Access to this full-text is provided by Springer Nature.
Content available from Nutrition Journal
This content is subject to copyright. Terms and conditions apply.
R E S E A R C H Open Access
Severely malnourished children with a low
weight-for-height have a higher mortality
than those with a low mid-upper-arm-
circumference: III. Effect of case-load on
malnutrition related mortality–policy
implications
Emmanuel Grellety
1*
and Michael H. Golden
2
Abstract
Background: Severe acute malnutrition (SAM) is diagnosed when the weight-for-height Z-score (WHZ) is <−3Z of
the WHO
2006
standards, or a mid-upper-arm circumference (MUAC) of < 115 mm or there is nutritional oedema.
Although there has been a move to eliminate WHZ as a diagnostic criterion we have shown that children with a
low WHZ have at least as high a mortality risk as those with a low MUAC. Here we take the estimated case fatality
rates and published case-loads to estimate the proportion of total SAM related deaths occurring in children that
would be excluded from treatment with a MUAC-only policy.
Methods: The effect of varying case-load and mortality rates on the proportion of all deaths that would occur in
admitted children was examined. We used the same calculations to estimate the proportion of all SAM-related
deaths that would be excluded with a MUAC-only policy in 48 countries with very different relative case loads for
SAM by only MUAC, only WHZ and children with both deficits. The case fatality rates (CFR) are taken from
simulations, empirical data and the literature.
Results: The relative number of cases of SAM by MUAC alone, WHZ alone and those with both criteria have a
dominant effect on the proportion of all SAM-related deaths that would occur in children excluded from treatment
by a MUAC-only program. Many countries, particularly in the Sahel, West Africa and South East Asia would fail to
identify the majority of SAM-related deaths if a MUAC only program were to be implemented. Globally, the
estimated minimum number of deaths that would occur among children excluded from treatment in our analyses
is 300,000 annually.
Conclusions: The number, proportion or attributable fraction of children excluded from treatment with any change
of current policy are the correct indicators to guide policy change. CRFs alone should not be used to guide policy
in choosing whether or not to drop WHZ as a diagnostic for SAM. All the criteria for diagnosis of malnutrition need
to be retained. It is critical that methods are found to identify those children with a low WHZ, but not a low MUAC,
in the community so that they will not remain undetected.
(Continued on next page)
* Correspondence: Emmanuel.Grellety.Bosviel@ulb.ac.be
1
Research Center Health Policy and Systems - International Health, School of
Public Health, Université Libre de Bruxelles, Bruxelles, Belgium
Full list of author information is available at the end of the article
© The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Grellety and Golden Nutrition Journal (2018) 17:81
https://doi.org/10.1186/s12937-018-0382-6
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
(Continued from previous page)
Keywords: Nutrition, Acute malnutrition, Severe acute malnutrition, SAM, Mid-upper-arm circumference, MUAC,
Weight-for-height, WHZ, Mortality, Case fatality rate, Wasting, Oedema, Kwashiorkor, Diagnosis, Case load, Prognosis,
Child, Human
Background
Severe acute malnutrition (SAM) is a lethal condition ac-
counting for about half to one million childhood deaths
[1] annually for children with a weight-for-height/length
(WHZ) below the recommended WHO cut-off. If children
with the other WHO definitions of SAM are added the
death toll is much larger. Identification and treatment of
all children with any of the current definitions of SAM
mandated by the World Health Organisation (WHO) is a
public health priority.
The WHO defines SAM using three independent cri-
teria, WHZ of <−3Z of the WHO
2006
growth standards,
an absolute mid-upper-arm circumference (MUAC) of <
115 mm or the presence of nutritional oedema [2,3].
Some children satisfy several of these criteria.
MUAC can be easily and quickly measured using a
simple coloured tape around the upper arm and oedema
can also be easily assessed in the field. On the other
hand, assessment of WHZ requires the weight and
height to be taken and the resulting numbers looked up
in tables. There is no doubt that MUAC is much easier
to assess than WHZ. For reasons of speed, convenience,
cheapness and simplicity MUAC has been used for many
years to assess malnutrition [4–8]. The ease of use
makes community screening for SAM with MUAC prac-
tical and has been a great advance in identifying affected
individuals in the community.
However, there has now arisen a concerted movement
to stop the assessment of WHZ altogether, even in hospi-
tals and clinics where it is routinely measured at present.
The advocates for only using MUAC are adamant that any
research to develop innovative methods to assess WHZ in
the community is a “waste of effort”as MUAC is the only
criterion that is needed [9–11]. We examined community
based survey data from 48 countries and find that only
16.5% of children who fulfil the WHO definitions of SAM
meet both the MUAC and WHZ criteria. If WHZ is aban-
doned as a criterion about 45% of children with SAM by
WHZ alone will fail to be identified because their MUAC
is above 115 mm. Which criterion identifies the majority
of SAM children varies dramatically from country to
country and the two criteria identify different individuals.
For these reasons we advocated that both MUAC and
WHZ continue to be routinely used to assess children for
SAM and, critically, that convenient and simple ways to
assess WHZ in the community to identify children with
only a deficit in WHZ but not MUAC has to be a major
research priority [12].
