Resuscitation 85 (2014) 1174–1178

Contents lists available at ScienceDirect

Resuscitation
journal homepage: www.elsevier.com/locate/resuscitation

Clinical Paper

Size does matter – Age-related weight estimation in “tall n’ thin” and

“tiny n’ thick” children and a new habitus-adapted alternative to the
EPLS-formula夽
Christian G. Erker ∗ , Mario Santamaria, Michael Moellmann
Department of Anesthesiology and Operative Intensive Care Medicine, Section Pediatric Anesthesiology, St. Franziskus-Hospital Muenster, Germany

a r t i c l e i n f o a b s t r a c t

Article history: Aim of the study: Weight in kilograms is a required parameter in the emergency medical care of children. In
Received 30 December 2013 emergent situations, obtaining an accurate weight is often not possible. In such situations, weight can be
Received in revised form 16 April 2014 estimated by using an age-dependent formula such as the EPLS-formula (age in years + 4) × 2. As recently
Accepted 18 April 2014
recognized for emergency tapes, the habitus of the child has a major influence on weight estimation. In
this study, the performance of various age-dependent formulas is to be investigated, with special regard
Keywords:
to children demonstrating non-normal growth.
Body weight
Methods: The performance of various formulas for weight estimation in children growing along the 5th,
Body weight and measures
Anthropometry
50th, and 95th percentile is investigated based on a mathematical model compared to the WHO and CDC
Emergency treatment reference percentiles using ICC and Bland–Altman methods. Additionally, a new formula for children
Development demonstrating non-normal growth is derived by regression analysis and tested: f × age in years + 6 with
Growth the factor f being 2 for “tall n’ thin”, 3 for normal and 4 for “tiny n’ thick” children.
Results: All previously published formulas lack precision when applied to children outside the 50th per-
centile. The new habitus-adapted formula shows a better performance for children growing along the
5th or 95th percentile.
Conclusions: The new formula provides enhanced precision in weight estimation and can help in reducing,
e.g. drug dosing errors. It should be used for weight estimation in children demonstrating non-normal
weight development and in situations when superior methods such as weighing or habitus-adapted
emergency tapes are not applicable.
© 2014 Elsevier Ireland Ltd. All rights reserved.

1. Introduction
Weight in kilograms (kg) is a required parameter in the emer-
gency medical care of sick or injured children.1 In cardiac arrest,
defibrillation energy and most drug doses are calculated based
on weight in kilograms. In non-critical medical situations, such
as planned surgery or in-patient care in hospitals, weight can and
should be measured. In life-threatening and prehospital situations,
weighing is commonly not an option as it is time-consuming and
requires repositioning the patient onto a scale. Parental weight esti-
夽 A Spanish translated version of the summary of this article appears as Appendix
in the final online version at http://dx.doi.org/10.1016/j.resuscitation.2014.04.032.
∗ Corresponding author at: St. Franziskus-Hospital Muenster, Department of
Anesthesiology and Operative Intensive Care Medicine, Hohenzollernring 72,
D-48145 Muenster, Germany.
E-mail address: christian.erker@sfh-muenster.de (C.G. Erker).
mation remembering the last weighing is the next best option,2
followed by weight estimation using a length-weight correla-
tion as in the Broselow tape.3,4 European pediatric life support
(EPLS) courses by the European Resuscitation Council (ERC) sug-
gest the use of an age-weight-correlation by an age-dependent
formula, commonly referred to as the EPLS-formula5 : [(age in
years + 4) × 2]. Several other formulas for weight estimation based
on age have been reported in the literature. The basic principle
of age-dependent formulas is the theory that a lone and sick or
injured, yet awake child might correctly tell the health provider his
age, but not his weight.
In pediatrics, growth is usually expressed by using percentiles.
The 50th percentile resembles the mean weight or length for
given age.11,12 All formulas have been developed by referencing
normal-sized and normal-weighted children.13 Therefore, they
lack accuracy in children demonstrating a non-normal growth
– specifically thinner-than-usual or thicker-than-usual children.
Weight estimation using a length-weight correlation, such as the

