Epi (version 2.56)

DMlate: The Danish National Diabetes Register.

Description

These two datasets each contain a random sample of 10,000 persons from the Danish National Diabetes Register. DMrand is a random sample from the register, whereas DMlate is a random sample among those with date of diagnosis after 1.1.1995. All dates are radomly jittered by adding a U(-7,7) (days).

Usage

data(DMrand)
       data(DMlate)

Format

A data frame with 10000 observations on the following 7 variables.

sex

Sex, a factor with levels M F

dobth

Date of birth

dodm

Date of inclusion in the register

dodth

Date of death

dooad

Date of 2nd prescription of OAD

doins

Date of 2nd insulin prescription

dox

Date of exit from follow-up.

Details

All dates are given in fractions of years, so 1998.000 corresponds to 1 January 1998 and 1998.997 to 31 December 1998.

All dates are randomly perturbed by a small amount, so no real persons have any of the combinations of the 6 dates in the dataset. But results derived from the data will be quite close to those that would be obtained if the entire 'real' diabetes register were used.

References

B Carstensen, JK Kristensen, P Ottosen and K Borch-Johnsen: The Danish National Diabetes Register: Trends in incidence, prevalence and mortality, Diabetologia, 51, pp 2187--2196, 2008.

In partucular see the appendix at the end of the paper.

Examples

data(DMlate)
str(DMlate)
dml <- Lexis( entry=list(Per=dodm, Age=dodm-dobth, DMdur=0 ),
               exit=list(Per=dox),
        exit.status=factor(!is.na(dodth),labels=c("DM","Dead")),
               data=DMlate )

# Cut the follow-up at insulin start, and introduce a new timescale,
# and split non-precursor states
system.time(
dmi <- cutLexis( dml, cut = dml$doins,
                      pre = "DM",
                new.state = "Ins",
                new.scale = "t.Ins",
             split.states = TRUE ) )
summary( dmi )