These suggestions met with a forceful criticism from
a multi-authored paper [9] which appears to have wide-
spread support by both agencies and donors [13]. The
putative basis of the opinion that WHZ should not be
used at all was that anything that diverts resources
from the widespread use of MUAC to identify SAM
would hinder its implementation and therefore WHZ
assessment must be suppressed [9,14]. The reasons
given against the use of WHZ did not simply emphasise
its inconvenience, with which we agree. The following
were asserted: 1) children with a low WHZ are healthy;
2) their low WHZ is due entirely to their having longer
legs so they do not require treatment; 3) WHZ is a poor
predictor of mortality in children; 4) MUAC is a good
predictor of mortality in children; 5) the two diagnostic
parameters are not complementary; and 6) addition of
WHZ does not improve the sensitivity or specificity of
future all-cause mortality prediction with MUAC.
These contentions were robustly refuted [15].
We have shown in the two preceding papers [16,17],
1) that WHZ < −3Z carries as high, or higher, risk of
death as MUAC < 115 mm; they are clearly not “healthy”
and undeserving of treatment. 2) That the two parame-
ters not only identify different children, and therefore
different risks, but also children satisfying both criteria
have a higher mortality showing the defects to be addi-
tive. 3) That “long legs”is an inadequate explanation for
the regional difference in SAM by WHZ [12,18,19]. 4)
That all the data previously analysed by comparison of
ROC curves, and relied upon to make the assertions of
MUAC’s superiority are severely biased because of math-
ematical coupling [20,21] as well as stochastic and other
problems of interpretation [15]. 5) Despite the flaws the
data actually show that WHZ carries a higher mortality
risk than MUAC when appropriately analysed [16]. In-
deed, there are abundant data to confirm that WHZ <
−3Z carries a substantial risk of death [22–26], but these
papers did not measure MUAC for comparison. Thus,
all the criticisms asserted by Briend et al., and repeated
[9,14,27,28] are, in our opinion, incorrect. Neverthe-
less, their advocacy has led most humanitarian agencies
and some Governments to abandon WHZ altogether.
We do agree that WHZ is more inconvenient and dif-
ficult to measure than MUAC; but this is the only legit-
imate criticism of widespread use of WHZ. The question
arises as to the potential fate of the ≈45% of children
who would not be identified if WHZ measurement was
omitted completely.
Grellety and Golden Nutrition Journal (2018) 17:81 Page 2 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Having shown that the case fatality rates (CFRs) are not
lower in children with only a deficit in WHZ, this paper
examines the practical programmatic differences between
a MUAC-only program and a complete program.
The object of this study was to estimate the proportion
and where possible the numbers of all SAM related
deaths that would occur in children who would be ex-
cluded from treatment if a MUAC-only program re-
placed a complete program.
Methods
Effect of case-load
We used a simple excel spreadsheet to demonstrate the
effect of variations of the proportions of the total
case-load comprised of children with SAM by MUAC,
WHZ and by both MUAC and WHZ with their corre-
sponding CFRs on the proportion of SAM deaths that
would occur in excluded children if a MUAC-only pro-
gram was used.
The total SAM-related-deaths is given by:
(M
CL
xM
CFR
+W
CL
xW
CFR
+B
CL
xB
CFR
)
Where M = children with MUAC < 115 mm and
WHZ > −3Z (S-muac): W = children with WHZ < −3Z
and MUAC > 115 mm (S-whz): B = children with “Both”
MUAC < 115 mm and WHZ < −3Z (S-both): subscript
CL = the proportions of the total case load of SAM that
are in categories M, W and B: subscript CFR = Case fa-
tality rates for children with M, W and B. The case load
always sums to 100% of SAM children (i.e. SAM due to
oedema, kwashiorkor with or without wasting, is not
considered in this calculation).
Then the proportion of total SAM-related-deaths that
would occur in children that would not be eligible for
admission and treatment if WHZ were to be dropped as
an admission criterion is given by:
1−(M
CL
xM
CFR
+B
CL
xB
CFR
)/(M
CL
xM
CFR
+
W
CL
xW
CFR
+B
CL
xB
CFR
).
For the simulation, the relative case-loads were varied
from zero children with S-muac to zero children with
S-whz. The remainder of the children either had the alter-
native criterion or had S-Both. The proportion of children
with S-both was varied from 10 to 30% (the limits we
found in representative nutritional surveys [12]). The
CFRs for S-muac, S-whz and S-both were examined by
changing the ratio of S-muac to S-whz mortality from half
to twice the mortality of the other to represent the likely
limits of the variation in mortality risk. S-both’s CFR was
set at the sum of the CFRs of S-muac and S-whz in ac-
cordance with the empirical data and most of the litera-
ture reports [16,17]. Variation of the overall CFR will
affect the total number of deaths, but the proportions of
SAM-related deaths which would be eligible or ineligible
for treatment is not affected by the absolute CFRs, only by
their ratios and relative case-loads. Thus, if the sizes of the
three CFRs and the proportions of the three case loads do
not change the percent of children that will become ineli-
gible for treatment does not change when the total
SAM-related death rate rises or falls.
