This is a very interesting question! A major problem with MAM treatment services is that they fail to achieve coverage. We have come a long way with SAM treatment programs. A decade ago we were routinely achieving < 5% coverage in center-based SAM treatment programs run by resource-rich NGOs. Now we routinely achieve coverage above 50% in resource-poor MoH programs supported by UNICEF and NGOs. I think that is has taken a long time for those in charge (in the broadest sense) of MAM treatment programs to understand the key role that coverage has in determining impact. Things are changing. Assessments of the coverage of MAM treatment programs have recently been done in Niger and Chad and an assessment is planned in Sudan. Some progress has also been made in southern Africa. This is progress. It is not entirely clear if this is a serious attempt to assess and improve program coverage or just box-ticking because a donor required a coverage assessment but it is a start. It is my understanding that the Coverage Monitoring Network has started looking at methods to assess SFP coverage. I think we have to think large changes like this through thoroughly. MAM treatment programs with sound logistics and good patient monitoring can be very useful as "bridging programs" for SAM treatment programs that may allow early discharge from TFP. Such programs may also provide family "protection" rations that help reduce intra-household sharing of RUTF. Both roles may reduce the cost of CMAM programs by reducing length of stay and RUTF volumes. The main problems with this approach are (1) SFP logistics (pipeline breaks are commonplace) and patient monitoring are often weak (it is quite common to find MAM cases who have become SAM cases in SFP and not detected and not doing well), and (2) good integration is required. Integration is difficult to achieve when MAM and SAM treatment are delivered by different agencies (all the way up and down the food chain). This fracturing does not seem to be amenable to fixing by current co-ordinating bodies (the clusters). I like your suggested one therapeutic program approach. I think that the use of RUTF throughout will simplify programs and, given the high nutritional density of RUTF, may not be any more expensive than current SFPs due to short lengths of stay and reduced transport and storage overheads. Here is a diagram (after an earlier diagram about CTC programming by Steve Collins and colleagues) of a one program approach: [img]http://www.brixtonhealth.com/oneProgram.png[/img] The main gap in this diagram is the "MAM protocol" ... I envisage something like the current SAM protocol but with RUTF reduced to something like half SAM the ration and fortnightly contact (here I am straying beyond my competence). I would have a MUAC < 125 mm threshold as the best evidence we have shows that mortality increases above baseline at about this level. I see no problem with keeping the 6 - 59 month age range as few older children will have MUAC < 125 mm and those that do will very likely benefit from treatment. This has been a long way of saying "Yes, we should be considering your proposal".

Dear Anon', you pose an excellent question. In my humble opinion, there certainly needs to be a revision in the management of MAM, most pertinently in environments affected by chronic emergencies. This pivots, not around the debate of efficacy, rather around scalability and ultimately sustainability of TSFPs in their current guise. As of 2010, UNICEF reported, 38% of countries (#23) that provide CMAM for SAM services were 100% dependant on UNICEF to provide the therapeutic foods and the remaining 62% of countries (#37) were 80% dependant on UNICEF. If you consider that your MAM case load is usually twice that of the SAM caseload (~2:1), there is little chance that any current government can afford or has the requisite budget allocation at their disposal for the integration of TSFP as a sustainable development initiative, thus many countries are reliant on the likes of WFP to fund/support such interventions now and going fwd. WFP alone certainly cannot meet the full demand and in turn themselves are vulnerable to external factors out of their control, namely fluxes in global annual yields and food prices. Therefore, the above gives ground to the argument that there needs to be a) some sort of modification in the way we currently manage MAM, to make curative MAM food-based interventions more affordable, scalable and thus sustainable, given the significant caseload and b) a greater investment in concurrent prevention based activities. This is of particular importance as the provision of hand-outs does not change an individual or HH's ability to meet their basic needs, it doesn't address the underlying causes, it merely treats the symptoms.

