This project was awarded “Best Medical Student Research Award” at the International Conference on Emergency Medicine (INTEM 2006) in New Delhi, November 2006. The study was presented at the International Conference on Emergency Medicine (INTEM 2006) in New Delhi, November 2006, and the Western Society for Academic Emergency Medicine conference in Portland, March 2007.
Internationalizing the Broselow Tape: How Reliable Is Weight Estimation in Indian Children
Version of Record online: 29 MAR 2008
© 2008 by the Society for Academic Emergency Medicine
Academic Emergency Medicine
Volume 15, Issue 5, pages 431–436, May 2008
How to Cite
Ramarajan, N., Krishnamoorthi, R., Strehlow, M., Quinn, J. and Mahadevan, S. V. (2008), Internationalizing the Broselow Tape: How Reliable Is Weight Estimation in Indian Children. Academic Emergency Medicine, 15: 431–436. doi: 10.1111/j.1553-2712.2008.00081.x
- Issue online: 29 MAR 2008
- Version of Record online: 29 MAR 2008
- Received November 1, 2007; revision received January 4, 2008; accepted January 8, 2008.
- body height;
- body weight;
- reproducibility of results;
Objectives: The Broselow pediatric emergency weight estimation tape is an accurate method of estimating children’s weights based on height–weight correlations and determining standardized medication dosages and equipment sizes using color-coded zones. The study objective was to determine the accuracy of the Broselow tape in the Indian pediatric population.
Methods: The authors conducted a 6-week prospective cross-sectional study of 548 children at a government pediatric hospital in Chennai, India, in three weight-based groups: <10 kg (n = 175), 10–18 kg (n = 197), and >18 kg (n = 176). Measured weight was compared to Broselow-predicted weight, and the percentage difference was calculated. Accuracy was defined as agreement on Broselow color-coded zones, as well as agreement within 10% between the measured and Broselow-predicted weights. A cross-validated correction factor was also derived.
Results: The mean percentage differences were −2.4, −11.3, and −12.9% for each weight-based group. The Broselow color-coded zone agreement was 70.8% in children weighing less than 10 kg, but only 56.3% in the 10- to 18-kg group and 37.5% in the >18-kg group. Agreement within 10% was 52.6% for the <10-kg group, but only 44.7% for the 10- to 18-kg group and 33.5% for the >18-kg group. Application of a 10% weight-correction factor improved the percentages to 77.1% for the 10- to 18-kg group and 63.0% for the >18-kg group.
Conclusions: The Broselow tape overestimates weight by more than 10% in Indian children >10 kg. Weight overestimation increases the risk of medical errors due to incorrect dosing or equipment selection. Applying a 10% weight-correction factor may be advisable.
Accurate estimation of a child’s weight is the critical first step in the management of pediatric emergencies, because most pediatric drug dosing and equipment sizes are determined using body weight. Given the impracticality of weighing children during emergencies, weights are often estimated. The effectiveness of emergent interventions, therefore, may ultimately depend on the accuracy of the weight estimation. Various methods, including visual assessments of weight1,2 and equations for drug dosing reliant upon age and height, have proven to be inefficient and prone to error.3–6 In fact, most medication errors and drug adverse effects are related to improper dosing.7–11 Drug dosing is heavily dependent on body weight estimation. The most common weight-estimating formulas are derived using Western growth references and are likely to overestimate weight when applied to the population of a developing country like India.
