Steps to measure MUAC

Current thinking is that estimating the mid-point "by eye" is all that is needed. This is what is now included in WHo material fro training community health workers and (I think) IMCI materials. The video at:

http://www.brixtonhealth.com/MUAC.FullQuality.mpg

also available at:

http://www.youtube.com/watch?v=3pQUtOsjjSY

reflects current thinking at the WHO.

You might notice that in the video the child has a mark at the mid-point of the upper-arm. This is there because the original video showed how to find the mid-point by measurement. This material was subsequently edited out at the request of the WHO.

This is absolutely no harm in finding the mid point by measurement.

As far as I know there has been no testing to compare the methods. The results of testing will depend on what you want to compare (raw values or class membership). If you are just interested in class membership then the evidence is that colour-banded straps reduce error more than anything else.

Mike is right to be concerned about the quality of measurement and the problem of rejected referrals. In CTC programs, rejected referrals are a real coverage killer and overcoming this was a key reason behind the adoption of MUAC as both a referral and admission criteria.

With CTC programs, I encourage the removal of the centre-based "filter" on admission. This "admission on referral" model means that any child referred to the program by a CHW, VHW, or community-based volunteer is admitted to the program even if only for a single day or to receive a limited package of goods (e.g. vitamin A and mebendazole and a consultation with a nurse). We note the source of the referral and take corrective action using outreach workers. In the past it was often GMP programs rather than community-based volunteers that were the source of most "inappropriate referrals". This has lessened with the use of MUAC since the problem was often a confusion between W/H and W/A.

A good case-finding method will pick up incident cases as soon as they become cases (you can check this by plotting a histogram of admission MUAC). This does happen in well-run CTC programs. For example, I recently saw a program in Kenya run by WVI (coverage was a little above 60% - the main problem was that SFP was not identifying response-failures) with an admission threshold of 110 mm and a median admission MUAC of 109 mm ... a child arriving with a MUAC below 100 mm was treated as a "critical incident" and lessons learned. This has implications for programs ... let's say you have a case-defining threshold of 115 mm and it is unlikely that you'll be measuring MUAC with an accuracy of better than 2 mm. A child might be referred with a MUAC of 114 mm but your centre staff might measure this as 116 mm. What to do? If you reject the referral then you risk damaging both timely case-finding and early treatment seeking (I would risk neither of these). If you admit you treat a "borderline" child who would probably be a case if you waited a few days but might not come back if you reject them. My inclination is to enrol such a child for a full course of treatment. Legalistic interpretation of case-defintions can be very harmful to program coverage.

If you are using MUAC to measure prevalence (and Mike's involvement with SMART means that he will probably have studied this more than I have) then errors are likely to be random so there should be little if any bias introduced.

In all cases (i.e. referral, admission, prevalence estimation), Mike is absolutely right ... "proper training and periodic supervision/assessment are important". You need to build this into admissions monitoring and into training and supervision in surveys.

I take issue with the often repeated "errors in measuring MUAC are greater than those with weight or height even by experienced anthropometrists" since the evidence is that when used with the level of staff you mention (CHWs) error is smaller for MUAC than for all other measures. We also have to bear in mind that height boards are quite rare outside of NGOs and IMCI does not cover the measurement of height. When using community based volunteers then MUAC measured with colour-banded straps is the best we can usually do. CTC programs typically use dozens of community-based volunteers (sometimes hundreds). Distributing height boards and scales and providing training to this number of staff is not usually a practical proposition. That said, Mike is right ... there are many sources of errors and these should be recognised and error minimised by training and supervision.

My preference is for colour-banded straps for most applications. Errors are lower with such straps than with numbered straps. They are used to ascertain class membership or diagnosis (e.g. SAM vs. not-SAM) rather than measurement. The appropriate measure of agreement is a generalised Cohen's Kappa (this has been discussed elsewhere in these forums).

Like Mike, I would have preferred that the change were tested (it may have been ... perhaps the forum administrator would like to contact WHO to check) but, truth be told, the more complicated measurement protocols were seldom used anyway. I'd been measuring MUAC using the current simple method for many years before I saw anyone use a string to ascertain the mid-point.

