Anaesthesia 2015, 70, 176–182 doi:10.1111/anae.

12860

Original Article
An aid to drug dosing safety in obese children: development of a
new nomogram and comparison with existing methods for
estimation of ideal body weight and lean body mass
L. C. Callaghan1 and J. D. Walker2

1 Clinical Fellow, Department of Anaesthesia, Alder Hey Children’s Hospital, Liverpool, UK

2 Consultant Anaesthetist, Department of Anaesthesia, Ysbyty Gwynedd, Bangor, UK

Summary
The risk of accidental over-dosing of obese children poses challenges to anaesthetists during dose calculations for
drugs with serious side-effects, such as analgesics. For many drugs, dosing scalars such as ideal body weight and lean
body mass are recommended instead of total body weight during weight-based dose calculations. However, the com-
plex current methods of obtaining these dosing scalars are impractical in the peri-operative setting. Arbitrary dose
adjustments and guesswork are, unfortunately, tempting solutions for the time-pressured anaesthetist. The study’s
aim was to develop and validate an accurate, convenient alternative. A nomogram was created and its performance
compared with the standard calculation method by volunteers using measurements from 108 obese children. The
nomogram was as accurate (bias 0.12 kg vs
0.41 kg, respectively, p = 0.4), faster (mean (SD) time taken 2.8 (1.0)
min (vs 3.3 (0.9) min respectively, p = 0.003) and less likely to result in mistakes (significant errors 3% vs 19%,
respectively, p = 0.001). We present a system that simplifies estimation of ideal body weight and lean body mass in
obese children, providing foundations for safer drug dose calculation.
.................................................................................................................................................................
Correspondence to: L. C. Callaghan
Email: leannecallaghan@doctors.org.uk
Accepted: 10 August 2014

Introduction Dose adjustment according to dosing scalars such

Obese children are at risk of overdose if dose calcula-
tions are made using total body weight (TBW).
Adjustment of body weight may be necessary, depend-
ing on the specific drug, due to the increased propor-
tion of fat and lean mass that complicates the
distribution, metabolism and excretion of drugs [1, 2].
Opioid-induced respiratory depression and paraceta-
mol-induced hepatic injury are examples of potential
harm due to accidental overdose. Obese children may
also have related co-morbidities such as obstructive
sleep apnoea or fatty infiltration of the liver [3].
as ideal body weight (IBW), lean body mass (LBM)
and TBW depends upon the specific drug [1, 2, 4, 5].
Examples include the use of IBW for intubation doses
of non-depolarising neuromuscular blocking drugs,
TBW for the dose of suxamethonium in rapid
sequence induction and LBM for induction doses of
propofol. The pharmacokinetics of many common
drugs such as paracetamol, morphine, local anaesthetic
agents, anticoagulants and parenteral fluids in child-
hood obesity are unclear and most of the established
guidance has been extrapolated from adult studies.

