Article Text

PDF

Leading article
Estimation of body weight in children in the absence of scales: a necessary measurement to insure accurate drug dosing
  1. Susan M Abdel-Rahman1,2,
  2. Anna Ridge3,
  3. Gregory L Kearns1,2
  1. 1Department of Pediatrics, The University of Missouri—Kansas City School of Medicine, Kansas City, Missouri, USA
  2. 2The Division of Pediatric Pharmacology and Therapeutic Innovation, The Children's Mercy Hospital, Kansas City, Missouri, USA
  3. 3Raigmore Hospital, Inverness-shire, UK
  1. Correspondence to Dr Gregory L Kearns, The Children's Mercy Hospital, 2420 Pershing, Third Floor, Kansas City, MI 64108, USA; gkearns{at}cmh.edu

Statistics from Altmetric.com

Body weight (BW) is one of most important measurements in paediatric medicine. In addition to being one of the first parameters announced at the birth of a child in the developed world, BW is used to determine intravenous fluid requirements, shock voltage administered during cardio-respiratory arrest, endotracheal tube size and to assess nutritional status. In prepubertal children, height, age and BW represent co-migrating anthropometric surrogates which are predictive of organ function including the liver and kidney1 (figure 1) and by inference, the function of physiologic processes which collectively determine drug disposition.2 It has been previously stated3 that the BW-based dosing approach has its origins in the Kleiber principle4 which makes the following basic assumptions: (1) that total BW correlates with organ size and hence function and (2) that basal metabolism is proportional to the BW raised to the 0.75 power (BW0.75). Old, previously applied standard approaches for paediatric drug dosing such as Clark's rule (eg, Infant dose=(BWinfant /BWadult) Adult dose) have been largely abandoned in that a simple proportionality of BW between a child and an adult does not accurately reflect the nonlinearity in the relationship between BW and the age-dependent changes in drug disposition.2

Figure 1

Relationship between renal total cortical volume and weight (A) and height (B) in children and adolescents. Adapted from Tan et al1 as deposited in the NIH Public Access database (http://www.nih.gov; doi: 10.1097/TP.0b013e318237053ef).

The safe administration of medicines to children relies on an ability to correctly calculate the drug dose5 and accurately measure and administer a given drug formulation. Given that the majority of paediatric drug doses are calculated on a milligram or microgram per kilogram bodyweight basis, it is essential that the prescriber and/or healthcare provider have an accurate determination of the child's weight. This poses a very real dilemma in resource constrained health care settings where parents or caregivers may have no idea what a given child's BW is6 and medical scales are often unavailable, inaccurate or broken.7 In these particular settings, the use of inaccurate estimates of BW can lead to therapeutic misadventures such as medicine overdosing or underdosing.8

The inaccessibility of sufficiently sensitive, functioning, calibrated scales in remote and resource-constrained regions has spawned the development of numerous strategies that permit healthcare providers to estimate weight. Each of these methods is based on one or more demographic and/or anthropometric variables and all are accompanied by unique advantages and limitations as discussed below.

Age-based methods

The majority of weight estimation methods published to date, rely upon age as the primary determinant of the child's weight (table 1).9–17 These methods encompass approaches ranging from a singular mathematical formula applied across a defined age range, methods that rely on multiple distinct formulae applied to discrete age groups, and the use of nomograms. To maximise their practicality, the majority of age-based methods employ ‘relatively’ simple mathematical operations (addition, multiplication, division) with the exception of one approach that requires the user to solve an exponential equation. The perceived simplicity of these calculations depends upon the education level of the person applying the method and the speed with which the user needs to arrive at the answer.

Table 1

Published age-based weight estimation methods

A detailed review of the studies that have evaluated these methods is beyond the scope of this commentary; however, they are referenced in table 1,18–25 and some general observations are provided as follows. The relative simplicity of the equations that are integrated into the age-based methods ignores the nonlinearity observed in weight-for-age that occurs early in the course of development. Further, these single-variable approaches, by default, necessitate a single ‘reference’ weight for each age. As a result, gross inaccuracies in prediction can occur in children who reside at the extremes of weight-for-age (figure 2). Finally, the median weight selected for the vast majority of these methods reflects those of a unique, often regional, population. Consequently, many of these approaches fail to effectively estimate weight in children that derive from different ethnic or racial backgrounds, or who may have nutritionally based alterations in body composition/habitus. Despite their limitations, age-based methods can be integrated with relative ease into clinical practice. However, these methods are rendered ineffectual in settings where the age of the child is unknown or undocumented.

Figure 2

Performance of representative weight estimation methods (age-based: ARC, length-based (device): Broselow, length-based (equation): Traub-Kichen, habitus-based: Cattermole, length-based and habitus-based: Mercy) in a population of children randomly sampled from the NHANES databases (1999–2007). Values that fall on the x-axis represent individuals for whom the method failed to generate a weight because the children fell outside the defined bounds of the estimation method. Performance of additional methods can be viewed in the article from which this figure is adapted (with permission).34

Length-based methods

The second most commonly used patient-derived variable in paediatric weight estimation methods is length (table 2).26–31 Original investigations describing experience with these published length-based methods are also denoted and referenced in table 2.11 ,17–21 25–28 32–39 Collectively, these methods are evenly split among those (1) that integrate length into a mathematical equation that the user must solve (both involve exponentials), (2) that use a reference table to provide the weight after length is obtained by traditional measures and (3) for which a measuring device has been adapted to incorporate the weight estimate. The practical limitation of these methods is that length can be difficult to obtain in an uncooperative or combative child. The scientific limitation of these methods is similar to that described for the age-based methods; namely that reliance on a single-variable necessitates a single ‘reference’ weight. For example, the most studied length-based weight estimation method, the Broselow tape, uses as its reference value, the ideal BW of US children. Accordingly, the method performs less than adequately in non-US settings and, within the US, performs poorly in children that are overweight or obese (figure 2). The Malawi tape28 represents a modification of the Broselow tape designed to accommodate the regional differences in height-for-weight observed in Malawi children. The result is a device with improved performance metrics for children in this country. However, an ideal weight estimation strategy should be universally applicable without the need of modification for each population to which it will be applied.