Dear Anon' Adding to what has been explained already, I would add a practical point of view to the discussion. In fact, back in 2005 MSF was implementing the approach you are mentioning in several countries, including Niger where I was working. Mainly because some of the products that are being used today did not exist or were used in GFD only. SAM was treated as per the common protocol, and MAM children were getting 1 sachet per day on a fortnightly basis. I am trying to retrieve my data from those times but can't find it now. However, I remember that the average weight gain was good for both SAM and MAM and average length of stay was not different from what is expected on the modern SFP programs. Summing to what Matt and Hatty explained, expanded admission criteria works. Nowadays the evolution of different products and the availability for the common RUTF supplier UNICEF is sketchy very often as a result of the expansion of IMAM treatment in more countries, and new approaches have developed. Last year, the problem of scarcity of supplies and poor supply chain became very evident in the South Sudan crisis, topped up with the high number of areas in need and the gap in actors in the field, and a movement to design an expanded admission criteria to cover areas where only MAM or SAM treatment were available was studied (Yet not finalized/approved) to ensure that whatever was the program available, both MAM and SAM would be given the present treatment during the gap until both are present. It is basically what you proposed and I explained during the times before RUSF. I am a supporter of that approach, which would solve many problems such as supply chain breakdown, use of the SFP ration to cover family needs in lack of GFD (in emergency settings) and constant changes of the ration as per WFP availability of goods. Having one product and a single program would simplify the task, reducing the kind of products and the number of suppliers required.

I worry about "arguments from caseload" as I feel they also argue against coverage and equitable access to services. A high coverage program will, by definition, have timely case-finding. Recruited children will then not to be very severely malnourished (in current CMAM terms we would see most cases admitted at MUAC = 114 mm with very few cases admitted with MUAC < 105 mm) and without medical complications. Such children respond well and quickly to treatment needing about than half the total RUTF and clinic time as more severe cases. In very high coverage programs, median length of stay can be as low as four or five weeks including a proof of cure visit. If we imagine a high coverage program admitting at MUAC < 125 mm with a median MUAC at admission of (e.g.) 123 mm we might expect lengths of stay as short as three weeks using half the RUTF per case. A back of the envelope calculation for a "single program" with a median LOS of 4 weeks and 40% RUTF savings per case would, with the same number of cases would cost: (4 / 8) * 0.6 = 0.3 (30%) that of a CMAM program with a median LOS of 8 weeks giving the full CMAM protocol. Such a program could treat a little over three times the number of cases as a conventional CMAM program at little or no additional costs. This is all imaginary (and might be nonsense). We need to work out what the "MAM protocol" should be, how well it works in uncomplicated MAM cases, and how to deliver at high coverage. That is, we are somewhat ignorant about if and how a "single program" would work. It is, however, a manageable prospectus of clinical and operational research. I am also worried about "hand-outs" arguments. Children with MUAC < 125 mm are generally at elevated (i.e. above baseline) risk of near term death (i.e. dead within a year). They are ill. If we recognise this we are treating sick children and "hand-out" becomes "life-saving treatment". We are treating a disease not its symptoms. Perhaps you refer to SAM And MAM as symptomatic of a disastrous political economy. I take that point but we are, on this forum, technicians not politicians. I agree that we need more than curative programs. To get these working we need to have quick, cheap, and simple tools that can identify context specific causal factors and associated attributable risks so we can design high impact preventive intervention. We do not yet have such tools (although they should be easy enough to make - other disciplines have done it). We might terms such programs "integrated nutrition sensitive programming" (INSP). The "nutrition sensitive" bit is about causes and their importance. The "integrated" bit reflects that inventions will often need to be multi-sectoral. When we have (e.g.) WASH and nutrition departments in UNOs and NGOs that barely speak with each other we might find this to be a stumbling block. We might add INSP tool development to a research prospectus. I think that a single program approach is what CTC and CMAM programs "want" to be. We got stuck with the MAM / SAM (WFP / UNICEF, TFP / SFP) divide because of were we started from rather than as evidence based programming.