Proportion of all SAM-related-deaths that would occur in
children ineligible for treatment with a MUAC-only
program by country
The literature and patient data reported in the first and
second papers [16,17] were subject to ascertainment
bias which made the proportions of the case load com-
ing from the different categories unrepresentative of
SAM in the community. In particular, the proportion of
children in the S-both category was much higher than
that found in the community. That is, the case load ra-
tios of S-muac: S-whz: S-both differed significantly from
that found in representative community surveys of mal-
nourished children [9]. For that reason the case load ratios
of S-muac, S-whz and S-both reported in papers I and II
[16,17] were not used in any calculation. To fairly repre-
sent the situation of SAM children in the community we
used the data previously published from representative
community surveys [12]. These ratios are derived from ana-
lysis of 48,697 SAM children out of a total surveyed popu-
lation of 1,384,068 children, 6–59 months, (1832 surveys)
from 48 countries.
The community-derived, proportionate case-load esti-
mates were then used to estimate the proportion of the
total deaths that would occur in SAM children with a
MUAC-only program; the residue of S-whz would be ex-
cluded. As the mortality rates for S-muac, S-whz and
S-both that would occur in untreated SAM-children in
the communities are unknown we used mortality rates
from 3 sources. First, those used in our theoretical simula-
tion; second, those found in paper 1 [16]; third, the rela-
tive risks of death derived from the meta-analysis of the
literature values where WHO criteria were used and
oedematous cases excluded [17]. The forest plots from the
meta-analyses comparing S-muac with S-whz and S-both,
using adjustment for study quality, [17] are given in
Additional file 1. The relative risks of death from S-muac,
S-whz and S-both were 1.00: 1.14: 2.70 respectively.
The calculations were the same as for the theoretical
simulation.
How many are affected?
In order to estimate the number of children excluded by
a MUAC-only program we examined data from a global
estimate of SAM-related deaths [1] and from India [29].
These estimates are minimum estimates because they
were based upon prevalence data rather than incidence
data and only on WHZ data. We used case loads of
S-muac, 39.5%, S-whz 44.0% and S-both 16.5% for the
global SAM-deaths estimate and S-muac 15.5%, S-whz
Grellety and Golden Nutrition Journal (2018) 17:81 Page 3 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
61.6% and S-both 22.9% for India [12]. The CFRs were
the same for the single deficits and double for S-both.
Ethical statement
This analysis used published data only thus no formal
ethical clearance was required.
Results
Theoretical considerations
Figure 1illustrates the effect of variation of the case-loads
of S-muac, S-whz and S-both and corresponding CFRs to
derive the percent of deaths occurring in children ex-
cluded by a MUAC-only program. The three lines in each
block show the effect of S-muac CFR being half, the same
or twice the CFR of S-whz. These CFRs represent the
likely relative risk and the outside limits. The three blocks
show the effect of an increasing S-both percentages.
If a WHZ-only program was used the results would be
the exact inverse of the percentage exclusion shown.
With a MUAC-only policy, if there are no children in
the community with S-whz (col 1) then all the SAM
children will be identified. If there are no children with
S-muac (last col) the only deaths of MUAC children will
be those who also have S-whz, i.e. those with S-both.
There will be slightly more deaths than the proportion
of overlap because of the higher mortality risk of S-both
children.
A likely scenario is given in block two, second row.
As the percentage of S-muac in the community de-
creases from 60 to 20% the deaths that occur in ex-
cluded children increases from 17 to 50%. To have
20% S-muac is frequently found in nutritional surveys
from some regions [12]. Figure 1also shows that the
exclusion rate is reduced with more S-both children.
For example, if 50% of the children have S-muac (col-
umn 5) as S-both increases from 10 to 20 to 30% the
relative proportion of deaths of excluded children de-
creases from 36, to 33 to 31% respectively. As the
proportion of S-muac decreases the effect of S-both
on excluded cases increases; thus, were there is 20%
of S-muac children (col 7) the percent of excluded
children falls from 64 to 50% to 38%.
Fig. 1 Percentage of SAM-related-deaths of children that would be excluded from treatment in a MUAC-only program. Simulation of the effect of
various case loads and case fatality rates for children with SAM by MUAC-only, WHZ-only and Both criteria. The data gives the percent of total
SAM-related-deaths that would occur in children excluded from treatment in a MUAC-only program. Case Loads and Case Fatality Rates for: S-
muac = MUAC < 115 mm with WHZ > = −3Z: S-whz = WHZ < −3Z with MUAC > 115 mm: S-both = MUAC < 115 mm and WHZ < −3Z. The colours
represent the percent of deaths occurring in cases that would be excluded from treatment in such a program: Red 75–100%: Pink 50–75%:
Orange 25–50%: green 10–25%: Blue 0–10%
Grellety and Golden Nutrition Journal (2018) 17:81 Page 4 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
In contrast, a change of CFR ratios from half to
twice has a relatively minor effect on the proportion
of excluded children (compare the 3 rows vertically).
Thus, a change of case-load ratio is more important
than a change in CRFs ratio within the ranges re-
ported [16,17] in determining the extent of exclusion
of S-whz children.
It should be emphasised that these simulations com-
pares the deaths in excluded children (S-whz) with all
the children that would be identified using a MUAC
measurement (i.e. S-muac plus S-both). In papers [16,17],
CFRs from S-muac and S-whz were compared. Here, by
combining S-both with S-muac in the calculations we rep-
licate the actual effect of only measuring MUAC on the
proportion of deaths related to SAM that would be ex-
cluded from treatment or considered in a coverage survey.
