growthscreener 1.22.0

• Extends the calculate_th() function with a support_missing_hgtf argument to allow for the calculation of target height from only the mother's height. This is useful when height from the biological father height is missing. It currently works only for Dutch populations. If mother's height is missing, the function will return NA, as before.
• Adds a silent safety trim of input arguments of calculate_th() to scalar values. This is to prevent the function from breaking when the input is a vector of length > 1. The function will now only use the first element of the input vector.
• Adds functions hermanussencole() and targetheight() from the chartdesigner package. In contrast to calculate_th() these functions are vectorised, so more suited for group applications rather than individual application.

v1.21.0

growthscreener 1.21.0

• Merges srm branch updates (v1.19.3) into main
• Undo the choice for weight for height guidelines for period 0-2 years
  throughout (made SvB 13jul20)
  This choice causes issues in practice as our advice does not line up
  with what the professional sees on their own diagrams. Sticking with 2012
  guidelines even if they are inconsistent.
• Makes list_screeners() aware of the language development guidelines
• Introduces lgd as abbreviation for language development

growthscreener 1.20.0

• Merges srm brnach into master
• Updates NEWS.md

growthscreener 1.19.0

• Version increases 1.16.0 --> 1.19.0 were made in srm branch
• Adds advice algorithm for future visit date (remind date guideline)
• Adds screener for language development, 2 yrs
• Changes all dates to yyyymmdd
• Merges with master

growthscreener 1.16.1

• Extrapolates the BMI cutoff table beyond 18 years to evade error when the child age is between 18 and 19 years
• Silences the Rm CMD CHECK error on global variable age
• Deprecates calculate_screening_doms()
• Removes lubridate dependency
• Makes sure ga is in completed weeks

growthscreener 1.16.0

• Merges the srm branch
• Generalises screening functions from two time points to any number of time points
• Assumes that the last time point is current
• Returns one advice for all time points combined (rather than m-1 retrospective pairs)
• Adds a new screening algorithm for language based on six Van Wiechen items
• Contains numerous other improvements
• Updates to roxygen version to 7.2.3
• Updates GitHub actions

growthscreener 1.15.0

• Update roxygen version to 7.2.1

growthscreener 1.14.0

• Siplify data structure to list with elements psn and xyz

growthscreener 1.13.0

• Adds support for the new target class of the bdsreader package

growthscreener 1.12.0

• Adds \dots argument to screen_curves_ind(), screen_curves_tgt() and list_screeners() for extendibility

growthscreener 1.11.0

• Adds list_screeners() to produce overview of screeners

growthscreener 1.10.0

• Relies now on new auto_format functionality from bdsreader
• Solves some smaller bugs
• Define stricter dependencies

growthscreener 1.9.0

• Breaking changes:
• Replaces the minihealth package by bdsreader
• Replaces the jamestest package by jamesdemodata

growthscreener 1.8.0

• Uses nlreferences package

growthscreener 1.7.0

• Major rewrite to replace the clopus by the centile package for clearer, easier and extendable reference selection and calculation

growthscreener 1.6.1

• Explicit declaration of clopus

growthscreener 1.6.0

• This version improves consistency in sieves and messages
• Adds checks on missing data for z1 in some edge cases where needed
• Adds messages **24 and **25 for edge cases where the Z-score cannot be calculated, even if all information is available
• Adds a test_gain argument to all screening functions that we may use to disable any checking on gain scores z1 - z0 (e.g. when there is only one observation for a child)
• Uses stricter scalar comparison operators && and ||
• Makes SD and SDS consistent in messages
• Replaces text "voorlaatste" to the more generic "eerdere"
• Adds test code that runs all json files in jamestest

growthscreener 1.5.0

This release updates the sieve for the guidelines for referral for under- and overweight.

• Introduces pick_reference_wgt() to select weight reference depending on sex, age, ga and etn
• Refreshes the sieve used for under- and overweight screening
• Adds code 2076 to indicate "severe underweight" (< -3SD), but - in conjunction with guidelines - we do not offer the strict refer advice
• Solves a bug in calculate_screening_doms() that prevented weight screening
• For clarity, reverts the weight screening code from wfh to wgt
• Updates to roxygen 7.1.1
• Evades testing problem by adding jamestest to suggests:

There is a discrepancy between the suggested references for under- and overweight. The underweight guidelines suggests "weight for age" for 0-1 years and "weight for height" for 1-2 years. The overweight guidelines suggests "weight for height" for 0-2 years. We followed the latter suggestion also for underweight.

We have interpreted that the "abnormal increase of 0.67 SD" in weight should occur above the +1 SD line, and deemed any such increases below the +1 SD line as normal variation.

v1.3.0

Important changes in this version:

• Renumbers the messages starting with 1000's
• Adds new screeners for weight and head circumference
• Replaces the AGD dependency by the clopus package
• Uses updated preterm WFH references
• Restructures internal computational flow to improve support for multiple screeners

Height screening rules

This initial release hold functions for screening of child height.