dtlg: A Performance-Focused Package for Clinical Trial Tables

Description

Create high-performance clinical reporting tables (TLGs) from ADaM-like inputs. The package provides a consistent, programmatic API to generate common tables such as demographics, adverse event incidence, and laboratory summaries, using 'data.table' for fast aggregation over large populations. Functions support flexible target-variable selection, stratification by treatment, and customizable summary statistics, and return tidy, machine-readable results ready to render with downstream table/formatting packages in analysis pipelines.

Author(s)

Maintainer: Ramiro Magno ramiro.morgado@ascent.io

Authors:

• Max Ebenezer-Brown max.ebenezer-brown@ascent.io

• Max Norman max.norman@ascent.io

• Xinye Li xinye.li@ascent.io

• Anja Peebles-Brown anna.peebles-brown@ascent.io

See Also

Useful links:

• https://AscentSoftware.github.io/dtlg/

Generate Core Safety Tables for Clinical Study Reports

Description

AET01_table() produces and combines the main safety summary tables typically found in Section 14.3.1 of a Clinical Study Report (CSR). It calculates patient counts and event totals for deaths, AE-related withdrawals, total AEs, and adverse events of special interest (AESIs).

Usage

AET01_table(
  adsl,
  adae,
  patient_var,
  treat_var,
  aesi_vars,
  aesi_heading = "Total number of patients with at least one",
  indent = "  "
)

Arguments

Value

A merged data.table summarising the main safety outcomes.

Examples

AET01_summary <- AET01_table(
  adsl = adsl,
  adae = aesi,
  patient_var = "USUBJID",
  treat_var = "ARM",
  aesi_vars = c("FATAL", "SER", "SERWD", "SERDSM", "RELSER",
                "WD", "DSM", "REL", "RELWD", "RELDSM", "SEV")
)
print(AET01_summary)

Create AET02-style AE summary table

Description

Summarises adverse events in a format similar to the AET02 table from a CSR, showing total AE counts, patients with AEs, and a breakdown by System Organ Class (SOC) and Preferred Term (PT).

Usage

AET02_table(
  adsl,
  adae,
  patient,
  treat,
  target = "AEDECOD",
  rows_by = "AEBODSYS",
  indent = nbsp(n = 4L)
)

Arguments

Value

A merged data.table containing AE summary.

Examples

# Create a AET02 table
AET02_table(
  adsl = adsl,
  adae = aesi,
  patient = "USUBJID",
  treat = "ARM",
  target = "AEDECOD",
  rows_by = "AEBODSYS",
  indent = "  "
)

Adverse Event Analysis Dataset example dataset

Description

adae is a re-export of the random.cdisc.data::cadae dataset, included in {dtlg} for function usage illustration and testing.

Usage

adae

Format

An object of class tbl_df (inherits from tbl, data.frame) with 1934 rows and 92 columns.

Examples

adae

ADaM Basic Data Structure (BDS) example dataset

Description

adlb is a re-export of the random.cdisc.data::cadlb dataset, included in {dtlg} for function usage illustration and testing.

Usage

adlb

Format

An object of class data.table (inherits from data.frame) with 8400 rows and 102 columns.

Examples

adlb

Subject-Level Analysis Dataset (ADSL) example dataset

Description

adsl is a re-export of the random.cdisc.data::cadsl dataset, included in {dtlg} for function usage illustration and testing.

Usage

adsl

Format

An object of class tbl_df (inherits from tbl, data.frame) with 400 rows and 55 columns.

Examples

adsl

Adverse Events of Special Interest (AESI) example dataset

Description

aesi is a modified version of the random.cdisc.data::cadae dataset, filtered to include only analysis-flagged records (ANL01FL == "Y") and extended with binary indicator variables corresponding to adverse events of special interest (AESIs).

Usage

aesi

Format

A data.frame with a subset of rows from cadae and additional derived columns including:

FATAL

Logical flag for fatal AEs (AESDTH == "Y").

SEV

Logical flag for severe AEs (AESEV == "SEVERE").

SER

Logical flag for serious AEs (AESER == "Y").