Country data
Because the mortality ratios in nearly all communities is
unknown, in Fig. 2we have used the CFRs from the
simulation and estimated from papers I and II [16,17].
These CFRs are then combined with the actual country
case-loads found in community nutrition surveys pub-
lished previously [12].
In Fig. 2we present, by country, the estimates of the
percentage of deaths of SAM children that would occur in
children excluded from treatment if only MUAC measure-
ments are taken. There is reasonable agreement between
the estimates based upon the different CFRs. For example,
in Senegal the three main CFR estimates indicate that 82,
81 and 83% of deaths occur in excluded children; whereas,
in Mozambique only 13, 12 and 15% of deaths occur in
excluded children. Taking the average of the empirical and
literature exclusion rates, 12 countries would exclude
more than three quarters and 34 would exclude more than
half of the SAM children that die. Only 3 of the 48 coun-
tries would include more than 80% of children who die
with SAM.
The corresponding analysis using a WHZ-only pro-
gram is given in Additional file 2. It is clear that a
WHZ-only program would also fail to identify a large
proportion of the children at high risk of death.
The countries are grouped by region in Fig. 3.Ifthe
countries of South & South East Asia and the Sahel were
to adopt a MUAC-only policy then a substantial propor-
tion of SAM children’sdeathswouldoccurinexcluded
children. The same applies in many of the countries in
West Africa. Some of these countries are characterised by
a dry Sahalian type interior and a wet, heavily populated
coast. These ecologically different areas may have different
levels of exclusion of S-whz so that the effect of only
measuring MUAC may be more deleterious in some areas
than others, and may give a within-country bias to nutri-
tional surveys aimed at establishing the prevalence of
SAM and the level of exclusion. The same conditions
apply to some of the East African countries. On the other
hand several countries in West, East and Central Africa as
well as Asia and Latin America would exclude less than
25% of children that contribute to SAM mortality.
Numbers of excluded SAM children who die
Conversion of relative case loads and CFRs into the
number of deaths of SAM children who will be ineligible
for any treatment under a MUAC-only program is
shown in Table 1. Using Black et al.’s[1] estimate of glo-
bal SAM deaths of 540,000 calculated from WHZ preva-
lence, the total deaths increases to over 800,000 when
we include S-muac deaths. Of these over 300,000 chil-
dren (38%) will die without the possibility of treatment if
WHZ is not measured. In India, although Black et al. es-
timated that there would be 145,000 deaths, Mohan &
Mohan [29] estimated the actual number of deaths due
to SAM to be 270.000; of these, more than half, 155,000,
would be excluded with a MUAC-only policy.
Discussion
If the primary objective of treating children with SAM is to
prevent death then it is logical to look at the percent of
deaths that occur in SAM children that would be excluded
from treatment with any change in policy. This should then
determine whether or not a policy change is unacceptable.
This information cannot be obtained from comparison of
CFRs by regression or areas under ROC curves.
The CFRs estimated from our empirical data [16] and
a meta-analysis of the literature [17] are consistent and
show that the CFRs for S-muac and S-whz are not suffi-
ciently different to affect the rate of exclusion when only
MUAC is used for SAM diagnosis. The dominant factor
is the case-load mix because even when the CFRs differ
substantially the numbers of children that are excluded
show relatively minor changes.
The present country data, by themselves, cannot be
used to determine the absolute numbers of children cal-
culated to die or that would be excluded because of a
change in policy. This requires our data to be combined
with the prevalence/incidence rates, population size and
community mortality rate for at least one of the diagnos-
tic groups. To then derive population attributable frac-
tions also needs the relative risks of death from SAM
children using the non-malnourished children in the
same community as the reference [30,31]. The results
comparing Black et al. [1] and Mohan & Mohan [29]
demonstrate the difficulties in arriving at accurate esti-
mates. Nevertheless, the numbers of children who die
with SAM who would be ignored, if only MUAC is used
is massive. We estimate this to be about 40% of all SAM
children’s deaths globally; but this is variable by country
and region as Fig. 3shows. Ignoring these children and
Grellety and Golden Nutrition Journal (2018) 17:81 Page 5 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
their deaths is the real cost of promoting a MUAC-only
program, and begs the question of “what is acceptable”
from a humanitarian point of view. Reliable and appropri-
ate estimates are essential if correct priorities and policies
are to be set by Governments to address SAM. Death is
not the only adverse effect of severe malnutrition. There
are other major health and long term consequences of
failing to identify and treat the very much larger number
of children with SAM that do not die, estimated to be 10
million in India by WHZ [29].
Fig. 2 Percentage of SAM-related-deaths of children that would be excluded for treatment with a MUAC-only program by country. Sim simulation data
from Fig. 1, representing the probable ratio of case fatality rates (CFRs) and likely extremes; All, IPF, OPT, SFC are the empirical case fatality rates of patients
under different modes of treatment [16]; Literature mortality rates from Additional file 1derived from the data in reference [17]; Case Loads S-muac = MUAC
< 115 mm with WHZ > = −3Z: S-whz = WHZ < −3Z with MUAC > 115 mm: S-both = MUAC < 115 mm and WHZ < −3Z; DRC Democratic Republic of the
Congo; CAR Central African Republic. The case loads per country are from reference [12]. The colours represent the percent of total SAM-related-deaths
occurring in cases that would be excluded from treatment in a MUAC-only program: Red 75–100%: Pink 50–75%: Orange 25–50%: green 10–25%: Blue 0–
10%. * These countries case load comes from a small sample size. ** The case load from Kenya comes from the North of Kenya (similar to Sahel)
Grellety and Golden Nutrition Journal (2018) 17:81 Page 6 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
One answer to the problem of excluded children could
be to increase the MUAC cut-off point. This is a simplistic
suggestion that is impractical as it would then include a
very large proportion of the whole childhood population
at a much lower risk of death [9,13,28,32,33] and divert
and dilute the attention needed for the high risk children.
