| Re: Nepal SPA 2021 [message #28747 is a reply to message #28691] |
Fri, 01 March 2024 20:55  |
SaraDHS
Messages: 46
Registered: December 2020
|
Member |
|
|
Hello,
I did not see the country-specific variable for "Fees paid for today's visit" in the recode file, so I think you need to use the raw file and merge the facility and ANC file in order to recreate the table. The strata variable is also a little difficult, as you need to recode the facility types to get the correct strata. See below the Stata code needed to recreate this table (3.40). Note that 1250 is the Total number of ANC clients, NOT the number of ANC users who reported paying for services. The % of the total that reported paying is 82.2%.
Best,
Sara
*Use Facility file
use "NPFC8AFLSP.dta", clear
*Drop if no ANC observations
drop if ancf==0 | ancf==.
*Save this Facility file
save "FC_NP_ancobs.dta"
*Merge Facility file with ANC observation file
use "NPAN8AFLSP.DTA", clear
merge m:1 facil using "FC_NP_ancobs.dta"
*Generate weights and strata for svyset
gen clwt= clientwt/1000000
recode factype(1/2=1 "Federal/provincial hospitals") (3=2 "Local hospitals") (5=3 "Private hospitals") (6=4 "PHCCs") (7/9=5 "Basic health centers") (10=6 "HTC standalone"), gen(fctype) //As you can see from the final report, the facility types were categorized in a different way for the strata than is used in factype, so you need to recode this variable to get the correct strata
egen strata = group(fctype province)
svyset facil [pw=clwt], strata(strata) singleunit(centered)
*The heading description of Table 3.40 tells us that only certain types of facilities were included, so drop facility types that were not included
drop if factype==2 | factype==5
*Tabulate the payment variable
svy: tab x204
Sara Riese, PhD
Senior Demographic and Health Researcher, DHS Program
|
|
|
|
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
