by Schneeweiss et.al. (2009) <doi:10.1097/EDE.0b013e3181a663cc> .

Dennis Robert

autoCovariateSelection

Automated Covariate Selection Using HDPS Algorithm

get_relative_risk function

The input <code>data.frame</code>. Ideally this should be the output <code>recurrence_data</code> from the
<code><a rd-options='autoCovariateSelection' href='get_recurrence_covariates'>get_recurrence_covariates</a></code> function. The first column should be the patient identifier column and
all other columns should be binary covariates. The values of these binary columns should be 1 indicating occurrence of covariate and 0 indicating
no occurrence of the covariate.

<code>get_relative_risk</code> function is a helper function used within the <code><a rd-options='autoCovariateSelection' href='get_prioritised_covariates'>get_prioritised_covariates</a></code> function.
This function computes the prevalence in the exposed and that in the unexposed and simply returns the relative risk for all the covariates in the
input <code>data.frame</code>

get_relative_risk: Compute relative risk for each of the covariates with respect to outcomes occurred

Description

Usage

Arguments

Value