The paper from Uttar Pradesh, India, by Kapil et al.
[34] is germane to addressing this suggestion. SAM in
Uttar Pradesh by WHZ was 2.2%; when MUAC was
added the prevalence increased to 2.5%. If the MUAC
cut-off was increased to 135 mm then 17% of all the
children in the population would need to be identified;
however, 12% of the S-whz children would still be missed
and the extra case load would only identify a further 15%
of S-whz. Five in 6 of the extra children would be “false
positives”for SAM. Our unpublished analysis from Africa
is in agreement with these figures. The cost, logistics, staff
time with inevitable disruption to other essential medical
services, add-on costs for the parents, possible family guilt
or stigma concerning the need to be checked for SAM
and risk of bringing the program into disrepute all miti-
gate against this policy. Elsewhere evidence shows that
Table 1 Estimation of the possible number of deaths from SAM that would be missed using a MUAC only program
WHZ deaths Total deaths WHZ only
(S-whz)
MUAC only
(S-muac)
Both criteria
(S-both)
MUAC-only
%missed
WHZ-only
%missed
Global 540,000
a
817,000 309,000 277,000 231,000 37.8 33.9
India 270, 000
b
309,000 155,000 39,000 115,000 50.2 12.6
The estimates of total deaths and proportions were derived as in methods. WHZ deaths deaths estimated by WHZ < −3Z (i.e. S-whz + S-both); Total deaths WHZ
deaths plus MUAC-only deaths based on ratios found in reference [12] for Global and India (S-both = 16.5 and 22.9%). Equal mortality risk for S-whz < −3Z and S-
muac < 115 mm and twice the morta lity risk for S-bot h is assumed.
a
Data from reference [1];
b
Data from reference [29]
Fig. 3 Percentage of SAM-related-deaths of children that would be excluded for treatment with a MUAC-only program by Region. Simulation see
Fig. 1;Literature data from [17]; Empirical data from [16]; Case load data from [12]; CFRs Case fatality rates; SE Asia South East Asia; S Asia South
Asia; DRC Democratic Republic of the Congo; CAR Central African Republic. The colour code is the same as Figs. 1and 2. * Case load ratios from
these countries is based on a small sample size. ** For Kenya the case load data comes from the North of Kenya (could be counted with Sahel).
In bold are some of the countries whose governments have officially adopted a MUAC-only program (2017)
Grellety and Golden Nutrition Journal (2018) 17:81 Page 7 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
there are several major difficulties when a false positive
rate exceeds the true positive rate [35,36]. Furthermore,
WHO now recommends “not to provide formulated sup-
plementary foods on a routine basis to children who are
moderately wasted”[37].
In view of the evidence, why has MUAC morphed
from a simple and effective community screening pro-
gram into a MUAC-only program? Is it really necessary
for total suppression of WHZ as a diagnostic for SAM
to legitimise the use of MUAC?
We suggest that there are several reasons. First, based
upon the data presented in the first two papers of this series
[16,17] as well as the present paper it is clear that an in-
appropriate statistical strategy has been exclusively used in
the past using ROC curve analysis of entire populations to
compare the relative CFRs [15]; when the risks are suffered
by different children entirely, the risks are also different,
making comparisons for diagnostic purposes largely mean-
ingless. Second, the exclusive focus on CFRs alone and the
notion that MUAC is a “superior”test; it can only be super-
ior if it identifies the same risk. Third, the neglect of case
load in determining the numbers of excluded children or
the calculation and use of further derived statistics such as
population attributable fraction. Fourth, repeated assertions
that it is safe to ignore S-whz children because they are
healthy when there are no data to support this contention
and abundant data to show that these children are at high
risk of death combined with misquotation and criticism of
any data that does not support the proposition [38]. Fifth,
by forceful advocacy to donors, research funding agencies,
UN agencies and many in the humanitarian organisations
[13]. Sixth, because of an understandable desire to make
everything as simple as possible whilst denying there is any
cost of excluding SAM children [26]. Simplification beyond
what is possible renders programs unworkable or unethical
(try removing the cold chain from vaccination services). An
effort to simplify SAM treatment by suppressing use of F75,
the initial diet designed for the most critically ill of children
[39] was dropped when the reasons for F75 were properly
explained to the agencies. Only using MUAC is certainly
simple, but it has a real cost, and that is measured in lives
lost. Last, because of cognitive biases, particularly confirm-
ation bias [40] among those subscribing to a MUAC-only
policy and those providing “confirmatory”low quality evi-
dence such as some of the papers reviewed in [17].
