The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Topics » Child Health » Measuring postnatal care for baby (M70)
Measuring postnatal care for baby (M70) [message #9007] Tue, 26 January 2016 16:34 Go to next message
lmcdougal
Messages: 17
Registered: March 2013
Member
I'm trying to perform a cross-national analysis of DHS data that includes a variable assessing whether or not a baby received postnatal care within a certain time period, using M70/M71. In looking through the recodes, M70 (baby postnatal check within 2 mos) seems to have a large number of "Not applicable" responses. I can't see where these responses are coming from based on the skip patterns in the questionnaires (they do not seem to be babies born only at home vs. in a facility as identified in M15, for example). This is the case in a number of the surveys I'm looking at, but Bangladesh 2011 and Benin 2011-12 are two examples.

Can anyone provide any clarification on where these "Not applicable"s are coming from? Thanks in advance.
Re: Measuring postnatal care for baby (M70) [message #9009 is a reply to message #9007] Tue, 26 January 2016 17:06 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3214
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:


These questions are only asked about the most recent birth. If, say, you open BDKR61FL.dta in Stata and enter "tab m70 bidx, m", you will see that all of the NA cases are for births with bidx>1.
Re: Measuring postnatal care for baby (M70) [message #9010 is a reply to message #9009] Tue, 26 January 2016 17:11 Go to previous messageGo to next message
lmcdougal
Messages: 17
Registered: March 2013
Member
Ah, I see - thank so much for the quick reply!
Re: Measuring postnatal care for baby (M70) [message #14562 is a reply to message #9007] Sat, 21 April 2018 19:34 Go to previous messageGo to next message
kingx025 is currently offline  kingx025
Messages: 95
Registered: August 2016
Location: Minneapolis. Minnesota
Senior Member
While I see that Tom Pullum answered your specific question, as a broader solution you may wish to check the "Universe" tab for variables of interest on the IPUMS-DHS website (at idhsdata.org). If you go to "select data" on the website, choose "children" as the unit of analysis, and then search for the variable m70, you can click on the variable name to get detailed documentation on the variable, including a "Universe" tab that specifies, for each sample, who is included in the coverage for this variable. In general, M70 is limited to last born children, but some samples include additional limitations.
For example, here's one of the more complicated universes for this variable:
Pakistan 2006: Last-born children under age 5 born to ever-married women age 15-49 who either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check after leaving the facility.

The IPUMS-DHS staff has empirically checked all reported universes for all samples in all variables included in the database, which now covers 117 samples and 28 countries. Checking the universes for variables and samples of interest on IPUMS-DHS's website can save you time and help you avoid inadvertent errors.

Miriam King


Dr. Miriam King
IPUMS-DHS Project Manager (www.idhsdata.org)
Re: Measuring postnatal care for baby (M70) [message #14626 is a reply to message #14562] Mon, 23 April 2018 10:25 Go to previous messageGo to next message
kingx025 is currently offline  kingx025
Messages: 95
Registered: August 2016
Location: Minneapolis. Minnesota
Senior Member
I had a request for posting the information about the universe for the M70 variable for additional countries, so I will do so here. However, anyone can check universes by consulting the documentation for the variable of interest on the IPUMS-DHS website.

Here are the universes for M70 (using children as the unit of analysis), for countries included in IPUMS-DHS at this point:
Bangladesh 2004: Last-born children under age 5 born to ever-married women age 10-49.
Bangladesh 2007: Last-born children under age 5 born to ever-married women age 15-49.
Bangladesh 2011: Last-born children under age 5 born to ever-married women age 12-49.
Bangladesh 2014: Last-born children under age 3 born to ever-married women age 15-49.
Benin 2011: Last-born children under age 5 born to women age 15-49.
Burkina Faso 2010: Last-born children under age 5 born to women age 15-49 either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check before leaving the facility.
Cameroon 2011: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Congo (Democratic Republic) 2013: Last-born children under age 5 born to women age 15-49.
Cote d'Ivoire 2011: Last-born children under age 5 born to women age 15-49.
Ethiopia 2011: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Ethiopia 2016: Last-born children under age 5 born to women age 15-49.
Ghana 2008: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Ghana 2014: Last-born children under age 5 born to women age 15-49.
Guinea 2012: Last-born children under age 5 born to women age 15-49.
India 2005: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
India 2015: Last-born children under age 5 born to women age 15-49.
Kenya 2008: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Kenya 2014: Last-born children under age 5 born to women age 15-49 whose households were selected for the men's survey.
Madagascar 2008: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Malawi 2010: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Mali 2012: Last-born children under age 5 born to women age 15-49.
Model 2015: Last-born children under age 5 born to women age 15-49.
Niger 2012: Last-born children under age 5 born to women age 15-49.
Nigeria 2008: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Nigeria 2013: Last-born children under age 5 born to women age 15-49.
Pakistan 2006: Last-born children under age 5 born to ever-married women age 15-49 who either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check after leaving the facility.
Pakistan 2012: Last-born children under age 5 born to ever-married women age 15-49.
Rwanda 2010: Last-born children under age 5 born to women age 15-49 either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check before leaving the facility.
Rwanda 2014: Last-born children under age 5 born to women age 15-49 either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check before leaving the facility.
Tanzania 2010: Last-born children under age 5 born to women age 15-49 and either a) delivered this child away from a medical facility, or b) delivered the child in a medical facility and did not receive a health check before they left the birth facility.
Tanzania 2015: Last-born children under age 5 born to women age 15-49 whose delivery of this child did not take place at a traditional birth attendant's space.
Uganda 2006: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Uganda 2011: Last-born children under age 5 born to women age 15-49.
Yemen 2013: Last-born children under age 5 born to ever-married women age 15-49.
Zambia 2007: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Zambia 2013: Last-born children under age 5 born to women age 15-49.
Zimbabwe 2005: Last-born children under age 5 born to women age 15-49 and delivered this child away from a health facility.
Zimbabwe 2010: Last-born children under age 5 born to women age 15-49 either a) delivered this child away from a medical facility, or b) if delivery took place in a medical facility, received a health check before leaving the facility.
Zimbabwe 2015: Last-born children under age 5 born to women age 15-49.


