Menu ENN Search
Language: English Français

Variance in number of children admitted to OTP using MUAC < and WHZ <-3SD

This question was posted the Prevention and treatment of severe acute malnutrition forum area and has 11 replies. You can also reply via email – be sure to leave the subject unchanged.

» Post a reply

Daniel Muhinja

Normal user

3 Jul 2010, 14:31

Since 2008, we have been screening and admitting children to outpatient therapeutic program using MUAC 11.0; in June 2009 we adopted MUAC< 11.5 as per the new WHO guidelines. However, we are expected to begin making admissions using WH Scores and MUAC. Before we begin implementation, I would be glad to know if there will be significant increase in OTP caseload after adoption of WHZ scores. Between MUAC and WHZ which one captures more children to OTP? am worried we could just require more staff due to extra workload of measuring children without any significant increase in caseload.* note this is among pastoral communities in Kenya

Mark Myatt

Frequent user

3 Jul 2010, 22:29

My STRONG ADVICE is to NOT use WHZ for admission (or for anything at all). Experiences with programs that use MUAC have repeatedly shown that programs that use W/H and MUAC for case-finding, referral, and admission tend to have low coverage and low impact. This has been repeatedly observed over the past decade and was even clear in the CTC research program.

Unless you are extremely careful, the effect on your case-load of introducing WHZ will be to reduce case-load. If we are to pretend that there are no practical implications of using WHZ and that MUAC with WHZ may capture more children. How many more is difficult to answer since W/H is strongly effected by body shape. Amongst sedentary agrarians or populations living at high altitudes or populations living at low temperatures the effect of adding WHZ will be very small. Amongst pastoralists with high milk consumption and in populations living in high temperatures (e.g. Afar) then the effect may be very large. Amongst Somali pastoralists you may see a 500% increase in case-numbers. Note that many of these cases are not "true cases" but older children with long legs.

I think the most likely effect will be to reduce case-load through "the problem of rejected referrals". So ... be very careful about how you deal with rejected referalls.


For more information.

I hope this helps.

Anonymous 585


Normal user

6 Apr 2011, 06:23


I would like to come back the post as there has been an article in the September 2010 FEX on the programming in Cambodia using both WHZ and MUAC for the same reasons.
I see the point of not including W/H in the active screening process, but think that W/H should be an admission criteria for children presenting themselves directly to the centers, either because of sickness or other reasons.
Or should admission criteria be really restricted to MUAC only and such children, if MUAC above 11.5 but W/H below -3, be turned away?

Are there new studies now clearly indicating mortality risk of MUAC versus W/H under WHO standards. What I have found in a quick research all refers back to NCHS. As well I would like to know if the studies mentioned elsewhere looking at MUAC only for discharge have already lead to some results?

Sorry, a lot of questions.

Anonymous 118

Nutrition Advisor

Normal user

6 Apr 2011, 17:50

In our own programs, we use MUAC and/or WHZ at the community and facility level. If a child meets either or both, he is admitted, so children who don't meet both are not excluded. If it's too complicated to have WHZ at the community level, it seems this could still apply using MUAC and/or WHZ at the facility level.

Mark Myatt

Frequent user

6 Apr 2011, 20:06

The issue is one of keeping W/H in its proper place. What has happened in some CMAM programs is that W/H takes over because it is seen (very wrongly) as the superior measure. This results in cases with MUAC < 115 mm being sent away because they have a WHZ > -3. This is very damaging to coverage. This is an unintended consequence. You add W/H because you don't want to exclude children but the negative effect that this has on coverage means that you effectively exclude many more children. Be careful.

A note on "sickness" : Peripheral muscle is preferentially catabolised during infection. This means that most cases of wasting due to infection will have low MUAC.

A note on body-shape : W/H is strongly effected by body shape so WHZ < 3 means different things in different places. MUAC is more universal. Discordancy usually occurs when you are working in populations with low SSR where the prognostic value of W/Z is lowest.

Studies of mortality risk are no longer possible since we now have very efficacious, very safe, and high coverage treatment protocol. A study would require (e.g.) finding children with MUAC < 115 and / or W/H < -3 and doing nothing. We can't do that. We have to work with the data that we have. We could, I suppose, get the datasets and recalculate WHZ using WGS and redo the analysis. I wonder of the data are available for this.

I know of no study of MUAC for discharge that has presented high quality follow-up data. All indications are that MUAC > 124 mmm is probably a safe discharge threshold.


Frequent user

7 Apr 2011, 06:56

This suggestion of using WFH and / or MUAC comes back again and again. It is based on the assumption that by using WFH you pick up some high risk children missed by WFH as both identify different children.

To test whether this approach makes sense, I went back with former colleagues to data we collected in rural Senegal in the 80's on the relationship between anthropometry and mortality based on a 6 month community follow up with very few children being treated. In those days, referral to the capital, 200 km away was the best we could propose.

We found in this reanalysis that the MUAC ROC curve was constantly above the WFH one, even when the new WHO standard was used. We then plotted the points WFH < -3 and/or MUAC < 115 mm. These two points were both below the MUAC ROC curve. Using WFH < -3 or MUAC < 115 mm indeed increased sensitivity, but this was at the expense of specficity. MUAC < 120 mm had about the same specificity level as WFH < -3 or MUAC < 115 but had a higher sensitivity.

Conclusion: using both diagnostic criteria is more complicated and no more effective than MUAC alone. If a higher sensitivity is needed, better use a higher MUAC cut off and forget WFH.

This study is submitted to a peer review journal. Hope it will be available soon.

Remark: all this is related to dectection of children with a high risk of death.

Joel Conkle

Nutrition / UNICEF Cambodia

Normal user

9 Apr 2011, 09:57

Hi everyone,

It looks like we will be removing W/H from identification criteria at the health center in Cambodia because of difficulty in measuring and time constraints. We will only use MUAC at community and health center.