As emergency care develops globally, devices such as the Broselow tape could be valuable in the care of acutely ill, pediatric patients. The Broselow tape allows emergency personnel to estimate a child’s weight and then determine standardized emergency drug dosages and equipment sizes using a color-coded zone system based on height–weight correlation. The Broselow tape has been validated in both ambulatory12–15 and simulated emergency situations16 in the United States and is believed to reduce complications arising from inaccurate drug dosing and equipment sizing. A recent report, however, suggests that the Broselow tape may be significantly underestimating weights in U.S. children due to rising childhood obesity.17 Studies have also been performed validating the Broselow tape in Switzerland,18 Australia,19 and South Korea.20
A recent study of Indian infants21 demonstrated that the Broselow tape was the most accurate method of weight estimation in children under 1 year of age. Unfortunately, weight estimation of older children has proven more problematic because the height–weight correlation begins to diverge significantly from that of Western references. No studies have been published on the accuracy of Broselow tape estimations in children 1 to 12 years of age in developing countries. Our research addresses this problem, specifically within India.
The current study evaluated the Broselow pediatric emergency weight estimation tape in Indian children ≤12 years of age. We hypothesized that the Broselow tape would overestimate weights in Indian children, leading to inaccurate medication dosing and equipment sizing.
This was a prospective, observational, hospital-based study of 548 children over a 6-week period from June 15, 2006, to July 31, 2006. Informed consent was obtained from all subjects or their guardians, and the research was approved by Stanford University’s Institutional Review Board for human subjects’ research.
Study Setting and Population
The study was conducted at The Institute of Child Health and Hospital for Children (ICH), a pediatric referral hospital of the Government of Tamil Nadu State, India. The hospital receives 2,000 pediatric patients per day from all socioeconomic backgrounds. Inclusion criteria were children aged 1 month to 12 years presenting to the ambulatory general pediatrics clinic. Exclusion criteria were children needing emergency care, children with severe dehydration or joint contractures, children with length <46 or >143 cm (prerequisite for using the Broselow tape), and children whose parents refused to allow participation.
Data were collected by two investigators. Subjects were weighed on a standing pediatric scale or a supine infant scale to the nearest 0.1 kg. Heavy outerwear and footwear were removed. Scales were calibrated daily according to manufacturer’s instructions (Braun). The Broselow tape–predicted weights were determined with the child lying supine by measuring from crown to heel. Both investigators measured, in a blinded fashion, 10% of the sample population to determine interrater reliability. Patient age and gender were recorded.
Key outcome measures for Broselow tape performance included mean percentage difference as an estimation of bias, standard deviation (SD) as an estimation of precision, and percentage agreement on color-coded zones as well as prediction of Broselow weight within 10% of measured weight as measures of tape accuracy supported by previous literature.12,17,18,20 Accuracies corrected by a cross-validated factor are an outcome measure for the validated tape.
Sample Size Estimation and Data Analysis
Weight categories were selected by dividing the tape’s nine color-coded zones into three groups to analyze the effectiveness of the Broselow tape for different pediatric size ranges. The data were distributed into three categories: <10, 10–18, and >18 with 175, 197, and 176 children in each group, respectively. Earlier literature on the tape divides children into <10-, 10- to 25-, and >25-kg groups.12,19,20 Our choice of thirds also allows easier comparison to children <12 months of age (approx. <10 kg), children from 1 to 5 years of age (approx. 10–18 kg), and children >5 years of age (approx. >18 kg). To detect a 3% difference between the Broselow-predicted weight and measured weight, each weight category required a sample size of at least 173 children, with a minimum aggregate of 519 total children.
We calculated the percentage difference between the Broselow-predicted weight and the measured weight as a measure of tape bias (100 × [measured weight − Broselow-predicted weight/measured weight]). The null hypothesis stated the Broselow-predicted weight is equivalent to the measured weight if the 95% confidence interval (CI) for the mean percentage difference included ±3%. The SD of the percentage difference was also calculated to estimate tape precision.
To maintain consistency with previous literature on the Broselow tape, a Bland-Altman analysis22 was performed. This method combines bias and precision to determine upper and lower “limits of agreement” by which the two methods differ.