http://dx.doi.org/10.1016/j.resuscitation.2014.04.032
0300-9572/© 2014 Elsevier Ireland Ltd. All rights reserved.
 C.G. Erker et al. / Resuscitation 85 (2014) 1174–1178
Fig. 1. Weight estimation using the various formulas described in the literature compared to the reference percentiles shows close estimation to the 50th percentile.
Broselow-tape, has the same limitation. This has recently been
recognized and led to the development of tools like the PAWPER-
tape6,7 or the Mercy-method-tape (ClinicalTrials NCT01709500)
that incorporate other parameters like habitus or mid-arm
circumference8 for a more precise and differentiated estimation.
The American Heart Association advocates in their actual guide-
lines for Pediatric Advanced Life Support (PALS) the use of tapes,
though stating that habitus might be an important consideration.9
The incorporation of habitus into an age-dependent formula has
yet to be introduced. This led the authors of this article to the
idea of developing a new as-simple-as-possible formula for age-
dependent weight estimation that can be also used in children
with non-normal weight. In this article, the performance of the
commonly used formulas in children with lean, normal and sturdy
habitus is examined followed by the presentation of a new formula
for weight estimation. This is based on the standard percentiles for
growth in childhood by the World Health Organization (WHO) and
the Center for Disease Control (CDC).
2. Materials and methods
For regression analysis, a linear regression model was applied
onto the same reference database in order to produce a formula
for weight as a function of age in years, separately for the 5th,
50th and 95th percentile. For goodness of fit testing, coefficient
of determination R2 and analysis of variance were performed.
The reference percentiles for children aged 1–12 years are
shown in Fig. 1. Using these data as reference, the authors compared
the results of various age-dependent formulas and specifically their
performance in children growing along the 5th, 50th and 95th per-
centile using “intraclass correlation coefficient” (ICC, exact values,
same observer) and the Bland–Altman-method (reference per-
centile as reference method, ratio). The correlation analysis using
the ICC is superior over the more common Pearson’s method when
comparing exact values. Correlations of above 95% were presumed
as good, 90–95% as moderate and below 90% as low. The results of
the newly derived habitus-adapted formulas were tested using the
same techniques against the original reference percentiles.
All calculations were performed using MedCalc13.0.4 and
Microsoft Excel 2010. When applicable, a p-value of 0.05 was con-
sidered statistically significant.
1175
3. Results
The performance of the various formulas described in the lit-
erature is shown in Fig. 1 and Table 1. The authors also added the
simplest formula [kg = 4 × years], which is a simplification of the
best guess method into analysis.
The displayed ICC-value represents the absolute correlation
between the reference values and the results of the shown for-
mula. The colored background of cell in Table 1 represents the
goodness of correlation as indicator for the quality of the formu-
las performance. The displayed Bland–Altman value represents an
indicator for quantitative difference of the formula’s estimation
compared to the reference percentile. A value below 1 signifies an
over-estimation by the formula. Concordantly, a value above 1 sig-
nifies under-estimation. For example, a value of 1.51 represents
a ratio of reference 1.51 vs. formula 1, meaning a formula result
roughly 66% of the reference value.
Most formulas show a sufficient performance when estimating a
normal child’s weight, but demonstrate less precision when applied
on thin or thick children. Noticeable is the fact that the simplest
formula, which is clearly the easiest to remember, does not perform
catastrophically.
In the regression analysis, the three reference percentiles can
be best described by the following three formulas using a linear
regression model. The goodness of fit of the regression equations is
sufficient, as shown by R2 and p-values.
5th percentile: kg = 1.9428 × years + 6.4444 R2 = 0.9810, p < 0.05
50th percentile: kg = 2.7117 × years + 6.6299 R2 = 0.9729, p < 0.05
95th percentile: kg = 4.2997 × years + 5.6315 R2 = 0.9671, p < 0.05
The authors suggest the following three formulas for estimat-
ing a child’s weight under consideration of the habitus. The linear
equations can still be calculated in mind without the need for a
calculator. The adaptations for habitus are easily been done by
choosing the factor 2, 3 or 4.
thin children (5th percentile): kg = 2 × years + 6
normal children (50th percentile) kg = 3 × years + 6
thick children (95th percentile): kg = 4 × years + 6
1176 C.G. Erker et al. / Resuscitation 85 (2014) 1174–1178