TheeaseofuseofMUACmakescommunityscreening
to identify children in need to SAM treatment practical and
has led to increasingly greater “coverage”rates for those
children with MUAC < 115 mm. Our data does not ques-
tion the utility of this, what is does demand is research to
find ways that are sufficiently simple to be applied in the
community so that children with WHZ < −3 can also be
included in treatment programs. Stereo-photography has
been used for many years [41]butwithmoderntechnology
this has become practical [42–44]. There are reports of low
cost scanning attachments to smart-phones that give pre-
cise measures of height, head-circumference and MUAC
[45]. Therefore, on the horizon are techniques that will
make the assessment of WHZ simple to use in the commu-
nity. Such studies must be properly funded, supported and
then implemented and make it premature to cease consid-
eringWHZasaproperdiagnosticforSAM.
Conclusions
Some within the nutritional community has been misled
by replicated but flawed analyses and assertions. They are
also attracted by the ease and low cost of MUAC screen-
ing; these practical aspects are clearly advantages to be
considered. MUAC-only programs fail to identify enor-
mous numbers of the most vulnerable children in many
societies. It may be difficult to identify S-whz children, but
that is not a reason to pretend these children do not exist
or to justify ignoring them by making false claims such as
that they are healthy. It is essential that they are included
in any program that claims to address the scourge of
SAM. In our opinion many of these programs should be
considered as contravening the dictates of Hippocrates.
Both a WHZ < −3Z and MUAC < 115 mm must be
retained and used wherever possible as diagnostic cri-
teria for SAM. The research priority must be to develop
innovative ways of assessing WHZ so that it can be ex-
tended to S-whz identification in the community.
Additional files
Additional file 1: Figure S1. Forest plots of papers 1–7of[17]to
determine the CRFs of children with S-muac, S-whz and S-both. The
meta-analyses were performed as in [17] using the quality of the study as
weighting (QE); only reports that used the recommended WHO diagnos-
tic criteria, and excluded oedematous (oed) cases were selected for this
analysis. IND India; NER Niger; SDN South Sudan; UGA Uganda; MWI
Malawi; SEN Senegal; RR relative risk; CI confidence intervals. (TIF 1732 kb)
Additional file 2: Figure S2. Percentage of SAM-related-deaths of children
that would be excluded from treatment by a WHZ-only program. Sim simula-
tion data from Fig. 1, representing the likely extremes and probable ratio of
case fatality rates (CFRs); All, IPF, OPT, SFC are the empirical case fatality rates of
patients under different modes of treatment [16]; Literature mortality rates from
Additional file S1, from reference [17]; Case Loads S-muac = MUAC < 115 mm
with WHZ > = −3Z: S-whz = WHZ < −3Z with MUAC > 115 mm: S-both =
MUAC < 115 mm and WHZ < −3Z; DRC Democratic Republic of the Congo;
CAR Central African Republic. The case loads per country are from reference
[12]. The colours represent the percent of total SAM-related-deaths occurring
in cases that would be excluded from treatment in a MUAC-only program:
Red 75–100%: Pink 50–75%: Orange 25–50%: green 10–25%: Blue 0–10%. *
These countries case load comes from a small sample size. ** The case load
from Kenya comes from the North of Kenya (similar to Sahel). (TIF 4541 kb)
Abbreviations
CFR: Case fatality rate; IPFs: In-patient treatment facilities; MAM: Moderate
acute malnutrition; MUAC: Mid-upper-arm-circumference; NGOs: Non-
governmental organisations; OTPs: Out-patient treatment programs; ROC
curve: Receiver operating characteristic curve; SAM: Severe acute
malnutrition; SFCs: Supplementary feeding centres; WHO: the World Health
Organisation; WHZ: Weight-for-height Z-score
Grellety and Golden Nutrition Journal (2018) 17:81 Page 8 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
Funding
Nutriset provided a PhD fellowship to Université Libre de Bruxelles in
support of EG. Nutriset had no role in any aspect of this research including
data collection, design, analysis, interpretation or writing the article. MHG
received no support.
Availability of data and materials
The datasets used and/or analysed during the current study are available
from the corresponding author on reasonable request.
Authors’contributions
EG & MHG were involved in all stages from the conception and design, data
acquisition, analysis and interpretation. Both authors approved the final
version of the article.
Ethics approval and consent to participate
This is a secondary analysis of anonymous published data. As no individual,
location or administrative district could be identified no formal ethical
clearance was required.
Consent for publication
Not applicable
Competing interests
The authors declare that they have no competing interests.
Publisher’sNote
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
Author details
1
Research Center Health Policy and Systems - International Health, School of
Public Health, Université Libre de Bruxelles, Bruxelles, Belgium.
2
Department
of Medicine and Therapeutics, University of Aberdeen, Aberdeen, Scotland.
Received: 24 May 2017 Accepted: 25 July 2018
References
1. Black RE, Victora CG, Walker SP, Bhutta ZA, Christian P, De Onis M, et al.
Maternal and child undernutrition and overweight in low-income and
middle-income countries. Lancet. 2013;382:427–51. http://www.
sciencedirect.com/science/article/pii/S014067361360937X
2. WHO, Unicef. WHO child growth standards and the identification of
severe acute malnutrition in infants and children: a joint statement by
the World Health Organization and the United Nations Children’sfund.
