The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Bangladesh » Postnatal Care (variable contain values that is not understandable)
Postnatal Care [message #19769] Wed, 12 August 2020 20:30 Go to next message
Samia Aziz is currently offline  Samia Aziz
Messages: 5
Registered: August 2020
Member
I am using data for postnatal care and looking for a variable "how many days or weeks after delivery did the first check up take place? the variable is coded as m51_1 in 2014 dataset. the variable contain value like 102, 103, 120 and so on. I am confused about what these values are referring in terms of time?
  • Attachment: Capture.JPG
    (Size: 117.12KB, Downloaded 36 times)
Re: Postnatal Care [message #19783 is a reply to message #19769] Fri, 14 August 2020 09:13 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2025
Registered: February 2013
Senior Member
The first digit is the unit the information was reported (1=hours, 2=days, 3=weeks, 9=special answers). The last two digits is the time after delivery that the first checkup took place in those units.
Re: Postnatal Care [message #19826 is a reply to message #19783] Tue, 18 August 2020 23:47 Go to previous messageGo to next message
Samia Aziz is currently offline  Samia Aziz
Messages: 5
Registered: August 2020
Member
Thanks for the reply. I am working on BDHS 2004, and looking at the variable m52_1 (who checked on your health after delivery). I am trying to recode the variable as 'medically trained providers' and 'others'. But the variable has no code for the outcome though the type is numeric (byte), i tried to generate the new variable considering as a string variable but it did not work out. can you please help.
  • Attachment: m52_1.JPG
    (Size: 164.90KB, Downloaded 37 times)
Re: Postnatal Care [message #19832 is a reply to message #19826] Wed, 19 August 2020 10:52 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 2025
Registered: February 2013
Senior Member

Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

What you see in the tabulation of the variables is the labels (Doctor, etc.) the variable itself is numeric. One way to see the codes (there are other ways) is with "tab m52_1" and then "tab m52_1, nolabel". You will then see the numericl codes that go with the labels. You should then be able to do the recode you want to do. Let us know if you still have trouble.
Re: Postnatal Care [message #19958 is a reply to message #19832] Tue, 08 September 2020 09:11 Go to previous messageGo to next message
Samia Aziz is currently offline  Samia Aziz
Messages: 5
Registered: August 2020
Member
I am using the BDHS 2014 dataset, I am interested to look into the women who received postnatal care. so I decided to select the sample of women who had given birth within three preceding years of the survey using variable if v417>0. Total number of women stands at 4494. However, when I am trying to look into the proportion of women who received PNC using the variable m66_1, the total number of respondents shows as 2806. I wonder was the question for all women who had a birth in the last three years of the survey? then why does the N vary? Please guide.
Re: Postnatal Care [message #19988 is a reply to message #19958] Fri, 11 September 2020 10:09 Go to previous messageGo to next message
Shireen-DHS is currently offline  Shireen-DHS
Messages: 45
Registered: August 2020
Location: USA
Member
Hello Samia,

If you are trying to match Table 9.10 of the BDHS 2014 Final report you need to perform several checks to match the Table. For instance, you would need to see the type of provider giving the postnatal check.

Please study the code below that will give you a match with the table.

Thank you.

Best,
Shireen Assaf
The DHS Program

*child's age, this is used to select for births in the last 3 years
gen age = v008 - b3_01

//PNC timing for mother
*did the mother have any check
gen momcheck = 0 if age<36
replace momcheck = 1 if (m62_1==1 | m66_1==1) & age<36

*PNC timing for mother
recode m51_1 (100/103 = 1 "<4hr") (104/123 200 = 2 "4-23hrs") (124/171 201/202 = 3 "1-2 days") ///
(172/197 203/206 = 4 "3-6 days") (207/241 301/305=5 "7-41 days") ///
(198/199 298/299 398/399 998/999 = 9 "dont know/missing") (242/297 306/397 = 0 "no pnc check") (else=0) if (age<36 & momcheck==1) , g(rh_pnc_wm_timing)
replace rh_pnc_wm_timing = 0 if m50_1==0 | m50_1==9
*recode as no PNC if not performed by a trained provider
replace rh_pnc_wm_timing = 0 if (m52_1>15 & m52_1<97) | m52_1==.
replace rh_pnc_wm_timing=. if age>=36 | bidx_01!=1
label var rh_pnc_wm_timing "Timing after delivery for mother's PNC check"

gen wt=v005/1000000

tab rh_pnc_wm_timing [iw=v005/1000000]
Re: Postnatal Care [message #20044 is a reply to message #19988] Mon, 21 September 2020 19:34 Go to previous messageGo to next message
Samia Aziz is currently offline  Samia Aziz
Messages: 5
Registered: August 2020
Member
I am using BDHS 2007 data and looking at variables of Postnatal care for mothers and trying to match the table 9.8.2 of 2007 report. However 'who checked on your health is not coded as m52_1 like other rounds(2011, 2014). I have found s434a_1 instead that gives information only for health checked by a qualified doctor. I want to create a variable of care provider as 'medicaly trained provider' and 'other providers'. but 2007 data contain a variable for each provider (s434a_1 to s434x_1) that made me confused on how to recode these variables as one variable.

I am also struggling in getting the value of 16% PNC coverage by a medically trained provider within 2 days of birth for the 2004 dataset (as mentioned in the 2014 report)

Can you please help me? I would highly appreciate your help or suggestion in this regard. hope to hear from you soon.
Re: Postnatal Care [message #20103 is a reply to message #20044] Fri, 25 September 2020 09:45 Go to previous message
Shireen-DHS is currently offline  Shireen-DHS
Messages: 45
Registered: August 2020
Location: USA
Member
Hello Samia,

You are correct in that the PNC variables have changed over the years. Please check the Guide to DHS Statistics to read about these changes: https://www.dhsprogram.com/Data/Guide-to-DHS-Statistics/inde x.htm#t=Postnatal_Care.htm
Check the section on "Changes over time".

Please also check our code that is posted on the GitHub site for this and other indicators. The code is mostly created to match recent surveys. We try to make corrections for older surveys when possible. The PNC chapter is Chapter 9. Please read the readme file and main files before trying to run any of this code. https://github.com/DHSProgram/DHS-Indicators-Stata

For the 2007 BDHS Postnatal checkup for women, the variable m51_1 is replaced by m51a_1. To get the provider breakdown you use the "s" variables as shown below. I followed the footnote of the table to know where to place the categories. Label 1 is for the qualified doctor in the table. I was a little off for the qualified doctor by 0.5 percentage points and I could not resolve this. The other provider categories match and the denominator matches. For age, if b19 variable is available in the dataset (this was added in recent surveys), please use that for age instead of how age is generated below.

Hope this helps. Please refer to the links I provided as well, these can be very useful resources.

Thank you.

Best,
Shireen
The DHS Program


*Stata code for table 9.8.1 of Bangladesh 2007 Final report

gen age = v008 - b3_01
gen m51_1=m51a_1
gen provider=0 if m51_1!=.
replace provider = 0 if m66_1==0 | m66_1==9
replace provider=4 if s434f_1==1 | s434g_1==1 |s434h_1==1 | s434i_1==1 | s434j_1==1 | s434x_1==1
replace provider=3 if s434d_1==1 | s434e_1==1
replace provider=2 if s434b_1==1 | s434c_1==1
replace provider=1 if s434a_1==1

Previous Topic: How to calculate mother's age at most recent birth (last birth) in BDHS 2011
Goto Forum:
  


Current Time: Sat Nov 28 12:17:44 Coordinated Universal Time 2020