Paq810: Filtered Paquid Sample Data Set for Illness-Death Modeling

Paq810R Documentation

Filtered Paquid Sample Data Set for Illness-Death Modeling

Description

A dataset derived from Paq1000 (originally a sample from the Paquid study available via the SmoothHazard package). This version excludes subjects where the age of dementia diagnosis or censoring ('r') was exactly equal to the age at study entry ('e'), ensuring valid observation intervals when using left truncation.

Usage

data(Paq810)

Format

A data frame with approximately 810 rows (original 1000 minus exclusions) and the following 8 columns:

dementia

Dementia status indicator: 0 = non-demented, 1 = demented.

death

Death status indicator: 0 = alive, 1 = dead.

e

Age at entry into the study (left-truncation time).

l

For demented subjects: age at the visit *before* the diagnostic visit. For non-demented subjects: age at the last visit.

r

Age at dementia diagnosis or censoring for dementia (event/censoring time for 0->1 transition). Guaranteed to be > 'e'.

t

Overall exit age. For dead subjects: age at death. For alive subjects: age at the latest news (censoring time for death).

certif

Primary school certificate indicator: 0 = without certificate, 1 = with certificate.

gender

Gender indicator: 0 = female, 1 = male.

Details

This dataset was created by filtering the Paq1000 data: Paq810 <- Paq1000[Paq1000$r > Paq1000$e, ]. This step is necessary to prevent issues with survival::Surv(e, r, dementia) which requires the stop time ('r') to be strictly greater than the start/entry time ('e').

The time variables 'e', 'l', 'r', and 't' are all ages in years. This dataset is suitable for fitting illness-death models with left truncation using functions like IllnessDeath.

Source

Derived from the Paq1000 dataset, which originates from the Paquid study and is included in the SmoothHazard package.

See Also

Paq1000, The SmoothHazard package.


frailtypack documentation built on Nov. 23, 2025, 1:09 a.m.