A response from malaria expert, Dr. Lia Florey,
Hi Nina,
It looks as though you are using the correct approach. When examining this particular dataset I see that a crosstab of hml33 and hml32 produces some unusual results. Same with the crosstab of hml33 and hml35. In order to replicate the estimates from the final report I would recommend the following changes to your code:
*Proportion of children 6-59 months old with malaria infection (microscopy) by Region
svy: mean micmalpos if hc1>=6 & hc1<=59 & hv103==1 & (hml32==0|hml32==1), over(hv024)
*Proportion of children 6-59 months old with malaria infection (mRDT) by Region
svy: mean rdtmalpos if hc1>=6 & hc1<=59 & hv103==1 & (hml35==0|hml35==1), over(hv024)
The hv103 specifies that children who slept in the household the night before the survey (de facto).
You might also try running tabulations instead of means:
svy: tab hv024 micmalpos if hc1>=6 & hc1<=59 & hv103==1 & (hml32==0|hml32==1), row ci per
svy: tab hv024 rdtmalpos if hc1>=6 & hc1<=59 & hv103==1 & (hml35==0|hml35==1), row ci per
I hope that this solves your problem.
Best,
Lia
The DHS Program User Forum
Members
Search
Help
Register
Login
Home
