The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Mortality » Neonatal Mortality in India NFHS-4
Neonatal Mortality in India NFHS-4 Wed, 30 May 2018 10:33
 tschoudhary Messages: 7Registered: January 2018 Member
Dear DHS team,

I am trying to calculate the neonatal mortality for India as reported in the National Report. The code used is as follows (taken from DHS User Forum):

***IABR73FL has been used for the analysis as mentioned at the DHS website. Analysis done using Stata 15.1.

gen age_death = .
replace age_death = 0 if b6<=106 & b5==0
replace age_death = 1 if b6>=107 & b6<=130 & b5==0
replace age_death = 2 if (b6> 130 & b6<=999) | b5==1
lab def age_death 0 "Early Neonatal Death" 1 "Late Neonatal Death" 2 "Survived Neonatal Period"
lab val age_death age_death
lab var age_death "Neonatal Mortality Status"
tab age_death [iw = v005/1000000]

The unweighted observations in the national report is 262210 and the weighted observations are 252723
The total number of children in the five years preceding the survey is 259627 as per the data file and with weight applied the number of observation is 249967.
The NMR reported is 29.5 as per the national report and we are getting a value of 29.4.

Kindly hep me identify what am i doing wrong.
Re: Neonatal Mortality in India NFHS-4 [message #15078 is a reply to message #15076] Wed, 30 May 2018 11:15
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
As neonatal mortality is calculated as one of five different mortality indicators, the approach to the calculation of neonatal mortality is a little more complicated than this. You can see example code in this thread.
Re: Neonatal Mortality in India NFHS-4 [message #15081 is a reply to message #15078] Wed, 30 May 2018 19:55
 tschoudhary Messages: 7Registered: January 2018 Member
Dear Trevor,

Thanks a lot for the prompt response. This is immensely helpful.

One additional query. Will the early neonatal mortality be calculated using the same technique of cohorts for DHS. Will b6 be the variable on which the calculation of early neonatal mortality be made (b6 <= 106)

Many Thanks,
Tarun.

[Updated on: Thu, 31 May 2018 06:19]

Report message to a moderator

Re: Neonatal Mortality in India NFHS-4 [message #15135 is a reply to message #15081] Thu, 07 June 2018 10:31
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
Early neonatal mortality is calculated in a simpler manner, and calculated separately from the approach used for the other child mortality indicators. We typically report it in table 8.4 on perinatal mortality. For this table, we calculate a simple proportion of births in the 0-59 months preceding the survey that resulted in an early neonatal death (using b6 <= 106).
Re: Neonatal Mortality in India NFHS-4 [message #15738 is a reply to message #15076] Sat, 08 September 2018 04:49
 SAK_Nepal Messages: 5Registered: March 2018 Location: Kathmandu Member
Dear Trevor,

Is it possible to create variables for "any perinatal death (Y/N)" "any infant death (Y/N)" "any child death(Y/N)"? I am trying to look at the association between institutional delivery and women who had any perinatal/infant/child death.

Your clarification would greatly be appreciated.

