There are no evidence-based discharge criteria based on W/H. The thresholds we use are arbitrary and based on the statistical properties of a narrow defined reference population rather than mortality or relapse risk. There are double-standards here. We cannot use MUAC because of an absence of evidence-based discharge criteria but we can use W/H despite an absence of evidence-based discharge criteria. I don't want to be misunderstood here ... we should be using evidence-based discharge criteria regardless of the indicator we use. There are internationally endorsed (i.e. by UNICF, WHO, &c.) discharge criteria for MUAC based on proportional weight gain. These are problematic in that they tend to allocate longer treatment to less wasted children. This is the opposite of what is needed. This is not a problem when a program has very effective and timely case-finding and recruitment but it is a problem at the start of a program or in a program with poor recruitment where the case-load varies considerably in the degree of wasting at admission. Some countries and NGOs have instituted "MUAC only" programming with admission and discharge decided by MUAC. The discharge criteria used is usually MUAC > 125 mm for two visits. This threshold was chosen based on an analysis of mortality risk from historical cohort data and from experiences from NGO programs in Africa (which suggest a threshold of MUAC > 123 mm). This is an evidence-based threshold. Cohort studies from many settings show that mortality is at baseline in children with MUAC > 125 mm and increases rapidly as MUAC drops below this threshold. This is better evidence than we have for W/H which has a very weak and location-specific association with mortality. The issue around the evidence for the MUAC > 125 mm threshold is that it is based on data from free-living populations and may not apply to recovering / recovered SAM cases. Recent evidence from Malawi suggests, however, that children discharged at MUAC > 125 mm from CTC / CMAM program are at lower risk of mortality than similar children that did not experience an episode of SAM. There are a number of studies planned looking at the safety of MUAC discharge criteria and to develop MUAC-based monitoring tools. I am PI on one study due to start in the next month or so and reporting in late 2012. If you contact me through the ENN administrator I will send you a copy of the study protocol. I hope this helps.

Dear Mark, As we are also looking at this issue of MUAC admission and discharge criteria and are planning operational research on the subject (on Asian population), I'm looking at what has been done already, so I would be very interested to see your study protocol. Thank you Pascale

OK ... anyone who wants the protocol please e-mail me at: mark*AT*brixtonhealth.com replacing the *AT* with the usual @

Dear Mark, I would like to say thank for your guidance and for clarifying further the issue on MUAC. I do understand everything you have written and really interested to read the study protocol. You are very right for the suggestion that is always better to use evidence based when we are all making recommendations. I wanted to say that very recently I have assisted in one of the nutrition forum in a country where actors in nutrition (mostly INGO) were pushing for the implementation of MUAC criteria as an admission but the questions that were raised from the MoH were about the use of 15 or 20% weight gain on discharge, furthermore, the MUAC discharge criteria at 125mm has been questioned in some situations mainly because of the extended length of stay since this has already some implications in regard to the number of RUTF that many children will have to consume until recovery...........

I think there are two issues here and they are related to each other through coverage ... Proportional weight gain : A very wasted child will have lost more weight than a less wasted child and, will therefore, need to gain more weight than a less wasted child. The proportional weight gain method means that the opposite can occur (i.e. we discharge the most severe cases too soon and the least severe cases too late). Length of stay (LOS) at MUAC > 125mm discharge : Data from CTC program indicate that the median MUAC gain of a SAM case receiving the full CTC protocol is 0.4 mm per day. If we admit at MUAC < 115 mm ... let us assume 114 mm then the patient must gain 12 mm circumference. This will take about 30 days. Add another visit to that and we get 37 to 44 days. That is not very long (I think that is just below the average of the early CTC programs). Of course if you admit at 95 mm then the LOS will be something like (126 - 95) / 0.4 = 78 days. Adding another visit takes this to 85 to 92 days. That's a long time and is only the median so you may have stays of 120 days or longer. The linking issue is timely case-finding and recruitment. If a program can get this right (and all high coverage programs do) then there is little variation in the degree of wasting of recruited cases and the degree of wasting is severe but not very severe (in a high coverage program admitting at MUAC < 115 mm (e.g.) the median MUAC at admission would be 114 mm with less than 5% of admissions below 105 mm and any case with a MUAC < 100 mm is treated as a critical incident). In this situation the proportional weight gain method does not tend to discharge non-recovered cases and the 125 mm threshold does not give rise to a long LOS. Last year I evaluated a CHS-delivered CMAM program in Bangladesh with 89% coverage and mean LOS was 30.44 days. Note that here we are talking of a running program. At start-up, program coverage will be low and there will be both prevalent and incident cases to deal with. At start up we will have a broad range of severity of cases and make some use of stabilisation centres (if available) for the most severe. I think that focussing on issue of the discharge criteria is missing the real point. It is directing attention away from real problems to their sequelae. The real problem is low coverage and one outcome of this is long LOS (other include high mortality, low cure, high defaulting, and even lower coverage). If you have with low coverage which always means you have a lot late treatment seeking then you will have problems with mortality, recruitment, retention, recovery, LOS, &c. which will lead to lower coverage which will lead to ... and so on. The cure to the problem of long LOS is NOT to complain about discharge criteria or, worse still, fiddle with discharge criteria to make LOS look reasonable. The cure is to run a program well and get a high coverage program that recruit and treat cases early. It is possible to to this. I have seen programs with coverage as high as 90%.I have seen program delivered through MoH PHC facilities with coverage as high as 70%-80%. It is, however, much easier to deliver a low coverage program. I have seen coverage as low as 8%. The high coverage programs have low LOS and the low coverage programs have high LOS. Of course ... it is much easier to debate issues of MUAC vs. W/H and proportional weight gain versus MUAC threshold but this does not deliver SAM services.

Thank you Mark. This is my forum of the year. Kind regards,

Samuel, Thank you for your kind comments. I have to point out that my answers can only be as relevant as the questions being asked. We all make this forum work. Mark