The paper cited by André (Velzeboer et al, 1983) is interesting since it discusses the problems with using common anthropometric indices (i.e. W/H, W/A, MUAC) with "minimally trained health workers". Mike may find the reported reliability for MUAC to be "woeful" but the reported reliability estimates for W/H and H/A are worse. MUAC may be the best of a bad bunch. This is not, however, a fair test for MUAC since MUAC is (generally) used as a classification tool (i.e. is this child severely wasted or not severely wasted) rather than as a 'continuous' measure. This means that a measure of error suited to 'continuous' measures (e.g. weight and height) are not appropriate. When we did the background work to see if it would be advisable to use MUAC as a joint referral and admission criteria for CTC program in Malawi and Ethiopia in 2002, 2003, and 2005 we found that reliability was poor. The main main source of error was in recording with (e.g.) a MUAC of 104 mm being recorded as 140 mm (a few errors like this will cause SD error to be as large as Mike calculates). In subsequent work we tested the ability of community based volunteers to classify whether a child was above or below a threshold (110 mm). In this work we found very few errors. Importantly, the errors were in 'borderline' subjects and in the direction of higher sensitivity (e.g. a child with MUAC = 111 mm was classified as being below 110 mm but a child with a MUAC = 109 mm was not classified as having a MUAC > 110 mm). Investigations in 2003 and 2005 found that carers put pressure on community-based volunteers to pull the MUAC strap tighter in order to facilitate admisison in 'borderline' cases. The work we did informed the design of the banded numberless (Brixton Health / VALID) MUAC strap design (2005).

Mike states that "The calcualtions [sic] of these parameters are being built into the next version of SMART ...". This needs to be a generalised Cohen Kappa. I think that both types of error should be catered for since raw MUAC may, in the future, be used for treatment monitoring.

Mike writes "Also some form of elastic insert that gives a constant pressure would be great - have been thinking how to do this simply for years and not come up with a good idea - perhaps some of the readers will have good ideas - or have engineer/materials science friends who could think about these problems". This mechanism has been available since 2005 with the numberless (Brixton Health / VALID) MUAC strap design. This functionality is easily added using a rubber band and the marks already on the strap (the dashed line and the two bars on the tail of the strap are used for this purpose) . As far as I know, very few people have used this 'feature' other than during training.

Further to my previous answer ... I have uploaded a document showing how to make an "equal pull" MUAC strap from Brixton Health / VALID International Ltd. (and other) design MUAC straps using very cheap and readily available materials (i.e. an elastic band, a staple, and a ball-point pen - you will also need a marker pen for other designs).

The "equal pull" MUAC strap is useful in early training for demonstrating a "not to loose but not too tight" pull when taking MUAC measurements. Each trainee has an "equal pull" strap for the first hour or so of practical training. After that that we remove the elastic band just as a parent removes training wheels on a child's bicycle.

You can find the document at:

http://tng.brixtonhealth.com/node/141

This document may be freely downloaded, copied, and distributed.

The left arm issue has been covered in this forum. For example, see André Briend's contribution above.

I would like to add a WARNING about the thinking underlying the referenced article:

http://pediatrics.aappublications.org/cgi/content/abstract/126/1/e195

This work assumes that there is a "gold-standard" for SAM and that this is W/H.

Let us put aside the MUAC vs. W/H debate for a moment and state the basic assumption more generally. It is that some measure of CRUDE ANTHROPOMETRY is the best available indicator for measuring the nutritional status of children. I think that this is a mistake.

The terms "nutritional status" and "anthropometric status" are often used interchangeably. Nutritional status refers to the internal state of an individual as it relates to the availability and utilisation of nutrients at the cellular level. This state cannot be observed directly so observable indicators are used instead. There are a range of observable indicators (biochemical, clinical, and anthropometric) of nutritional status, none of which taken alone or in combination are capable of providing a full picture of an individual's nutritional status. There is, therefore, no single "gold-standard" indicator of nutritional status.

While W/H may not be a "gold-standard" it may be a "good" standard. It may even be a better standard than MUAC. We don't have to guess about this. There is plenty of evidence available.

