get_table_history_est {LTASR}

Stratify Person Table with Time Varying Co-variate

Description

get_table_history_est reads in a data.frame/tibble (persondf) containing basic demographic information for each person of the cohort as well as a data.frame/tibble (historydf) containing time varying exposure information and stratifies the person-time and deaths into 5-year age, 5-year calendar period, race, sex and exposure categories. Additionally, average cumulative exposure values for each strata and each exposure variable are included. These strata are more crudely calculated by taking regular steps (such as every 7 days) as opposed to evaluating every individual day. See Details for information on how the person file and history file must be formatted.

Usage

get_table_history_est(
  persondf,
  rateobj,
  historydf,
  exps,
  strata = dplyr::vars(),
  step = 7,
  batch_size = 25 * step
)

Arguments

persondf

data.frame like object containing one row per person with the required demographic information.

rateobj

a rate object created by the parseRate function, or the included rate object us_119ucod_19602021.

historydf

data.frame like object containing one row per person and exposure period. An exposure period is a period of time where exposure levels remain constant. See Details for required variables.

exps

a list containing exp_strata objects created by exp_strata().

strata

any additional variables contained in persondf on which to stratify. Must be wrapped in a vars() call from dplyr.

step

numeric defining number of days to jump when calculating cumulative exposure values. Exact stratification specifies a step of 1 day.

batch_size

a number specifying how many persons to stratify at a time.

Details

The persondf tibble must contain the variables:

• id,

• gender (character: 'M'/'F'),

• race (character: 'W'/'N'),

• dob (date),

• pybegin (date),

• dlo (date),

• rev (numeric: values 5-10),

• code (character: ICD code)

The historydf tibble must contain the variables:

• id,

• begin_dt (date),

• end_dt (date),

• <daily exposure levels>

Value

A data.frame with a row for each strata containing the number of observed deaths within each of the defined minors/outcomes (⁠_o1⁠-⁠_oxxx⁠) and the number of person days.

Examples

library(LTASR)
library(dplyr)

#Import example person file
person <- person_example %>%
mutate(dob = as.Date(dob, format='%m/%d/%Y'),
         pybegin = as.Date(pybegin, format='%m/%d/%Y'),
         dlo = as.Date(dlo, format='%m/%d/%Y'))

#Import example history file
history <- history_example %>%
  mutate(begin_dt = as.Date(begin_dt, format='%m/%d/%Y'),
         end_dt = as.Date(end_dt, format='%m/%d/%Y'))

#Import default rate object
rateobj <- us_119ucod_19602021
#Define exposure of interest. Create exp_strata object.The `employed` variable
#indicates (0/1) periods of employment and will be summed each day of each exposure
#period. Therefore, this calculates duration of employment in days. The cut-points
#used below will stratify by person-time with less than and greater than a
#year of employment (365 days of employment).
exp1 <- exp_strata(var = 'employed',
                   cutpt = c(-Inf, 365, Inf),
                   lag = 0)

#Stratify cohort by employed variable.
py_table <- get_table_history_est(persondf = person,
                                  rateobj = rateobj,
                                  historydf = history,
                                  exps = list(exp1))

#Multiple exposures can be considered.
exp1 <- exp_strata(var = 'employed',
                   cutpt = c(-Inf, 365, Inf),
                   lag = 0)
exp2 <- exp_strata(var = 'exposure_level',
                   cutpt = c(-Inf, 0, 10000, 20000, Inf),
                   lag = 10)

#Stratify cohort by employed variable.
py_table <- get_table_history_est(persondf = person,
                                  rateobj = rateobj,
                                  historydf = history,
                                  exps = list(exp1, exp2))