discrepancies between `anthro` R package and `zscorer` results

[ernestguevarra]

from en-net.org (https://www.en-net.org/question/3787.aspx#lastpost):

Dear Mark and Ernest, thanks for developing the zscorer package; it is a great help.

I would like to ask your help, as when I compare the results I get from Zscorer to the one from Anthro in R , there are some discrepancies. Specifically, with Zscorer, the number of children defined as SAM, according to their WHZ, is lower than the number I get in anthro.

The sample includes 200 children aged 6 to 60 months old.

The commands we used in R were: 

• For Zscorer:

fullbaseline <- addWGSR(data=fullbaseline, sex= "sex_child", firstPart= "weight_average", secondPart= "height_average", index = "wfh", standing = "height_position", output = "WHZ")

age was counted in days according to the formula: 

fullbaseline$age_days<-fullbaseline$child_age * (365.25 / 12)

With this command, we get :

table(fullbaseline$sam_mam)

MAM :107 ; SAM : 93 

• For Anthro:

with (fullbaseline, anthro_zscores ( sex = sex_child, age = age_child_months, is_age_in_month = TRUE, weight = weight_average, lenhei = height_average, measure= height_position, armc = MUAC_average, triskin = triceps, subskin = subscapula ) )

With this command, we get :

table(fullbaseline$sam_mam)

MAM:  92 ;  SAM  : 107 

With Anthro, we have one missing data as one child was 60 months and a few days at the enrolment.

Here, I report some examples of children defined SAM in Antrho and MAM in Zscorer

AgeMonths	AgeDays	WEIGHT	HEIGHT_Lenght	H.Position	Oedema	WHZ ZSCORER	WHZAnthro
33.4	1016.6125	8.9	83.6	1/L	0	-2.88	-3.07
32.8	998.3500	9.5	86.6	1/L	0	-2.88	-3.07
36.8	1120.1000	8.9	83.6	1/L	0	-2.88	-3.07
24.5	745.7188	7.5	74.5	1/L	0	-2.84	-3.03
32.2	980.0875	8.0	78.1	1/L	0	-2.92	-3.08

Thank you in advance for your assistance!

Kind regards,

Silvia

Issue related to #121