- Email: [email protected]
Validation of the mid-arm-based weight estimation formula (the Cattermole formula) for Korean children
Resuscitation 132 (2018) 13–16
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Simulation and education
Validation of the mid-arm-based weight estimation formula (the Cattermole formula) for Korean children☆
T
⁎
Jea Yeon Choia,1, Dongbum Suhb,1, Do Kyun Kimc, , Young Ho Kwakc, Jae Yun Jungc, Jin Hee Leeb, Jin Hee Jeongd, Hyuksool Kwonb, So Hyun Paeke a
Department of Emergency Medicine, Gachon University Gil hospital, Incheon, Republic of Korea Department of Emergency Medicine, Seoul National University Bundang Hospital, Gyeonggido, Republic of Korea c Department of Emergency Medicine, Seoul National University Hospital, Seoul, Republic of Korea d Department of Emergency Medicine, Seoul National University Boramae Medical Center, Seoul, Republic of Korea e Department of Emergency Medicine, CHA Bundang Medical Center, CHA University, Gyeonggido, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Body weight Pediatrics Resuscitation Weights and measures
Background: A mid-arm circumference-based weight estimation formula has recently been proposed. The Cattermole formula, which is suggested for children aged 1–11 years, is calculated as (mid-arm circumference in cm − 10) × 3 kg. The objective of this study was to externally validate the Cattermole formula using a Korean national survey database. Methods: The data collected from a national survey conducted in 2005 by the Korean Pediatric Society included body weight, height, age and mid-arm circumference. To determine whether the Cattermole formula accurately predicted the actual weights of Korean children, we compared the performance of the Cattermole formula with those of other age-based and height-based weight estimation methods. Results: A total of 91,147 children aged 1–11 years were assessed for eligibility. The mean age was 5.3 years, and 52.6% of the children were boys. The mean measured weight was 22.4 kg, and the mean mid-arm circumference was 17.9 cm. The Cattermole formula resulted in a mean difference of 1.35 kg and a mean percentage error of 11.5% with 95% limits of agreement of −8.1 to +10.8. A total of 37.9% of the estimates were within 10% of the measured weights. Conclusion: Across all ages, the Cattermole formula was not more accurate than the other methods. However, in children older than 6 years, the Cattermole formula results were more accurate than those of the age-based formulae with the exception of the Broselow tape. Clinical trial registration: None.
Introduction Accurately measuring the weight of a critically ill child is essential during pediatric resuscitation. The child’s weight is used to calculate the defibrillation energy and most drug doses and to determine the sizes of equipment used. However, during resuscitation, there is insufficient time to weigh the child, and rapid and reliable methods of estimation must be applied. Several methods that aim to accurately estimate a child’s weight have been devised. Commonly, weight can be estimated from age-based formulae, such as the Advanced Pediatric Life Support (APLS) formula [1], or from length-based methods using the Broselow tape [2]. Length-
based methods, most prevalently the Broselow tape, have long been used and are generally accepted as more accurate than age-based formulae [3,4]. However, the Broselow tape is significantly limited by its length, so a recent review presented that the newer length- and habitusbased systems are superior to the Broselow tape [5]. In response to increasing concerns about wrong measurement of actual weights with length-based methods due to child obesity, other anthropometric measurements have been proposed [6]. Some of these measures are based on other physical features such as foot length [7]. A simple method that has been studied is weight estimation by the parents or a clinician [8]. The mid-arm circumference (MAC) has been used for many years in
☆
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.08.026 Corresponding author at: Department of Emergency Medicine, Seoul National University Hospital, 101 Daehang-no, Jongno-gu, Seoul 03080, Republic of Korea. E-mail address: [email protected] (D.K. Kim). 1 These authors contributed equally to this work. ⁎
https://doi.org/10.1016/j.resuscitation.2018.08.026 Received 8 June 2018; Received in revised form 31 July 2018; Accepted 24 August 2018 0300-9572/ © 2018 Published by Elsevier B.V.
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
the assessment of malnutrition in the developing world [9]. However, several studies have suggested the use of weight estimation methods that employ the MAC [10–12]. A formula that includes the MAC and knee height has been validated in geriatric patients in the emergency department [13]. In 2010, Cattermole et al. suggested a new formula that is calculated as (MAC in cm − 10) × 3 for children aged 1–11 years. They reported that this formula performed better than the Broselow tape and the APLS aged-based formula in older children [14]. However, this method has not been validated in other populations. The objectives of this study were to externally validate the Cattermole formula using a Korean national survey database and to compare the utilities of other existing weight estimation methods.
Table 1 Methods of child weight estimation.
Methods
MAC: Mid-arm circumference, APLS: Advanced pediatric life support.
Study setting and population
(LOA), which measure the spread around the MPE, and the percentage of patients whose weight estimation was within 10% of the measured weight are also reported. The statistical analyses were performed with SPSS software (version 18.0; SPSS Inc., Chicago, IL, USA), and the level for statistical significance was set a p < 0.05.
