The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in Stata » Estimating preterm birth
Estimating preterm birth [message #18709] Sat, 01 February 2020 11:05 Go to next message
Ghose is currently offline  Ghose
Messages: 8
Registered: May 2018
Member
Hi,

I'm working with IR datasets from several countries in sub-Saharan Africa. I'd like to ask if there is any way to estimate preterm birth from women's daraset, as this variable doesn't seem to appear anywhere.

Thanks in advance for your suggestions!
Ghose
Re: Estimating preterm birth [message #18735 is a reply to message #18709] Fri, 07 February 2020 11:04 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1932
Registered: February 2013
Senior Member

Following is a response from Senior DHS Specialist, Kerry MacQuarrie:

Please look at the b20 variables in the IR file. This is duration of pregnancy. The suffix _0x refers to the birth index number (birth order).

Re: Estimating preterm birth [message #19149 is a reply to message #18735] Tue, 28 April 2020 09:16 Go to previous messageGo to next message
lillo?S is currently offline  lillo?S
Messages: 24
Registered: December 2015
Member
Hello, I understand that the variable b20_0x can take values up to 10. I was wondering what these values mean. Does they refer to 30/31-day months or 28-day months? I would like to construct a preterm birth variable and have doubts on how to construct it; preterm is defined as babies born alive before 37 weeks of pregnancy are completed, so I think here it is important to know how this variable was constructed. Thank you very much for clarifying this.
Re: Estimating preterm birth [message #19152 is a reply to message #18735] Thu, 30 April 2020 01:13 Go to previous messageGo to next message
mianrashid is currently offline  mianrashid
Messages: 13
Registered: February 2016
Location: Pau, France
Member
Hello,

How we calculate the day of the first union in given month and year?


MianRashid
Re: Estimating preterm birth [message #19160 is a reply to message #19149] Fri, 01 May 2020 10:11 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1932
Registered: February 2013
Senior Member
This is a multiple occurring variables, and each occurrence of every multiple variable must have its own name because statistical packages do not generally support the use of arrays or subscripts. For example, the third occurrence of the variable M18 would be named M18$03 in SPSS, or M18_03 in SAS and STATA.

You might want to use the Child file (KR) or the Births file (BR), because in these files, you will have just one variable for each observation.
Re: Estimating preterm birth [message #19161 is a reply to message #19160] Fri, 01 May 2020 10:24 Go to previous messageGo to next message
lillo?S is currently offline  lillo?S
Messages: 24
Registered: December 2015
Member
Hi Brigette, I am using the child file (KR). The issue was related to what the duration of pregnancy means. If preterm is defined as babies born alive before 37 weeks of pregnancy are completed, do we define a preterm birth as duration of pregnancy (b20) being smaller than 9 or smaller than 10, where smaller is < (not ≤)?
Re: Estimating preterm birth [message #19197 is a reply to message #19161] Fri, 08 May 2020 13:34 Go to previous messageGo to next message
lmallick is currently offline  lmallick
Messages: 22
Registered: June 2019
Member
Dear user,

We do not have precise information on preterm births (see post: https://userforum.dhsprogram.com/index.php?t=msg&goto=11 535&S=Google). We ask about the number of months a woman was pregnant in calendar months (not 28 days). However, it is difficult to assess gestational age is limited in low-resource settings (see this article, for example: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5997365/#S1). That said, I would use less than 9 months as an approximation.

You could also use low birthweight as a proxy, as low birthweight and preterm birth are associated, although in places where birthweight is not often assessed or recorded on cards that mothers have possession of to present during the interview, perceived size at birth is also subject to error ( https://bmjopen.bmj.com/content/bmjopen/9/12/e031986.full.pd f).

Thanks,
Lindsay Mallick
Lead Research Analyst
Re: Estimating preterm birth [message #19198 is a reply to message #19197] Fri, 08 May 2020 14:29 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 1932
Registered: February 2013
Senior Member
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

When you have the month and year of an event, but not the day, and some calculation depends on knowing the day, we usually assign the 15th of the month. That will minimize the error. I think you may find exceptions, in which the 1st of the month is assigned (or is implied by the way the calculation is done). But usually the 15th.
Previous Topic: BMI
Next Topic: Multilevel analysis-Pakistan DHS
Goto Forum:
  


Current Time: Tue Sep 29 22:47:29 Eastern Daylight Time 2020