Description Usage Arguments Details Value Author(s) See Also Examples

From records of drug purchase and possibly known treatment intensity, the time since first drug use and cumulative dose at prespecified times is computed. Optionally, lagged exposures are computed too, i.e. cumulative exposure a prespecified time ago.

1 2 3 4 5 6 7 8 9 10 |

`purchase` |
Data frame with columns |

`id` |
Character. Name of the id variable in the data frame. |

`dop` |
Character. Name of the |

`amt` |
Character. Name of the |

`dpt` |
Character. Name of the |

`fu` |
Data frame with |

`doe` |
Character. Name of the |

`dox` |
Character. Name of the |

`breaks` |
Numerical vector of dates at which the time since first exposure, cumulative dose etc. are computed. |

`use.dpt` |
Logical: should we use information on dose per time. |

`push.max` |
Numerical. How much can purchases maximally be pushed forward in time. See details. |

`rm.dose` |
Logical. Should the dose from omitted period of
exposure (due to the setting of |

`lags` |
Numerical vector of lag-times used in computing lagged cumulative doses. |

`lag.dec` |
How many decimals to use in the construction of names for the lagged exposure variables |

`lag.pre` |
Character string used for prefixing names of lagged
exposure variables. Aimed to facilitate the use of |

`pred.win` |
The length of the window used for constructing the
average dose per time used to compute the duration of the last
purchase. Only used when |

The intention of this function is to generate covariates for a particular drug for the entire follow-up of each person. The reason that the follow-up prior to first drug purchase and post-exposure is included is that the covariates must be defined for all follow-up for each person in order to be useful for analysis of disease outcomes.

The functionality is described in terms of calendar time as underlying
time scale, because this will normally be the time scale for drug
purchases and for entry and exit for persons. In principle the
variables termed as dates might equally well refer to say the age
scale, but this would then have to be true *both* for the
purchase data, the follow-up data and the `breaks`

argument.

Drug purchase records (in `purchase`

) are used to construct
measures of drug exposure at prespecified timepoints (in
`breaks`

) in follow-up intervals (in `fu`

). Each person may
have more than one follow-up interval. They should be disjoint, but
this is not checked.

If `use.dpt`

is `TRUE`

then the dose per time information is
used to compute the exposure interval associated with each purchase.
Exposure intervals are stacked, that is each interval is put after any
previous. This means that the start of exposure to a given purchase
can be pushed into the future. The parameter `push.max`

indicates
the maximally tolerated push. If this is reached by a person, the
assumption is that some of the purchased drug may not be counted in
the exposure calculations — see `rm.dose`

.

The `dpt`

can either be a constant, basically translating each
purchased amount into exposure time the same way for all persons, or
it can be a vector with different treatment intensities for each
purchase. In any case the cumulative dose is computed taking
`dpt`

into account, unless `rm.dose`

is `FALSE`

in
which case the actual purchased amount is cumulated. The latter is
slightly counter-intuitive because we are using the `dpt`

to push
the intervals, and then disregard it when computing the cumulative
dose. The counter argument is that if the limit `push.max`

is
reached, the actual dosage may be larger than indicated the
`dpt`

, and is essentially what this allows for.

If `use.dpt`

is `FALSE`

then the exposure from one purchase
is assumed to stretch over the time to the next purchase, so we are
effectively allowing different dosing rates (dose per time) between
purchases. Formally this approach conditions on the future, because
the rate of consumption (the accumulation of cumulative exposure) is
computed based on knowledge of when next purchase is made. Moreover,
with this approach, periods of non-exposure does not exist, except
after the last purchase where the future consumption rate is taken to
be the average over the period of use (or a period of length
`pred.win`

), and hence defines a date of cessation of drug.

Finally, if `use.dpt`

is `FALSE`

, at least two purchase
records are required to compute the measures. Therefore persons with
only one drug purchase record are ignored in calculations.