Thanks
SAK

SAK
Re: Neonatal Mortality in India NFHS-4 [message #15761 is a reply to message #15738] Tue, 11 September 2018 10:26
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
Using the IR file, try the following code:
```* Any stillbirths
gen stillbirths = regexm(vcal_1,"TPPPPPP")

gen early_neonatal_deaths = 0
gen infant_deaths = 0
gen child_deaths = 0

* rename birth variables for use in forvalues loop
rename b*_0* b*_*

* Any early neonatal deaths, infant deaths, child deaths
forvalues b = 1/20 {
* early neonatal - days 0-6
replace early_neonatal_deaths = 1 if inrange(b6_`b',100,106)
* infant - from birth trough age 11 months
replace infant_deaths = 1 if inrange(b6_`b',100,211)
* child from 12 months up through 4 completed years
replace child_deaths = 1 if inrange(b6_`b',212,304)
}
* perinatal mortality includes stillbirths and early neonatal deaths
gen perinatal_deaths = (stillbirths | early_neonatal_deaths)

tab stillbirths
tab early_neonatal_deaths
tab infant_deaths
tab child_deaths
tab perinatal_deaths
```
Re: Neonatal Mortality in India NFHS-4 [message #17620 is a reply to message #15761] Sun, 28 April 2019 23:55
 Sara Ali Messages: 15Registered: April 2019 Member
hello sir, i am working with PDHS 2017-18 data for measuring the determinants of neonatal mortality. kindly guide me how to generate the dependent variable neonatal?
Re: Neonatal Mortality in India NFHS-4 [message #17627 is a reply to message #17620] Mon, 29 April 2019 10:03
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
If you are working with the KR or BR file, then
gen neonatal_death = (inrange(b6,100,130))
Re: Neonatal Mortality in India NFHS-4 [message #17650 is a reply to message #17627] Thu, 02 May 2019 04:45
 Sara Ali Messages: 15Registered: April 2019 Member
M using KR file, but i have some confusions, for generating neonatal, you have taken the variable b6, i.e age at deaths but can't we use hw1 variable i.e child age in month?
Re: Neonatal Mortality in India NFHS-4 [message #17655 is a reply to message #17650] Thu, 02 May 2019 11:09
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
hw1 is age of the child in months for living children that are included in the anthropometry sample or subsample (that is it isn't necessarily set for all children). Most importantly it is only living children, which doesn't help you for neonatal mortality.
Re: Neonatal Mortality in India NFHS-4 [message #17659 is a reply to message #17655] Fri, 03 May 2019 03:09
 Sara Ali Messages: 15Registered: April 2019 Member
and i have checked the association of dependent variable i.e neonatal with some independent variables but i am not sure that which variable should i select for mothers education, (v106 or v107)?,and for poverty s816h can be used or not? similarly for child birth weight (m19 or m19a)?
Re: Neonatal Mortality in India NFHS-4 [message #17662 is a reply to message #15076] Fri, 03 May 2019 11:26
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
Please read the Recode Manual and the Guide to DHS Statistics for answers to these and other similar questions.
For education, v106 is the highest grade of education attended and v107 is the highest year or grade completed WITHIN that level. You would not use v107 without also using v106. Please also see v133 which provides the total years of education, based on v106 and v107, and v149 which provides a further categorization of the levels of education.
For poverty we use the wealth index quintiles in v190 (see the Wealth Index topic page).
For weight at birth, m19 provides the weight, but is typically only reported by about half of respondents - other simply don't know it. m19a is not the weight, but provides information about the source of the birth weight.
You might also look at m18 - size at birth, which is typically provided by almost all women, and gives an estimate of the size of the child.
Re: Neonatal Mortality in India NFHS-4 [message #17669 is a reply to message #17662] Sun, 05 May 2019 10:36
 Sara Ali Messages: 15Registered: April 2019 Member
Thanks alot sir, but i am confused between some other variables after reading the manual. antenatal care (m57q or m57f), family planning (s325), partner's or husband's education level (v701 or v702), breastfeeding is also not clear (m4 or v3a08g) and the last is the type of place of residence (v025 or v102).
Re: Neonatal Mortality in India NFHS-4 [message #17671 is a reply to message #15076] Sun, 05 May 2019 16:17
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
It is not clear what your confusion is with these variables, please provide more information. m57e-m57l are public sector facilities for antenatal care, while m57m-m57r are private sector facilities. For partner's education, v701 is partner's highest level of education attended, and v702 is the highest grade completed at that level (just like v106 and v107). For breastfeeding, see m4 (v3a08g is a reason given for not using contraception). v025 and v102 are identical - you can use either. s325 is the last source of family planning. usually v312 is the variable that you want to look at concerning family planning.
Re: Neonatal Mortality in India NFHS-4 [message #17672 is a reply to message #17671] Mon, 06 May 2019 04:17
 Sara Ali Messages: 15Registered: April 2019 Member
Thanks again sir, one more query. which variable should i use for infections as infections are the major cause of neonatal mortality. ?
Re: Neonatal Mortality in India NFHS-4 [message #17677 is a reply to message #15076] Mon, 06 May 2019 10:44
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
There is no variable for infections available.
Re: Neonatal Mortality in India NFHS-4 [message #17685 is a reply to message #17627] Tue, 07 May 2019 07:34
 Sara Ali Messages: 15Registered: April 2019 Member
Sir what does this 100 and 130 shows? why do we use this range ??
Re: Neonatal Mortality in India NFHS-4 [message #17692 is a reply to message #15076] Tue, 07 May 2019 09:55
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
See the Recode Manual page 46 for an explanation of the coding of b6. The first digit means the age at death was reported in days, and the last two digits means the number of days, thus this is between 0 days and 30 days.
Re: Neonatal Mortality in India NFHS-4 [message #17693 is a reply to message #17692] Tue, 07 May 2019 11:56
 Sara Ali Messages: 15Registered: April 2019 Member
okay sir, Thank you so much.
Re: Neonatal Mortality in India NFHS-4 [message #17703 is a reply to message #17662] Tue, 07 May 2019 23:52
 Sara Ali Messages: 15Registered: April 2019 Member
sir i need stata commands for replicating table 1.1 (Results of the household and individual interviews Number of households, number of interviews, and response rates, according to residence (unweighted), Pakistan DHS 2017-18) and table 3.1.1 (Background characteristics of respondents Percent distribution of ever-married women and men age 15-49 by selected background characteristics, Pakistan DHS 2017-18)
Re: Neonatal Mortality in India NFHS-4 [message #17762 is a reply to message #17692] Thu, 23 May 2019 05:35
 Sara Ali Messages: 15Registered: April 2019 Member
which variable should i use for diarrhea (h11 or any), fever(h22 or any), cough (h31 or any), and HIV (don't know) affecting the neonatal mortality? IS there any other disease which must be added affecting the neonatal mortality in the data ?
Re: Neonatal Mortality in India NFHS-4 [message #17767 is a reply to message #17762] Thu, 23 May 2019 12:13
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
Data on diarrhea, fever, and cough are only collected on children who are alive at the time of interview and related to the 2 weeks preceding the survey.
Re: Neonatal Mortality in India NFHS-4 [message #17772 is a reply to message #17767] Sat, 25 May 2019 07:54
 Sara Ali Messages: 15Registered: April 2019 Member
ok, then which variable should i use for infections? as infection is the major independent variable causing neonatal deaths
Re: Neonatal Mortality in India NFHS-4 [message #17773 is a reply to message #15076] Sat, 25 May 2019 15:12
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
There is no information about infections for the neonatal deaths. The survey was not designed to collect this information.
Re: Neonatal Mortality in India NFHS-4 [message #17815 is a reply to message #17773] Tue, 11 June 2019 08:52
 Sara Ali Messages: 15Registered: April 2019 Member
In univariate analysis, how could i plot Percentage Distribution of Mother's by Age. v012 is respondents current age, but i recode it to MA as 15-19, 20-24, 25-29, >29. which variable should i use v012 or MA ? and what command in R or stata to plot a bar chart.
Re: Neonatal Mortality in India NFHS-4 [message #17829 is a reply to message #17655] Thu, 20 June 2019 03:46
 Sara Ali Messages: 15Registered: April 2019 Member

