The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Countries » Ethiopia » creating place of ANC variable in 2005, 2011 and 2016 EDHS (Place of ANC)
creating place of ANC variable in 2005, 2011 and 2016 EDHS [message #24677] Fri, 17 June 2022 17:43 Go to next message
gebretsh@gmail.com is currently offline  gebretsh@gmail.com
Messages: 17
Registered: June 2022
Member
Hello dear DHS experts,
I am attempting to create a variable "place of ANC" for the four rounds of the EDHS (2005-2019, the variable was not collected in 2000). I wanted to make this available in the IR file.
I tried creating the variable using a syntax shown below. Yet, I am not 100% sure whether I followed the right approach. I customized the syntax used to create ANC providers posted on github.com. But the problem in this variable is that there are mothers who are reported to have more than one place of ANC ( e.g: home and health facilities) and I could not figure this issue out. Can I ignore these women from my analysis (who had more than 1 place of ANC)? I am very grateful for your response and possibly, you could provide me with the correct stata code to make this variable.

Thanks.


Place of ANC service

gen rh_anc_place = 20 if m57f_1! = .
replace rh_anc_place = 11 if m57x_1 == 1 (other)
replace rh_anc_place = 10 if m57t_1 == 1 (NGO other health facility)
replace rh_anc_place = 9 if m57s_1 == 1 (NGO health facility)
replace rh_anc_place = 8 if m57o_1 == 1 (other private medical)
replace rh_anc_place = 7 if m57n_1 == 1 (private clinic)
replace rh_anc_place = 6 if m57m_1 == 1 (private hospital)
replace rh_anc_place = 5 if m57h_1 == 1(other public health facility)
replace rh_anc_place = 4 if m57g_1 == 1(public health post)
replace rh_anc_place = 3 if m57f_1 == 1(public health center/station)
replace rh_anc_place = 2 if m57e_1 == 1(public hospital)
replace rh_anc_place = 1 if m57a_1 == 1| m57b_1==1 (respondent home/ other home)
replace rh_anc_place = 9 if m57a_1-m57x == 9
replace rh_anc_place = . if age>=60
NB: Other m2 are missing, no value.
label define rh_anc_place 1 "home" 2 "public health hospital" 3 "public health center/station" 4 "public health post" 5 "other public health facility" 6 "private hospital" 7 " private clinic" 8 "other private medical" 9 "NGO health facility" 10 "NGO other health facility" 11 "other"
label val rh_anc_place rh_anc_place
Re: creating place of ANC variable in 2005, 2011 and 2016 EDHS [message #24691 is a reply to message #24677] Wed, 22 June 2022 10:42 Go to previous messageGo to next message
Janet-DHS is currently offline  Janet-DHS
Messages: 899
Registered: April 2022
Senior Member
Following is a response from DHS Research & Data Analysis Director, Tom Pullum:

Your Stata code looks good but I have a couple of comments. (1) You initialize rh_anc_place at 20, but your label does not include 20. Perhaps you want to have a numeric code for women who did not have any ANC visits. If so, that code should be included in the label. (2) When converting a multiple options variable such as m57* to a set of mutually exclusive categories, DHS recommends sequencing them so that higher quality options come later. I think that's what you have done, except that "home" is last on your list. It should be first, because it is the lowest quality option. (But I'm surprised that "home" is even an option for ANC visits.)
Re: creating place of ANC variable in 2005, 2011 and 2016 EDHS [message #24695 is a reply to message #24691] Wed, 22 June 2022 12:24 Go to previous message
gebretsh@gmail.com is currently offline  gebretsh@gmail.com
Messages: 17
Registered: June 2022
Member
Thanks so much dear Dr. Now, your beautiful responses have cleared my confusion away. I will keep you posted should I do not figure this issue out
with my current knowledge.
Regards,
Previous Topic: Ethnicities 2019
Next Topic: Ethnicity values across surveys
Goto Forum:
  


Current Time: Sat Dec 7 16:52:41 Coordinated Universal Time 2024