A data frame with one record per person and follow-up date
(`breaks`

). Date of entry and date of exit are included too; but
only follow-up in the intersetion of `range(breaks)`

and
`range(fu$doe,fu$dox)`

is output.

`id`

person id.

`dof`

date of follow up, i.e. start of interval. Apart from possibly the first interval for each person, this will assume values in the set of the values in

`breaks`

. All other variables refer to status as of this date.`dur`

the length (

`dur`

ation) of interval.`tfi`

`t`

ime`f`

rom first`i`

nitiation of drug.`off`

Logical, indicating whether the person is

`off`

drug. So it is`FALSE`

if the person is exposed at`dof`

.`doff`

`d`

ate of latest transition to`off`

drug. Note that tis defined also at dates after drug exposure has been resumed.`tfc`

`t`

ime`f`

rom latest`c`

essation of drug.`ctim`

`c`

umulative`tim`

e on the drug.`cdos`

`c`

umulative`dos`

e.`ldos`

suffixed with one value per element in

`lags`

, the latter giving the cumulative doses`lags`

before`dof`

.

Bendix Carstensen, b@bxc.dk. The development of this function was supported partly through a grant from the EFSD (European Foundation for the Study of Diabetes)

`Lexis`

,
`cutLexis`

,
`mcutLexis`

,
`addCov.Lexis`

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 | ```
# Example data for drug purchases in 3 persons --- dates (dop) are
# measured in years, amount purchased (amt) in no. pills and dose per
# time (dpt) consequently given in units of pills/year. Note we also
# include a person (id=4) with one purchase record only.
n <- c( 10, 18, 8, 1 )
hole <- rep(0,n[2])
hole[10] <- 2 # to create a hole of 2 years in purchase dates
# dates of drug purchase
dop <- c( 1995.278+cumsum(sample(1:4/10,n[1],replace=TRUE)),
1992.351+cumsum(sample(1:4/10,n[2],replace=TRUE)+hole),
1997.320+cumsum(sample(1:4/10,n[3],replace=TRUE)),
1996.470 )
# purchased amounts mesured in no. pills
amt <- sample( 1:3*50 , sum(n), replace=TRUE )
# prescribed dosage therefore necessarily as pills per year
dpt <- sample( 4:1*365, sum(n), replace=TRUE )
# collect to purchase data frame
dfr <- data.frame( id = rep(1:4,n),
dop,
amt = amt,
dpt = dpt )
head( dfr, 3 )
# a simple dataframe for follow-up periods for these 4 persons
fu <- data.frame( id = 1:4,
doe = c(1995,1992,1996,1997)+1:4/4,
dox = c(2001,2003,2002,2010)+1:4/5 )
fu
# Note that the following use of gen.exp relies on the fact that the
# purchase dataframe dfr has variable names "id", "dop", "amt" and
# "dpt"" and the follow-up data frame fu has variable names "id",
# "doe" and "dox"
# 1: using the dosage information
dposx <- gen.exp( dfr,
fu = fu,
use.dpt = TRUE,
breaks = seq(1990,2015,0.5),
lags = 2:4/4,
lag.pre = "l_" )
format( dposx, digits=5 )
# 2: ignoring the dosage information,
# hence person 4 with only one purchase is omitted
xposx <- gen.exp( dfr,
fu = fu,
use.dpt = FALSE,
breaks = seq(1990,2015,0.5),
lags = 2:3/5 )
format( xposx, digits=5 )
# It is possible to have disjoint follow-up periods for the same person:
fu <- fu[c(1,2,2,3),]
fu$dox[2] <- 1996.2
fu$doe[3] <- 1998.3
fu
# Note that drug purchase information for the period not at risk *is* used
dposx <- gen.exp( dfr,
fu = fu,
use.dpt = TRUE,
breaks = seq(1990,2015,0.1),
lags = 2:4/4 )
format( dposx, digits=5 )
``` |

Embedding an R snippet on your website

Add the following code to your website.

For more information on customizing the embed code, read Embedding Snippets.