Background In paediatric resuscitation, for a rapid and accurate estimate of children's weight, the Broselow tape can be used in children who are 46–144 cm tall. The Broselow tape has previously been found to provide the most accurate estimate of children's weight internationally, but it is not known how many fall outside the range of the tape, or whether such children can be assumed to be of adult weight, or how otherwise to estimate the weight of these children.
Objectives To determine what proportion of children in different age groups falls outside the limits of the Broselow tape, how their weight compares with that of the adults and what correlates most strongly with weight in these children.
Methods This was a population-based prospective observational study of Chinese children up to 12 years old, from schools in Hong Kong. Weight was measured to the nearest 0.2 kg, and the height, foot-length and mid-arm circumference (MAC) were measured to the nearest 0.1 cm.
Results 40% of 10-year olds and 70% of 11-year olds were too tall for the tape. Their median weight was 41.9 kg. This was significantly less than the median weight of 18-year olds (55 kg, p<0.0001) in Hong Kong. The strongest correlate with weight in these children was MAC.
Conclusions The Broselow tape is inappropriate for use in most children over 10 years old. Children too tall for the tape cannot be assumed to be of adult weight; to do so would imply an average overestimate of 30%. Weight estimates in older children could be based on MAC.
- paediatric resuscitation
- paediatric emergency med
Statistics from Altmetric.com
In paediatric resuscitation, rapid and accurate estimation of the children's weight is necessary in order to provide appropriate drug and fluid doses, equipment selection and ventilator settings. A commonly used method of weight estimation is the Broselow tape (BT),1 which is laid alongside the child from the heel to the top of the head to measure the height. The tape is marked in kilograms according to the estimated weight, for a child of that height, and divided into colour-coded zones, which correspond to the lists of drug doses and equipment sizes appropriate for children of that weight.
The accuracy of the BT was originally described in 1988.1 The tape was constructed from the median values of the weight-for-height centile curves, using data collected by the US National Center for Health Statistics between 1963 and 1975. Subsequent editions of the tape were based on the data from successive National Health and Nutrition Examination Survey (NHANES) and have therefore taken account of the changing weight of American children.
Although many studies (including the original study)1–7 have considered an acceptable estimate to be within 10% of the actual weight, this is an arbitrary figure. Drugs with a narrower therapeutic dose range will require more precise weight estimation than those with a wider range. It has been argued that in adult resuscitation, there is far more variability in drug dosing, as exemplified by giving the same dose to persons weighing 50 or 100 kg.8 However, children are not little adults, and there is no evidence that it is safe to treat older children with doses significantly different from their ‘correct’ weight-dependent dose.
The BT is used internationally, and is recommended by paediatric resuscitation courses from the USA9 and the UK.10 It has been described as the ‘gold standard’ of weight estimation methods.11 Many studies have assessed its accuracy in local populations in comparison with other methods. The most commonly used alternative methods of weight estimation use age-based formulae, as in the UK Advanced Paediatric Life Support (APLS) course.10 In Western populations, the BT has consistently been found to be more accurate than age-based formulae. However, there is a tendency to underestimate weight.1–4 11–14 This is also true in Korean,5 Hong Kong,6 Pacific Island and Maori children.15 In India, it appears to overestimate.7 ,16 In all these studies, the BT was more accurate in smaller and lighter children, especially those who are under 25 kg. New length-based weight estimation tools, which incorporate a body habitus element have been shown to outperform the BT,17 ,18 but in all length-based methods, the size and the range of the estimation error increase as the age and the size of the child increase.
Other suggested methods include staff or parental estimation2 or tools based on mid-arm circumference (MAC) or foot length.19 Different MAC-based tools have recently been derived and validated.20 ,21 Our own MAC formula, weight (kg)=(MAC (cm)−10)×3, was at least as accurate and precise as the BT in school-age children, but it was inadequate in preschool children.20
Not only does the BT become less accurate the taller the child, but it is also unclear what should be done with the children who are too tall for the tape. The BT suggests that a child taller than the limit of the tape is considered a ‘medium adult’. In our previous paper,20 comparing our proposed MAC-based weight estimation rule with the BT, 12.5% of our total sample of children aged 1–11 years on their last birthday were too tall to be measured by the BT. Other papers have found similar results.3 ,4 ,17 ,21 However, no study specified the proportions according to age group or described the weights of those children whose height exceeded the length of the tape. It is therefore not clear as to how many children of a given age are likely to be too tall for the tape, or whether there is any justification for assuming that such children can be considered of ‘medium adult’ weight, or whether there is a more accurate method of weight estimation in this particular group of children.