Accuracy of the tape was analyzed with respect to color-coded zone prediction and weight prediction. First, percentage agreement on the same color-coded zone by the Broselow tape and the measured weight, as well as percentage overestimation by one or two color-coded zones, was calculated. Second, the percentage of times the Broselow-predicted weight was within 10% of the measured weight was also determined. Finally, a correction factor for the Broselow-predicted weight was then derived by serially testing corrections until the accuracy within 10% for each weight group was maximized. The correction factor was then tested by cross-validation against a random half of the sample as well as by linear regression. This correction factor was applied to the original Broselow-predicted weights, and new corrected accuracies were obtained. All analysis was conducted using SPSS 14.0 (SPSS Inc., Chicago, IL).
Of the 548 children who met the study criteria, 50.9% (279) were males and 49.1% (269) were females. A wide range of ages (1–156 months) and weights (2.7–37 kg) were represented. Combined results for boys and girls are presented because separate gender analysis yielded similar conclusions. Age and gender data are summarized in Table 1. Interrater agreement on concordant color-coded zone selection for blinded samples was 96% (54/56 children), which yields an unweighted kappa coefficient of 0.95 (95% CI = 0.90 to 1), representing substantial agreement.
|Broselow-predicted Weight Groups (kg)|
|Age range (months)||1–28||8–114||49–156|
|Age (months), mean ± SD||9.3 ± 5||43.4 ± 20.2||108.4 ± 21.9|
|Gender, number female (%)||87 (49.7%)||96 (48.7%)||86 (48.8%)|
Bias and Precision of the Broselow Tape in Indian Children
Figure 1 illustrates the Broselow-predicted weight for a given measured weight in Indian children. The regression line indicates an overestimation of weight by the tape. The accuracy of the Broselow-predicted weights decreased with increasing patients’ weights.
Table 2 displays the mean percentage difference between the Broselow-predicted weight and measured weight as well as the SD. The mean percentage difference was significantly greater in the 10- to 18-kg (−11.3%) and the >18-kg (−12.9%) groups than the <10-kg (−2.4%) group.
|Broselow-predicted Weight Groups (kg)|
|<10 (n = 175)||10–18 (n = 197)||>18 (n = 176)|
|Measures of bias and precision|
|Mean percentage difference ± CI||−2.4 ± 1.9%||−11.3 ± 1.5%||−12.9 ± 2%|
|SD of % difference||12.6||10.9||13.5|
|Bland-Altman limits of agreement (kg), lower, upper; ± CI||−1.8, 1.5; ± 0.1||−3.8, 1.4; ± 0.2||−8.8, 3.6; ±0.4|
|Accuracy by weight|
|Accurate within 10% ± CI||52.6 ± 7.4%||44.7 ± 6.9%||33.5 ± 7.0%|
|Accurate within 10% ± CI (U.S. children – comparison values)*||55 ± 2.0%||59.9 ± 1.9%||51.1 ± 1.8%|
|Accurate within 10% after correction ± CI||NA||77.1 ± 5.9%||63.0 ± 7.1%|
|Accuracy by color zone|
|Color-coded zone agreement ± CI||70.8 ± 6.7%||56.3 ± 6.9%||37.5 ± 7.2%|
|Color-coded zone agreement after correction ± CI||NA||64.5 ± 6.7%||61.4 ± 7.2%|
To estimate concordance between the Broselow-predicted weight and measured weight, we used the Bland-Altman method. The limits of agreement (Table 2) suggest that the tape is most concordant in the <10-kg weight category (−1.8 to 1.5; ±0.1 kg) compared to the 10- to 18-kg category (−3.8 to 1.4; ±0.2 kg) and the >18-kg category (−8.8 to 3.6; ±0.4 kg).
Accuracy of Broselow Tape Color-coded Zones and Weight Estimates
The frequency of accurate color-coded zone prediction and weight prediction within 10% by the Broselow tape is shown in Table 2. The tape predicts the correct color zone 70.8% of the time in children <10 kg, but only 56.3% of the time in the 10- to 18-kg group. In the >18-kg category, the tape is more likely to overestimate by one zone (52.8%) than predict the correct color-coded zone (37.5%). Disagreement by two color zones occurred in 4% of patients >18 kg. The Broselow tape predicted weights within 10% of measured weight in children <10, 10–18, and >18 kg a total of 52.6, 44.7, and 33.5% of the time, respectively.