Table 1
Results of the various formulas compared to the reference percentiles using ICC and Bland–Altman methods for children aged 1–12 years.

name formula references ICC 5th pc (95%-KI) ICC 50th pc (95%-KI) ICC 95th pc (95%-KI)
BA - mean diff. (+-1,96 SD) BA - mean diff. (+-1,96 SD) BA - mean diff. (+-1,96 SD)
5
EPLS / „old“ kg = (years + 4) x 2 0,95 (0,06-0,99) 0,88 (0,34-0,96) 0,48 (-0,10-0,77)
APLS 23
0,90 (0,81-0,99) 1,12 (0,96-1,29) 1,51 (1,11-1,92)
17
“new” APLS 1 to 5 years kg = (2 x years) + 8 0,70 (-0,04-0,89) 0,96 (0,92-0,98) 0,79 (-0,04-0,94)
6 to 12 years kg = (3 x years) + 7 0,79 (0,59-0,98) 0,98 (0,78-1,17) 1,30 (1,11-1,50)
24
Luscombe kg = (years x 3) + 7 0,66 (-0,07-0,89) 0,96 (0,54-0,99) 0,80 (0,05-0,93)
25
0,74 (0,61-0,88) 0,93 (0,79-1,06) 1,24 (1,04-1,44)
26
Best Guess ≤12 months: kg = (months + 9) / 2 0,62 (-0,08-0,87) 0,93 (0,31-0,98) 0,87 (0,04-0,96)
1 to 5 years: kg = (2 x years) + 10 27
0,72 (0,60-0,84) 0,90 (0,79-1,00) 1,20 (1,05-1,35)
6 to 14 years: kg = 4 x years
28
Park ≤12 months: kg = (months + 9) / 2 0,67 (-0,07-0,88) 0,96 (0,78-0,98) 0,83 (-0,03-0,95)
1 to 5 years: kg = (2 x years) + 9 0,76 (0,61-0,90) 0,94 (0,81-1,07)
>6 years: kg = 4 x years - 1
simplified kg = 4 x years - 0,67 (0,18-0,84) 0,92 (0,88-0,94) 0,85 (-0,03-0,96)
Best-guess- 0,94 (0,06-1,82) 1,16 (0,14-2,18) 1,52 (0,38-2,66)
method
29
Shann 1 to 9 years: (2 x years) + 9 0,89 (-0,01-0,97) 0,94 (0,80-0,97) 0,55 (-0,08-0,81)
>9 years: 3 x years 0,84 (0,77-0,92) 1,06 (0,90-1,21) 1,42 (1,04-1,80)
30
Nelson 3 to 12 months: kg = (months + 9) / 2 0,87 (0,05-0,96) 0,98 (0,55-0,99) 0,63 (-0,08-0,87)
1 to 6 years: kg = 2 x (years + 4) 31
0,86 (0,76-0,96) 1,07 (0,98-1,16) 1,44 (1,21-1,66)
7 to 12 Jahre: kg = [(years x 7) - 5] / 2
new habitus- tall n‘thin: 2 x years + 6 this arcle 0,99 (0,99-0,99)
adapted 1,01 (0,90-1,13)
formula normal: 3 x years + 6 0,98 (0,93-0,99)
0,97 (0,79-1,15)
ny n‘thick: 4 x years + 6 0,97 (0,94-0,99)
1,06 (0,85-1,26)
pc = percentile; ICC = intraclass correlation coefficient; BA = Bland–Altman mean difference from reference percentile as ratio [reference/formula (±1.96 standard deviation)].
Color-coding: ICC > 95% green, 90–95% orange, below 90% red.