Name
Applicability
Method of calculation
Cattermole Traditional APLS Updated APLS
1–11 years 1–10 years < 1 year 1–5 years 6–12 years
Park
< 12 months 1–4 years 5–14 years Height 46–143.4 cm
Weight (kg) = (MAC − 10) ×3 Weight (kg) = (age in years + 4) × 2 Weight (kg) = (0.5 × age in months) +4 Weight (kg) = (2 × age in years) + 8 Weight (kg) = (3 × age in years) + 7 Weight (kg) = (age in months + 9)/2 Weight (kg) = 2 × (age in years) + 9 Weight (kg) = 4 × (age in years) − 1 Read weight from measuring tape
Broselow
We collected data that included body weight, height, age and MAC from a national survey that was conducted by the Korean Pediatric Society (KPS) under the supervision of the Ministry of Health and Welfare in 2005 for the development of the new 2007 Korean national growth charts. In 2005, a KPS project measured 150,000 children nationwide. The measurements were performed by 6 trained measurement teams and 28 general hospital teams between April and July of 2005 and August and September of 2005. The study population was composed of children between 0 and 20 years of age. The data were collected from daycare centers, childcare counseling rooms, kindergartens, elementary, middle and high schools and colleges from 16 assigned areas. To avoid the result is distorted by extreme values, we standardized collected data by eliminating outliers. We define outliers as the weight for height is more than 3 × (high standard deviation) or less than 3 × (low standard deviation) and eliminated from analysis. Consequently, the physical measurements of approximately 142,000 children were included. The original data were provided by the Center for Disease Control and Prevention, which is an affiliate of the Ministry of Health and Welfare. Based on the data included above, only data from children 1–11 years of age were ultimately used. The Cattermole formula developed with 1–11 years old age. Thus, we selected the same range of age to facilitate comparison with the results.
Ethics This study complied with the Declaration of Helsinki and was approved by the Seoul National University Hospital Institutional Review Board with a waiver of informed consent. Results Out of 142,945 children aged 0–20 years from the KPS project, 6029 were excluded due to the loss of data about age, heights, weights or MAC. Among 136,916 children, a total of 91,147 subjects aged 1–11 years were included in this study. The number of male subjects was 47,977 (52.6%), and the mean age of all subjects was 5.3 (SD 2.8) years. The mean weight was 22.4 (SD 9.9) kg, and the MAC was 17.9 (SD 2.8) cm. The participants included 55,730 (61.1%) 1–5-year-old, 19,561 (21.5%) 6–8-year-old and 15,856 (17.4%) 9–11-year-old children. The mean MACs of the 1–5, 6–8, and 9–11-year-old groups were 16.6 (SD 1.7) cm, 18.6 (SD2.4) cm and 21.6 (SD 3.1) cm, respectively. Using the Cattermole formula, we found that only 37.9% of all patients had estimated weights that were within 10% of their actual weights. Overall, the Cattermole formula had an MD of 1.35 kg and an MPE of 11.48% (SD 26.28, 95% CI 11.31–11.65). The COV was 20, and the 95% LOA were −8.08 to 10.77 kg. The MPE of the older APLS formula showed the highest value, and the MPE of the Cattermole formula was the second highest relative to the MPEs that were calculated for all other methods. The COV of the traditional APLS formula showed the poorest result. In terms of the proportions of subjects with estimated weights that were within 10% of the actual weights, the Broselow tape method exhibited the best performance (Table 2). In the 1–5-year-old group, the MD of the Cattermole formula was 3.17 kg, and the MPE was 21.3% (SD 27.3, 95% CI = 21.08–21.53); the latter value was the highest among the estimation methods. The Broselow tape method and Park et al.’s formula produced lower MPE values than the Cattermole formula. In this age group, 30.3% of the estimates based on the Cattermole formula were within 10% of the measured weight, which was the poorest performance among the three age groups (Table 2). In the 6–8-year-old group, the MPE and MD of the Cattermole formula were 0.33% (SD 14.91, 95% CI 0.12–0.54) and 0.18 kg, respectively, and these values indicated the best accuracy among all age groups. The proportion of measures within 10% of the actual weight was 53.5%, which also represented the best performance of the Cattermole formula among all age groups. However, among all estimation methods, the Broselow tape still performed best in terms of the accuracy according to the proportion of estimates that were within 10%