Table 2

Published length-based weight estimation methods

To alleviate the problems that arise with the use of a single reference value, the innovators of the devised weight estimation method (DWEM) included a subjective assessment of body habitus in their tool. In addition to length, the user evaluates the child's build as ‘slim,’ ‘average’ or ‘heavy’ and reads the weight that corresponds with length and build from a reference table. The addition of this second subjective parameter reduced bias across a broader spectrum of weights but was still associated with an increase in bias at the extremes of weight. Part of this persistent bias may be associated with variability in the user's ability to accurately assess body habitus. In one of our recent studies (SM Abdel-Rahman, unpublished, 2014), we observed that healthcare providers were able to accurately categorise children that were underweight (body mass index, BMI<5th centile) as ‘slim,’ but failed to categorise the majority (>75%) of overweight and obese children as ‘heavy.’ Notably, the most recent iteration of the Broselow tape now incorporates a similar subjective assessment of habitus. However, the application of this subjective assessment is unlikely to substantially alter the utility of this tool as >30% of the paediatric population in much of the developed world exceeds the bounds of the Broselow tape.

Habitus-based methods

A few methods40–42 estimate paediatric weight-based circumferential measures (table 3), predominantly mid-upper arm circumference (MUAC) which represents an anatomic site where the bone:muscle:fat ratio approximates that of the whole body. These methods serve as an extension of the work in the field of nutrition where the relationship between MUAC and weight has been used to assess nutritional status in critically ill children and also, for mass screening programmes of children in developing countries.43 These methods are typically developed using normative data from single site studies and perform best in a relatively narrow range of ages (figure 2). They require obtaining the circumference of the midupper arm or other anatomic site (using a tape measure or an insertion tape) followed by a combination of relatively simple mathematical operations (eg, addition, subtraction, multiplication). Apart from the limitations associated with relying on a single anthropometric variable, these methods are unlikely to experience broad uptake for use as weight estimation tools without expansion beyond the narrow age ranges currently recommended for their application and validation in children of different racial/ethnic backgrounds.

Table 3

Published Habitus-based weight estimation methods

Dual length-based and habitus-based method

In an attempt to address the limitations nested in the single-variable weight estimation methods, a dual variable method based on surrogates of total body length (humeral length) and body habitus (MUAC) was developed (table 4).44 The two measurements are made in the same anatomic location to facilitate ease of performance and the reliance on a single limb enhances measurement accuracy in a child who may not otherwise cooperate with a total body length measure. Both anthropometric measures are integrated into this method as continuous variables thereby markedly improving the predictive performance of the method irrespective of where the child falls on the weight-for-age spectrum (figure 2). As demonstrated by the investigations performed to validate this method,45 ,46 there are no height restrictions for its use and it is applicable over a very broad range of ages.

Table 4

Published Length-based plus Habitus-based weight estimation method

A paper-based device (the Mercy TAPE) was also developed and subsequently validated to facilitate application of the method.46 Irrespective of whether the method is applied using a standard tape measure or the Mercy TAPE, measurements are made to obtain the humeral length and MUAC. Each value is assigned a fractional weight which can be obtained from a reference table (when using a standard tape measure) or read directly from the device. To arrive at the estimated weight of the child, the two numbers are simply added together. Comparable performance in populations that differ from the population in which the method was developed (SM Abdel-Rahman, unpublished) confirms that there is far less variation in predicted weight among children of varying race and ethnicity when the combination of length and girth are considered together than for either variable alone.

Concluding remarks

In many settings, paediatric weight estimation methods have been eschewed because of problems with bias, precision and restrictions in the population to whom they can be applied. This has led to a general assertion that weight estimation strategies should be avoided, even in settings where no consistently reliable alternatives exist for determining weight. Such an opinion has the real potential to disadvantage and even harm children. As weight estimation strategies evolve to integrate several anthropometric variables (eg, Mercy and Mercy TAPE), their predictive performance improves dramatically. The result is the availability of low-tech, inexpensive, portable and highly accurate tools that can be easily integrated into virtually any clinical setting. Serious consideration should be given to the use of such methods in settings where time and/or resources limit access to calibrated scales.

References

View Abstract

Footnotes

  • Contributors SMA-R performed all analyses required in comparison with different methods for prediction of body weight and contributed these sections of the manuscript. She also contributed sections of the manuscript, created the illustrations and performed final copy editing of the manuscript prior to submission. AR contributed to the literature review and the creation of the summaries concerning the various weight estimation methods reviewed in the manuscript. She participated in the drafting of the manuscript and reviewed (and approved) the final draft prior to submission. GLK conducted the review of the literature regarding the use of weight as a physiological surrogate. He contributed to the drafting of the background and conclusions sections of the paper and took responsibility for preparing the final draft of the manuscript for submission.

  • Funding Supported in part by resources received from the WHO (contracts 200326848, 200376118, and 200474476), the Eunice Kennedy Shriver National Institute of Child Health and Human Development (Pediatric Trial Network contracts: HHSN2752010000031 (HHSN27500008) and HHSN275200900012C), the USA Food and Drug Administration (grant 1 U01 FD004249-01) and the Children's Mercy Hospital. The Mercy Tape represents intellectual property solely owned by the Children's Mercy Hospital (US patent number 8 590 168).

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.

Request permissions

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.