action_levels: Compute Action Levels

In pediatric.zcalc: Z-Score Calculator for Biomarkers: Childhood to Young Adulthood

Compute Action Levels

Description

Assigns monitoring or action levels based on individual percentile ranks using standard IDEFICS thresholds.

Usage

action_levels(
  df,
  sex = NULL,
  lvl_name = c("none", "monit", "action"),
  perc_level = c(0.9, 0.95),
  append = FALSE,
  filter = NULL
)

Arguments

df	A data frame containing percentile columns such as 'waist_percentile', 'sbp_percentile', 'hdl_percentile', etc.
sex	Character vector. Same length as 'age', 'height', and 'values'. Accepts "f" for female or "m" for male.
lvl_name	Character vector of level labels. Defaults to 'c("none", "monit", "action")'.
perc_level	Numeric vector of two percentiles used as cutoffs. Defaults to 'c(0.9, 0.95)' for 90th and 95th percentile.
append	Logical. If 'TRUE', appends action level columns to 'df'. If 'FALSE', returns only the computed levels.
filter	Character. Optional. Limits calculation to a specific domain: '"adiposity"', '"blood_pressure"', '"blood_lipids"', '"blood_glu_insu"', or '"overall"'.

Details

Action levels are derived using 'cut()' on percentile values. For example, a value > 95th percentile maps to '"action"'. HDL is reversed ('1 - hdl_percentile') since low HDL values are considered unhealthy.

Value

A list (or a modified data frame if 'append = TRUE') containing action level classifications for each domain.

Examples

df <- data.frame(waist_percentile = c(0.85, 0.96))
action_levels(df)

df <- data.frame(
  hdl_percentile = c(0.1,0.5),
  homa_percentile = c(0.4,0.9),
  trg_percentile = c(0.6,0.5),
  waist_percentile = c(0.9,0.99),
  sbp_percentile = c(0.8,0.01)
)
action_levels(df)

df <- data.frame(
  #sex = c("m", "m"),
  hdl_percentile = c(0.1,0.5),
  homa_percentile = c(0.4,0.9),
  trg_percentile = c(0.6,0.5),
  crp_percentile = c(0.95, 0.9),
  waist_percentile = c(0.9,0.99),
  sbp_percentile = c(0.8,0.01)
)
action_levels(df, sex = c("m", "m"))

pediatric.zcalc documentation built on Sept. 30, 2026, 5:13 p.m.