| Stillbirth [message #30192] |
Fri, 11 October 2024 08:51  |
Majani
Messages: 1
Registered: October 2024
|
Member |
|
|
Hi,
I would love to determine the exposure-response relationship between heat stress and stillbirth (for single births only) in Kenya, Tanzania and Uganda using DHS-8 (2022) data for Kenya and Tanzania and DHS-2018-19-MIS data for Uganda. My main problem is how do I estimate the exact date of stillbirth (variable p32), including the date, month and year of stillbirth for all single births for use in my analysis.
Best,
Elizabeth.
|
|
|
|
| Re: Stillbirth [message #30212 is a reply to message #30192] |
Tue, 15 October 2024 14:01  |
Janet-DHS
Messages: 893
Registered: April 2022
|
Senior Member |
|
|
Following is a response from DHS staff member, Tom Pullum:
The Kenya and Tanzania surveys were part of DHS-8, as you say, and they have NR files for pregnancies in the past 3 years (p19=0 through 35). Open one of the NR files and enter "describe p*". You will see that the day, month, and year of the end of the pregnancy are given by p17, p1, and p2, respectively. Single pregnancies are identified by p0=0. You probably want to restrict to cases with p32=1 or 2 (live births and stillbirths) and try to fit Pr(p32=2).
If you want a longer time period than the past 3 years, for example the past 10 years, you can reshape the p variables in the IR file.
The Uganda survey did not include pregnancy histories. I do not have time to check, but if this survey included a contraceptive calendar, you can identify stillbirths as terminations (T) preceded by 6 or more months of pregnancy (P), and the month and year when they occurred, but it is much more work and you can't go back more than 5 years. The DHS Guide to Statistics and calendar tutorial would help you analyze the calendar.
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
