shift_table: Create a shift table

In flextable: Functions for Tabular Reporting

Create a shift table

Description

Create a shift table ready to be used with tabulator().

The function is transforming a dataset representing some 'Laboratory Tests Results' structured as CDISC clinical trial data sets format to a dataset representing the shift table.

Shift tables are tables used in clinical trial analysis. They show the progression of change from the baseline, with the progression often being along time; the number of subjects is displayed in different range (e.g. low, normal, or high) at baseline and at selected time points or intervals.

Usage

shift_table(
  x,
  cn_visit = "VISIT",
  cn_visit_num = "VISITNUM",
  cn_grade = "LBNRIND",
  cn_usubjid = "USUBJID",
  cn_lab_cat = NA_character_,
  cn_is_baseline = "LBBLFL",
  baseline_identifier = "Y",
  cn_treatment = NA_character_,
  grade_levels = c("LOW", "NORMAL", "HIGH"),
  grade_labels = c("Low", "Normal", "High")
)

Arguments

x	Laboratory Tests Results data frame.
cn_visit	column name containing visit names, default to "VISIT".
cn_visit_num	column name containing visit numbers, default to "VISITNUM".
cn_grade	column name containing reference range indicators, default to "LBNRIND".
cn_usubjid	column name containing unique subject inditifiers, default to "USUBJID".
cn_lab_cat	column name containing lab tests or examination names, default to "LBTEST".
cn_is_baseline	column name containing baseline flags, default to "LBBLFL".
baseline_identifier	baseline flag value to use for baseline identification. Its default is "Y".
cn_treatment	column name containing treatment names, default to NA.
grade_levels	levels to use for reference range indicators
grade_labels	labels to use for reference range indicators

Value

the shift table as a data.frame. Additionnal elements are provided in attributes:

• "VISIT_N": count of unique subject id per visits, labs and eventually treatments. This element is supposed to be used as value for argument hidden_data of function tabulator().

• "FUN_VISIT": a utility function to easily turn visit column as a factor column. It should be applied after the shift table creation.

• "FUN_GRADE": a utility function to easily turn grade column as a factor column. It adds "MISSING/Missing" and "SUM/Sum" at the end of the set of values specified in arguments grade_levels and grade_labels. It should be applied after the shift table creation.

Examples

library(data.table)
library(flextable)

# data simulation ----
USUBJID <- sprintf("01-ABC-%04.0f", 1:200)
VISITS <- c("SCREENING 1", "WEEK 2", "MONTH 3")
LBTEST <- c("Albumin", "Sodium")

VISITNUM <- seq_along(VISITS)
LBBLFL <- rep(NA_character_, length(VISITNUM))
LBBLFL[1] <- "Y"

VISIT <- data.frame(
  VISIT = VISITS, VISITNUM = VISITNUM,
  LBBLFL = LBBLFL, stringsAsFactors = FALSE
)
labdata <- expand.grid(
  USUBJID = USUBJID, LBTEST = LBTEST,
  VISITNUM = VISITNUM,
  stringsAsFactors = FALSE
)
setDT(labdata)

labdata <- merge(labdata, VISIT, by = "VISITNUM")

subject_elts <- unique(labdata[, .SD, .SDcols = "USUBJID"])
subject_elts <- unique(subject_elts)
subject_elts[, c("TREAT") := list(
  sample(x = c("Treatment", "Placebo"), size = .N, replace = TRUE)
)]
subject_elts[, c("TREAT") := list(
  factor(.SD$TREAT, levels = c("Treatment", "Placebo"))
)]
setDF(subject_elts)
labdata <- merge(labdata, subject_elts,
  by = "USUBJID", all.x = TRUE, all.y = FALSE
)
labdata[, c("LBNRIND") := list(
  sample(
    x = c("LOW", "NORMAL", "HIGH"), size = .N,
    replace = TRUE, prob = c(.03, .9, .07)
  )
)]

setDF(labdata)




# shift table calculation ----

SHIFT_TABLE <- shift_table(
  x = labdata, cn_visit = "VISIT",
  cn_grade = "LBNRIND",
  cn_usubjid = "USUBJID",
  cn_lab_cat = "LBTEST",
  cn_treatment = "TREAT",
  cn_is_baseline = "LBBLFL",
  baseline_identifier = "Y",
  grade_levels = c("LOW", "NORMAL", "HIGH")
)

# get attrs for post treatment ----
SHIFT_TABLE_VISIT <- attr(SHIFT_TABLE, "VISIT_N")
visit_as_factor <- attr(SHIFT_TABLE, "FUN_VISIT")
range_as_factor <- attr(SHIFT_TABLE, "FUN_GRADE")

# post treatments ----
SHIFT_TABLE$VISIT <- visit_as_factor(SHIFT_TABLE$VISIT)
SHIFT_TABLE$BASELINE <- range_as_factor(SHIFT_TABLE$BASELINE)
SHIFT_TABLE$LBNRIND <- range_as_factor(SHIFT_TABLE$LBNRIND)

SHIFT_TABLE_VISIT$VISIT <- visit_as_factor(SHIFT_TABLE_VISIT$VISIT)

# tabulator ----

my_format <- function(z) {
  formatC(z * 100,
    digits = 1, format = "f",
    flag = "0", width = 4
  )
}

tab <- tabulator(
  x = SHIFT_TABLE,
  hidden_data = SHIFT_TABLE_VISIT,
  row_compose = list(
    VISIT = as_paragraph(VISIT, "\n(N=", N_VISIT, ")")
  ),
  rows = c("LBTEST", "VISIT", "BASELINE"),
  columns = c("TREAT", "LBNRIND"),
  `n` = as_paragraph(N),
  `%` = as_paragraph(as_chunk(PCT, formatter = my_format))
)

# as_flextable ----

ft_1 <- as_flextable(
  x = tab, separate_with = "VISIT",
  label_rows = c(
    LBTEST = "Lab Test", VISIT = "Visit",
    BASELINE = "Reference Range Indicator"
  )
)

ft_1

flextable documentation built on Oct. 30, 2024, 9:15 a.m.