SERWD

Serious AE leading to withdrawal (AESER == "Y" & AEACN == "DRUG WITHDRAWN").

SERDSM

Serious AE leading to dose modification/interruption.

RELSER

Serious and related AE.

WD

AE leading to withdrawal.

DSM

AE leading to dose modification/interruption.

REL

Related AE.

RELWD

Related AE leading to withdrawal.

RELDSM

Related AE leading to dose modification/interruption.

Details

These derived flags include seriousness, severity, fatality, relatedness, and treatment consequence (e.g., dose modification or withdrawal), and are used to illustrate key safety summaries in clinical reporting.

Each derived variable is labeled using with_label() for compatibility with tabulation functions.

This dataset is included in {dtlg} to support function testing, usage examples, and reproducible safety analyses.

See Also

random.cdisc.data::cadae, multi_event_true()

Examples

aesi

Convert a TableTree to a dtlg table

Description

as_dtlg_table() reformats a TableTree object into a format close to that of dtlg's data.table.

Usage

as_dtlg_table(tt, .label_col = "stats")

Arguments

Value

A data.table.

Examples

vars <- c('AGE', 'RACE', 'ETHNIC', 'BMRKR1')
var_labels <- c("Age (yr)", "Race", "Ethnicity", "Continuous Level Biomarker 1")

# Summary statistics table split by ARM with custom labels.
(tt <- tern_summary_table(
  adsl,
  target = vars,
  treat = 'ARM',
  target_name = var_labels
))

# Format as a dtlg table
as_dtlg_table(tt)

Calculate counts of a categorical variable

Description

calc_counts() counts observations of a categorical variable (target) by another (treat) and reports summary statistics in clinical trial reporting format.

Usage

calc_counts(
  dt,
  target,
  target_name = target,
  treat,
  indent = nbsp(n = 4L),
  .total_dt = NULL,
  pct_dec = 1
)

Arguments

Value

A list containing a data.table formatted as follows:

• First column is named stats and contains the target variable name indicated in target in the first row. Subsequent rows contain the levels of target.

• Other columns are for the levels of the grouping variable (treat).

• All columns are of character type.

This table is structured for easy integration with Shiny output widgets.

Examples

calc_counts(dt = adsl, "RACE", treat = "ARM", indent = "  ")[[1]]

Calculate descriptive summary statistics for a numeric variable

Description

calc_desc() summarises a numeric variable (target) by another (treat) and reports summary statistics in clinical trial reporting format. The following statistics are calculated for target, per group, i.e. by variable treat levels:

• n: number of observations

• Mean (SD): mean and standard deviation of target

• Median: median of target

• ⁠Min, Max⁠: minimum and maximum of target

• Missing: number of missing target values

Usage

calc_desc(
  dt,
  target,
  target_name = target,
  treat,
  indent = nbsp(n = 4L),
  pct_dec = 1
)

Arguments

Value

A list containing a data.table formatted as follows:

• First column is named stats and contains the target variable name indicated in target in the first row. Subsequent rows contain the summarised statistics labels.

• Other columns are for the levels of the grouping variable (treat).

• All columns are of character type.

This table is structured for easy integration with Shiny output widgets.

Examples

# Calculate summary statistics for the age of the subjects in each region.
calc_stats(dt = adsl, "AGE", treat = "REGION1")[[1]]

# Calculate summary statistics for biomarker 1 in each of the three arms
# (`ARM`).
calc_stats(dt = adsl, "BMRKR1", treat = "ARM")[[1]]

Calculate summary statistics for a variable

Description

calc_stats() calculates summary statistics for a variable on groups. This is a generic function; note that it dispatches based on the class of target (second argument), not dt (first argument).

Usage

calc_stats(
  dt,
  target,
  target_name = target,
  treat,
  indent = nbsp(n = 4L),
  .total_dt = NULL,
  pct_dec = 1
)

Arguments

Value

A data.table of summary statistics. The format depends on the type of the target variable:

• If the target variable is categorical, i.e. type character, factor or logical then the output is that of calc_counts().