Study protocol and data analysis The subjects were classified into the following 3 groups that were based on the groups used in the Cattermole et al. study [14], which in turn were based on the Advanced Trauma Life Support course: 1–5, 6–8, and 9–11-year-olds [15]. This classification was performed to facilitate comparison of the results. The primary outcome of interest was the agreement between the weights estimated by the Cattermole formula and the measured weights of Korean children. The secondary outcomes were the comparisons of the estimation performance of the Cattermole formula with those of four other methods, i.e., the traditional APLS formula, the updated APLS formula, the Broselow tape and Park et al.’s formula (Table 1) [16]. The Broselow tape itself was not used. The children's weights were estimated by applying the measured height to the corresponding interval of the 2011 version of the Broselow tape. If the age of a subject deviated from the limit required by each formula, or if the height of a subject was less than 46 cm or more than 143.4 cm, a missing value was entered. The Cattermole formula and the measured weight were assessed via calculations of the mean difference (MD: estimated weight − measured weight) and the mean percentage error (MPE: 100 × (estimated weight − measured weight)/measured weight). The coefficient of variation (COV) was calculated as the standard deviation of the differences between the estimated and actual weights divided by the overall means of the estimated and actual weights. The 95% limits of agreement 14
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
Table 2 Comparison of weight-estimation methods across all ages and within each age group. Cattermole N = 91,147
Traditional APLS N = 91,147
Updated APLS N = 91,147
Broselow N = 84,609
Park N = 91,147
MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (kg) COV (%) ≤10% > 10% > 20% > 30%
1.35 4.71 1.32, 1.38 11.48 26.28 11.31, 11.65 −8.07, 10.77 20 37.9 27.7 15.1 19.3
−3.76 5.71 −3.80, −3.72 −12.38 14.55 −12.47, −12.28 −15.18, 7.66 28 40.6 31.7 16 11.7
−0.94 4.92 −0.98, −0.91 −3.07 −3.18 −2.95, 17.80 −10.78, 8.90 22 41.5 32.9 16.6 9.0
−0.44 2.93 −0.47, -0.42 −0.46 10.87 −0.54, −0.40 −6.30, 5.42 14 69.1 24.0 5.6 1.3
−0.24 4.71 −0.27, −0.21 0.62 16.73 0.51, 0.73 −9.66, 9.18 21 48.9 30.3 13.3 7.5
Age 1–5 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤10%
N = 55,730 3.17 3.74 3.14, 3.20 21.30 27.31 21.08, 21.53 −4.31, 10.65 20 30.3
N = 55,730 −1.84 2.48 −1.86, −1.82 −9.32 12.21 −9.43, −9.22 −6.80, 3.12 16 46.9
N = 55,730 −1.84 2.48 −1.86, −1.82 −9.32 12.21 −9.43, −9.22 −6.80, 3.12 16 46.9
N = 55,729 −0.17 1.81 −0.19, −0.16 0.02 9.65 −0.06, 0.10 −3.79, 3.45 11 74.6
N = 55,730 −0.84 2.48 −0.86, −0.82 −3.00 13.11 −3.11, −2.90 −5.80, 4.12 15 56.7
Age 6–8 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤10%
N = 19,561 0.18 3.79 0.13, 0.23 0.33 14.91 0.12, 0.54 −7.4, 7.76 15 53.5
N = 19,561 −3.91 5.14 −3.98, −3.84 −12.18 15.13 −12.39, −11.97 −14.19, 6.37 22 40.1
N = 19,561 1.94 5.00 1.87, 2.01 11.33 18.98 11.06, 11.60 −8.06, 11.94 19 31.8
N = 19,479 −0.84 3.67 −0.89, −0.79 −1.49 12.22 −1.66, −1.31 −8.18, 6.5 15 61.6
N = 19,561 0.79 4.99 0.72, 0.80 6.28 18.16 6.02, 6.53 −9.19, 10.77 19 39.5
Age 9–11 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤ 10%
N = 15,856 −3.58 4.81 −3.65, −3.50 −9.30 12.33 −9.49, −9.10 −13.2, 6.04 13 45.2
N = 15,856 −10.34 8.76 −10.47, −10.20 −23.36 16.01 −23.61, −23.12 −27.86, 7.18 26 18.7
N = 15,856 −1.36 8.58 −1.49, −1.23 1.17 20.99 0.84, 1.49 −18.52, 15.8 23 34.6
N = 9,401 −1.16 5.34 −1.27, −1.06 −1.23 14.11 −1.52, −0.95 −11.84, 9.52 16 51.7
N = 15,856 0.62 8.48 0.49, 0.75 6.39 21.87 6.05, 6.73 −16.34, 17.58 22 32.9
APLS: Advanced pediatric life support, MD: Mean difference, SD: Standard deviation, CI: Confidence interval, MPE: Mean percentage error, LOA: Limits of agreement, COV: Coefficient of variation.
The Cattermole formula was based on the correlation between the MAC and body weight. Our study is the largest to assess the accuracy and precision of a MAC-based formula and also the first to report an external validation of this method. In the original study by Cattermole et al. [14], the Cattermole formula was found to be less suitable for children under 6 years of age, but it out-performed the Broselow tape and the traditional APLS age-based formula in older children [14]. However, when we tested the Cattermole formula on a large number of Korean children, this rule was less accurate than when it was developed and evaluated. The finding that the actual weight was well estimated in the 6–8year-old age group is similar to the results of a previous study. However, the accuracy and precision across the entire age range were found to be lower than those reported in the original study. Our MPE for all ages was 11.48%, whereas Cattermole et al. [14] reported a bias (MPE) of 2.9%. The overall COV was 20.0 in this study, which is comparable with the COV of 16.1 that was reported in the original study. We found the formula to be accurate (within 10%) only 37.9% of the time, which was inferior to the previously reported accuracy of 44.2%. The LOA method was used to evaluate the precision, and this method was also used in the previous study. We found that the LOAs for
of the actual weight. In the 9–11-year-old group, the MPE and MD of the Cattermole formula were −9.30% (SD 12.33, 95% CI = −9.49 to −9.10) and −3.58 kg, respectively. The Cattermole formula produced a negative MD and MPE, which indicated that this weight estimation method underestimated the actual weight. In this oldest group, the MPE of the updated APLS formula was 1.17% (SD 20.99, 95% CI 0.84–1.49), and the MPE of the Broselow tape was −1.23% (SD 14.11, 95% CI = −1.52 to −0.95). The proportions of estimates within 10% of the actual weight were 51.7% according to the Broselow tape and 45.2% according to the Cattermole formula.