Dr. Miriam King
IPUMS-DHS Project Manager (www.idhsdata.org)
Re: Measuring postnatal care for baby (M70) [message #15047 is a reply to message #9007] Mon, 28 May 2018 08:16 Go to previous messageGo to next message
Vaskset is currently offline  Vaskset
Messages: 12
Registered: May 2018
Member

Dear Sir/Madam,
I am comparing postnatal care for newborns in 2008 and 2014 in Ghana. I do not see "postnatal care for newborns' in the 2008 GHDS report and so I calculated it using inputs from some post on this forum but will like to know if the syntax is correct.

*###########conditions for postnatal care####
* age in months
gen age = v008-b3
lab var age "Age in months"
ta age
sort caseid bidx
* keep if under 24 months and living with mother
keep if age<25 & b9==0
codebook b9
* ... and keep the last born of those. If caseid is the same as the prior case, then not the last born.
drop if _n > 0 & caseid == caseid[_n-1]

gen posnababy=m71
replace posnababy=1001 if m71<=202 & m71!=.
replace posnababy=1000 if m71>202 & m71<=316
replace posnababy=1000 if m71==998 |m71==999
ta posnababy
Thank you.
Re: Measuring postnatal care for baby (M70) [message #15058 is a reply to message #15047] Tue, 29 May 2018 09:28 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3214
Registered: February 2013
Senior Member
Following is a response from Senior DHS Stata Specialist, Tom Pullum:

Usually, v008-b3 will be the same as hw1, and you can use hw1. However, I see that in this file there are 15 cases in which v008-b3=hw1+1. Why, I don't know (I will look into that), so go ahead and use v008-b3.

Your method for selecting the youngest child living with the mother is not quite correct. After "keep if age<25 & b9==0" you could enter these lines, which will sort each woman's children who are living with her by age, and then select the youngest one:

sort age
egen sequence=seq(),by(caseid)
tab sequence
drop if sequence==1
drop sequence
I don't understand what you are trying to do with your recode of m71. Why are you recoding it into 1000 and 1001?

Re: Measuring postnatal care for baby (M70) [message #15095 is a reply to message #15058] Fri, 01 June 2018 10:00 Go to previous messageGo to next message
Vaskset is currently offline  Vaskset
Messages: 12
Registered: May 2018
Member
Hello Bridgette,
Thanks for the response.
1000 and 1001 represent 0 and 1 respectively.

I made some changes in the syntax but it still does not give the expected result (6.3%) and the number of observations keep changing.
Your further assistance will be appreciated.

Below is the code.
gen age = v008-b3
lab var age "Age in months"
ta age
sort age
egen sequence=seq(),by(caseid)
tab sequence
drop if sequence==1
drop sequence

drop if m71==.
gen postnatal=m71
replace postnatal=1 if m71==100|m71==101| m71==201 |m71==202
recode postnatal 102=1
replace postnatal=0 if m71>202
ta postnatal
Re: Measuring postnatal care for baby (M70) [message #15151 is a reply to message #15095] Fri, 08 June 2018 14:17 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3214
Registered: February 2013
Senior Member
Following is a response from DHS Senior Research Associate, Lindsay Mallick:

Please see program (attached and below) for calculating PNC. PNC is calculated for mothers and babies for the most recent births in the two years preceding the survey. Be sure to exclude women who could not identify the provider of the check. Also, note the way we collect information and calculate PNC has become more precise in DHS-7 surveys. See program attached and below.