176 © 2014 The Association of Anaesthetists of Great Britain and Ireland

Callaghan and Walker | A nomogram to aid dosage calculations in obese children
In local audits, we have observed the widespread
use of arbitrary, unscientific dose reduction by clini-
cians for this patient subgroup, presumably in an effort
to prevent overdose. We have also observed a lack of
clear guidance in obtaining IBW and LBM values for
obese children within clinical and therapeutic work-
place reference resources.
Ideal body weight is a concept, that cannot be
directly measured. It represents the weight associated
with the longest life expectancy for age, sex and height
and is based on Cole et al.’s data, which pre-dates the
current obesity surge [6]. Ideal body weight can be
derived using the reverse body mass index (BMI) method.
This is the ideal BMI, corrected for height, as follows:
Ideal body weight ¼ BMI50 height2
where ideal BMI (BMI50) is the BMI value on the 50th
centile of a UK BMI chart for the child’s age and sex
[4–8].
Other methods exist for estimation of IBW, such
as Moore’s method, McLaren’s method and linear
equations such as [2 9 (age + 4)]. These methods
have shown to be useful for certain age subgroups of
children, but the reverse BMI method is superior
across a wider range of ages and heights [9].
Lean body mass is defined as the mass of the lean
tissues (skeletal, organs, etc.) remaining once the adi-
pose tissue mass is removed. Obese children generate
additional lean mass, particularly in the legs, as a result
of the demands of carrying excess weight. They are also
taller on average for their age [10]. Radiological and lab-
oratory based techniques such as dual energy X-ray
absorptiometry (DEXA) scanning and K-Dilution have
been used to obtain this measurement, with varying suc-
cess [11]. The disadvantages of such investigations
include expense, time, complexity and lack of immediate
availability. The validity of the measurement is transient
and is lost as the child grows. An alternative mathemati-
cal estimate for LBM has been recommended, which
makes use of the observation that a mean of 29% of the
excess weight carried by an obese child is lean tissue [4,
7, 8]. This can be expressed as the equation:
LBM ¼ IBW þ 0:29ðTBW
IBWÞ
It is important to note that this method requires an
IBW, and may not be appropriate for non-ambulatory
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2015, 70, 176–182
children whose immobility prevents lean tissue devel-
opment.
Calculation of IBW and LBM using Equations (1)
and (2) requires a number of detailed steps and is only
possible if suitable BMI charts are easily available. The
process is potentially time consuming, with each step car-
rying the possibility of undetected error [12]. We believe
we can minimise the temptation of arbitrary or ‘educated
guess’ dose adjustment if the process of estimating IBW
and LBM can be simplified without compromising accu-
racy, enhancing drug safety in obese children.
The study’s aims were to design and validate a
new nomographic method and compare its perfor-
mance with the existing calculation methods for esti-
ð1Þ mation of IBW and LBM in obese children. Our
primary outcome measure was the accuracy of the cal-
culations when compared with the same calculation
carried out by a computer spreadsheet. Our secondary
outcome measures were the speed of the methods and
the incidence of mistakes.
Methods
Creation of a new nomographic method for
calculation of IBW and LBM
Nomography is a graphical method of calculation first
used in the nineteenth century. Calculations using
nomograms can be accurate to three significant figures,
and before the advent of modern computers they were
widely used in diverse fields such as engineering, chemistry
and medicine [13]. Nomograms are particularly useful in
healthcare settings, where access to IT is often limited [14].
Our first step was to define a mathematical function
that would return ideal BMI. The Child Growth Founda-
tion gave us permission to use their 50th centile BMI data.
These data follow a sigmoid curve above the age of 5 years.
Microsoft Excel was used to fit a four-parameter logistic
(4PL) curve to the data from age 5 upwards [15, 16].
This resulted in two equations:
8:91
BMI50 ðboysÞ ¼ 24:27
  ð3Þ
age 4:40
1þ
15:78
and
7:51
BMI50 ðgirlsÞ ¼ 22:82
  ð4Þ
ð2Þ age 4:44
1þ 13:46
These equations (both for age >5) show a very
high degree of fit with the original data, as is shown in
177
Anaesthesia 2015, 70, 176–182 Callaghan and Walker | A nomogram to aid dosage calculations in obese children
Table 1. By combining these with Equations (1) and
(2), we have the ability to calculate IBW and LBM.
The calculation we wish the nomogram to perform
has the following variables: age; sex; height; IBW;
LBM; and TBW. In addition, we also have the value of
BMI50, calculated from age and sex. To construct the
nomogram, we decided to combine two separate
nomograms: one for age, sex, height and IBW; the
other for IBW, LBM and TBW. By aligning the (com-
mon) IBW scale, we can combine the two nomograms
into one. In addition, we can superimpose the two
sex-specific age scales onto one axis. There are a num-
ber of ways to ‘multiply’ using a nomogram. A ‘Z-
chart’ (so called, because the middle axis – if projected
– will cross the two outer scales in a ‘Z’ or ‘N’) has
the advantage that the product is presented on a linear
scale, and thus can ‘feed’ into the next section of the
nomogram [17, 18]. Equation (1) is already in the
required format for this.
The second part of the nomogram can be simply
modelled as a parallel-scale nomogram. This can be
achieved by rewriting (2) in the form:
LBM ¼ 0.71 IBW þ 0.29 TBW ð5Þ
It is interesting to note at this point that LBM is a
weighted average of IBW and TBW. A first draft of
the nomogram was created by hand, using standard
techniques, to ensure that any subsequent software
output looked ‘as expected’. We then used Pynomo
(an open-source software package available from
http://www.pynomo.org) to draft the final chart. The
final result is shown in Fig. 1.
Comparison of the new nomogram with the
standard calculation method
The project and consent process were approved by the
Research and Development Department, Alder Hey
Table 1 Measures of fit data for the median BMI
(BMI50) equations.
Emax SSE R2
BMI50 (boys) 0.05 0.14 1.000
BMI50 (girls) 0.006 0.04 1.000
Emax is the maximum absolute error, SSE is sum of squared
errors.
178
Children’s NHS Trust. The study was designed to sim-
ulate the clinical scenario for which the nomogram
was created, i.e. a peri-operative or urgent drug dose
calculation in an obese child. Obesity was defined
using the UK clinical cut-off, i.e. a BMI greater than
or equal to the 98th centile of a UK BMI chart.
The performance of the nomogram (referred to
from here on as ‘the nomographic method’) was com-
pared with the existing standard calculation method
i.e. Equations (1) and (2) above (referred to from here
on as ‘the equation method’). Volunteers used both
methods to perform IBW and LBM calculations using
obese children’s measurements and the results were
compared with a computerised spreadsheet of bench-
mark ‘correct’ calculations.
We assembled an anonymised database of age, sex,
height and weight from 108 obese children from inpa-
tient peri-operative records during local audits and ser-
vice evaluations between 2008 and 2012, and from a
local research database.
Volunteers were recruited from the staff of Alder
Hey Children’s Hospital. Informed consent was
obtained from all volunteers before any study involve-
ment. The 32 volunteers were 24 doctors and eight
paediatric pharmacists (Table 2) whose routine sphere
of practice included the calculation, prescribing or
checking of drug doses for obese children in the peri-
operative or acute care period. Each volunteer per-
formed IBW and LBM calculations for six patients,
using three patients’ data for each of the methods. The
order in which the two methods were used was ran-
domly assigned by the volunteer by selecting one of
two unlabelled packs. The six children’s measurements
were taken in a sequential manner from the anony-
mised database and when the final child was reached,
the list returned to the start and the data were used
again by different volunteers. The time for completion
of each method was recorded, and an average was
taken to reflect the time to calculate IBW and LBM
for one child. The nomogram was scaled to A4 paper
format. Volunteers were allowed to use the calculator
they would normally use for dose calculation and pre-
scribing, or a desktop calculator provided by the
researcher (Casio MS-80TV).
An Excel spreadsheet was used to calculate a
benchmark value for both IBW and LBM for each
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Callaghan and Walker | A nomogram to aid dosage calculations in obese children Anaesthesia 2015, 70, 176–182

