gigs

KO | EN

gitlite — search

#python #react #ai #typescript #go #android #machine-learning #rust #docker #kubernetes #pytorch #mcp #ios #open-source

(★ 10)

Convert between measurements and z-scores/percentiles for INTERGROWTH-21st and WHO Child Growth Standards, and classify growth patterns.

gigs Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# gigs <img src="man/figures/logo.png" align="right" height="138" />

[![Project Status: Active – The project has reached a stable, usable
state and is being actively
developed.](https://www.repostatus.org/badges/latest/active.svg)](https://www.repostatus.org/#active)
[![R-CMD-check](https://github.com/ropensci/gigs/actions/workflows/R-CMD-check.yaml/badge.svg)](https://github.com/ropensci/gigs/actions/workflows/R-CMD-check.yaml)
[![pkgcheck](https://github.com/ropensci/gigs/workflows/pkgcheck/badge.svg)](https://github.com/ropensci/gigs/actions?query=workflow%3Apkgcheck)
[![codecov](https://codecov.io/github/ropensci/gigs/graph/badge.svg?token=FZcgqIpQL4)](https://codecov.io/github/ropensci/gigs)
[![CRAN
version](https://www.r-pkg.org/badges/version/gigs)](https://www.r-pkg.org/pkg/gigs)
[![ropensci](https://badges.ropensci.org/626_status.svg)](https://github.com/ropensci/software-review/issues/626)
[![status](https://joss.theoj.org/papers/c9986578f8e3d93f6195f03e8d23308b/status.svg)](https://joss.theoj.org/papers/c9986578f8e3d93f6195f03e8d23308b)

## Overview

Produced as part of the Guidance for International Growth Standards
project at the London School of Hygiene & Tropical Medicine, **gigs**
provides a single, simple interface for working with the WHO Child
Growth standards and outputs from the INTERGROWTH-21st
project. You will find functions for converting from anthropometric
measures (e.g. weight or length) to z-scores and centiles, and the
inverse. Also included are functions for classifying newborn and infant
growth according to literature-based cut-offs.

**gigs** is of use to anyone interested in fetal and child growth,
including child health researchers, policymakers, and clinicians. This
package is best suited to growth data where the gestational age (GA) of
each child is known, as the use of the growth standards included in
**gigs** is GA-dependent. We recommend you check out the [available
standards](#available-international-growth-standards) section to see if
your anthropometric measurements can be converted to z-scores/centiles
by **gigs**. We recommend using **gigs** to generate continuous or
categorical measures of fetal/newborn/child growth, which can then be
used in downstream analyses.

## Installation

You can install the development version of **gigs** from
[GitHub](https://www.github.com/ropensci/gigs) with `pak`, or from the
[ropensci R-universe](https://ropensci.r-universe.dev/builds):

``` r
# install.packages("pak")
pak::pkg_install(pkg = "ropensci/gigs")
```

``` r
install.packages("gigs", repos = "https://ropensci.r-universe.dev")
```

## Rationale + terminology

When working with growth measurements from individual children, it is
helpful to compare those measurements to a growth standard, which
represents average growth for an population of children. This allows
assessment of individual growth - for example, that a baby was born
small, but later caught up to its peers in size. It also allows you to
compare measurements from different children.

These standards of growth can be made in different ways, but the best
studies utilise international samples made of up thousands of healthy
fetuses, newborns, or children. In **gigs**, you’ll find different
international growth standards, allowing you to compare growth measures
from children with different ages, weights, heights, and more.

In general, you’ll use **gigs** to transform raw growth measures to
*z-scores* or *centiles*. Z-scores and centiles represent the location
of a measurement within a normal distribution of values, such that:

- A *z-score* is the number of standard deviations from the mean for a
 given anthropometric measurement (e.g. height or weight).
- A *centile* represents the proportion of measurements in some
 distribution which we would expect to be lower than a measurement
 we’ve taken. In *gigs*, these are represented as a value between `0`
 and `1`. For example, `0.5` corresponds to the 50th centile
 (i.e. the mean), whereas `0.75` corresponds to the 75th
 centile.

In growth data, z-scores and centiles represent the size a fetus,
newborn, or child relative to its peers. Size here is considered
relative to a standardising variable, which is usually age but could
also be another variable such as their length. By tracking a child’s
relative size as they grow, you can see if they are achieving their
growth potential or not. If not, this may indicate underlying issues
such as ill health or undernutrition.

## Classification functions

**gigs** includes a number of functions which permit fast identification
of at-risk infants through classification of suboptimal growth. The
cut-offs used are sourced from research literature; you can check the
function documentation to see these sources.