Nutritional status can be usefully defined at the individual, as opposed to the cellular, level as the ratio of nutrient reserves (muscle and fat) to the nutrient requirements of organs (brain, liver, heart, kidneys, lungs, &c.). It is generally recognised that muscle plays a special role as a nutrient reserve during infection and that infection is a major aetiological factor in (and often coincidental with) acute undernutrition. W/H expresses the relationship between weight and height. In children, about 4% of weight is nutrient reserves in muscle. About 96% of weight is, therefore, unrelated to important nutrient reserves. Height is almost completely unrelated to the nutrient requirements of organs. MUAC, however, is directly related to muscle mass and is, therefore, a direct measure of important nutrient reserves.

The evidence that is currently available suggests that an index known as the lean-mass ratio (LMR), the ratio of the estimated mass of the limbs to the estimated mass of the trunk, is the best ANTHROPOMETRIC indicator of nutritional status. The available evidence suggests that MUAC uncorrected for age or height is strongly associated with LMR and is a better indicator of nutritional status than all other practical indicators and that W/H is NOT associated with LMR and is the worst practical indicator of nutritional status.

An alternative to examining the association between an anthropometric indicator and nutritional status is to examine the prognostic or predictive value (i.e. of predicting death) of various indicators. When this has been done, W/H has been consistently shown to be least effective predictor of mortality and that, at high specificities, MUAC is superior to both height-for-age and weight-for-age. It should be noted that this makes the assumption that we are interested in child survival rather than treating "thinness".

W/H is known to depend on climate (lower in warmer climates), altitude (higher in mountains), dietary composition (e.g. higher when milk is consumed), and ethnicity. This means that the meaning of W/H variaes from place to place and person to person. This really rules it out as a "gold-standard". MUAC may also be similarly effected but (limited evidence) to a far lesser extent.

To summarise ... the available evidence consistently shows that W/H is worse than all other practical anthropometric indicators. Any other practical indicator (e.g. H/A, W/A, MUAC, MUAC/A, MUAC/H) performs measurably better than W/H. MUAC looks to be better than the others but also wins out on simplicity, cost, coverage, &c. You might want to look at:

http://www.unu.edu/unupress/food/FNB_v27n3_suppl.pdf

for some detail and references.

In terms of indicators that are practical to collect in developmental and emergency settings, MUAC has the best claim to be a practicable "gold-standard" of nutritional status.

Looked at from this perspective we can interpret the referenced article to show that the WHO version of W/H may improve on the NCHS version but is, unfortunately, still not in very close agreement with MUAC and has all the inherent disadvantages of W/H.

I have a copy of an article by Brambilla et al which uses LMR and DEXA ... it is a late proof with some reviewer comments in the margin. I can let you have this as a PDF if you contact me directly by e-mail:

mark@brixtonhealth.NOSPAM.THANK.YOU.com

remove the ".NOSPAM.THANK.YOU".

You might also like to contact André Briend about this.

As far as I know ... there is no evidence base for recommending the use of the left arm in children (WHO do not recommend this in their IMCI material). It is probably just a superfluous transfer of the procedure used in adults.

Thanks André but do you have a hard link to that document ... I get a session timed out message when I try to connect and then I get a search page.

Yes. Thanks!

I have been thinking and working just a little bit on the issue of dropping of the left arm rule.

I did a little exercise measuring MUAC without asking for the left arm and found that the left arm was presented in the vast majority of cases. This was because carers tended to hold the child with their left arm leaving their most dextrous (right) hand free. The child's free arm (i.e. the arm measured) is then the left arm.

With the left arm only rule we measure the favoured arm of left-handed children. I suppose that we think this OK because few people (global estimates range from 10% to 15%) are left-handed but, if there is a difference in muscle mass between the two arms with the most favoured arm having the bigger mass, we are discriminating against left handed children.

If the mother is left handed then we will get the right arm instead. There is the standard "nature vs. nurture" arguments about whether handedness is inherited or learned (it seems to be a bit of both). Either way the child of a left handed mother is more likely to be left-handed than the child of a right handed mother. I think this means that we may be more likely to measure the least favoured arm of both left and right handed children if we drop the left arm only rule.

It seems to me that we should drop the left arm only rule.

Does this make sense?

I think that if we drop the left arm only rule we will end up (most of the time) measuring the least favoured arm because that is the arm that is presented to us.