Andre, I look forward to reading your study. Did you disaggregate by age? For me, it's difficult to interpret much of the work done on the risk of mortality because MUAC selects younger children here.

I'm left wondering does MUAC show equal or more risk for mortality when compared to W/H in children <5 years simply because younger children have a higher mortality risk; and if age does influence this, what is mortality risk using MUAC compared to W/H or even no criteria for children < 24 months.

On a related topic, we are now integrating treatment for SAM into IMCI. I'm in touch with colleagues in Ethiopia. Does anyone know of another country where this integration has happened?


Frequent user

14 Apr 2011, 10:25

Dear Joel,

We did not analyse the data by age. In practice, you don't separate children by age either when deciding which child should get treatment. You have to select for treatment children of different ages and different nutritional status pooled together.

MUAC used with a fixed cut off does select younger children everywhere. This may explain why it is better in identifying high risk children, but so what ? Let's be pragmatic, we want something effective to identify high risk children (and who respond to treatment), whatever the reason.

MUAC is clearly related to muscle mass, something which is important for survival. See: Wolfe RR. The underappreciated role of muscle in health and disease. Am J Clin Nutr. 2006 Sep;84(3):475-82. Freely available at:

Whether age or the muscle effect is the most relevant to determine risk in relation to MUAC remains an open question. Maybe these interpretation are not really opposed, as body proportions, and in particular the percentage of muscle in body weight, increases with age. Young children and malnourished children have something in common in terms of body composition.

Anonymous 512

Nutrition Specialist

Normal user

18 Apr 2011, 13:42

Hi All

Thanks for the invaluable information and discussion. Just to add on the discussion, in some of the countries where we undertake CMAM programmes such as South Sudan, the national protocol recommends to use the following as admission criteria WHZ and MUAC. Hence, screening may be done using MUAC but admission should be based on either. As stated by Mark Myatt this affects the coverage of the programme but on the other hand, there is no much that could be done since MoH national protocol should also be followed. Thus, in this situation what will you advice?


Mark Myatt

Frequent user

18 Apr 2011, 15:03

In programs that I have evaluated recently in a number of countries in Africa and Asia the implementing / supporting agency have argued for MUAC-only CMAM programming and they have got their way despite national guidelines including W/H.

The general argument is that the guidelines are now out of date and are better suited to outmoded and lower coverage modes of service delivery (i.e. TFCs).

The detailed argument is that most primary care facilities do not have height boards, IMCI guidelines and training do not cover height measurement or calculation / lookup of WHM or WHZ, and height boards are not in essential clinic packs. This means that W/H cannot be used in CMAM because CMAM relies on using the most proximate to beneficiary mode of delivery (e.g. primary care facilities or community health workers). We could use W/H in TFCs because because these were either dedicated units based at secondary level health facilities or as stand-alone vertical program units (e.g. NGO-run TFCs). CMAM is about proximity to beneficiary and integration into services delivered at primary care facilities. We can't use W/H in such settings. Also, MUAC cannot be used by most community-based case-finders except in a two-stage model which has been shown (time and again) to lead to "the problem of rejected referrals" which was identified by the CTC Research Program as a major barrier to coverage. Using these arguments it is usually possible to convince governments that W/H in CMAM is impractical and counterproductive.

If reason does not win the day then we have resorted to pretence ... We say that we will use both MUAC and W/H but just don't use W/H because there are no height boards in clinics. I have heard of one case where the few height boards that were in clinics were taken away "for calibration" and never returned.

The middle-way is to say that we use MUAC for case-finding and admission and if a child arrives at clinic with MUAC above admission criteria we will check W/H if possible. With this approach you need to be careful to keep W/H in a very secondary role in order to protect coverage.

I urge you to allow discretionary admissions on a clinician's referral (e.g. for a child recovering from a serious infection to be put on a "high protein / high energy diet), "visible severe wasting", disability, &c.

Regine Kopplow

Health&Nutrition EU Aid Volunteer Concern Worldwid

Normal user

18 Apr 2011, 15:15

Admissions into OTP/SC should be done by either, MUAC and/or WHZ. If one of the criteria is met, the child should be admitted. If MUAC is used for screening at community level but admission at the health facility is based on WHZ only, mothers will increasingly refuse to bring their SAM (by MUAC) children for treatment due to the negative feedback spreading fast. The distance to the health facilities is often large, mothers time is limited, the workload already high. If referred based on MUAC, the admission of the child has to happen. This requires accurate MUAC screening at community level, transparent referral criteria, a close by treatment facility, and use of the same SAM criteria at communtiy and health facility level. However if equipment and time allows, WHZ should be used for admision at health facility level (preferably during routine health interventions/ IMCI consultations) in addition. During coverage surveys it is most likely that a number of missed out cases will be SAM by WHZ only, but having been screened by community volunteers with a MUAC above the SAM cut-off point. However using these admission protocols the most critical children will have the best chance of being detected and referred due to MUAC best predicting a high risk of mortality.

Mark Myatt

Frequent user

18 Apr 2011, 15:34

I'd just like to say that I agree with all of that.

I'd also advise admitting referrals from community-based-volunteers (CBV) with MUAC above the admission threshold for a two week period in order to avoid the problem of rejected referrals. If a CBV keeps making inappropriate referrals then something must be done about it that doesn't involve sending mothers away empty-handed and displeased with a bad opinion of the CBV and of the program which they might spread. IF RUTF is is short supply then I recommend the child see a nurse and get Vitamin A with the mother getting a bar of soap and a packet of biscuits (something like this). This is a "placebo" in the original sense of "I shall please". We need to be very careful about negatively effecting coverage.

Back to top

» Post a reply