Age (years)
Boys Girls Ideal body weight (kg) Lean body mass (kg)
5 5 85 115
6 6
Total body weight (kg)
185
7
7 110
8 180
80
175
9
8 Body mass nomogram 105
170
75 165
9 100
10 160
95 155
70
Height (metres) 150
11 10
90 145
2.00
1.95
65 140
85
135
12
11 1.90 130
60 80
1.85
125

1.80
75 120
13 1.75
55 115
12
70 110
1.70 105
1.65 50 65
100
14
13 1.60 60 95

1.55 45
90
55 85
1.50
15 1.45 40 80
14 50
75
Instructions: Draw a line from the age through 1.40 70
1.35 45
the height to meet the Ideal Body Weight scale. The 35
65
16 15
ideal weight is read at this point. A second line is
drawn from the ideal weight to the actual weight on 1.30 40 60
the Total Body Weight scale. The Lean Body Mass 1.25
30 55
is read from its scale where this line crosses it. 35
1.20 50
16 In the example shown, an eleven year old boy
17 1.15 45
who is 1.42 m tall and who weighs 71 kg has an ideal 25 30
weight of 34 kg and a lean body mass of 45 kg. 1.10 40
17 1.05 35
25
NB: Lean Body Mass calculations are only valid
18 20 30
for overweight patients, i.e. for those cases where ac- 1.00
18 0.95
tual weight is higher than ideal weight. 20 25

