The DHS Program User Forum
Discussions regarding The DHS Program data and results
Home » Data » Dataset use in Stata » Bangladesh 2014 - Antibiotics for ARI
Bangladesh 2014 - Antibiotics for ARI [message #11611] Mon, 16 January 2017 15:37 Go to next message
fewerling is currently offline  fewerling
Messages: 19
Registered: November 2016
Location: Brazil
Member
Dear all,

We are trying to reproduce the % of children with ARI symptoms that were given antibiotics for Bangladesh 2014 (Table 10.8 in the country report).
We are using the following codes to calculate the results:

** KR dataset **
gen abari=.
replace abari=0 if h31b==1 & inlist(h31c,1,3)
egen byte _abari=anymatch(h32a h32b h32c h32d h32e h32f h32g h32h h32i h32j h32l h32n h32s h32t h32u h32v h32w), v(1)
recode abari (0=1) if _abari==1
svyset v021 [pweight=v005]
svy: prop abari



Survey: Proportion estimation

Number of strata = 1 Number of obs = 406
Number of PSUs = 267 Population size = 416668524
Design df = 266

------------------------------------------------------------ --
| Linearized
| Proportion Std. Err. [95% Conf. Interval]
-------------+---------------------------------------------- --
abari |
0 | .9158814 .0160748 .8783446 .9425932
1 | .0841186 .0160748 .0574068 .1216554
------------------------------------------------------------ --

We found an estimate of 8.4% and a weighted sample size of 417, however the results in the report are 34.2% and sample size is the same (417).

Are we missing antibiotic variables in the code?

Thanks!

Best regards,
Fernanda
Re: Bangladesh 2014 - Antibiotics for ARI [message #11621 is a reply to message #11611] Tue, 17 January 2017 16:20 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 DHS Senior Research Associate, Shireen Assaf:

The following code will match the report for prevalence of ARI and for treatment of ARI. However, like you, we (the analysis staff) cannot match the estimate for antibiotics. We are forwarding the question to our Data Processing staff to confirm the number in the report. If you do not hear from us within two weeks, please ask again.
*ari, matches
 gen ari=0
 replace ari=1 if h31b==1 & (h31c==1 | h31c==3)
 replace ari =. if age>=60 | b5==0
 
*treatment of ari, matches
 gen trtari=0 if ari==1
 replace trtari=1 if ari==1 & (h32a==1 | h32b==1 | h32c==1 | h32d==1 | h32e==1 | h32f==1 | h32g==1 | h32h==1 | h32i==1 |     h32j==1 | h32l==1 | h32n==1 | h32s==1 | h32t==1 | h32u==1 | h32v==1 | h32w==1 )
 label var trtari "received treatment for ari"

* antibiotics, does not match, I get the correct denominator but the estimate is 8.4% and not 34% as in the final report. 
 gen antibio=0 if ari==1
 replace antibio=1 if ari==1 & (s538g==1 | s538h==1 |s538i==1 |s538j==1 |s538k==1 | s538l==1 |s538m==1 )
            

[Updated on: Tue, 17 January 2017 16:56]

Report message to a moderator

Re: Bangladesh 2014 - Antibiotics for ARI [message #11626 is a reply to message #11621] Wed, 18 January 2017 00:44 Go to previous messageGo to next message
fewerling is currently offline  fewerling
Messages: 19
Registered: November 2016
Location: Brazil
Member
OK, thank you.
Re: Bangladesh 2014 - Antibiotics for ARI [message #17500 is a reply to message #11626] Tue, 02 April 2019 04:32 Go to previous message
abdullahkafi is currently offline  abdullahkafi
Messages: 1
Registered: February 2019
Member

Similar I am facing about the reported estimates of Antibiotic for ARI (34.2%).
Please response anyone from the forum or DHS data processing staff to help us to find this figure.


Previous Topic: Need Guidance- Predictors of stunting
Next Topic: hypertension
Goto Forum:
  


Current Time: Tue Sep 29 22:49:55 Eastern Daylight Time 2020