# New version of R package 'zscorer' available

This question was posted the Assessment and Surveillance forum area and has 21 replies.

### Mark Myatt

Consultant Epidemiologist

Frequent user

5 Nov 2019, 14:55

Ernest and I have been working on the zscorer package for the R Language for Data Analysis and Graphics. v0.3.1 of the zscorer package calculates and adds nutritional anthropometry z-scores to survey data:

- Weight-for-length (wfl) z-scores for children with lengths between 45 and 110 cm
- Weight-for-height (wfh) z-scores for children with heights between 65 and 120 cm
- Length-for-age (lfa) z-scores for children aged less than 24 months
- Height-for-age (hfa) z-scores for children aged between 24 and 228 months
- Weight-for-age (wfa) z-scores for children aged between zero and 120 months
- Body mass index-for-age (bfa) z-scores for children aged between zero and 228 months
- MUAC-for-age (mfa) z-scores for children aged between 3 and 228 months
- Triceps skinfold-for-age (tsa) z-scores for children aged between 3 and 60 months
- Sub-scapular skinfold-for-age (ssa) z-scores for children aged between 3 and 60 months
- Head circumference-for-age (hca) z-scores for children aged between zero and 60 months

The z-scores are calculated using the WHO Child Growth Standards for children aged between zero and 60 months or the WHO Growth References for school-aged children and adolescents.

**New in this version** is that MUAC-for-age (mfa) z-scores for children aged between 60 and 228 months are calculated using the new MUAC-for-age growth reference developed by Mramba et al. (2017) using data from the USA and Africa:

Mramba L, Ngari M, Mwangome M, Muchai L, Bauni E, Walker AS, et al. *A growth reference for mid upper arm circumference for age among school age children and adolescents, and validation for mortality: growth curve construction and longitudinal cohort study*. BMJ. 2017

This reference has been validated with African school-age children and adolescents.

The zscorer comes packaged with the WHO Growth References data and the new MUAC-for-age reference data.

The package contains a Shiny app that provides a web interface to the package's functions. You can see an example of this functionality here.

The package can be downloaded and installed in R in the usual manner. The package website is here.

A manual is inlcuded in the package and is available here.

### Carlos Grijalva-Eternod

UCL Institute for Global Health

Normal user

5 Nov 2019, 19:54

This R package is very neat and useful.

Thank you Mark and Ernest.

### Ernest Guevarra

Katilingban

Frequent user

5 Nov 2019, 20:15

Thanks for sharing this to the forum Mark!

Mark and I are pleased to be able to share this package along with the accompanying web interface with the forum and to everyone who does nutrition studies and surveys. We think this fills a gap that we noted from some of the questions/issues raised in this thread with regard to tools for z-score calculation.

For those who already use R Language for Data Analysis and Graphics, then installation and usage will be straightforward using the links that Mark shared documenting the functions in the zscorer package.

For those that don't use R or are still in the early stages of learning R, the web interface allows for anyone to perform the calculations using familiar and easy to learn point and click user interface.

This package and the web interface is in active development. Mark and I are happy to receive feedback/comments/issue or bug reports/feature requests either via this EN-NET forum or through our GitHub development portal found here. Mark and I also embrace open source development and as such our codebase is available for anyone to review and give feedback to or build on. This is available, again, through our GitHub development portal.

### Anonymous 27728

Étudiant

Normal user

5 Nov 2019, 22:13

Excuse me, for it's not an answer but rather a question: please tell me if you know a way to calculate the target weight and height of a child without using the WHO Zscore table.

I would like to know whether there is another method to find the target weight of a child without using the WHO z-score table, because after taking the anthopomteric measurements for a child with W/H -3 we use the WHO table to find the target weight. My question is to find out whether there is another method of calculating this?

Thank you for taking my question into consideration, Tidiani Cisse Student

### Sharon Cox

Assoc Professor

Normal user

6 Nov 2019, 01:32

Thanks Ernest and Mark. I agree that this fills a gap. I will be sharing with my students.

### Mbaye Diop

Cellule de lutte contre la malnutrition

Normal user

7 Nov 2019, 13:59

Hello,

There are expected target weights for 1 month periods as part of growth monitoring and promotion (GMP).

Try to get acquainted with these tables, but they are for children from 0-23 months.