19 0.90 20
19 0.85 15
15
15
20
10
10
20 10

Figure 1 Nomogram for calculation of ideal body weight and lean body mass.

patient using Equations (1) and (2) and the median Table 2 Breakdown of the study volunteers by spe-
values provided by the Child Growth Foundation. The cialist branch of paediatric care and grade. Values are
database was used in such a way that each calculation number.
was done twice by different people using the same Registrar Fellow Consultant Total
method. This allowed us to look at repeatability, but it Anaesthesia 1 3 13 17
also gave an additional check on the benchmark val- Emergency 2 1 3
medicine
ues. Whenever a mistake was identified, the spread-
Paediatric ICU 1 1 2
sheet calculation was also rechecked. Surgery 2 2
Bias, limits of agreement and repeatability were Pharmacist 8
Totals 3 6 15 32
calculated for LBM, for both methods. Bias was com-
pared between both methods using an unpaired t-test.
A mistake was defined as an error > 5 kg (we an alpha-value for statistical significance at p = 0.005,
assumed that there was no relationship between the based on a Bonferroni correction.
magnitude of a mistake and the weight of the child, Although both IBW and LBM were both calcu-
because of the linear scales used in the construction of lated, we took LBM as the outcome measure on which
the nomogram). The incidence of mistakes was to base the statistical analyses. This reduced the num-
compared for the two methods using the chi-squared ber of statistical tests needed and in doing so reduced
test. the magnitude of the Bonferroni correction. However,
Speed of use of the nomographic method and the we did wish to ensure that the interim calculation was
equation method was compared using a paired t-test. correct; it is not impossible to derive a nomogram that
As multiple statistical analyses were performed, we set would give a correct final result but an incorrect