*Generate weight
iw=v005/1000000

*DHS-7 and later surveys will have the b19 variable, age in months 
*To create age in months in older surveys, run the following lines
               capture confirm variable b19
               if _rc { // b19 does not exist, so create equivalent for old calculation method
                              gen b19 = v008 - b3
                              label variable b19 "Age of child in months or months since birth"
               }



*********************************************************

*PNC - 2008- DHS surveys
               
               *pnc timing         
               recode m71 ( 100/202 =2 "PNC within 2 days") (203/305= 1 "PNC 3-41 days") (900/1000 =9 "Don't know, missing") (else=0 "No PNC or after 41 days") if bidx==1, gen(pnctimebaby) 
                recode pnctimebaby (1 0 9= 0 "No PNC, pnc 2+days or dk/miss") (2=1 "PNC within 2 days") if bidx==1, gen(pnc2dbaby) 
                lab var pnctimebaby "Timing of PNC for child"
               lab var pnc2dbaby "Postnatal care within 2 days"
               gen pnc2dbaby = 0
               ta pnc2dbaby if bidx==1 & b19<24 [iw=wt], m
               
               *provider of pnc 
                recode m72 (11/29 = 1 "Skilled provider") (96 =0 "Non-skilled provider") ( else = 9 "Don't know, missing") if bidx==1, gen (pncprovbaby) 
                gen pncprovbaby = 0 
               
               *pnc within 2 days
               *PNC calculated if it was given by provider                           
               gen skillpnc2dbaby = 0 
               replace skillpnc2dbaby = 1 if pnc2dbaby==1 & pncprovbaby==1 
               lab var skillpnc2dbaby "Postnatal care within 2 days by skilled provider"
               lab val skillpnc yesno
               ta skillpnc2dbaby if bidx==1 & b19<24 [iw=wt],m

**************

*For DHS7 surveys:

*!RUN ABOVE CODE AND THE FOLLOWING

               *generate general variable for if they had a check at all or not
               gen check =0 
               replace check =1 if (m70==1 |m74==1)
               
               *for those newer surveys
               replace skillpnc2dbaby = 0 if check ==0 
               
               *Now look at babies born in a facility, if checked before discharge             
               *provider of pnc check - health facility-born babies
               recode m76 (11/29 = 1 "Skilled provider") (96 =0 "Non-skilled provider") ( else = 9 "Don't know, missing")  if check==1 , gen (pncprovbaby_hf)

               *code timing of check if health facility delivery
               recode m75 ( 100/202 =2 "PNC within 2 days") (203/305= 1 "PNC 3-41 days") (900/1000 =9 "Don't know") (else=0 "No PNC or after 41 days")  if check==1, gen(pnctimebaby_hf)
               ta pnctimebaby_hf  [iw=wt]
               
               recode pnctimebaby_hf (1 0 9= 0 "No PNC, pnc 2+days or dk/miss") (2=1 "PNC within 2 days") if check==1 , gen(pnc2dbaby_hf)
               lab var pnc2dbaby_hf "Postnatal care within 2 days"
               ta pnc2dbaby_hf  bidx if b19<24 [iw=wt], m

               *PNC calculated if it was given by medically trained provider                       
               gen skillpnc2dbaby_hf = 0 if bidx==1 
               replace skillpnc2dbaby_hf = 1 if pnc2dbaby_hf==1 & pncprovbaby_hf==1 &check==1
               lab var skillpnc2dbaby_hf "Postnatal care within 2 days by skilled provider"
               ta skillpnc2dbaby_hf bidx if b19<24 [iw=wt], m

               ************
               *combine two timing variables
               gen newpnctimebaby =  skillpnc2dbaby_hf
               *if they delivered at home or only got checked after discharge
               replace newpnctimebaby = 1 if skillpnc2dbaby==1 & (facdel==0 | facdel == 9 | (facdel ==1 & m74==0 ))
               label var newpnctimebaby "Postnatal care within 2 days by skilled provider"
               tab newpnctimebaby bidx if b19<24  [iw=wt], col
               
Re: Measuring postnatal care for baby (M70) [message #15171 is a reply to message #15151] Wed, 13 June 2018 00:27 Go to previous message
Hassen
Messages: 121
Registered: April 2018
Location: Ethiopia,Africa
Senior Member
Thank you very much,Bridgette and Lindsay Mallick!! I have learned a lot from your posts.
Respectfully,Hassen


Hassen Ali(Chief Public Health Professional Specialist)
Previous Topic: Definition of ARI symptoms
Next Topic: Vaccination coverage using 2013 ndhs data set: Nigerian & dhs
Goto Forum:
  


Current Time: Thu Dec 26 09:27:50 Coordinated Universal Time 2024