The performance of these suggested formulas is tested using
ICC and Bland–Altman, the results are displayed in Table 1. The
correlation against the reference percentiles is displayed in Fig. 2.
Of all described formulas, only this three-parted formula shows
good correlation in all three subgroups with the lowest quantitative
difference from the reference percentiles.
4. Discussion
The EPLS-formula is the most commonly referred age-
dependent formula. Imprecise estimations, especially in popula-
tions with growth differing from the United States [13–15], have led
to the development of several other formulas.16 The “Advanced Life
Support Group” (ALSG) changed their recommendation for an age-
dependent formula in the 5th edition of their “advanced pediatric
life support” (APLS) course manual17 from the EPLS-formula to a
more complex one (see Table 1). Each of the various formulas shows
superiority in specific collectives; some are easy to remember and
some are not.18 Formulas have the advantage over tapes that they
are always available, even in low-resource settings. Also, they aren’t
time-consuming and do not interfere with physical interventions,
making them practicable especially for resuscitation scenarios. Age
changes less frequently than weight and is an important parame-
ter in every day’s language for children and parents. In stressful
situations, such as cardiac arrest, age can be told more precisely
than weight by accompanying parents. It is also easier for health

70

60

50
weight [kg]

40 new tall n'thin

new normal
new ny n'thick
30 weight 5th
weight 50th
weight 95th
20

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12
age [years]

