mtvc {mtvc}

(mtvc) Multiple Time Varying Covariates

Description

Restructure dataset into counting process format to model time varying variables

Usage

mtvc(data, dates, origin = "1970-01-01", start, stop, event, complications)

Arguments

data

Dataframe to be restructured. Has to be in wide format, with a line for each individual.

dates

Name of the columns that contain dates that point out when the variables of interest change value. If an individual does not experience the event of interest, then the respective date should be either a missing value or the origin date.

origin

Day from which the function starts counting days to convert into dates.

start

Date of first contact with the individual (i.e. first medical visit).

stop

Date of death or last visit of the follow-up.

event

Binary variable that indicates if the individual has experienced the event.

complications

Name of the columns that contain values of time varying covariates.

Details

Time varying variables are covariates that might change during the follow-up, so it is fundamental to apply the counting process structure to the data frame of interest, in order to allocate properly the right amount of time that each patient has contributed to the study in each health status.

Value

Dataset in counting process format.

References

1. F. W. Dekker, et al., Survival analysis: time-dependent effects and time-varying risk factors, Kidney International, Volume 74, Issue 8, 2008, Pages 994-997.

Examples

data(simwide)
cp.dataframe=mtvc(data=simwide,
origin='1970-01-01',
dates=c(FIRST_CHRONIC,FIRST_ACUTE,FIRST_RELAPSE),
complications=c(CHRONIC,ACUTE,RELAPSE),
start=DATETRAN,
stop=DLASTSE,
event=EVENT)