Dear All, Thanks for having replied to my question in an open and constructive manner. I agree that TSFP coverage rate is likely very low, despite the huge number of MAM cases admitted in TSFP. Sorry for being a bit provocative, but what is the expected impact of TSFP with a coverage rate less than 20 % ? Referring to the provision of “protection ration” for preventing intra family sharing and to the need for addressing the underlying causes, the recent publication of Langendorf and al, Sept 2014, on the prevention of acute malnutrition in Niger, shows that “preventive distributions combining a supplementary food and cash transfer had a better preventive effect on MAM and SAM than strategies relying on cash transfer or supplementary food alone.” This result looks promising for informing a change in strategy. I also think there are, at least, anecdotal evidence that conditional cash transfers can have positive outcomes on household dietary diversity. I also love the idea of “integrated nutrition sensitive programming “. In fact, my suggestion of setting admission criteria for acute malnutrition treatment programme at 120 mm is, partly, based on a recent discussion I had with a nutrition partner who told me that programme outcomes/outputs using this threshold have been tested. However, I could not find any literature on this. So I would have two questions on the use of MUAC threshold at 125 mm: 1) Is there any information on the proportion of children with MUAC comprised between 120mm and 125 mm compared to those with MUAC<120 mm in the reference population? and in emergency nutrition situation? 2) Dear Mark, you said that “children with MUAC <125 mm are generally at elevated risk of near death”. Is the increase risk of near term mortality, in children with 120mm125 mm, sufficiently significant to justify, the admission of children with MUAC<125 mm in treatment services for acute malnutrition (rather than restricting access to children with MUAC<120mm only), especially if a preventive supplementary feeding activity (targeting all children with MUAC>120 mm) is available? My point is that, the food cost of 2 months MAM treatment using RUSF (18 – 20 USD/ chil) is, more or less, equivalent to the food cost of 12 months preventive supplementary feeding with SC+ (daily ration of 100g/child/day). I know that many other factors need to be taken into account to support a strategic shift in the management of acute malnutrition, and it looks like more robust evidence are needed to create a momentum strong enough to bring down the existing barriers. Let's hope that this topic can be discussed at a global level.

Dear all. You may be interested in the results of a study conducted as part of an evaluation of Cambodian policy on screening and treating malnutrition. The study was based on datasets from UNICEF, l’Institut de Recherche pour le Developpement, WFP, World Vision and International Relief and Development. From a total of 11,818 children, the current cut-offs of MUAC for diagnosing acute malnutrition (MUAC<125 mm), MAM (115mm = MUAC <125 mm) and SAM (MUAC <115 mm) compared poorly to the same categories as defined by WHZ. The study found an optimal MUAC cut off to detect acute malnutrition of 138 mm and to detect SAM of 133 mm. The study proposed a two-step screening procedure for SAM screening: 1. MUAC, using a broader cutoff of <133 mm (instead of the current 115 mm cutoff), which would identify over 65% of the children with a WHZ<-3. This would be done by community health volunteers. 2. Diagnosis of SAM using both MUAC (diagnostic cut-off: <115 mm) and WHZ (<-3) independently of each other at a primary health service. This approach should improve the cost-effectiveness of screening programs and the treatment because sensitivity would be significantly improved. Next steps are to pilot the proposition and to establish MUAC thresholds for different age groups.

Thank you (Annon) for flagging this study and the findings, especially your point that 2mths of MAM Tx using RSFP is comparable to 12m of preventative SF. This is also outlined in the aforementioned study, ‘Limited data suggest that community-wide preventive approaches have a greater impact on reducing childhood acute malnutrition than programs targeting already acutely malnourished children [8],[9]. However, much less is known about which prevention strategies work best in a given context, especially when targeting large populations with limited resources and integrated approaches that combine multiple interventions with different aims’. This last point is particularly pertinent when considering the study findings. The study was conducted in a district where 12.9% of the population were declared to be at severe risk of food insecurity, whereby we can assume a more significant proportion were moderately food insecure and presumably some HH’s not deemed to be high risk. There were also no other food interventions in the area. The finding of this study is very interesting for ministries/agencies looking to support high impact, cost effective initiatives. However, we should always be mindful not to apply such findings in a sweeping fashion to all contexts struggling with chronic or cyclical ‘hunger gaps’, simply assuming the outcome “preventive distributions combining a supplementary food and cash transfer had a better preventive effect on MAM and SAM than strategies relying on cash transfer or supplementary food alone”, without considering a) the degree of food insecurity in the target population b) the causes of MN in the target population and c) the importance each cause played in contributing to the outcome of acute MN. Indeed the targeting was based not on an index of HH food insecurity, rather a district known to be affected by this parameter and then on HH’s having at least one child measuring 60 cm–80 cm (at any time during the study period whatever their nutritional status). Mark – I agree that re. curative care that we’re treating a disease (sorry semantics on my part) and indeed we are technicians not politicians. Alas, operationally we work in a political context, with limited financial resources, so the decision to direct money towards curative and/or preventative initiatives plays a significant role.