Thank you

### Martin Njoroge

Epidemiologist

Normal user

13 Nov 2019, 08:57

Thanks Mark & Ernest. Quite informative and helpful.

### Severine Frison

Normal user

15 Apr 2020, 11:57

Hello Mark,

I've been using the package, really great work. Thanks a lot.

I have one question though, it seems like zscores are rounded to 2 digits only. Is there a way to change that option?

Many thanks in advance

Sev

### Cazes Cécile

Normal user

22 May 2020, 23:24

Hello,

The weight-for-height calculation table used in the DRC corresponds to the WHO unisex boy table but does not take into account the child's age. So when I compare the classification of children according to the W / H ratio according to the DRC table and according to a variable calculated with the package zscorer or other package R such as anthro or igrowup, I observe differences in classifications.

For example, a 29 month old child with a weight of 7.2 kg and a height of 73.5 cm has a W / H ratio = -3 according to the DRC table but a W / H ratio = - 3.27 according to the WHO table for boys. So when I check the classification by the nurses of this child in two categories, SAM or Non-SAM, my calculated variable indicates a SAM classification while the child is non-SAM according to the DRC table, so there is no error according to the table used in the DRC. I found 1.5% of discrepancy of this type due to the fact that the calculation of the weight / height index according to the WHO 2006 standards is based on the age or the taking of the standing or lying height of the child.

The ideal would be that I can integrate the DRC weight / height table into an R program to avoid these discrepancies. What do you think? Do you think this is doable? How? Or maybe there are other alternatives?

Thank you in advance for your help.

### Cazes Cécile

Normal user

25 May 2020, 11:33

Thank you very much for your answer.

I would like to understand how the DRC table is designed.

According to my research, this is the WHO table for boys, but I noticed that the values correspond to either children under 24 months of age (lying down measurement) or children of + 24 months of age (standing measurement).

If I understand your answer correctly, you mean that the DRC table corresponds to the WHO boys table with a systematic correction made on the height: + 0.7 cm compared to the actual height? is it this?

so I should apply the argument: index = wfh height = height + 0.7 cm systematically to get the values from the DRC table?

Thanks for your help

### Cazes Cécile

Normal user

25 May 2020, 14:34

again, Thank you for your suggestions, I will apply your instructions.

One last question, do you have a bibliographic reference regarding the 0.7 cm correction?

Thank you again.

### Cazes Cécile

Normal user

26 May 2020, 11:50

Hello again,

After checking, I still have the same problem of discordant values between the WHO table for the WFL boy table (0-2years old, 45 to 110 cm) and the WFL value calculated by the package.

All of the children below are under 24 months of age with a length measurement while lying down. The WHO boy WFL table indicates a wfl z-score strictly equal to -3; these children are classified as not severe.

On the other hand, the variable calculated with zscorer indicates a wfl z-score <-3; these children are in the severe class. I found the same result using the R anthro and igrowup packages.

With zscorer:

getWGSR (sex = 1, firstPart = 7.2, secondPart = 73.5, index = "wfl", standing = "2")

[1] -3.078397

getWGSR (sex = 1, firstPart = 6.8, secondPart = 71.1, index = "wfl", standing = "2")

[1] -3.073756

getWGSR (sex = 1, firstPart = 5.7, secondPart = 65.0, index = "wfl", standing = "2")

[1] -3.078341

getWGSR (sex = 1, firstPart = 5.3, secondPart = 63.2, index = "wfl", standing = "2")

[1] -3.199316

getWGSR (sex = 1, firstPart = 5.4, secondPart = 63.5, index = "wfl", standing = "2")

[1] -3.10421

### Cazes Cécile

Normal user

26 May 2020, 14:42

I apologize for the confusion on the term "size" which comes from the translation from French / English I think (only one word in French).

Your explanations are of great help, thank you very much.

I had this assumption on rounding of size or weight but I would not have been able to confirm them as you did. thank you so much.

some children are classified in the -3 SD column when they are severe and miss out on treatment with RUTF it is a shame. In the specific case of these 5 children, the MUAC was between 115 and 122, therefore children at high risk of mortality who escape treatment with RUTF according to the national protocol.

Thanks for your help.

If you have any problem posting a response, please contact the moderator at post@en-net.org.