Discussions The MAC-based formula does not require any special tools, and the upper arm is always accessible in situations involving the critically ill. Thus, the MAC-based formula has been considered useful for estimating a child victim’s weight when the parents are not available and the child’s age is unclear. However, we validated the performance of the Cattermole formula using the MAC and found that it was less accurate than the Broselow tape and age-based formulae. 15
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
Funding
the MPEs were wide both overall and in each age category. In the Cattermole et al. study, the estimation accuracy of the Broselow tape was good overall for all children 1–11 years of age [14], and our study produced similar results. In the previous study, Cattermole et al. reported that the MAC-based formula was superior to the age-based APLS method and found no difference in performance between the Cattermole formula and the Broselow tape for children 6–8 years of age. These authors also demonstrated that the Cattermole formula was superior for children 9–11 years of age [14]. However, in our validation study, the proportion of estimated weights that were within 10% of the actual weights was lowest for the Cattermole formula, and this formula was less accurate than other age-based formulae, such as the APLS and Park et al.’s formulae. In the 1–5-year-old group, all parameters (i.e., MPE, AE and the within 10% value) of the performance of the Cattermole formula were inferior to those of the Broselow tape, APLS formula and Park et al.'s formula. In the 6–8-yearold and 9–11-year-old age groups, the Cattermole formula did not outperform the Broselow tape. However, with the exception of the Broselow tape, the Cattermole formula was superior to the other two agebased formulae for the 6-11-year-old age group. Some differences existed between our study and the original Cattermole et al. study. First, the data collection time was different. The previous study was based on data collected in 1999–2002, and this study was based on data collected in 2005. Thus, the Cattermole formula may not reflect the increase in child obesity. In the previous study, the authors used the 1998 version of the Broselow tape, whereas we used the 2011 version; thus, a change in the weight application interval according to height should have been included. Additionally, we did not calculate weights for children who were taller than the length of the Broselow tape, whereas Cattermole et al. entered a value of 55 kg, i.e., the weight of a mid-size adult. This approach may have affected the accuracy of the Broselow tape. This study has some limitations. First, our study included only Korean children who were aged 1–11 years. The Cattermole formula has only been examined with Chinese children; therefore, it is necessary to determine whether it is reliable for other races and for infants and adults. Second, outliers were eliminated from the original database prior to the analysis. About 8000 children who were excluded may be that are hardest to estimate using age and height estimation. This may have affected the outcome. Third, the Broselow tape was not applied to the actual children; only the weights on the tape relative to the measured heights were compared. Therefore, our estimates may be somewhat different from the estimates that would have resulted from actually using the Broselow tape.
None. Author contributions Drs. Choi, Suh and Kim had full access to all of the data in the study and take responsibility for the integrity of the data, as well as for the accuracy of the data analysis. Study concept and design: Dr. Kim. Acquisition, analysis, and interpretation of the data: Drs. Choi, Suh and Kim. Drafting of the manuscript: Drs. Choi and Suh. Critical revision of the manuscript for important intellectual content: Drs. Lee, Jung, Jeong, Kwon, Paek and Kwak. Statistical analysis: Drs. Suh, Kwon and Kim. Administrative, technical, or material support: Drs. Paek, Lee and Kwak. Study supervision: Dr. Kim. All the authors contributed substantially to its revision. Acknowledgment None. References [1] Advanced Life Support Group. Advanced paediatric life support: the practical approach. West Sussex, UK: John Wiley & Sons Ltd; 2011. [2] Lubitz DS, Seidel JS, Chameides L, Luten RC, Zaritsky AL, Campbell FW. A rapid method for estimating weight and resuscitation drug dosages from length in the pediatric age group. Ann Emerg Med 1988;17:576–81. [3] 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. [4] Black K, Barnett P, Wolfe R, Young S. Are methods used to estimate weight in children accurate? Emerg Med (N Y) 2002;14:160–5. [5] Wells M, Goldstein LN, Bentley A, Basnett S, Monteith I. The accuracy of the Broselow tape as a weight estimation tool and a drug-dosing guide—a systematic review and meta-analysis. Resuscitation 2017;121:9–33. [6] Milne WK, Yasin A, Knight J, Noel D, Lubell R, Filler G. Ontario children have outgrown the Broselow tape. CJEM 2012;14:25–30. [7] Grivas TB, Mihas C, Arapaki A, Vasiliadis E. Correlation of foot length with height and weight in school age children. J Forensic Legal Med 2008;15:89–95. [8] Krieser D, Nguyen K, Kerr D, Jolley D, Clooney M, Kelly AM. Parental weight estimation of their child’s weight is more accurate than other weight estimation methods for determining children’s weight in an emergency department? Emerg Med J 2007;24:756–9. [9] Jelliffe DB, Jelliffe EFP. Community nutritional assessment. Oxford: Oxford University Press; 1989. [10] Carroll W, Jay N, Alexander J. Towards better weight estimation in the seriously ill child–a comparison of methods. Arch Dis Child 2001;84:A12. [11] Larson LL. Relationship of upper arm circumference and body weight. J Emerg Nurs 1985;11:246–8. [12] Utley R. Mid-arm circumference. Estimating patients’ weight. Dimens Crit Care Nurs 1990;9:75–81. [13] Lin BW, Yoshida D, Quinn J, Strehlow M. A better way to estimate adult patients’ weights. Am J Emerg Med 2009;27:1060–4. [14] Cattermole GN, Leung PY, Mak PS, Graham CA, Rainer TH. Mid-arm circumference can be used to estimate children’s weights. Resuscitation 2010;81:1105–10. [15] American College of Surgeons Committee on Trauma. Advanced trauma life support student course manual. Chicago: American College of Surgeons; 2008. [16] Park J, Kwak YH, Kim DK, Jung JY, Lee JH, Jang HY, et al. A new age-based formula for estimating weight of Korean children. Resuscitation 2012;83:1129–34.
Conclusion The application of the Cattermole formula, which is a weight estimation method based on the MAC, to data from Korean children produced results that differed from those of previous studies that estimated weights with the Broselow tape. The results of this study were not as accurate as those of the original study. While the Cattermole formula produced better results for children aged 6 years and older and was most accurate for children aged 6–8 years, among the existing weight estimation methods, the performance of the Broselow tape was best overall and was also similar for all age groups. Conflicts of interest None of the authors have any competing interests to declare.