• If the target variable is numeric, then the output is that of calc_desc().

Examples

# Calculate summary statistics of a numeric variable, e.g. `AGE`.
calc_stats(dt = adsl, "AGE", treat = "ARM")[[1]]

# Calculate summary statistics of a categorical variable, e.g. `SEX`.
calc_stats(dt = adsl, "SEX", treat = "ARM")[[1]]

Convert a contingency table to a long-format observation-level data frame

Description

cross_tab_to_obsv_tab() expands a contingency table or matrix of counts into a long-format data frame where each row represents one observation. The output contains one column per dimension of the input, with repeated rows according to the frequency counts.

Usage

cross_tab_to_obsv_tab(cross_tab, strings_as_factors = TRUE)

Arguments

Value

A data.frame in long format with one row per implied observation and one column per dimension of the input table.

Examples

dim_names <- list(Sex = c("Male", "Female"),
                  Response = c("Yes", "No"))
cross_tab <- matrix(c(2, 1, 3, 4), nrow = 2, dimnames = dim_names)
cross_tab_to_obsv_tab(cross_tab)

Get or set data.table copy semantics

Description

These functions control how maybe_copy_dt() decides whether to return a data.table by reference (in place) or by value (as a deep copy).

Usage

dt_copy_semantics()

set_dt_copy_semantics(dt_copy_semantics = c("reference", "value"))

Arguments

Details

The copy semantics are stored in the global option dtlg_dt_copy_semantics. The option can take two values:

• "reference" (default): inputs are treated with reference semantics.

  • If the input is already a data.table, it is returned unchanged and aliases are preserved.

  • If the input is a data.frame, it is converted to a data.table in place via data.table::setDT(), mutating the caller’s object.

• "value": inputs are treated with value semantics.

  • The input is converted to a data.table (if necessary) and a deep copy is returned, leaving the original unchanged.

Value

• dt_copy_semantics() returns the current semantics as a string, "reference" or "value".

• set_dt_copy_semantics() sets the semantics, returning the previous semantics invisibly.

See Also

maybe_copy_dt()

Examples

# Get current semantics (defaults to "reference")
dt_copy_semantics()

# Switch to value semantics
old <- set_dt_copy_semantics("value")
dt_copy_semantics()

# Restore previous semantics
set_dt_copy_semantics(old)

Count events

Description

event_count() counts events defined by predicate expressions passed in .filters.

Usage

event_count(
  dt,
  patient,
  treat,
  label,
  .filters = NULL,
  .total_dt = dt,
  pct_dec = 1
)

Arguments

Value

A one-element list, where the element is a data.table.

Examples

# Count deaths per arm.
event_count(
  adsl,
  patient = "USUBJID",
  treat = "ARM",
  label = "Total number of deaths",
  .filters = "DTHFL == 'Y'"
)[[1]]

# Count patients withdraw from study due to an adverse event.
withdrawn_lbl <- "Total number of patients withdrawn from study due to an AE"
event_count(
  adsl,
  patient = "USUBJID",
  treat = "ARM",
  label = withdrawn_lbl,
  .filters = "DCSREAS == 'ADVERSE EVENT'"
)[[1]]

# Count patients with at least one adverse event.
# NB: When `.filters` is `NULL` (i.e., omitted), all records in `dt` are used
# for counting events.
event_count(
  adae,
  patient = "USUBJID",
  treat = "ARM",
  label = "Total number of patients with at least one AE",
  .filters = "ANL01FL == 'Y'",
  .total_dt = adsl
)[[1]]

Summarise adverse events by arm and other grouping variables

Description

event_count_by() creates a tabular summary of adverse events grouped by a higher-level classification variable (e.g., system organ class), and counts both the number of events and the number of unique patients per treatment arm.