2009. http://www.who.int/nutrition/publications/severemalnutrit ion/
9789241598163_eng.pdf
3. WHO. Guideline: Updates on the management of severe acute malnutrition
in infants and children. Geneva, World Health Organization; 2013. http://
www.who.int/nutrition/publications/guidelines/updates_management_
SAM_infantandchildren/en/.
4. Jelliffe DB. The assessment of nutritional status of the community. Geneva:
WHO; 1966. PMID: 4960818
5. Jelliffe EFP, Jelliffe DB. The arm circumference as a public health in dex
of protein-calorie malnutrition of early childhood. J Trop Pediatr. 1969;
32:1527–30.
6. Jelliffe DB. Arm circumference in children. Lancet. 1970;1(7641):305–6.
7. Shakir A, Morley D. Measuring malnutrition. Lancet. 1974;1:758–9.
8. Shakir A. Arm circumference in the surveillance of protein-calorie
malnutrition in Baghdad. Am J Clin Nutr. 1975;28:661–5.
9. Briend A, Alvarez JL, Avril N, Bahwere P, Bailey J, Berkley JA, et al. Low mid-
upper arm circumference identifies children with a high risk of death who
should be the priority target for treatment. BMC Nutr. 2016;2:63. https://
bmcnutr.biomedcentral.com/articles/10.1186/s40795-016-0101-7
10. EN-Net. WFH versus MUAC. 2015. Emergency Nutrition Network. http://
www.en-net.org/question/1915.aspx
11. EN-Net. Only MUAC for admission and discharge? 2015. Emergency
Nutrition Network. http://www.en-net.org/question/1922.aspx
12. Grellety E, Golden MH. Weight-for-height and mid-upper-arm circumference
should be used independently to diagnose acute malnutrition: policy
implications. BMC Nutr 2016, 2: 10. https://bmcnutr.biomedcentral.com/
articles/10.1186/s40795-016-0049-7
13. Bailey J, Chase R, Kerac M, Briend A, Manary M, Opondo C et al. Combined
protocol for SAM/MAM treatment. The ComPAS study. Field exchange 2016,
53: 44. http://www.ennonline.net/fex/53/thecompasstudy
14. Hammond W, Badawi AE, Deconinck H. Detecting severe acute malnutrition in
children under five at scale. The Challenges of Anthropometry to Reach the
Missed Millions. Ann Nutr Disord Ther. 2016;3:1030. http://austinpublishinggroup.
com/nutritional-disorders/fulltext/andt-v3-id1030.php
15. Grellety E, Golden MH. Response to Briend et al “low mid-upper-arm-
circumference identifies children with a high risk of death and should be
the priority target for treatment”. BMC Nutr. 2016;2-63:1–12. https://bmcnutr.
biomedcentral.com/articles/10.1186/s40795-016-0101-7
16. Grellety E, Golden MH. Severely malnourished children with a low weight-
for-height have a higher mortality than those with a low mid-upper-arm-
circumference: I. Empirical data demonstrates Simpson’s paradox. Nutr J.
2018. https://doi.org/10.1186/s12937-018-0384-4
17. Grellety E, Golden MH. Severely malnourished children with a low weight-
for-height have a higher mortality than those with a low mid-upper-arm-
circumference: II. Systematic literature review and meta-analysis. Nutr J.
2018. https://doi.org/10.1186/s12937-018-0383-5
18. Post CL, Victora CG. The low prevalence of weight-for-height deficits in
Brazilian children is related to body proportions. J Nutr 2001;131:1290–1296.
http://jn.nutrition.org/content/131/4/1290.full
19. Roberfroid D, Huybregts L, Lachat C, Vrijens F, Kolsteren P, Guesdon B.
Inconsistent diagnosis of acute malnutrition by weight-for-height and mid-
upper arm circumference: contributors in 16 cross-sectional surveys from
South Sudan, the Philippines, Chad, and Bangladesh. Nutr J. 2015;14:1.
https://nutritionj.biomedcentral.com/articles/10.1186/s12937-015-0074-4
20. Archie JP Jr. Mathematic coupling of data: a common source of error. Ann
Surg. 1981;193:296. http://journals.lww.com/annalsofsurgery/abstract/1981/
03000/mathematic_coupling_of_data__a_common_source_of.8.aspx
21. Tu YK, Maddick IH, Griffiths GS, Gilthorpe MS. Mathematical coupling can
undermine the statistical assessment of clinical research: illustration from
the treatment of guided tissue regeneration. J Dent. 2004;32:133–42.
https://doi.org/10.1016/j.jdent.2003.10.001
22. Puffer RR, Serrano CV. Patterns of mortality in childhood: report of the inter-
American investigation of mortality in childhood, Paho scientific publication
no. 262, 1–470. Washington: Pan American Health Organization; 1973.
http://www.popline.org/node/518356
23. Puffer RR, Serano CV. The role of nutritional deficiency in mortality: findings
of the inter-American investigation of mortality in childhood. Bol Ofic Sanit
Panam. 1973;7:1–25.
24. Pelletier DL. The relationship between child anthropometry and mortality in
developing countries: implications for policy, programs and future research.
J Nutr. 1994;124(10 Suppl):2047S–81S.
25. O’Neill SM, Fitzgerald A, Briend A, Van Den Broeck J. Child mortality as predicted
by nutritional status and recent weight velocity in children under two in rural
Africa. J Nutr. 2012;142:520–5. https://doi.org/10.3945/jn.111.151878.
