1) Currently, there is no internationally agreed cut-off for maternal MUAC. If you work in a country using other than 210mm standard, report the results according to the local as well as the 210mm standard.
2) Your sample should also include pregnant and lactating women. If your capacity allows, consider using stratified random sampling focusing on three main groups: pregnant women, lactating women and not pregnant or lactating women. For more information on MUAC cut-off for pregnant women, please read FANTA's publication below.
3) Acute malnutrition is prone to significant seasonal differences in many countries. Therefore, if you need to compare your baseline and endline data to assess the result of your work, ensure that the data is collected at the same time of a year, otherwise you will receive two sets of data which say very little about the change your project has (not) achieved.
4) Since the differences in the prevalence of acute malnutrition are often very small (e.g. from 7% to 5%), MUAC-based surveys need to use a very small margin of error (2-3%), using a large sample of women. With a larger team of measurers, MUAC data for a survey can be collected within 6 - 8 working days (training incl. piloting can be done in 1-2 days). To ensure maximum quality, use the guidance recommended below; if required, contract an in-country or headquarters-based advisor to design methodology, train your team and supervise the survey quality.