16
Contents lists available at ScienceDirect
Resuscitation journal homepage: www.elsevier.com/locate/resuscitation
Simulation and education
Validation of the mid-arm-based weight estimation formula (the Cattermole formula) for Korean children☆
T
⁎
Jea Yeon Choia,1, Dongbum Suhb,1, Do Kyun Kimc, , Young Ho Kwakc, Jae Yun Jungc, Jin Hee Leeb, Jin Hee Jeongd, Hyuksool Kwonb, So Hyun Paeke a
Department of Emergency Medicine, Gachon University Gil hospital, Incheon, Republic of Korea Department of Emergency Medicine, Seoul National University Bundang Hospital, Gyeonggido, Republic of Korea c Department of Emergency Medicine, Seoul National University Hospital, Seoul, Republic of Korea d Department of Emergency Medicine, Seoul National University Boramae Medical Center, Seoul, Republic of Korea e Department of Emergency Medicine, CHA Bundang Medical Center, CHA University, Gyeonggido, Republic of Korea b
A R T I C LE I N FO
A B S T R A C T
Keywords: Body weight Pediatrics Resuscitation Weights and measures
Background: A mid-arm circumference-based weight estimation formula has recently been proposed. The Cattermole formula, which is suggested for children aged 1–11 years, is calculated as (mid-arm circumference in cm − 10) × 3 kg. The objective of this study was to externally validate the Cattermole formula using a Korean national survey database. Methods: The data collected from a national survey conducted in 2005 by the Korean Pediatric Society included body weight, height, age and mid-arm circumference. To determine whether the Cattermole formula accurately predicted the actual weights of Korean children, we compared the performance of the Cattermole formula with those of other age-based and height-based weight estimation methods. Results: A total of 91,147 children aged 1–11 years were assessed for eligibility. The mean age was 5.3 years, and 52.6% of the children were boys. The mean measured weight was 22.4 kg, and the mean mid-arm circumference was 17.9 cm. The Cattermole formula resulted in a mean difference of 1.35 kg and a mean percentage error of 11.5% with 95% limits of agreement of −8.1 to +10.8. A total of 37.9% of the estimates were within 10% of the measured weights. Conclusion: Across all ages, the Cattermole formula was not more accurate than the other methods. However, in children older than 6 years, the Cattermole formula results were more accurate than those of the age-based formulae with the exception of the Broselow tape. Clinical trial registration: None.
Introduction Accurately measuring the weight of a critically ill child is essential during pediatric resuscitation. The child’s weight is used to calculate the defibrillation energy and most drug doses and to determine the sizes of equipment used. However, during resuscitation, there is insufficient time to weigh the child, and rapid and reliable methods of estimation must be applied. Several methods that aim to accurately estimate a child’s weight have been devised. Commonly, weight can be estimated from age-based formulae, such as the Advanced Pediatric Life Support (APLS) formula [1], or from length-based methods using the Broselow tape [2]. Length-
based methods, most prevalently the Broselow tape, have long been used and are generally accepted as more accurate than age-based formulae [3,4]. However, the Broselow tape is significantly limited by its length, so a recent review presented that the newer length- and habitusbased systems are superior to the Broselow tape [5]. In response to increasing concerns about wrong measurement of actual weights with length-based methods due to child obesity, other anthropometric measurements have been proposed [6]. Some of these measures are based on other physical features such as foot length [7]. A simple method that has been studied is weight estimation by the parents or a clinician [8]. The mid-arm circumference (MAC) has been used for many years in
☆
A Spanish translated version of the abstract of this article appears as Appendix in the final online version at https://doi.org/10.1016/j.resuscitation.2018.08.026 Corresponding author at: Department of Emergency Medicine, Seoul National University Hospital, 101 Daehang-no, Jongno-gu, Seoul 03080, Republic of Korea. E-mail address: [email protected] (D.K. Kim). 1 These authors contributed equally to this work. ⁎
https://doi.org/10.1016/j.resuscitation.2018.08.026 Received 8 June 2018; Received in revised form 31 July 2018; Accepted 24 August 2018 0300-9572/ © 2018 Published by Elsevier B.V.
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
the assessment of malnutrition in the developing world [9]. However, several studies have suggested the use of weight estimation methods that employ the MAC [10–12]. A formula that includes the MAC and knee height has been validated in geriatric patients in the emergency department [13]. In 2010, Cattermole et al. suggested a new formula that is calculated as (MAC in cm − 10) × 3 for children aged 1–11 years. They reported that this formula performed better than the Broselow tape and the APLS aged-based formula in older children [14]. However, this method has not been validated in other populations. The objectives of this study were to externally validate the Cattermole formula using a Korean national survey database and to compare the utilities of other existing weight estimation methods.
Table 1 Methods of child weight estimation.
Methods
MAC: Mid-arm circumference, APLS: Advanced pediatric life support.
Study setting and population
(LOA), which measure the spread around the MPE, and the percentage of patients whose weight estimation was within 10% of the measured weight are also reported. The statistical analyses were performed with SPSS software (version 18.0; SPSS Inc., Chicago, IL, USA), and the level for statistical significance was set a p < 0.05.
