mixed.biased

List of <code>ipd</code>, patient-level data, and <code>meta</code>, study-level
data. 5 trials each with 200 observations for each trial. The
Weibull-based model had a -.2 log(hazard) main effect for the centered
covariate <code>x</code> and -.4 interaction effect between the centered
covariate and treatment factor. There was no additional contribution
for the mean of the covariate for main or interaction effect. This
mimics an ecological bias in which there is a patient-level main and
interaction effect but no effects at the study-level.Intended for illustrating use of MLMA with separation of study-level
and patient-level estimation.

Mixed effects IPD meta-analysis, mixed-level meta-analysis
        with patient level data and mulivariate survival estimates for
        aggregate studies

S. Kovalchik

ipdmeta

Tools for Individual patient data and Mixed-level Meta-analysis

mixed.biased function

For <code>ipd</code>:<tabular>ll</tabular>{
<code>group</code> 	Cluster factor, 1 to 10<code>trt</code> 	Treatment indicator, 1 or 0<code>x</code> 	Continuous covariate<code>x.bar</code> 	Mean of x by <code>group</code> <code>time</code> 	Survival time<code>event</code> 	Indicator if event was observed, 1 or 0<code>trt.group</code> 	Factor of treatment and group
}For <code>meta</code>:<tabular>ll</tabular>{
<code>time</code> 	Time corresponding to survival estimate<code>surv</code> 	Survival estimate from treatment.group KM curve<code>sigma2</code> 	Corresponding variance of KM survival estimate<code>group</code> 	Cluster factor, 1 to 10<code>x</code> 	Treatment within study sample mean of covariate<code>x.bar</code> 	Same as x <code>sub.group</code> 	Factor of treatment and group
}

mixed.biased: Simulated Mixed-Level Meta-Analysis Survival Data with Ecological Bias

Description

Arguments

format

See Also