Usage

event_count_by(
  dt,
  patient,
  treat,
  rows_by,
  target,
  .total_dt = dt,
  indent = nbsp(n = 4L),
  pct_dec = 1
)

Arguments

Value

A data.table with the following structure:

• One row per combination of rows_by and target

• One row per group total (⁠Total number of events⁠)

• One row per patient-level total (⁠Total number of patients with at least one event⁠)

Columns include:

• stats: character column with labels

• one column per level of the treat variable, formatted as "n (x%)"

See Also

event_count(), calc_stats(), total_events()

Examples

event_count_by(
  dt = adae[adae$ANL01FL == "Y"],
  patient = 'USUBJID',
  treat = 'ARM',
  rows_by = 'AEBODSYS',
  target = 'AEDECOD',
  .total_dt = adsl,
  indent = ' '
)

Format count(s) and percentage(s) (⁠n (pct%)⁠)

Description

format_n_pct() formats counts (n) and respective percentages (pct) as "n (pct%)".

Usage

format_n_pct(n, pct, .pct_digits = 1L)

Arguments

Value

A character vector of strings following the format "n (pct%)", except if n is zero, then the format is simply "0".

Examples

# Simple cases.
format_n_pct(n = 25, pct = 18.66)
format_n_pct(n = 25, pct = 18.66, .pct_digits = 2)

# If both `n` and `pct` are zero then the format is special, i.e. simply "0".
format_n_pct(n = 0, pct = 0)

# `format_n_pct()` is vectorised over `n` and `pct` but their length must
# match.
format_n_pct(n = c(20, 50), pct = c(10, 25))

# Missing values result in empty strings in the output.
format_n_pct(n = c(20, NA), pct = c(10, 25))
format_n_pct(n = c(20, 50), pct = c(10, NA))

Add indentation to strings

Description

indent() prefixes a string with a sequence of HTML non-breaking spaces, to effectively work as indentation.

Usage

indent(x, n = 4L, indentation = nbsp(n = n))

Arguments

Value

A character vector of the same length as x.

Examples

# Default is to indent by four non-breaking spaces.
indent("Mean")

# Choose a different indentation level.
indent("Mean", n = 2L)

# `indent()` is vectorised over `x`
indent(c("Mean", "Median", "Max, Min", "Missing"))

Retrieve the label of an object

Description

label() gets the attached label to an object.

Usage

label(x)

Arguments

Value

The label attribute (string) associated with object passed in x or NULL if the label attribute does not exist.

See Also

with_label()

Examples

label(1)
label(with_label(1, "my label"))

Return a data.table by reference or by value

Description

maybe_copy_dt() returns its input as a data.table, with behaviour controlled by the global copy semantics option dt_copy_semantics().

Usage

maybe_copy_dt(x)

Arguments

Details

• If the semantics are "reference" (default):

  • If x is already a data.table, it is returned unchanged. Aliasing holds, so mutations with ⁠:=⁠ will affect both input and output.

  • If x is a data.frame, it is converted to a data.table in place via data.table::setDT(), mutating the caller’s object. The returned object is a data.table with the same contents. For efficiency, the column vectors are reused without a deep copy.

• If the semantics are "value":

  • x is converted to a data.table (if necessary) and a deep copy is returned. Mutating the result does not affect the input.

Value

A data.table. Whether the return value aliases the input depends on the semantics:

• "reference": input is mutated in place, aliasing guaranteed if x is already a data.table.

• "value": a fresh copy is returned, independent of the input.

See Also

dt_copy_semantics(), set_dt_copy_semantics()

Examples

# Default: reference semantics
df <- data.frame(a = 1:3)
out <- maybe_copy_dt(df)
data.table::is.data.table(df) # TRUE, converted in place

# Switch to value semantics
old <- set_dt_copy_semantics("value")
dt <- data.table::data.table(a = 1:3)
out2 <- maybe_copy_dt(dt)
out2[, b := 99L]
"b" %in% names(dt)  # FALSE, original unchanged

# Restore previous semantics
set_dt_copy_semantics(old)

Merge a list of list-wrapped data.tables into one data.table

Description

This function is typically used to combine multiple reporting tables, each produced by event_count(), total_events(), or multi_event_true(), into a single summary table. These intermediate tables are often returned as one-element lists containing a data.table.