26. Olofin I, McDonald CM, Ezzati M, Flaxman S, Black RE, Fawzi WW, et al.
Associations of suboptimal growth with all-cause and cause-specific
mortality in children under five years: a pooled analysis of ten prospective
studies. PLoS One. 2013;8:e64636.
27. Briend, A. Use of MUAC for severe acute malnutrition. CMAM forum 2012.
http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.662.303&rep=
rep1&type=pdf
28. Briend A, Maire B, Fontaine O, Garenne M. Mid-upper arm circumference
and weight-for-height to identify high-risk malnourished under-five
children. Matern Child Nutr. 2012;8:130–3. http://onlinelibrary.wiley.com/doi/
10.1111/j.1740-8709.2011.00340.x/full
29. Mohan P, Mohan SB. Management of Children with severe acute
malnutrition in India: we know enough to act, and we should act now.
Indian Pediatr. 2017;54(10):813–4. http://indianpediatrics.net/oct2017/813.pdf
30. Rockhill B, Newman B, Weinberg C. The use and misuse of population
attributable fractions. Am J Public Health. 1998;98:2119–21.
31. Hanley JA. A heuristic approach to the formulas for population attributable
fraction. J Epidemiol Community Health. 2001;55:508–14. http://jech.bmj.
com/content/55/7/508.full
32. Shepherd S, Becquet R. Integrated Treatment Protocol for Acute
Malnutrition: A Non Inferiority Trial in Burkina Faso (MUAC-Only).
ClinicalTrials.gov 2017. https://clinicaltrials.gov/ct2/show/NCT03027505
Grellety and Golden Nutrition Journal (2018) 17:81 Page 9 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
33. Maust A, Koroma AS, Abla C, Molokwu N, Ryan KN, Singh L, et al. Severe
and moderate acute malnutrition can be successfully managed with an
integrated protocol in Sierra Leone. J Nutr. 2015;145(11):2604–9. https://doi.
org/10.3945/jn.115.214957
34. Kapil U, Pandey RM, Bansal R, Pant B, Varshney AM, Yadav CP, et al. Mid-
upper arm circumference in detection of weight-for-height Z-score below−
3 in children aged 6–59 months. Public Health Nutr. 2018;5:1–6. https://doi.
org/10.1017/S1368980017004165
35. Wald NJ, Hackshaw AK, Frost CD. When can a risk factor be used as a
worthwhile screening test? BMJ: Br Med J. 1999;319:1562–5. https://doi.org/
10.1136/bmj.319.7224.1562
36. Rosenblatt RA. The perinatal paradox: doing more and accomplishing less.
Health Aff. 1989;8:158–68. https://doi.org/10.1377/hlthaff.8.3.158
37. World Health Organization. Guideline: assessing and managing children at
primary health-care facilities to prevent overweight and obesity in the
context of the double burden of malnutrition. Updates for the Integrated
Management of Childhood Illness (IMCI) Geneva (Switzerland): World Health
Organization; 2017. http://apps.who.int/iris/bitstream/10665/259133/1/
9789241550123-eng.pdf
38. Murray C. How to accuse the other guy of lying with statistics. Stat Sci.
2005;20:239–41. http://www.jstor.org/stable/20061179
39. Briend A. Management of severe malnutrition: efficacious or effective? J Pediatr
Gastroenterol Nutr. 2001;32:521–2. http://journals.lww.com/jpgn/Fulltext/2001/
05000/Management_of_Severe_Malnutrition__Efficacious_or.5.aspx
40. Kahneman D. Thinking, Fast and Slow. Macmillan, 2011. http://www.math.
chalmers.se/~ulfp/Review/fastslow.pdf
41. Piebson WR. Monophotogrammetric determination of body volume.
Ergonomics. 1961;4:213–8. http://www.tandfonline.com/doi/abs/10.1080/
00140136108930521
42. Wells JCK, Ruto A, Treleaven P. Whole-body three-dimensional photonic
scanning: a new technique for obesity research and clinical practice. Int J
Obes. 2008;32:232–8. http://www.nature.com/ijo/journal/v32/n2/abs/
0803727a.html
43. Yu W. Development of a three-dimensional anthropometry system for
human body composition assessment. The University of Texas at Austin;
2008. https://repositories.lib.utexas.edu/bitstream/handle/2152/17838/yuw.
pdf%3Bjsessionid%3D61F89DE75AA927339A6A0909C615A7FE?
sequence%3D2
44. Mikat RP. Chest, waist, and hip circumference estimations from stereo
photographic digital topography. J Sports Med Phys Fitness. 2000;40:58.
45. Conkle J, Ramakrishnan U, Flores-Ayala R, Suchdev PS, Martorell R.
Improving the quality of child anthropometry: manual anthropometry in
the body imaging for nutritional assessment study (BINA). PLoS One. 2017;
12(12):e0189332. http://journals.plos.org/plosone/article?id=10.1371/journal.
pone.0189332
Grellety and Golden Nutrition Journal (2018) 17:81 Page 10 of 10
Content courtesy of Springer Nature, terms of use apply. Rights reserved.
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Content uploaded by Emmanuel Grellety Bosviel
Author content
All content in this area was uploaded by Emmanuel Grellety Bosviel on Sep 15, 2018
Content may be subject to copyright.