Fig. 2. The new habitus-adapted formulas compared to the reference percentiles demonstrate close estimation to the corresponding percentile for each of the three habitus
categories.
 C.G. Erker et al. / Resuscitation 85 (2014) 1174–1178
care providers to estimate age than weight. Age is also a required
parameter for the use of formulas for estimation of endotracheal
tube length and size in children.5
All formulas perform more or less well when estimating a nor-
mal grown child’s weight. All formulas perform badly in children
with a non-standard habitus. The difference in estimation in thick
children can be as large as 52%, which has the potential for rel-
evant problems for example in drug dosing.19 The most critical
errors involve decimal errors and consecutively tenfold under-
or overdosage, which is not a problem associated with weight
estimation.20 The demonstrated dosage differences of approxi-
mately 50% due to wrong weight estimation can supposedly lead
to a clinical relevant dosing error in drugs with a low therapeutic
threshold. This involves for example antiarrhythmics in periar-
rest situations or sedatives in seizure therapy or status epilepticus,
which is the drug group affected most by medication errors.20
High-dose epinephrine has been shown to have adverse effects and
negative influence on outcome,21,22 although most studies exam-
ined tenfold dosage increments. Ultimately, an absolutely exact
weight is presumably not necessary in life-and-death situations as
the exact dosing of most resuscitation drugs in childhood is still
unknown.9
Using raw data of the percentiles of the World Health Orga-
nization (WHO)10 for children under 24 months age11 and the
percentiles of the Center for Disease Control (CDC)12 for children
above 24 months age, the authors developed a mathematical model
for the 5th, 50th, and 95th percentile (weight-for-age) which is
gender-unspecific. This mathematical approach is preferred over
the collection of individual measurements because of the far supe-
rior database of the percentiles that are based on thousands of
individual observations. The selected collective for the use of for-
mulas in the context of this article is children between 1 and
12 years age. Infants show a great interindividual variability, and
are normally accompanied by a caretaker. Thus, the use of age-
dependent formulas might not make sense in infants. Children
above 12 years age are usually able to report their weight correctly
and the weight differences to younger adults are smaller than in
younger childhood. Also, the entry into puberty results in a great
variability of growth and consecutively weight. This is in line with
most of the established formulas, the EPLS-formula being used for
the ages 1–10 and the APLS-formula for the ages 1–12.
When developing a new formula for clinical use that can be used
in all three categories of habitus, it is necessary to consider the
following: (1) a new formula should be as simple as possible so
that it can be remembered in a stressful situation, (2) the necessary
adaptations for thin and thick children must not be too difficult in
order to keep it simple, (3) the performance of the new formula
must still be good after simplification. The authors preferred a lin-
ear model over a logarithmic or exponential model because of the
practical simplicity. The displayed R2 -values demonstrate an ade-
quate goodness of fit for this approach. One possible pitfall is the
non-standardized and highly subjective categorization of the child
as thin, thick or normal. While the extremes are easy to identify, the
misjudgment of a normal grown child into another category leads to
wrong weight estimation and potentially to drug dosing errors. The
highest differences that can occur using the new formulas is 22%
when misjudging a normal child as thick and 28% when misjudging
a normal child as thin.
The additional consideration of the child’s habitus has
shown improvement of weight estimation using length-weight-
correlations by emergency tapes. When using an age-weight-
correlation, the consideration of the habitus as demonstrated in
this article leads to a more precise estimation of weight in chil-
dren demonstrating non-normal weight. Clearly, it is not as easy
to remember as a single formula. But also in normal weighing
children, the new formula performs superior to others. This might
1177
aid the introduction of this new formula, as the adaptation needed
for thin or thick children is only the replacement of the factor with
2, 3 or 4. The additional precision aids the healthcare provider for
example by reducing drug dosing errors, especially for drugs with a
low therapeutic threshold. Nevertheless, correct weighing by scales
or the use of habitus-adapted emergency tapes still is the superior
method, when applicable. The use of this habitus-adapted for-
mula makes sense in resuscitation situations and life-threatening
emergencies, when weighing and other technical methods are not
available or too time-consuming to apply and an approximated
weight is needed instantly.
Conflict of interest statement
The authors state that they have no conflicts of interest.
References
1. Bartelink IH, Rademaker CMA, Schobben AFAM, van den Anker JN. Guidelines
