sample of 20 clusters with an average cluster size of 20

data

A user friendly function 'crrcbcv' to compute bias-corrected variances for competing risks regression models using proportional subdistribution hazards with small-sample clustered data. Four types of bias correction are included: the MD-type bias correction by Mancl and DeRouen (2001) <doi:10.1111/j.0006-341X.2001.00126.x>, the KC-type bias correction by Kauermann and Carroll (2001) <doi:10.1198/016214501753382309>, the FG-type bias correction by Fay and Graubard (2001) <doi:10.1111/j.0006-341X.2001.01198.x>, and the MBN-type bias correction by Morel, Bokossa, and Neerchal (2003) <doi:10.1002/bimj.200390021>.

Xinyuan Chen

crrcbcv

Bias-Corrected Variance for Competing Risks Regression with
Clustered Data

Fan Li

cls function

A data frame containing 20 clusters with an average cluster size of 20 and the following five variables.
Simulation is detailed in the paper Finite-sample adjustments in variance estimators for
clustered competing risks regression. Chen, Li. 2022. Under Review. Statistics in Medicine.<dl class="dl-horizontal">
<dt><code>I</code></dt><dd>id of clusters</dd>
<dt><code>X_1</code></dt><dd>a cluster-level covariate generated from the standard normal distribution</dd>
<dt><code>X_2</code></dt><dd>an individual-level covariate generated from the standard normal distribution</dd>
<dt><code>eps</code></dt><dd>event type. 0=censored, 1, 2</dd>
<dt><code>T_obs</code></dt><dd>observed event time</dd>
</dl>

Format

Clustered competing risks simulated data — cls

A data frame containing 20 clusters with an average cluster size of 20 and the following five variables.
Simulation is detailed in the paper Finite-sample adjustments in variance estimators for
clustered competing risks regression. Chen, Li. 2022. Under Review. Statistics in Medicine.<dl class='dl-horizontal'>
<dt><code>I</code></dt><dd>id of clusters</dd>
<dt><code>X_1</code></dt><dd>a cluster-level covariate generated from the standard normal distribution</dd>
<dt><code>X_2</code></dt><dd>an individual-level covariate generated from the standard normal distribution</dd>
<dt><code>eps</code></dt><dd>event type. 0=censored, 1, 2</dd>
<dt><code>T_obs</code></dt><dd>observed event time</dd>
</dl>

cls: Clustered competing risks simulated data

Description

Usage

Arguments

Format

References

Examples