Here is an article about recent work on improving estimates of the numbers of children With SAM from both cohort and survey data.

The main points of interest in this article are:

(1) It feeds into / can be used to improved methods used to estimate caseload from survey (e.g. SMART) data.

(2) It corrects to much quoted 19 million SAM cases to include expected incident cases. That 19 million turns into over 110 million. That should be useful for advocacy and focus our minds on the scale of the problem.

The PDF include a useful independent commentary on the method and findings.

I hope this is of some use to someone.