gen neonatal_death = (inrange(b6,100,130))
this command is not working for generation of new variable, what would be the reason.? the error is
"type mismatch"

Re: Neonatal Mortality in India NFHS-4 [message #17912 is a reply to message #17773] Mon, 15 July 2019 01:14
 Sara Ali Messages: 15Registered: April 2019 Member
what is the age of children taken in PKBR71FL data file ? means children less than 59 months or only neonatal ?
Re: Neonatal Mortality in India NFHS-4 [message #17913 is a reply to message #17912] Mon, 15 July 2019 09:59
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
PKBR71FL includes all children born to the respondents of any age. PKKR71FL is limited to children age 0-59 months.
Re: Neonatal Mortality in India NFHS-4 [message #17919 is a reply to message #17913] Thu, 18 July 2019 02:08
 Sara Ali Messages: 15Registered: April 2019 Member
is there any information regarding neonatal infections in PDHS 2017-18 dataset?
Re: Neonatal Mortality in India NFHS-4 [message #17926 is a reply to message #17919] Thu, 18 July 2019 10:07
 Trevor-DHS Messages: 721Registered: January 2013 Senior Member
Not to my knowledge
Re: Neonatal Mortality in India NFHS-4 [message #17969 is a reply to message #17919] Fri, 02 August 2019 10:23
 Riyaan ahmd Messages: 2Registered: July 2019 Member
Yes
Dear sara
In pdhs survey have this information
You can contact to me
Re: Neonatal Mortality in India NFHS-4 [message #17970 is a reply to message #17912] Fri, 02 August 2019 13:07
 Riyaan ahmd Messages: 2Registered: July 2019 Member
You can take help from me
Re: Neonatal Mortality in India NFHS-4 [message #19906 is a reply to message #17970] Wed, 26 August 2020 19:08
 Matilda chinonye Messages: 1Registered: August 2020 Member
Please I am confused on how to use cox ph model for neonatal, infant mortality and child mortality. I have gotten this variables form github but I don't know how to fit these variables as survival data. Please I need help
Re: Neonatal Mortality in India NFHS-4 [message #19912 is a reply to message #19906] Thu, 27 August 2020 14:33
 Shireen-DHS Messages: 15Registered: August 2020 Location: USA Member
Dear user,

These are aggregate rates and cannot be used for survival analysis. Survival analysis is at the individual level.

Thank you.

Shireen
 Previous Topic: Mortality rates and binary variable Next Topic: Previous pregnancy birth interval in months
Goto Forum:

Current Time: Sat Sep 26 00:37:43 Eastern Daylight Time 2020