Name
Applicability
Method of calculation
Cattermole Traditional APLS Updated APLS
1–11 years 1–10 years < 1 year 1–5 years 6–12 years
Park
< 12 months 1–4 years 5–14 years Height 46–143.4 cm
Weight (kg) = (MAC − 10) ×3 Weight (kg) = (age in years + 4) × 2 Weight (kg) = (0.5 × age in months) +4 Weight (kg) = (2 × age in years) + 8 Weight (kg) = (3 × age in years) + 7 Weight (kg) = (age in months + 9)/2 Weight (kg) = 2 × (age in years) + 9 Weight (kg) = 4 × (age in years) − 1 Read weight from measuring tape
Broselow
We collected data that included body weight, height, age and MAC from a national survey that was conducted by the Korean Pediatric Society (KPS) under the supervision of the Ministry of Health and Welfare in 2005 for the development of the new 2007 Korean national growth charts. In 2005, a KPS project measured 150,000 children nationwide. The measurements were performed by 6 trained measurement teams and 28 general hospital teams between April and July of 2005 and August and September of 2005. The study population was composed of children between 0 and 20 years of age. The data were collected from daycare centers, childcare counseling rooms, kindergartens, elementary, middle and high schools and colleges from 16 assigned areas. To avoid the result is distorted by extreme values, we standardized collected data by eliminating outliers. We define outliers as the weight for height is more than 3 × (high standard deviation) or less than 3 × (low standard deviation) and eliminated from analysis. Consequently, the physical measurements of approximately 142,000 children were included. The original data were provided by the Center for Disease Control and Prevention, which is an affiliate of the Ministry of Health and Welfare. Based on the data included above, only data from children 1–11 years of age were ultimately used. The Cattermole formula developed with 1–11 years old age. Thus, we selected the same range of age to facilitate comparison with the results.
Ethics This study complied with the Declaration of Helsinki and was approved by the Seoul National University Hospital Institutional Review Board with a waiver of informed consent. Results Out of 142,945 children aged 0–20 years from the KPS project, 6029 were excluded due to the loss of data about age, heights, weights or MAC. Among 136,916 children, a total of 91,147 subjects aged 1–11 years were included in this study. The number of male subjects was 47,977 (52.6%), and the mean age of all subjects was 5.3 (SD 2.8) years. The mean weight was 22.4 (SD 9.9) kg, and the MAC was 17.9 (SD 2.8) cm. The participants included 55,730 (61.1%) 1–5-year-old, 19,561 (21.5%) 6–8-year-old and 15,856 (17.4%) 9–11-year-old children. The mean MACs of the 1–5, 6–8, and 9–11-year-old groups were 16.6 (SD 1.7) cm, 18.6 (SD2.4) cm and 21.6 (SD 3.1) cm, respectively. Using the Cattermole formula, we found that only 37.9% of all patients had estimated weights that were within 10% of their actual weights. Overall, the Cattermole formula had an MD of 1.35 kg and an MPE of 11.48% (SD 26.28, 95% CI 11.31–11.65). The COV was 20, and the 95% LOA were −8.08 to 10.77 kg. The MPE of the older APLS formula showed the highest value, and the MPE of the Cattermole formula was the second highest relative to the MPEs that were calculated for all other methods. The COV of the traditional APLS formula showed the poorest result. In terms of the proportions of subjects with estimated weights that were within 10% of the actual weights, the Broselow tape method exhibited the best performance (Table 2). In the 1–5-year-old group, the MD of the Cattermole formula was 3.17 kg, and the MPE was 21.3% (SD 27.3, 95% CI = 21.08–21.53); the latter value was the highest among the estimation methods. The Broselow tape method and Park et al.’s formula produced lower MPE values than the Cattermole formula. In this age group, 30.3% of the estimates based on the Cattermole formula were within 10% of the measured weight, which was the poorest performance among the three age groups (Table 2). In the 6–8-year-old group, the MPE and MD of the Cattermole formula were 0.33% (SD 14.91, 95% CI 0.12–0.54) and 0.18 kg, respectively, and these values indicated the best accuracy among all age groups. The proportion of measures within 10% of the actual weight was 53.5%, which also represented the best performance of the Cattermole formula among all age groups. However, among all estimation methods, the Broselow tape still performed best in terms of the accuracy according to the proportion of estimates that were within 10%
Study protocol and data analysis The subjects were classified into the following 3 groups that were based on the groups used in the Cattermole et al. study [14], which in turn were based on the Advanced Trauma Life Support course: 1–5, 6–8, and 9–11-year-olds [15]. This classification was performed to facilitate comparison of the results. The primary outcome of interest was the agreement between the weights estimated by the Cattermole formula and the measured weights of Korean children. The secondary outcomes were the comparisons of the estimation performance of the Cattermole formula with those of four other methods, i.e., the traditional APLS formula, the updated APLS formula, the Broselow tape and Park et al.’s formula (Table 1) [16]. The Broselow tape itself was not used. The children's weights were estimated by applying the measured height to the corresponding interval of the 2011 version of the Broselow tape. If the age of a subject deviated from the limit required by each formula, or if the height of a subject was less than 46 cm or more than 143.4 cm, a missing value was entered. The Cattermole formula and the measured weight were assessed via calculations of the mean difference (MD: estimated weight − measured weight) and the mean percentage error (MPE: 100 × (estimated weight − measured weight)/measured weight). The coefficient of variation (COV) was calculated as the standard deviation of the differences between the estimated and actual weights divided by the overall means of the estimated and actual weights. The 95% limits of agreement 14