Dear Bidni, I suspect you refer to the study below : http://www.ncbi.nlm.nih.gov/pubmed/24983995 This study, however, evaluates MUAC against WFH used as a gold standard to define malnutrition. This is inappropriate for programmes aiming at reducing malnutrition associated mortality. This has been discussed many times in different forums, including this one, meetings and publications. MUAC is used more and more frequently based on its capacity to identify high risk children as it is more closely linked to the risk of death than WFH. Based on this criteria, WFH is a poor indicator and should not be used as gold standard. The reasons why MUAC is more closely associated with risk of death than WFH are not clear. Association with muscle mass, young age or stunting, all of which are related to survival may play a role. These possible mechanisms are not mutually exclusive and may all be part of the explanation of this association. MUAC also makes possible community screening, an option which is not possible with WFH. This increases coverage and makes possible early detection and treatment with improved prognosis.

All, I am very impressed at the responses in this forum. I think we agree that something needs to be done. How can we move forward with this? Óscar, Thanks for your report. It shows that these sorts of program can work. Anonymous 2822, It is not provocative to state facts. Impact is: impact = coverage * effectiveness With coverage of 20% the maximum impact that can be had is 20%. Coverage is the key to impact. 20% coverage is better than no coverage bit if falls very far short of SPHERE minimum standards. I have not heard iof TSFP programs achieving minimum coverage standards apart from over small sub-areas of overall program areas. WRT The proportion MUAC < 120 mm and the proportion MUAC < 125 mm ... MUAC is (pretty much) normally distributed. This means that there will be an exponential increase in numbers with increasing MUAC when MUAC is below the mean (120 mm and 125 mm should be well below the mean). This means that a small change in the MUAC threshold can lead to a large change in potential program numbers. Under a normal model with reasonable guesses for the mean (145 mm) and SD (11 mm) of the MUAC distribution we would expect there to be about 2 cases between 120 and 125 mm for each case below 120 mm (i.e. a three fold increase in case-numbers). Children with MUAC < 125 mm and untreated are at significant risk of near-term mortality. Morality risk increases as MUAC decreases. The increase is not linear: [img]http://www.brixtonhealth.com/muacMortality.png[/img] This figure is for untreated cases. I favor 125 mm because this is also a safe discharge criteria. It has the simplicity of allowing the same threshold to be used for admission and discharge in all curative programs. It is also "conservative" in the sense that MUAC < 125 mm is a well accepted case-definition for MAM. Looking at mortality in the plot above, I think there is some room for adjusting the threshold to meet resource availability. If prevalence is high and resources limited then we could use MUAC < 115 mm (we would have treated this as MAM before about 2010 ... some programs still do). If prevalence is a bit lower or resources are available we could use MUAC < 120 mm. In low prevalence contexts we might want to use MUAC < 135 mm (which has been a definition of GAM in community programs in development contexts). To avoid complications I think that MUAC < 120 mm might be safe with MUAC < 125 mm as a "watch-listing" criteria. I think you are right to take a health economics perspective. I would go further and calculate costs per death averted. Bindi Borg, I agree that treatment coverage starts with coverage of screening services. For a acute condition coverage should be also be frequent as well as being everywhere. I am very impressed by the work of ALIMA / BENFET on this. See [url=https://www.youtube.com/watch?v=WRxmmRUS_To]this video[/url]. MUAC is pretty much independent of age when you look at the relationship between MUAC and mortality. Use of MUAC/A and MUAC/H does not improve performance (most studies show worse performance). Perhaps you mean MUAc thresholds for (e.g.) children aged between 5 and 10 years. This has been discussed briefly [url=http://www.en-net.org/question/436.aspx] elsewhere on these forums[/url]. I think that prevalence is low in children age > 30 month and very low in children age > 59 months so the work required may not be worth it and may not be ethical. Hatty / Anonymous 2822, My concerns with case-transfer programs is that (I think) insufficient thought has been given to protection issues. To put it very crudely ... if someone were minded to have sex with children I think they would be more likely to get that with money than with a bag of CSB/Oil/Sugar premix. We already know that this can be a significant issue in the humanitarian sector. Perhaps I worry unnecessarily. André, I agree with all of that.