Not so much "a panel of experts" but "a panel of experts and field practitioners". The impetus for change came from (a) field practitioners noting the problem with percentage weight gain, and (b) the desire for MUAC only programming that allows CMAM to be delivered by community health workers. I am a little confused about what you write ... I do not think we should claim a child has recovered from wasting when it is, by an objective measure, still wasted. The old W/H discharge criteria (I've seen 80% WHM, 85% WHM, -2 WHZ and -1.5 WHZ by NCHS and WGS used in different programs) discharged after MAM, the % weight gain indicator at 15% corresponded to the majority of patients exceeding 80% WHM (by WGS) at discharge (18% was about 85% WHM by WGS) estimated from a set of CMAM patient cohorts (mostly CTC programs in Ethiopia and Malawi). The proposed MUAC approach is, therefore, no change in this regard. The practice of discharging at the SAM / MAM threshold is only sensible if you can discharge into a well-functioning SFP with a good OTP-SFP interface and good monitoring in SFP (I don't often see these criteria met). In some programs OTP and SFP are integrated but this is seldom the case as (e.g.) the RUTF is supplied by UNICEF and the CSB by WFP which tends to vertical programming. Looking at your points ... (1) If we RAISE (are we really doing this?) the discharge criteria we can expect length of stay (LOS) to increase but LOS depends on more than discharge criteria. To list the most obvious ... admission MUAC, compliance at / by clinic, compliance at home, intra-household sharing of RUTF. These can all be controlled by proper program functioning. Admission MUAC is a coverage issue and we know that high coverage CMAM is possible. In these programs we find (e.g.) median admission MUAC at 114 mm. These children can be in and out in 30 days. Compliance at clinic (Do staff stick to protocol? Do they give antimicrobials?, Is RUTF continually available?, &c.) is under program control. Compliance at home is usually considered a "consumer deficit" (i.e. the fault of the patient) but with careful recruitment of community-based volunteers, co-operation with traditional health practitioners, appropriate and ongoing community mobilisation, and patient monitoring can be controlled (i.e. this is also under the control of the program - in the CTC research program we did not see a lot of household compliance issues because we designed the program to prevent them). I think you also need to see the "balance" of the reform. The current weight-gain method means that healthier children get to stay in a long time and sicker children get to stay in a short time. The intention of the reform is to reverse this. There is a reasonable expectation that little will change regarding LOS (i.e. we will just distribute the patient-days in program on a more rational basis) ... this may not be the case (see [url=http://www.brixtonhealth.com/MUAC.Response.CMAM2008.pdf]this presentation[/url]). (2) We know that defaulting varies from program to program in similar settings. This is another of those issues that we tend to think of as a consumer deficit when it is (to a very large extent) a program deficit. Well run programs have high coverage and, as part of that, low default rates. One of the reasons for using MUAC for admission (and there are many good reasons) is that it is intuitive and helps raise awareness of SAM. The simplicity could be an advantage as we can explain that "yellow" (115 mm - 125 mm) is "recovering but still at risk" and "green" is recovered. Removing weight (and height) from programs should improve patient throughput. This should go some way to reducing defaulting. I have to stress that defaulting is usually a problem with how a program is run. We have used W/H, % weight gain, and MUAC in CMAM program and the good ones have low defaulting rates and the bad ones have high defaulting rates (if they even bother to measure this properly) regardless of the discharge criteria used. This (strongly) suggests that default is not strongly associated with the discharge criteria used. (3) Very low MUAC at admission should be treated as a critical incident and reasons investigated and programs reformed. This is a simple coverage failure. You should not tolerate more than a handful of cases with MUAC < 100 mm after the first few months of program operation. Fix the late presentation issue and the problem goes away. Don't fix it and you have a bad program and that badness will be reflected in the non-response rate. If you don't want to look bad then do a better job. BTW ... this is another of those program deficits. Late presentation is usually a failure of case-finding / referral, community mobilisation, or a bad program with rude staff, long waiting times, periodic RUTF stock-outs, poor clinic compliance, &c. (4) See comment on LOS (above). I would be wary of complicating the protocol. Using MUAC for entry and exit is a welcome simplification. It would be a shame to add complication. There is work underway exploring these issues. Expect publications from a few teams over the next few months. The 125 mm threshold is also being investigated. The 125 mm threshold is considered to be safe (i.e. in terms of post-discharge relapse or death). A lower threshold may be just as safe. I hope this is of some use.

I forgot to say that the 15% weight gain threshold was further tested against data from over 500 nutritional anthropometry surveys (over half a million children) from 39 countries and against the WGR reference dataset. The original analysis was reported [url=http://www.brixtonhealth.com/FNB_27_3_2006_B.pdf]in this article[/url] which also proposed the percentage weight gain discharge criteria (yes ... it is my mistake and I put my hands up to that). The more detailed analysis is presented in this [url=http://www.brixtonhealth.com/JointMay2009.pdf]statement[/url].

Where has critique against the % weight gain been presented? I understand that NUGAG is preparing to publish regarding admission and discharge criteria, but is it so far a matter of internal discussion?

I'm away from my office and do not have the papers to hand. There have been some recent publications by MSF / EpiCentre staff using data from their programs in Niger, Mali, Chad and Sudan. Perhaps Susan Shepherd or Nancy Dale could provide exact references. I have a publication co-authored with Nancy Dale and André Briend on this issue that is currently "in press". This uses data from an MSF program in Sudan. The first time I saw the % weight gain problem presented in a public forum was by Susan Shepherd (MSF) in a WHO meeting in Geneva in early 2010. The problem was clear and, once pointed out, painfully obvious. The possibility of using MUAC for discharge was first discussed in a public forum at the 2008 CMAM conference in Washington using data collected by CONCERN and SC-US in Ethiopia (presentation link in my earlier message in this thread). A FANTA funded / VALID run study arising from that meeting is just closing up (the last few patients are now in last month of follow-up) and results will be available by the end of this year. So ... this has been an open debate and collaboration between many agencies ... very much not an "internal discussion".

Thanks again Mark. My difficulty is with identifying recent results and discussions on these matters, as much of it won't show up on a database search on e.g Pubmed/Medline. I want to look into OTP outcomes where WHZ has ben omitted and am trying to navigate between published material and the discussion between different actors and agencies which to some extent has to be viewed as internal. The input here is invaluable.