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
Table 2 Comparison of weight-estimation methods across all ages and within each age group. Cattermole N = 91,147
Traditional APLS N = 91,147
Updated APLS N = 91,147
Broselow N = 84,609
Park N = 91,147
MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (kg) COV (%) ≤10% > 10% > 20% > 30%
1.35 4.71 1.32, 1.38 11.48 26.28 11.31, 11.65 −8.07, 10.77 20 37.9 27.7 15.1 19.3
−3.76 5.71 −3.80, −3.72 −12.38 14.55 −12.47, −12.28 −15.18, 7.66 28 40.6 31.7 16 11.7
−0.94 4.92 −0.98, −0.91 −3.07 −3.18 −2.95, 17.80 −10.78, 8.90 22 41.5 32.9 16.6 9.0
−0.44 2.93 −0.47, -0.42 −0.46 10.87 −0.54, −0.40 −6.30, 5.42 14 69.1 24.0 5.6 1.3
−0.24 4.71 −0.27, −0.21 0.62 16.73 0.51, 0.73 −9.66, 9.18 21 48.9 30.3 13.3 7.5
Age 1–5 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤10%
N = 55,730 3.17 3.74 3.14, 3.20 21.30 27.31 21.08, 21.53 −4.31, 10.65 20 30.3
N = 55,730 −1.84 2.48 −1.86, −1.82 −9.32 12.21 −9.43, −9.22 −6.80, 3.12 16 46.9
N = 55,730 −1.84 2.48 −1.86, −1.82 −9.32 12.21 −9.43, −9.22 −6.80, 3.12 16 46.9
N = 55,729 −0.17 1.81 −0.19, −0.16 0.02 9.65 −0.06, 0.10 −3.79, 3.45 11 74.6
N = 55,730 −0.84 2.48 −0.86, −0.82 −3.00 13.11 −3.11, −2.90 −5.80, 4.12 15 56.7
Age 6–8 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤10%
N = 19,561 0.18 3.79 0.13, 0.23 0.33 14.91 0.12, 0.54 −7.4, 7.76 15 53.5
N = 19,561 −3.91 5.14 −3.98, −3.84 −12.18 15.13 −12.39, −11.97 −14.19, 6.37 22 40.1
N = 19,561 1.94 5.00 1.87, 2.01 11.33 18.98 11.06, 11.60 −8.06, 11.94 19 31.8
N = 19,479 −0.84 3.67 −0.89, −0.79 −1.49 12.22 −1.66, −1.31 −8.18, 6.5 15 61.6
N = 19,561 0.79 4.99 0.72, 0.80 6.28 18.16 6.02, 6.53 −9.19, 10.77 19 39.5
Age 9–11 years MD (kg) SD 95% CI MPE (%) SD 95% CI LOA (%) COV (%) ≤ 10%
N = 15,856 −3.58 4.81 −3.65, −3.50 −9.30 12.33 −9.49, −9.10 −13.2, 6.04 13 45.2
N = 15,856 −10.34 8.76 −10.47, −10.20 −23.36 16.01 −23.61, −23.12 −27.86, 7.18 26 18.7
N = 15,856 −1.36 8.58 −1.49, −1.23 1.17 20.99 0.84, 1.49 −18.52, 15.8 23 34.6
N = 9,401 −1.16 5.34 −1.27, −1.06 −1.23 14.11 −1.52, −0.95 −11.84, 9.52 16 51.7
N = 15,856 0.62 8.48 0.49, 0.75 6.39 21.87 6.05, 6.73 −16.34, 17.58 22 32.9
APLS: Advanced pediatric life support, MD: Mean difference, SD: Standard deviation, CI: Confidence interval, MPE: Mean percentage error, LOA: Limits of agreement, COV: Coefficient of variation.
The Cattermole formula was based on the correlation between the MAC and body weight. Our study is the largest to assess the accuracy and precision of a MAC-based formula and also the first to report an external validation of this method. In the original study by Cattermole et al. [14], the Cattermole formula was found to be less suitable for children under 6 years of age, but it out-performed the Broselow tape and the traditional APLS age-based formula in older children [14]. However, when we tested the Cattermole formula on a large number of Korean children, this rule was less accurate than when it was developed and evaluated. The finding that the actual weight was well estimated in the 6–8year-old age group is similar to the results of a previous study. However, the accuracy and precision across the entire age range were found to be lower than those reported in the original study. Our MPE for all ages was 11.48%, whereas Cattermole et al. [14] reported a bias (MPE) of 2.9%. The overall COV was 20.0 in this study, which is comparable with the COV of 16.1 that was reported in the original study. We found the formula to be accurate (within 10%) only 37.9% of the time, which was inferior to the previously reported accuracy of 44.2%. The LOA method was used to evaluate the precision, and this method was also used in the previous study. We found that the LOAs for
of the actual weight. In the 9–11-year-old group, the MPE and MD of the Cattermole formula were −9.30% (SD 12.33, 95% CI = −9.49 to −9.10) and −3.58 kg, respectively. The Cattermole formula produced a negative MD and MPE, which indicated that this weight estimation method underestimated the actual weight. In this oldest group, the MPE of the updated APLS formula was 1.17% (SD 20.99, 95% CI 0.84–1.49), and the MPE of the Broselow tape was −1.23% (SD 14.11, 95% CI = −1.52 to −0.95). The proportions of estimates within 10% of the actual weight were 51.7% according to the Broselow tape and 45.2% according to the Cattermole formula.
Discussions The MAC-based formula does not require any special tools, and the upper arm is always accessible in situations involving the critically ill. Thus, the MAC-based formula has been considered useful for estimating a child victim’s weight when the parents are not available and the child’s age is unclear. However, we validated the performance of the Cattermole formula using the MAC and found that it was less accurate than the Broselow tape and age-based formulae. 15
Resuscitation 132 (2018) 13–16
J.Y. Choi et al.