In the case of disabled children ... I think that using W/H is not a solution. Kyphosis, scoliosis, inability to stand, weekness, limblessness &c. all affect the height / length measurement. Also, most PHCs in developing countries do not have height boards and IMCI does not cover height measurement of W/H look-up and calculation. Better, I think, to treat disabled children as discretionary admissions or "visible severe wasting" admissions. We have the issue of monitoring and discharge ... I suggest monitoring weight and discharging of (e.g.) 15% weight gain.

Regarding whether to measure the left arm routinely when measuring MUAC in young children, it is probably not important. Hand dominance does not begin to be expressed until age 2-3 years, and is not fully established until 5-6 years of age, so there is little reason to think one arm would be larger in circumference than the other in this age group. Nonetheless, the left arm only rule does remove a decision from the measurer, which is, in general, a good thing.

Indi's report is interesting. I may have this wrong. Doing this at an OTP site with mothers feeding the child with RUTF (with their right hand) may have biased my findings. Any other experiences?

Yes. The upper arm and thigh have similar composition. A lot of work has been done with different measures including calf and thigh circumferences. I did some work (a long tome ago as a junior researcher) on calf circumference in the elderly ago. It was a promising indicator being strongly related to activities of daily living (ADL) and other quality of life (QOL) indicators.

The problems, I think, with using the thigh are (1) accessibility in older males and (2) acceptability (I'm not sure I'd like a stranger to put his hand up my wife's and daughters' dresses AND I'm not sure that I would fell comfortable or safe measuring this indicator in the field). These have meant that we have tended to avoid thigh circumference.

BTW ...legs may be favoured. In sports science they call this "bilateral deficit" and use gait analysis and training to correct this in athletes. Acrobats and gymnasts lead with different legs for different "tricks" ... box-splits (e.g.) are usually done with the stronger leg leading. The old saw that "lost people walk in circles" is true but it is not clear if this is due to favoured legs or some form of navigational confusion (blindfolded people tend to walk in circles but the direction of the turn is not consistent). The favoured leg hypothesis may be nonsense ... although I can think of an evolutionary "just so" story about lost children not wandering too far from their last known position being advantageous in the "struggle for life".

We'd have to get the trousers off the older males but we sometimes have to remove shirts for MUAC. It is known that the opportunities of unsupervised contact with children offered by working in the humanitarian sector attracts pedophiles to NGOs. I am concerned that the use of thigh circumference (TC) blurs the boundary between acceptable and unacceptable touching making child-protection difficult.

This article (e.g.) :

http://tropej.oxfordjournals.org/content/27/5/267.abstract

compares reports MUAC and TC that thigh circumference has some advantages over MUAC. I think that these advantage may be outweighed by child-protection problems with TC.

Try the following Pubmed searches:

http://www.ncbi.nlm.nih.gov/pubmed?term=thigh%20circumference%20malnutrition

http://www.ncbi.nlm.nih.gov/pubmed?term=thigh%20circumference%20nutrition

I hope this is of some use.

Hervé ... You are right ... adherence to considered guidelines and proper supervision is essential. We are getting there with guidelines and monitoring procedures but (IMO) we still have a very long way to go.

Carlos ... I am taken aback by you being taken aback. Here some documentation ...

http://www.hapinternational.org/pool/files/you-feel-like-you're-nothing.pdf

http://www.hapinternational.org/pool/files/no-one-to-turn-to-embargoed-till-27-5.pdf

http://www.hapinternational.org/pool/files/fmr15-8.pdf

http://www.hapinternational.org/pool/files/no-one-to-turn-to-embargoed-till-27-5.pdf

http://www.savethechildren.org.uk/en/docs/sexual_violence_and_exploitation.pdf

http://www.hapinternational.org/pool/files/bso-handbook.pdf

Plenty more available. I though this was a widely known problem.

You may be right. Sexual psychopathy, situational or otherwise, is not my field. Probably a bit if both. Either way ... we need to be careful to avoid it.

Check the local UNICEF office. MUAC straps are a standard UNICEF supply item wherever they support CMAM. UNICEF produce their own straps and distribute them from their own stocks as (I believe) an essential commodity. See:

https://supply.unicef.org/unicef_b2c/mimes/catalog/images/MID_UPPER_ARM_CIRCUMFERENCE_MEASURING_TAPES.pdf

and:

https://supply.unicef.org