The aim of this study therefore was to determine, first, what proportion of children in different age groups fall outside the limits of the BT; second, whether the weight of these children is significantly different from that of the adults; and third, which of these—age, height, foot length or MAC—correlates most strongly with weight in these children.
This was a subgroup analysis of our previously published population-based observational study of 1370 healthy Chinese children aged 0–12 years, conducted in primary schools and kindergartens in Hong Kong.20
Height was measured to the nearest 0.1 cm with a stadiometer (Harpenden portable stadiometer, Holtain, UK). Weight was measured to the nearest 0.2 kg using electronic scales (Compact precision scale C200H, Conair Far East Ltd, Hong Kong). MAC was measured with the child's right arm relaxed in 90° of flexion at the elbow. The midpoint between olecranon and acromion was identified, and an inelastic tape measure was used to determine MAC to the nearest 0.1 cm. Foot length was measured from the right heel to the tip of the hallux, with the child's supine and the ankle at 90° with the foot flat against the base of a rigid box, the heel resting on the inside wall of the box.
Children were measured without shoes, wearing the lightest school uniform including socks. School uniforms were weighed separately and subsequent adjustment was made to the measured weight of the child.
Adult weight was determined from historical data from a recent Hong Kong study of 6–18-year olds.22
Proportions of children outside the height range of the BT were grouped by age. Weight was not normally distributed but was positively skewed, so the median values are presented. Mann–Whitney U test was used for comparison with adult weight; p value less than 0.05 was considered statistically significant. Pearson correlation coefficients (r), with 95% CIs, were calculated to determine the relationship with weight for each of the other parameters. The MAC formula was assessed using Bland–Altman analysis: the bias indicates the mean percentage difference between estimated and actual weight, and the limits of agreement indicate in what range 95% of the differences between estimated and actual weights will fall. We also determined the proportion of estimates that were within 10, 20, and 30% of the actual weights.
MedCalc V. 10.4.0.0 (Frank Schoonjans) and Statview V. 5.0 for Windows (SAS Institute Inc, Cary, North Carolina, USA) were used for the analysis.
Ethical approval was obtained from the Clinical Research Ethics Committee of the Chinese University of Hong Kong. Written parental consent was obtained for all subjects more than a week prior to the school visit. Children who were unwilling to participate on the day of the study were excluded.
No children under 7 years old were too tall for the BT. Data from 782 children aged 7–11 years, on their last birthday (55% boys), were analysed for this study.
A large proportion of children aged over 10 years were too tall for the BT (table 1).
The median weight of the children too tall for the tape was significantly less than the adult weight in the comparison group: 41.9 kg (IQR 36.7–47.3) versus 55 kg (IQR 53.9–55.7, n=756), difference 13.1 kg (31%). This difference was highly statistically significant (p<0.0001), and also likely to be clinically important (figure 1).
In subjects too tall for the tape, the strongest correlate with weight was MAC. Height and foot length were poor correlates and age did not correlate with the weight at all (table 2).
When the MAC formula was applied to the subset of children too tall for the BT, Bland–Altman bias was −12.8% (on average, the formula underestimated weight), with limits of agreement −35.0 to 9.3%. About 44% of the estimates were within 10% of the actual weight; 79% were within 20%; and 97% were within 30%.
This study has demonstrated that a large proportion of 10-year and 11-year olds (approximately 40 and 70%, respectively), and a smaller but significant number of 9-year olds (10%) are too tall to allow the use of BT.