### Growth classification in `data.frame`-like objects

Use the `classify_growth()` function to quickly compute growth
indicators in `data.frame`-like objects. All `classify_*()`-style
functions in **gigs** use
[data-masking](https://rlang.r-lib.org/reference/args_data_masking.html),
so you provide a `data.frame`-like object in the `.data` argument and
then refer to your column names directly. In `classify_growth()`, you
can also use the `.analyses` argument to specify which growth indicators
you want to classify.

``` r
life6mo_newborns <- gigs::life6mo[life6mo$age_days == 0, ]

# Use classify_growth() to get multiple growth indicators at once
life6mo_classified <- classify_growth(
 .data = life6mo_newborns,
 gest_days = gestage,
 age_days = age_days,
 sex = as.character(sex),
 weight_kg = wt_kg,
 lenht_cm = len_cm,
 id = as.factor(id),
 .outcomes = c("svn", "stunting")
)
#> ! Unused factor levels kept after small vulnerable newborn categorisation:
#> "Term AGA" and "Term LGA".
#> ! Unused factor levels kept after stunting categorisation: "outlier".
#> 
#> ── `gigs::classify_growth()` ───────────────────────────────────────────────────
#> ✔ Small vulnerable newborns: Success
#> ✔ Stunting: Success

head(life6mo_classified, n = 4)
#>    id gestage sex visitweek pma age_days wt_kg   len_cm headcirc_cm  muac_cm
#> 1   1     273   M         0 273        0  2.30 42.06667    33.26667 9.433333
#> 28  4     250   F         0 250        0  1.50 42.03333    30.03333 8.066667
#> 36  5     238   F         0 238        0  2.39 43.46667    33.63333 9.166667
#> 56  8     240   F         0 240        0  1.80 41.73333    31.46667 8.033334
#>    birthweight_centile         svn       lhaz        stunting stunting_outliers
#> 1          0.010765424    Term SGA -3.5406446 stunting_severe   stunting_severe
#> 28         0.002833163 Preterm SGA -2.2854751        stunting          stunting
#> 36         0.756367868 Preterm AGA -0.6087434    not_stunting      not_stunting
#> 56         0.126440075 Preterm AGA -1.6989568    not_stunting      not_stunting
```

When using `classify_growth()`, you will be informed which of the
analyses you wanted to run were successful. In the example below,
because `lenht_cm` is not specified, stunting indicators cannot be
computed.

``` r
life6mo_classified <- classify_growth(
 .data = life6mo_newborns,
 gest_days = gestage,
 age_days = age_days,
 sex = as.character(sex),
 weight_kg = wt_kg,
 id = as.factor(id),
 .outcomes = c("svn", "stunting")
)
#> ! Unused factor levels kept after small vulnerable newborn categorisation:
#> "Term AGA" and "Term LGA".
#> 
#> ── `gigs::classify_growth()` ───────────────────────────────────────────────────
#> ✔ Small vulnerable newborns: Success
#> ! Stunting: Not computed (`lenht_cm` not supplied)

head(life6mo_classified, n = 4)
#>    id gestage sex visitweek pma age_days wt_kg   len_cm headcirc_cm  muac_cm
#> 1   1     273   M         0 273        0  2.30 42.06667    33.26667 9.433333
#> 28  4     250   F         0 250        0  1.50 42.03333    30.03333 8.066667
#> 36  5     238   F         0 238        0  2.39 43.46667    33.63333 9.166667
#> 56  8     240   F         0 240        0  1.80 41.73333    31.46667 8.033334
#>    birthweight_centile         svn
#> 1          0.010765424    Term SGA
#> 28         0.002833163 Preterm SGA
#> 36         0.756367868 Preterm AGA
#> 56         0.126440075 Preterm AGA
```

You can also use `classify_*()` functions which are specific to the
growth indicator you’d like to calculate, for example `classify_svn()`
to get small, vulnerable newborn classifications for each infant:

``` r
# Small vulnerable newborns - note no ID parameter, as it is assumed that all
# measures are taken at birth
life6mo_svn <- classify_svn(
 .data = life6mo_newborns,
 weight_kg = wt_kg,
 gest_days = gestage,
 sex = as.character(sex)
)
#> ! Unused factor levels kept after small vulnerable newborn categorisation:
#> "Term AGA" and "Term LGA".

head(life6mo_svn, n = 4)
#>    id gestage sex visitweek pma age_days wt_kg   len_cm headcirc_cm  muac_cm
#> 1   1     273   M         0 273        0  2.30 42.06667    33.26667 9.433333
#> 28  4     250   F         0 250        0  1.50 42.03333    30.03333 8.066667
#> 36  5     238   F         0 238        0  2.39 43.46667    33.63333 9.166667
#> 56  8     240   F         0 240        0  1.80 41.73333    31.46667 8.033334
#>    birthweight_centile         svn
#> 1          0.010765424    Term SGA
#> 28         0.002833163 Preterm SGA
#> 36         0.756367868 Preterm AGA
#> 56         0.126440075 Preterm AGA
```

## Conversion functions

### Available international growth standards

**gigs** facilitates the proper use of international growth standards,
which are growth charts developed using international samples of healthy
singleton children born to mothers that had their health needs met
during pregnancy. They represent an international standard of ‘optimal’
growth. **gigs** implements international growth standards from the WHO
and INTERGROWTH-21st project:

- `ig_nbs` - INTERGROWTH-21st Newborn Size standards
 (including very preterm)
 <details>