The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » General Data Questions » IDHS 2017 (Date of birth)
IDHS 2017 [message #29356] Wed, 05 June 2024 11:21 Go to next message
Patriciaje is currently offline  Patriciaje
Messages: 7
Registered: April 2024
Member
Hello,

I'm conducting a research using the 2017 Indonesia DHS data and I really need the child's date of birth in days / century day code (CDC) for my analysis (I believe it's B17 and B18). Unfortunately, this information is marked as NA in the dataset. Is there any way I can access this data? It is crucial for my research as i can't progress without it.

Thank you so much for your help.

Regards,

[Updated on: Wed, 05 June 2024 12:24]

Report message to a moderator

Re: IDHS 2017 [message #29363 is a reply to message #29356] Thu, 06 June 2024 11:26 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3082
Registered: February 2013
Senior Member

Following is a response from Senior DHS staff member, Tom Pullum:

I just looked at the PR and KR files, which normally would include day of birth for children under 5. You are right, day of birth is missing, but day is included for the date of vaccinations.

The omission of day of birth for children under 5 may be due to the fact that this survey did not include anthropometry. For the calculation of the HAZ, WAZ, and WHZ it is essential to be able to calculate days of age (as date of measurement minus date of birth). But I wish it had been included! Soon after that survey, DHS started including day of birth for ALL births in the birth history.

I confirmed that age in years (b8) was calculated as int((v008-b3)/12). Here, v008 is the cmc of interview and b3 is the cmc of birth. This calculation ignores the day of the interview (even though it is available) as well as the day of birth (which was omitted). The assumption in this calculation is that the day of birth is between 1 and the day of the interview, inclusive--that is, if the month of interview is the same as the month of birth, the child has already had the birthday by the date of interview.

I could suggest a way to impute the day of birth, and then calculate a cdc, but I hesitate to do this. The hepatitis B birth dose is supposed to be within 24 hours of birth. The day of that dose may well provide the best estimate of the day of birth. An imputation procedure could not legally produce a birth day that was AFTER the date of the first immunization. Sorry I can't be more helpful.
Re: IDHS 2017 [message #29364 is a reply to message #29363] Thu, 06 June 2024 13:37 Go to previous messageGo to next message
Patriciaje is currently offline  Patriciaje
Messages: 7
Registered: April 2024
Member
Thank you so much for the reply,

This might be a bit off topic but how did you calculate the percentage of hepatitis B birth dose vaccination in table 10.2? It has the percentage of vaccination at 0 days (at birth) and more than 1 days (in the indonesian 2017 DHS report). How did you calculate this without the date of birth? I believe there's no specific question about it and only the date of vaccination are available. Did you calculate it with CMC or is there any other method? Can you show me how to do it in stata/spss?

Than you so much, I really appreciate for your assistance

[Updated on: Thu, 06 June 2024 13:39]

Report message to a moderator

Re: IDHS 2017 [message #29374 is a reply to message #29364] Mon, 10 June 2024 07:19 Go to previous messageGo to next message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3082
Registered: February 2013
Senior Member

Following is a response from Senior DHS staff member, Tom Pullum:

I just looked at Table 10.2. There is a footnote (4) for the Hepatitis B birth dose. In the questionnaire, I found Q513A (on page 486 of the final report). It asks specifically about the timing of the birth dose. None of the other immunizations are specified in terms of days, and for them "months of age=v008-b3" is sufficient.
Re: IDHS 2017 [message #29375 is a reply to message #29374] Mon, 10 June 2024 08:09 Go to previous messageGo to next message
Patriciaje is currently offline  Patriciaje
Messages: 7
Registered: April 2024
Member
Thank you again for the response,

Yes, i also noticed Q513A but i am not sure which variables it reflects into in the dataset, was it H50?

Also i need to confirm about the footnote in the english version of the IDHS2017 report, does it means that the children whose information is based on the mother's report (H50 code 2) received vaccine within 24 hours, and children who have the vaccine recorded in card (H50 code 1 ) is also considered to have received the vac within 24 hours? Then, how about those who have vaccination marked card but no date stated (H50 code 3)? And what fell into the scope of more than 1 days?

I want to match this part that I highlighted in the Indonesian version of the report (I attached the picture). I'm sorry that I keep asking about this, but it's still not clear to me how to get the percentage in Table 10.2 in the Indonesian version of the IDHS2017 that divides HB0 percentage into 0 days and more than 1 day.
  • Attachment: IMG_2202.jpeg
    (Size: 703.61KB, Downloaded 13 times)
Re: IDHS 2017 [message #29376 is a reply to message #29375] Mon, 10 June 2024 11:01 Go to previous message
Bridgette-DHS is currently offline  Bridgette-DHS
Messages: 3082
Registered: February 2013
Senior Member

A couple of clarifications from Senior DHS Staff Member, Guillermo Rojas:

• Question Q513A is asked to mothers that reported that the child didn't have a vaccination card or other document with vaccines registered. This question is taken into consideration in tables reporting vaccinations for this and other vaccines registered in questions Q513A to Q524A.
• If the child has a vaccination card or other document, the information is taken from Q508A and questions Q513A to Q524A are skipped. In this case as vaccines are registered in the card/document, essentially interviewers copy the data from the card/document to the questionnaire. We check the dates at secondary editing time with the date of birth (in this case just month and year), but we trust the data registered. In CAPI (this survey wasn't CAPI) we check dates at data collection and secondary editing times. In other words, if the date is registered in the card, we report that the vaccine was given.
Previous Topic: Validation challenges: v169b with DHS Statcompiler
Next Topic: Birth registration for all household members
Goto Forum:
  


Current Time: Sun Jun 16 15:03:15 Coordinated Universal Time 2024