Hi, I am glad you were able to use the Sampling and Weight YouTube video to help guide your initial analysis, however, Madagascar is very unique in that the sampling strategy is designed around malaria endemicity zones. In looking at the "ÉCHANTILLONNAGE" section of Appendix A on page 93 of the final report you will see that the 4 epidemiological facies were stratified into urban and rural, with the sub-desert facie composed only of rural areas. This creates a total of 7 sampling strata. Since this is not a pre-made variable in the Madagascar 2013 dataset you will have to create one to use.
Here is a Stata command to help you create that variable. You can use the same logic to create a variable in SAS.

g strata=shfacies
replace strata = shfacies*100 if hv025==2
svyset [pw=hhwgt], psu(hv021) strata(strata)