The most common mistake is to overestimate SAM prevalence, always better use a lower prevalence than the latest data (generally use the average between prevalence and lower C.I). I dont know how far from the minimum sample size you are but it could still be ok, on the other hand, if you really want to reach the sample size the only option is to resample villages. Do you really need to estimate coverage or barriers/boosters identified are enough to inform the programme and the necessary actions to be taken?

Dear all
Thank you Dr. LEO for sharing with us this aspect of prevalence. we often use the prevalence of SMART surveys at the regional level for SQUEAC at district level so we need to adjust the prevalence according to the geography area , seasonality etc ...
it's a challenge.

Another source of error is the average size of a geographical area that will be investigated (eg the village). if the list of villages is not exhaustive and if the population data are not update the average size is wrong and we can select less village than necessary.

sometime estimate coverage is not so important, the most important is barriers and booster investigation ..... but we need coverage value for donors or for physiological satisfaction.

hope this help