The simplest interpretation of this data is that you have an immediate need of at least 67 SAM cases (urgent treatment required) and 265 MAM cases (SFP, watch-listed for SAM, or ...), screening should be continued, preventative measured should be identified and implemented. I assume that you want to be able to take this data and produce a prevalence estimate. That is possible but you have to be aware that "mass" screening can often be biased samples. In some cases the healthy refuse or do not turn up. This gives an upward bias to the prevalence estimate. In other cases the unwell infants are "hidden". This gives a downward bias to the prevalence estimate. I saw both of these effects when we tested case-finding methods for SAM coverage surveys. That said, you can produce an estimate: P(SAM) = 67/1010 = 6.6% P(MAM) = 265/1010 = 26.2% P(GAM) = (67 + 265) / 1010 = 32.9% These look very hight to me. If you report these then you should say that hey refer to your 11 villages (which you may have selected because you suspected a very poor situation) and that the data came from screening not survey. Some estimate of the direction and magnitude of bias should be made. Also see [url=http://www.en-net.org/question/328.aspx]here[/url] and [url=http://www.en-net.org/question/679.aspx]here[/url]. I hope this is of use.

Hello, Could you specify whether you used MUAC or WFH to estimate SAM and MAM? Severine

If you are interested in further analysis, I would recommend that you analyze the data by measurement. Like how many of the SAM are identified by MUAC < 115mm or by WHZ < -3 or both. Similarly for MAM. You can also look at rate of SAM and MAM by age groups (6 - 23m vs 24 - 59m) or by Sex. In general, do you think distribution of age and sex is acceptable? Since you have screened more than 1000 Under five children from 11 sites, you can also look at the rate of SAM and MAM by sites, is there any village which is contributing more compared to others.

Thanks so much Guys.. Thank you Mark I really appreciate.

Happy to help.