This is a very interesting approach. It has been tried by ALIMA. I have a little of their data (regarding performance of mothers at taking the measurement) and I understand that they consider this to be very promising approach. You question seems to be about performance with an indicator: number of SAM kids referred by MUAC measured by mother ------------------------------------------------------ total number of prevalent and incident SAM cases The numerator may be tricky to get as "source of referral" is seldom collected exhaustively on OTP records. The numerator can only be estimated very approximately. This makes for an inaccurate and imprecise indicator. What you want is difficult to measure. I think you would be OK if you did a small survey of mothers of cases and more than a low proportion stated that they had come to the program because of measuring their own child's MUAC (or because a neighbour had). You could be more formal and collect data as above and on changes in the MUAC at admission and on raw admissions. You could then used mortality data from the historical cohorts and life tables to work out a likely number of averted years of life lost (YLL) for SAM this is a very close estimate of DALYs. Divide YLL by the cost of the 900 straps and their distribution and you have a cost per DALY. If this is below US$100 then the intervention is usually though to be worthwhile. I am interested in how you got to 900 straps. Do you have only 900 mothers with children ages 6-59 months in your camp? I would expect you will also give MUAC straps to CHWs, THPs, TBAs, drug vendors, &c. too. I hope this helps.

An article looking at the technical performance of mothers using MUAC to screen their own children for MAM and SAM has been published and is available [url=http://www.archpublichealth.com/content/73/1/26/abstract]here[/url].