Usage

merge_table_lists(dt_l)

Arguments

Details

This helper unwraps and merges them, row-wise, to produce a consolidated safety report table — commonly used in clinical study reports or data monitoring reviews.

Value

A single merged data.table, row-bound from all input tables.

Examples

# Count deaths by treatment arm
death_table <- event_count(
  adsl,
  patient = "USUBJID",
  treat = "ARM",
  label = "Total number of deaths",
  .filters = "DTHFL == 'Y'"
)

# Count study withdrawals due to adverse events
withdrawal_table <- event_count(
  adsl,
  patient = "USUBJID",
  treat = "ARM",
  label = "Total number of patients withdrawn from study due to an AE",
  .filters = "DCSREAS == 'ADVERSE EVENT'"
)

# Count patients with at least one adverse event
patients_with_ae_table <- event_count(
  adae,
  patient = "USUBJID",
  treat = "ARM",
  label = "Total number of patients with at least one AE"
)

# Count total number of adverse events (not patients)
total_ae_events_table <- total_events(
  dt = adae,
  treat = "ARM",
  label = "Total number of AEs"
)

# Summarise AESIs (e.g., serious, related, severe, etc.)
aesi_vars <- c("FATAL", "SER", "SERWD", "SERDSM", "RELSER",
               "WD", "DSM", "REL", "RELWD", "RELDSM", "SEV")

aesi_table <- multi_event_true(
  dt = aesi,
  event_vars = aesi_vars,
  patient = "USUBJID",
  treat = "ARM",
  heading = "Total number of patients with at least one",
  .total_dt = adsl,
  indent = "  "
)

# Combine all safety tables into a single summary table
safety_summary <- merge_table_lists(list(
  patients_with_ae_table,
  total_ae_events_table,
  death_table,
  withdrawal_table,
  aesi_table
))

safety_summary

Summarise multiple AESI-like events per treatment arm

Description

multi_event_true() generates a summary table showing the number and percentage of patients with at least one event across multiple binary indicator variables (e.g., flags for adverse events of special interest).

Usage

multi_event_true(
  dt,
  event_vars,
  patient,
  treat,
  heading,
  label = NULL,
  .total_dt = NULL,
  indent = nbsp(n = 4L),
  pct_dec = 1
)

Arguments

Details

Each event is counted only once per patient. This function is typically used for summarising Adverse Events of Special Interest (AESIs) or other derived flags (e.g., SER, FATAL, RELDSM) that are binary (TRUE/FALSE).

Value

A one-element list containing a data.table with one row per event plus one header row. The first column is "stats" (row labels), and subsequent columns are one per treatment arm, with values in "n (x%)" format.

See Also

event_count(), total_events()

Examples

aesi_vars <- c(
  "FATAL", "SER", "SERWD", "SERDSM", "RELSER",
  "WD", "DSM", "REL", "RELWD", "RELDSM", "SEV"
)

heading <- "Total number of patients with at least one AE"

multi_event_true(
  dt = aesi,
  event_vars = aesi_vars,
  patient = "USUBJID",
  treat = "ARM",
  heading = heading,
  .total_dt = adsl,
  indent = "  "
)[[1]]

Generate Non-Breaking Spaces for HTML Output

Description

nbsp() generates a string of HTML non-breaking spaces (⁠ ⁠).

Usage

nbsp(n = 1L)

Arguments

Value

A character string containing n HTML non-breaking spaces (⁠ ⁠).

Examples

# One non-breaking space
nbsp()
nbsp(1)

# Several non-breaking spaces
nbsp(n = 2)
nbsp(n = 5)

# When `n` is zero, an empty string is returned.
nbsp(n = 0)

Print a dtlg table

Description

A convenience wrapper around print() for printing dtlg tables with consistent formatting options.

Usage

print_dtlg(
  dt,
  row.names = FALSE,
  trunc.cols = TRUE,
  class = FALSE,
  nrows = Inf,
  justify = "left"
)

Arguments

Value

Invisibly returns the printed object.

