| 2013 Nigeria DHS analysis on IPTp uptake [message #8864] |
Tue, 05 January 2016 19:10  |
Chinazo
Messages: 2
Registered: January 2016
|
Member |
|
|
Hi,
I am trying to analyze the 2013 Nigerian DHS dataset on uptake of IPTp-sp and had the following issues.
I could not find the variable for the uptake of IPTp-3. Uptake of IPTp-3 is included in the 2013 final report. Kindly provide a clue on the variable used for the IPTp-3 on the report.
I also tried to generate a variable for the ANC visits to be able to calculate the percentage of women who took 1st, 2nd and 3rd dose of SP during antenatal using variable m14_1. The denominator I got is 12,931 and is different from what we have in the 2013 Nigerian DHS report which is 12,473. The data was weighted before this analysis.
I also tried cross tabulating some variables after setting the complex weighting and did not have the level of significance in the output. I got a note that the missing test statistics is because of stratum with single sampling unit.
Kindly advice on how to resolve these issues.
Thanks
Chinazo
|
|
|
|
| Re: 2013 Nigeria DHS analysis on IPTp uptake [message #8866 is a reply to message #8864] |
Wed, 06 January 2016 14:03  |
lflorey
Messages: 1
Registered: January 2016
Location: Rockville, MD USA
|
Member |
|
|
Hello,
You will need to create a variable representing 3+ doses of SP for IPTp. The standard recode variables that you will need are the variable specifying which drug the woman took to prevent malaria during her most recent pregnancy ending in a live birth (m49a_1), the variable specifying the number of doses taken (ml1_1) and the variable specifying the source of the medicine (ml2_1) to verify that the medicines were received during an ANC visit.
The variable m14_1 is not actually used in the calculation of IPTp coverage. Instead the variable mentioned above (ml2_1) is used to verify that the medicines that women reported receiving for prevention of malaria during pregnancy were received via an ANC visit. In the standard DHS6 questionnaire, this question is only asked of women who received SP and who attended ANC. I would recommend reviewing the questions in the women's question pertaining to IPTp to examine the skip patterns. In order to calculate the correct denominator you need to restrict your tabulation to women who had a live birth in the past two years. I would recommend using the variables for the date of the interview (v008) and the for the date of birth of the most recently born child (b3_01). Both of these variables are in Century-Month-Code (CMC) format. When you subtract the date of birth from the date of interview you have the number of months between these two events (or the time since birth in months). agemnth=v008-b3_01. Using this information you should be able to generate the correct denominator.
If you are using Stata, when you svyset your data you can add an option which will adjust for strata with single sampling units.
singleunit(method); method may be missing, certainty, scaled, or centered. Please see help svyset in Stata for more details.
The code would look something like: svyset [pw=wgt], psu(v001) strata(v022) singleunit(centered).
Good luck!
|
|
|
|
| Re: 2013 Nigeria DHS analysis on IPTp uptake [message #8877 is a reply to message #8866] |
Sun, 10 January 2016 10:38 |
Chinazo
Messages: 2
Registered: January 2016
|
Member |
|
|
Hi,
Thanks a lot. I used the variables and recommended codes and it worked for the socio-demographic variables and the complex weighting.
The outcome variable still has the denominator of 5,180 which is the number of women that received IPTp and not 12,743 which is the total number of women that have a life birth in the last two years preceding the survey.
Will be glad to receive your assistance once again.
Chinazo
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