© 2014 The Association of Anaesthetists of Great Britain and Ireland 179

Anaesthesia 2015, 70, 176–182 Callaghan and Walker | A nomogram to aid dosage calculations in obese children
interim one. To this end we also calculated bias and
limits of agreement for the nomogram for IBW.
Results
A total of 32 volunteers took part in the study; each per-
formed three calculations using the nomographic
method and three with the equation method. This
resulted in a total of 96 calculations using each method.
Overall bias was similar for both techniques. Bias
and limits of agreement [19] for both methods are
shown in Table 3. The nomographic method had a
much lower variance in error (p = 0.002, Fig. 2), and
a better coefficient of repeatability (p < 0.003). The
nomographic method had significantly fewer mistakes
(three compared with 18 for the equation method,
p = 0.001).
The nomographic method was faster than the
equation method: mean (SD) time per calculation was
2.8 (1.0) min for the nomogram method and 3.3 (0.9)
min for the equation method (p = 0.003).
Analysis of the nomogram reading of IBW
revealed that in all cases where LBM was correct, IBW
was correct also. Of the three mistakes made in read-
ing LBM, two had been correct at the interim calcula-
tion. No other mistakes were found.
Discussion
The nomographic method was quicker to use, and less
likely to result in a mistake, than the equation method.
We were unable to demonstrate any significant differ-
ence in accuracy between the two methods. This is not
unexpected: the accuracy of the calculation method is
limited by the user’s ability to read an accurate ideal
BMI from the 50th centile of a BMI chart, the x-axis
of which only provides 6-month intervals for age.
The standard equation method for IBW and LBM
estimation is cumbersome, involving a series of steps,
and relies on having a calculator and an appropriate
age and sex-specific BMI chart available. Mistakes were
more likely compared with the nomographic method
(19% vs 3%).
The study has several limitations. First, a flaw in
our study design meant that, regrettably, we did not
measure speed of improvement during the three calcu-
lations using each method. Study volunteers com-
mented that they had fully grasped the use of the
180
Table 3 Comparison of Bland–Altman analysis of lean
body mass in kg, as calculated by the nomogram and
equation methods, both as compared with the bench-
mark value. Bias is the mean error, and limits of
agreement are bias 1.96 9 SD. Coefficient of repeat-
ability is 1.96 9 SD of the differences in subsequent
readings. A lower coefficient of repeatability indicates
better agreement between repeated measurements.
Values are number.
Nomogram Equation p value
Error
Bias; kg 0.12
0.41 0.4*
SD; kg 1.98 5.72 0.002†
Limits of
3.77 to 4.0
11.63 to 10.81
agreement; kg
Mistakes 3 18 0.001‡
Repeatability
Coefficient of 6.0 16.15 0.003†
repeatability; kg
*t-test.
†Levene’s test for unequal variances.
‡Chi-squared test.
nomogram by the third attempt; which seemed faster
than the first. This feature was not shared with the cal-
culation method, which was felt by subjects to be
equally complex during all three calculations. Although
there was no difference in accuracy between the first
and third calculations for either method, further study
may tease out an even greater speed benefit of the
nomogram than we have been able to demonstrate.
Second, the nomogram is appropriate for use in the
UK population only. The UK ideal BMI data and UK
obesity definitions are not identical to those of the
World Health Organization. A nomogram alteration
would be necessary for international use.
The nomogram has linear weight scales because of
the underlying mathematical equations and construc-
tion methods. This leads to the possibility of having a
greater relative error at lower weights (an error of 1 kg
is far more significant in a child of 5 kg than in a
60-kg teenager). To account for this, we have set the
nomogram to have a minimum age of 5 years and
have created a separate nomogram, using similar tech-
niques, for children under five. This has not been pre-
sented here as it remains to be validated.
Knowledge of the clinical implications of obesity
in children is limited, although obesity is associated
with an increased risk of serious airway and respira-
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Callaghan and Walker | A nomogram to aid dosage calculations in obese children
30.00
Difference of calculated and benchmark
20.00
10.00
LBM (kg)
0.00
20 30 40 50
–10.00
–20.00
–30.00
Mean of calculated and benchmark LBM (kg)
Figure 2 Combined Bland–Altman plot showing per-
formance of the nomographic method (○) and the
equation method (●) against the benchmark values.
The equation method errors are more widely spread
than the nomogram errors.
tory complications during and after anaesthesia [3].
Obese children are more likely to develop co-morbidi-
ties such as diabetes, fatty liver infiltration [20],
obstructive sleep apnoea, hypertension and asthma [3].
Weight-based drug dose calculation and pharmacoki-