Funding
the MPEs were wide both overall and in each age category. In the Cattermole et al. study, the estimation accuracy of the Broselow tape was good overall for all children 1–11 years of age [14], and our study produced similar results. In the previous study, Cattermole et al. reported that the MAC-based formula was superior to the age-based APLS method and found no difference in performance between the Cattermole formula and the Broselow tape for children 6–8 years of age. These authors also demonstrated that the Cattermole formula was superior for children 9–11 years of age [14]. However, in our validation study, the proportion of estimated weights that were within 10% of the actual weights was lowest for the Cattermole formula, and this formula was less accurate than other age-based formulae, such as the APLS and Park et al.’s formulae. In the 1–5-year-old group, all parameters (i.e., MPE, AE and the within 10% value) of the performance of the Cattermole formula were inferior to those of the Broselow tape, APLS formula and Park et al.'s formula. In the 6–8-yearold and 9–11-year-old age groups, the Cattermole formula did not outperform the Broselow tape. However, with the exception of the Broselow tape, the Cattermole formula was superior to the other two agebased formulae for the 6-11-year-old age group. Some differences existed between our study and the original Cattermole et al. study. First, the data collection time was different. The previous study was based on data collected in 1999–2002, and this study was based on data collected in 2005. Thus, the Cattermole formula may not reflect the increase in child obesity. In the previous study, the authors used the 1998 version of the Broselow tape, whereas we used the 2011 version; thus, a change in the weight application interval according to height should have been included. Additionally, we did not calculate weights for children who were taller than the length of the Broselow tape, whereas Cattermole et al. entered a value of 55 kg, i.e., the weight of a mid-size adult. This approach may have affected the accuracy of the Broselow tape. This study has some limitations. First, our study included only Korean children who were aged 1–11 years. The Cattermole formula has only been examined with Chinese children; therefore, it is necessary to determine whether it is reliable for other races and for infants and adults. Second, outliers were eliminated from the original database prior to the analysis. About 8000 children who were excluded may be that are hardest to estimate using age and height estimation. This may have affected the outcome. Third, the Broselow tape was not applied to the actual children; only the weights on the tape relative to the measured heights were compared. Therefore, our estimates may be somewhat different from the estimates that would have resulted from actually using the Broselow tape.
None. Author contributions Drs. Choi, Suh and Kim had full access to all of the data in the study and take responsibility for the integrity of the data, as well as for the accuracy of the data analysis. Study concept and design: Dr. Kim. Acquisition, analysis, and interpretation of the data: Drs. Choi, Suh and Kim. Drafting of the manuscript: Drs. Choi and Suh. Critical revision of the manuscript for important intellectual content: Drs. Lee, Jung, Jeong, Kwon, Paek and Kwak. Statistical analysis: Drs. Suh, Kwon and Kim. Administrative, technical, or material support: Drs. Paek, Lee and Kwak. Study supervision: Dr. Kim. All the authors contributed substantially to its revision. Acknowledgment None. References [1] Advanced Life Support Group. Advanced paediatric life support: the practical approach. West Sussex, UK: John Wiley & Sons Ltd; 2011. [2] Lubitz DS, Seidel JS, Chameides L, Luten RC, Zaritsky AL, Campbell FW. A rapid method for estimating weight and resuscitation drug dosages from length in the pediatric age group. Ann Emerg Med 1988;17:576–81. [3] 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. [4] Black K, Barnett P, Wolfe R, Young S. Are methods used to estimate weight in children accurate? Emerg Med (N Y) 2002;14:160–5. [5] Wells M, Goldstein LN, Bentley A, Basnett S, Monteith I. The accuracy of the Broselow tape as a weight estimation tool and a drug-dosing guide—a systematic review and meta-analysis. Resuscitation 2017;121:9–33. [6] Milne WK, Yasin A, Knight J, Noel D, Lubell R, Filler G. Ontario children have outgrown the Broselow tape. CJEM 2012;14:25–30. [7] Grivas TB, Mihas C, Arapaki A, Vasiliadis E. Correlation of foot length with height and weight in school age children. J Forensic Legal Med 2008;15:89–95. [8] Krieser D, Nguyen K, Kerr D, Jolley D, Clooney M, Kelly AM. Parental weight estimation of their child’s weight is more accurate than other weight estimation methods for determining children’s weight in an emergency department? Emerg Med J 2007;24:756–9. [9] Jelliffe DB, Jelliffe EFP. Community nutritional assessment. Oxford: Oxford University Press; 1989. [10] Carroll W, Jay N, Alexander J. Towards better weight estimation in the seriously ill child–a comparison of methods. Arch Dis Child 2001;84:A12. [11] Larson LL. Relationship of upper arm circumference and body weight. J Emerg Nurs 1985;11:246–8. [12] Utley R. Mid-arm circumference. Estimating patients’ weight. Dimens Crit Care Nurs 1990;9:75–81. [13] Lin BW, Yoshida D, Quinn J, Strehlow M. A better way to estimate adult patients’ weights. Am J Emerg Med 2009;27:1060–4. [14] Cattermole GN, Leung PY, Mak PS, Graham CA, Rainer TH. Mid-arm circumference can be used to estimate children’s weights. Resuscitation 2010;81:1105–10. [15] American College of Surgeons Committee on Trauma. Advanced trauma life support student course manual. Chicago: American College of Surgeons; 2008. [16] Park J, Kwak YH, Kim DK, Jung JY, Lee JH, Jang HY, et al. A new age-based formula for estimating weight of Korean children. Resuscitation 2012;83:1129–34.
Conclusion The application of the Cattermole formula, which is a weight estimation method based on the MAC, to data from Korean children produced results that differed from those of previous studies that estimated weights with the Broselow tape. The results of this study were not as accurate as those of the original study. While the Cattermole formula produced better results for children aged 6 years and older and was most accurate for children aged 6–8 years, among the existing weight estimation methods, the performance of the Broselow tape was best overall and was also similar for all age groups. Conflicts of interest None of the authors have any competing interests to declare.
16