on paediatric dosing on the basis of developmental physiology and pharmacoki-
netic considerations. Clin Pharmacokinet 2006;45:1077–97.
2. Partridge RL, Abramo TJ, Haggarty KA, et al. Analysis of parental and nurse weight
estimates of children in the pediatric emergency department. Pediatr Emerg
Care 2009;25:816–8.
3. Luten RC, Wears RL, Broselow J, et al. Length-based endotracheal tube and emer-
gency equipment in pediatrics. Ann Emerg Med 1992;21:900–4.
4. Luten R. Error and time delay in pediatric trauma resuscitation: addressing
the problem with color-coded resuscitation aids. Surg Clin North Am
2002;82:303–14, vi.
5. ERC: EPLS – European Pediatric Life Support: ERC manual for health care workers
dealing with paediatric resuscitation; 2010.
6. Wells M, Coovadia A, Kramer E, Goldstein L. The PAWPER tape: a new concept
tape-based device that increases the accuracy of weight estimation in chil-
dren through the inclusion of a modifier based on body habitus. Resuscitation
2012;84:227–32.
7. Wells M, Kramer E. Optimizing emergency drug dosing in children. Acad Emerg
Med 2008;15:1325.
8. Cattermole G, Leung P, Mak P, Graham C, Rainer T. Mid-arm circumference can
be used to estimate children’s weights. Resuscitation 2010;81:1105–10.
9. Kleinman ME, Chameides L, Schexnayder SM, et al. Part 14: Pediatric Advanced
Life Support: 2010 American Heart Association Guidelines for Cardiopulmonary
Resuscitation and Emergency Cardiovascular Care. Circulation 2010;122:S876.
10. de Onis M, Garza C, Onyango AW, Rolland-Cachera M. WHO growth standards
for infants and young children. Arch Pediatr 2009;16:47–53.
11. de Onis M, Garza C, Onyango AW, Borghi E. Comparison of the WHO child growth
standards and the CDC 2000 growth charts. J Nutr 2007;137:144–8.
12. de Onis M, Onyango AW. The Centers for Disease Control and Prevention 2000
growth charts and the growth of breastfed infants. Acta Paediatr 2003;92:413–9.
13. Lulic I, Kovic I. Comparing methods for weight estimation of children: a prelim-
inary study. Resuscitation 2010;81:S77.
14. House DR, Ngetich E, Vreeman RC, Rusyniak DE. Estimating the weight of chil-
dren in Kenya: do the Broselow tape and age-based formulas measure up? Ann
Emerg Med 2013;61:1–8.
15. Ali K, Sammy I, Nunes P. Is the APLS formula used to calculate weight-for-age
applicable to a Trinidadian population? BMC Emerg Med 2012;12:9.
16. Seddon C, Lockitt L, Dhanjal S, Eisenhut M. Validation of advanced paediatric
life support formulas for weight calculation in a multiethnic population. ISRN
Pediatr 2012;2012:1–4.
17. Samuels M. Advanced paediatric life support: the practical approach. 5th ed.
Chichester, West Sussex, UK: Wiley-Blackwell; 2012.
18. Graves L, Chayen G, Peat J, O’Leary F. A comparison of actual to estimated weights
in Australian children attending a tertiary children’s’ hospital, using the original
and updated APLS, Luscombe and Owens, Best Guess formulae and the Broselow
tape. Resuscitation 2013;85:392–6.
19. Lim CAE, Kaufman BJ, O’Connor J, Cunningham SJ. Accuracy of weight estimates
in pediatric patients by prehospital Emergency Medical Services personnel. Am
J Emerg Med 2013;31:1108–12.
20. Ghaleb MA, Barber N, Franklin BD, Yeung VWS, Khaki ZF, Wong ICK. Sys-
tematic review of medication errors in pediatric patients. Ann Pharmacother
2006;40:1766–76.
21. Berg RA, Otto CW, Kern KB, et al. High-dose epinephrine results in greater early
mortality after resuscitation from prolonged cardiac arrest in pigs: a prospective,
randomized study. Crit Care Med 1994;22:282–90.
22. Perondi MBM, Reis AG, Paiva EF, Nadkarni VM, Berg RA. A comparison of high-
dose and standard-dose epinephrine in children with cardiac arrest. N Engl J
Med 2004;350:1722–30.
23. Geduld H, Hodkinson PW, Wallis LA. Validation of weight estimation by age and
length based methods in the Western Cape, South Africa population. Emerg Med
J 2011;28:856–60.
1178 C.G. Erker et al. / Resuscitation 85 (2014) 1174–1178
24. Kelly A, Nguyen K, Krieser D. Validation of the Luscombe weight formula for
estimating children’s weight. Emerg Med Australas 2011;23:59–62.
25. Luscombe MD, Owens BD, Burke D. Weight estimation in paediatrics: a com-
parison of the APLS formula and the formula ‘Weight = 3(age) + 7’. Emerg Med J
2011;28:590–3.
26. Thompson MT, Reading MJ, Acworth JP. Best Guess method for age-based weight
estimation in paediatric emergencies: validation and comparison with current
methods. Emerg Med Australas 2007;19:535–42.
27. Tinning K, Acworth J. Make your best guess: an updated method for paediatric
weight estimation in emergencies. Emerg Med Australas 2007;19:528–34.
28. Park J, Kwak YH, Kim DK, et al. A new age-based formula for estimating weight
of Korean children. Resuscitation 2012;83:1129–34.
29. Theron L, Adams A, Jansen K, Robinson E. Emergency weight estimation in
Pacific Island and Maori children who are large-for-age. Emerg Med Australas
2005;17:238–43.
30. Kliegman RM, Nelson WE. Nelson textbook of pediatrics. 19th ed. Philadelphia,
PA: Saunders; 2011.
31. Varghese A, Vasudevan VK, Lewin S, Indumathi CK, Dinakar C, Rao SDS. Do the
length-based (Broselow) Tape, APLS, Argall and Nelson’s formulae accurately
estimate weight of Indian children? Indian Pediatr 2006;43:889–94.