netics are poorly understood. Serious complications and
death have been reported when drugs have been
inappropriately dosed according to an obese child’s
TBW [21]. Virtually all the available anaesthetic and
peri-operative obesity drug dosing information has been
extrapolated from studies in adults, and the behaviour
of each drug needs to be studied in obese children.
This process may take many years, so in the
interim we must attempt to use the methods available
as accurately as possible. To date, the majority of work
in this field has attempted to simplify weight-based
dosing by limiting it to three dosing scalars: IBW;
LBM; and TBW. The paucity of guidance and the
cumbersome nature of the existing methods to calcu-
late these values makes guesswork the easiest option
available to anaesthetists, pharmacists and paediatric
doctors. This cannot be encouraged, particularly where
potentially harmful drugs are in use.
We conclude that the nomographic method for
estimation of IBW and LBM in ambulatory obese chil-
dren aged 5 years and over is both accurate and rapid.
It can be accessed as a desktop PDF file or wall poster
and can be filed in the patient’s case notes, providing
a record of calculations made over time. Our ongoing
research includes validation of a nomogram for the
© 2014 The Association of Anaesthetists of Great Britain and Ireland
Anaesthesia 2015, 70, 176–182
under-five age group, and the development of a smart-
phone ‘app’.
Acknowledgements
We thank Dr Richard Sarginson, Consultant Paediatric
60 70 80 90
Anaesthetist at Alder Hey Children’s Hospital, for his
valuable contribution in reviewing this manuscript.
Leif Roschier, developer of Pynomo, provided assis-
tance with the alignment of the age scales on the
nomogram. Data access and use for the local research
database was granted by the Medicines for Children
Research Network Co-director for Merseyside, Dr Jo
Blair. BMI 50th centile data: ©copyright Child Growth
Foundation, used with permission.
Competing interests
No competing interests or funding declared.
References
1. Kendrick JG, Carr RR, Ensom MHH. Pharmacokinetics and drug
dosing in obese children. Journal of Pediatric Pharmacology
and Therapeutics 2010; 15: 94–109.
2. Ingrande J, Lemmens HJM. Dose adjustment of anaesthetics in
the morbidly obese. British Journal of Anaesthesia 2010; 105
(Suppl 1): i16–23.
3. Tait AR, Voepel-Lewis T, Burke C, Kostrzewa A, Lewis I. Inci-
dence and risk factors for perioperative adverse respiratory
events in children who are obese. Anesthesiology 2008; 108:
375–80.
4. Mortensen A, Lenz K, Abildstrøm H, Lauritsen TLB. Anesthetiz-
ing the obese child. Pediatric Anesthesia 2011; 21: 623–9.
5. Veyckemans F. Child obesity and anaesthetic morbidity.
Current Opinion in Anesthesiology 2008; 21: 308–12.
6. Cole TJ, Freeman JV, Preece MA. Body mass index reference
curves for the UK, 1990. Archives of Disease in Childhood
1995; 73: 25–9.
7. Smith HL, Meldrum DJ, Brennan LJ. Childhood obesity: a chal-
lenge for the anaesthetist? Pediatric Anesthesia 2002; 12:
750–61.
8. Forbes GB, Welle SL. Lean body mass in obesity. International
Journal of Obesity 1983; 7: 99–107.
9. Phillips S, Edlbeck A, Kirby M, Goday P. Ideal body weight in
children. Nutrition in Clinical Practice 2007; 22: 240–5.
10. Kleber M, Schwarz A, Reinehr T. Obesity in children and
adolescents: relationship to growth, pubarche, menarche, and
voice break. Journal of Pediatric Endocrinology and Metabo-
lism 2011; 24: 125–30.
11. Campanozzi A, Dabbas M, Ruiz JC, Ricour C, Goulet O. Evalua-
tion of lean body mass in obese children. European Journal of
Pediatrics 2008; 167: 533–40.
12. Thimbleby H, Oladimeji P, Cox A. Number entry interfaces and
their effects on errors and number perception. Proceedings
IFIP Conference on Human-Computer Interaction - INTERACT
2011, Vol. 4. Berlin: Springer Verlag, 2011: 178–85.
13. Evesham H. The History and Development of Nomography.
Boston: Docent Press, 2011.
181
Anaesthesia 2015, 70, 176–182 Callaghan and Walker | A nomogram to aid dosage calculations in obese children
14. Williams DJ, Walker JD. A nomogram to calculate the Physio-
logical and Operative Severity Score for the enUmeration of
Mortality and morbidity (POSSUM). British Journal of Surgery
2014; 101: 239–45.
15. Healy MJR. Statistical analysis of radioimmunoassay data. Bio-
chemical Journal 1972; 130: 207.
16. Blattner P. Using Microsoft Excel 2000. Indianapolis, IN: Que,
1999.
17. Allcock HJ, Jones JR. The Nomogram, 4th edn. London: Pitman
& Sons, 1950.
182
18. Levens AS. Nomography, 2nd edn. New York: Wiley & Sons,
1959.
19. Bland JM, Altman DG. Statistical methods for assessing agree-
ment between two methods of clinical measurement. Lancet
1986; 327: 307–10.
20. Roberts EA. Nonalcoholic steatohepatitis in children. Current
Gastroenterology Reports 2003; 5: 253–9.
21. Newgreen D. Review of Non-prescription Analgesics. 1998: 15–6.
Therapeutic Goods Administration. http://www.tga.gov.au/pdf/
archive/review-analgesics-9802.pdf (accessed 30/04/2014).
© 2014 The Association of Anaesthetists of Great Britain and Ireland