Examples

calc_stats(dt = adsl, "AGE", treat = "ARM", indent = "  ")[[1]] |>
  print_dtlg()

Rounded percentage

Description

round_pct() returns the rounded percentages of x values.

Usage

round_pct(x, digits = 1L, method = c("round", "round_sum"))

Arguments

Value

A numeric vector of the same length as x with rounded percentages.

Examples

x <- c(1 / 3, 1 / 3, 1 / 3)

# Default method ensures precise rounding but total might not be 100%.
round_pct(x = x)
sum(round_pct(x = x))

# You can trade off rounding precision for precision on the total with the
# method `"round_sum"`.
round_pct(x = x, method = "round_sum")
sum(round_pct(x = x, method = "round_sum"))

# Vary the number of decimal places, e.g. increase to three.
round_pct(x = x, digits = 3, method = "round_sum")

# Missing values are ignored.
x <- c(1, 2, NA)
round_pct(x = x, digits = 3)

Rounds numbers while preserving the total sum

Description

round_sum() rounds a numeric vector of non-negative values to a specified number of decimal places while ensuring that the sum of the rounded value remains as close as possible to the original total.

Usage

round_sum(x, digits = 0L)

Arguments

Value

A numeric vector of the same length as x, with values rounded in such a way that the total sum is preserved.

Examples

# Rounds to integers, preserving the sum of 100.
x <- c(33.3333, 33.3333, 33.3334)
(y <- round_sum(x))
identical(sum(x), sum(y))

# Rounds to integers, preserving the sum of 1002.
x <- c(100.5, 200.25, 300.75, 400.5)
(y <- round_sum(x))
identical(sum(x), sum(y))

# Rounds to one decimal place, preserving the total sum.
x <- c(12.345, 67.890, 19.765)
(y <- round_sum(x))
identical(sum(x), sum(y))

Summary Table

Description

summary_table() summarises clinical variables into a report table using data.table as backend.

Usage

summary_table(
  dt,
  target,
  treat,
  target_name = target,
  indent = nbsp(n = 4L),
  .total_dt = dt,
  pct_dec = 1,
  treat_order = NULL,
  skip_absent = TRUE
)

Arguments

Value

A data.table of summary statistics. The format depends on the type of the target variable:

• If the target variable is categorical, i.e. type character, factor or logical then the output is that of calc_counts().

• If the target variable is numeric, then the output is that of calc_desc().

See Also

tern_summary_table()

Examples

dmg_vars <- c("AGE", "RACE", "ETHNIC")
dmg_var_lbls <- c("Age (yr)", "Race", "Ethnicity")

# Demographics table (DMT01)
summary_table(
  adsl,
  target = dmg_vars,
  treat = 'ARM',
  target_name = dmg_var_lbls
)

# Demographics table (DMT01) with continuous variable (e.g., BMRKR1)
summary_table(
  adsl,
  target = c(dmg_vars, "BMRKR1"),
  treat = 'ARM',
  target_name = c(dmg_var_lbls, "Biomarker 1")
)

Create a summary table using multiple rows for grouping on one target column

Description

Create a summary table using multiple rows for grouping on one target column

Usage

summary_table_by(
  dt,
  target,
  treat,
  rows_by,
  indent = nbsp(n = 4L),
  .total_dt = dt,
  pct_dec = 1,
  treat_order = NULL,
  skip_absent = TRUE
)

Arguments

Value

The same output as summary_table() except that folded by variables indicated in rows_by.

Examples

summary_table_by(adlb, target = "AVAL", treat = "ARM", rows_by = c("PARAM","AVISIT"))

Create a summary table using multiple rows for grouping on two target column ideal for creating change from baseline tables

Description

Create a summary table using multiple rows for grouping on two target column ideal for creating change from baseline tables

Usage

summary_table_by_targets(
  dt,
  target,
  treat,
  rows_by,
  indent = nbsp(n = 4L),
  .total_dt = NULL,
  pct_dec = 1,
  treat_order = NULL,
  skip_absent = TRUE
)

Arguments

Value

data.table

Examples

adlb <- random.cdisc.data::cadlb|>dplyr::filter(AVISIT != "SCREENING")
labs <- summary_table_by_targets(adlb, c('AVAL','CHG'), 'ARM', c('PARAM','AVISIT'), '  ', NULL)

Generate Core Safety Tables (CSR Section 14.3.1) using tern/rtables

Description

tern_AET01_table() produces a consolidated safety summary table using rtables and tern. It mirrors the output and interface of AET01_table(), generating standard adverse event summaries (e.g. death, withdrawal, AESIs) for Clinical Study Reports (CSR) Section 14.3.1.

