progdismodel {changeLOS}R Documentation

progressive disability model

Description

Change in LOS and impact of an intermediate event on mortality can also be investigated in a so-called progressive disability model. This multi-state model can be described and the Aalen-Johansen estimator for transition probabilities can be computed. Further will be computed the mortality, attributable mortality and the population attributable mortality.

Usage

progdismodel(model,observ, max.time)

Arguments

observ a data.frame of the form data.frame(id, from, to, time, oid):
id:
id (patient id, admision id)
from:
the state from where a transition occurs
to:
the state to which a transition occurs
time:
time of the transition
oid:
the observation id
model an object of class 'msmodel', which describes the progressive disability model with two transient and four absorbing states.
max.time the last time point of interest

Details

mortality: P( death, t ), P(death | risk factor absent,t) and P(death | risk factor present,t).

attributable mortality: P(death | risk factor present,t) - P(death | risk factor absent,t).

population attributable mortality: (P(death,t) - P(death | risk factor absent,t))/P(death,t).

Value

A list of

trans an object of class trans
aj an object of class aj
times.par the time points for which the following values are computed.
PAR population attributable mortality
AR attributable mortality
death P(death,t)
death.given.rfa P(death | risk factor absent,t)
death.given.rfp P(death | risk factor present,t)

Author(s)

Matthias Wangler mw@imbi.uni-freiburg.de

Examples

data(los.data)

p <- prepare.progdismodel(data=los.data)

pdm <- progdismodel(p$model, p$observ)

[Package changeLOS version 2.0.9-2 Index]