Although no other studies have specifically described the proportions of children who exceed the length of the BT according to age, other studies have similarly found that large proportions of children are not suitable for the BT. Lubitz et al's1 original study intended to recruit children from 1 week to 12 years old, but the oldest child included was only 10 years 11 months. Comment was made that ‘a number of children’ exceeded the range of the tape, and data for those greater than 40 kg were discarded. The actual numbers or proportions were not provided. In Hofer's study,3 24.6% of the children exceeded the range of the tape and were excluded from the statistical analysis. The oldest child included was 11 years 10 months. In Nieman’s study,4 of the 8265 children under the age of 12 years, 546 (6.6%) were excluded from the analysis because they were too tall for the tape. About 17.5% of the Yamamoto's sample were too tall, but it is unclear what the upper age limit was.17 Abdel-Rahman and Ridge21 found that only 62.6% of the children aged 2 months to 16 years could be measured by the BT, and Knight et al23 had to exclude 53% of their sample of paediatric trauma patients aged up to 16 years.23 In the Hong Kong study,6 21 of 930 children aged 7 days to 11 years 11 months were excluded for not fitting the tape, but it was not specified how many were too tall and how many too short. Only 15% of the subjects included in that study were over 8 years old. Theron et al15 commented that in Pacific Island and Maori children, the tape was too short for those over 10 years old, and these children were therefore not recruited. The Indian, Korean and PAWPER studies5 ,7 ,16 ,18 also excluded all children who exceeded the length of the tape.
This study did not attempt to determine the accuracy of the BT; that has already been demonstrated in numerous other studies internationally. However, it is clear that the accuracy of the BT is only applicable to those whom it fits. When comparing its accuracy with other methods, all the above studies have effectively excluded from comparison what we have now shown to be a large proportion of the population whose weight we are trying to estimate. This would be acceptable only if those who are too tall for the tape can safely be considered to be of adult size.
This study is also the first to describe how much these children weigh and it is clear that it is not justifiable to assume they are the same as adults. With a median weight of less than 42 kg, they are significantly lighter than adults from the same population (55 kg). This finding implies that to treat these children as adults would result in an overestimate of ≥30%. To administer drugs with this degree of overdose would be clinically unacceptable.
This leaves us with a dilemma. The BT has been shown to be the most accurate method of weight estimation in those children within its height limits, but we have shown that many children do not fit the tape and cannot be assumed to be adults. Therefore, how should we estimate the weight of the children under 12 s who do not fit the tape?
Our third finding is that the strongest correlate with weight in this group of children is MAC, far stronger than age, height or foot length. Application of our own MAC formula in this subgroup produced estimates of weight more precise than any age-based methods, even in younger children,24 but less precise than the BT is in those it can measure.20 A large study of the NHANES database has externally validated the MAC formula in 6–11-year olds, with better bias and precision than 12 other previously published weight estimation methods, including the BT and the latest APLS age-based formula.21 The only method that outperformed the MAC formula was the Mercy Tape, which was derived in that study. The Mercy Tape is also based on MAC, together with a measure of humeral length. It does seem that the most promising options for weight estimation development will include MAC, and further work needs to be done to establish the best method, especially in older children and adolescents.
This is a subgroup analysis of our previous study, and so makes no claim to validate the formula we derived in that study.
We only studied children under 12 years old. This was consistent with the original study by Lubitz et al1 and most of the other BT studies mentioned earlier, and with the definition of ‘schoolchildren’ used in paediatric life-support courses.10 From the historical data set that we used for the weight of 18-year olds,22 we found that 90.4% of 12-year olds and 98.5% of 13-year olds were outside the limits of the tape.
We studied only Chinese children in Hong Kong. However, comparison between published data for 2006 shows that mean height in boys and girls is approximately 2%–3% greater in England25 than in Hong Kong22 for each year of age between 7 and 11 years. This suggests that the results of this study would be even more marked in England, with an even greater proportion of children too tall for the tape.
A large proportion of 10-year and 11–year olds are too tall for the BT to be used, but they weigh significantly less than adults and to treat them as may overestimate weight by an average of 30%. A different method of weight estimation is therefore needed for these children, and MAC correlates very strongly with weight in this group. We suggest that to avoid systematically overestimating the weight of older schoolchildren, weight estimation methods based on the MAC should be further developed and validated.
Contributors All authors were involved in the planning of the study; ML and GC conducted the field work; all authors were involved in the analysis of the data. GC wrote the manuscript with contributions from the other three authors.
Funding We received an internal grant of HK$72 000 (approximately US$10 000) from the Chinese University of Hong Kong to conduct the ‘Healthy Children's Vital Signs and USCOM Study’.
Competing interests None.
Ethics approval Clinical research ethics committee of the Chinese University of Hong Kong. .
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement Other aspects of the study have been published elsewhere, as part of the ‘Healthy Children's Vital Signs and USCOM study’.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.