Usage

tern_AET01_table(
  adsl,
  adae,
  patient_var,
  treat_var,
  aesi_vars,
  aesi_heading = "Total number of patients with at least one",
  indent = "  "
)

Arguments

Details

The function returns a single formatted rtables table summarising core safety endpoints by treatment arm.

Value

A TableTree object from the rtables package.

Examples

tern_AET01_table(
  adsl = adsl,
  adae = aesi,
  patient_var = "USUBJID",
  treat_var = "ARM",
  aesi_vars = c("FATAL", "SER", "SERWD", "SERDSM", "RELSER",
                "WD", "DSM", "REL", "RELWD", "RELDSM", "SEV")
)

Generate AET02-style AE summary using tern and rtables

Description

This function builds a System Organ Class (SOC) and Preferred Term (PT) adverse event summary table, following the AET02 CSR format, using the tern and rtables packages.

Usage

tern_AET02_table(
  adsl,
  adae,
  patient,
  treat,
  target = "AEDECOD",
  rows_by = "AEBODSYS",
  indent = "  "
)

Arguments

Value

A TableTree object with AE summary by SOC/PT.

See Also

AET02_table()

Create a clinical reporting table with tern/rtables

Description

tern_summary_table() is a convenience wrapper around {rtables} and {tern} commands to generate a clinical reporting summary statistics tables whilst using a similar interface as summary_table(). This can be helpful for side by side comparisons of the two functions.

Usage

tern_summary_table(dt, target, treat, target_name = target)

Arguments

Value

A data.table of summary statistics. The format depends on the type of the target variable:

• If the target variable is categorical, i.e. type character, factor or logical then the output is that of calc_counts().

• If the target variable is numeric, then the output is that of calc_desc().

See Also

summary_table()

Examples

dmg_vars <- c("AGE", "RACE", "ETHNIC")
dmg_var_lbls <- c("Age (yr)", "Race", "Ethnicity")

# Demographics table (DMT01)
tern_summary_table(
  adsl,
  target = dmg_vars,
  treat = 'ARM',
  target_name = dmg_var_lbls
)

# Demographics table (DMT01) with continuous variable (e.g., BMRKR1)
tern_summary_table(
  adsl,
  target = c(dmg_vars, "BMRKR1"),
  treat = 'ARM',
  target_name = c(dmg_var_lbls, "Biomarker 1")
)

Count total events

Description

total_events() counts the number of observations in dt in each group defined by treat levels. Counts are returned in wide format, i.e. one column per level in treat.

Usage

total_events(dt, treat, label)

Arguments

Value

A list wrapping a one-row data.table of 1 + n variables, where n is the number of levels in treat. First variable is stats, character type, whose value is the argument passed in as label. Following variables are of integer type and provide the counts.

Examples

# In the absence of pre-filtering, `total_events()`, actually, just counts
# observations in `dt`.
total_events(dt = adsl, treat = "ARM", label = "Subjects")[[1]]

# If `dt` is pre-filtered, e.g. with a condition matching an event, then
# `total_events()` can be used to (effectively) count events.
total_events(dt = adsl[adsl$DTHFL == 'Y'], treat = "ARM", label = "Deaths")[[1]]

# Another example using the complement predicate condition.
total_events(dt = adsl[adsl$DTHFL == 'N'], treat = "ARM", label = "Lives")[[1]]

Add a label attribute to an object

Description

Add a label attribute to an object

Usage

with_label(x, label)

Arguments

Value

x labeled by label.

See Also

label()

Examples

label(1)
label(with_label(1, "my label"))