Correction for Children >10 kg
To maximize the accuracy of the tape in Indian children >10 kg, we derived a correction factor as a percentage of Broselow-predicted weight for each weight group. The correction factors derived for maximum accuracy were 9.6% (10–18 kg) and 12.2% (>18 kg). Using a rounded correction factor of 10% improved accuracy to values that were statistically similar to maximal correction. The original accuracy for the weight group and accuracy after applying the 10% correction are shown in Table 2.
The color-coded zone agreement improved significantly after applying the 10% correction to the Broselow-predicted weight (Figure 2). Overestimation by one zone decreased from 38.6 to 19.3% in the 10- to 18-kg group and from 52.8 to 18.8% in the >18-kg group. Furthermore, overestimation by two zones disappeared (4%–0%) after correction. Agreement on the correct color zone increased significantly in the >18-kg category from 37.5 to 61.4%. Although underestimation of weight in each group increased slightly, the overall accuracy of the tape improved after application of the 10% correction factor.
In this study, the Broselow tape consistently overestimated weights of Indian children >10 kg by 10% or more, thereby placing pediatric patients at risk for medication dosing and equipment sizing errors. In children >18 kg, the Broselow tape estimation was more likely to be incorrect by one color-coded zone (52.8%) than predict the correct zone (37.5%). Overestimation by two color-coded zones occurred in 4% of children >18 kg.
The Broselow tape is recommended for use by Pediatric Advanced Life Support (PALS) guidelines, which are currently being taught by the Indian Academy of Pediatrics.23 When using the Broselow tape in Indian children, weight overestimation is a reason for serious concern. Lower weights in Indian children are most likely caused by undernutrition and are a combination of decreased lean body mass as well as decreased adipose tissue. Overestimation of weight could lead to the administration of higher-than-recommended doses of potentially harmful emergency medications such as epinephrine.
During emergency care, accurate determination of equipment size is also critical. Overestimation of equipment size by the Broselow tape may lead to injury or potentially life-threatening time delays during procedures.
Our 10% correction factor allows the Broselow tape to estimate weights that are more representative of the general Indian pediatric population. Adjusting weights by 10% would lead to a downward shift in Broselow color-coded zones, dropping children at zone margins one color-coded weight class. For instance, a child with an estimated weight of 24 kg would currently fall within the orange zone. However, applying the 10% correction factor would reduce the estimated weight to 21.6 kg, and the child would be reclassified in the blue zone.
Recent research into the accuracy of the Broselow tape in the United States has brought up concerns regarding underestimations of drug dosages secondary to the obesity epidemic in children.17 Comparing the Broselow tape’s accuracy in Indian children to the recent validation study done in American children demonstrates that the tape is significantly less accurate in the Indian population. The age-based accuracy data for American children were matched with the weight-based accuracy in Indian children (<12 months old to <10 kg; 12–59.9 months old to 10–18 kg; and >60 months old to >18 kg). Although age and weight categories cannot be compared precisely, the ranges of age and weights in our data sets are similar. The Broselow tape’s accuracy (within 10% of actual weight) in Indian children compares favorably to American children in the <10-kg group, but is significantly lower in the 10- to 18- and >18-kg weight groups, indicating that the Broselow tape in India is not performing with the same accuracy as in the United States. However, after application of the 10% correction factor, we found the accuracy of the Broselow tape in Indian children to be higher (77.1% vs. 59.9% in children weighing 10–18 kg and 63.0% vs. 51.1% in children weighing >18 kg) than in American children.
The Broselow tape and color-coded zone system was developed based off the 50th percentile curves on the National Center for Health Statistics (NCHS) growth reference in the 1970s. Exclusive breastfeeding24,25 and nutritional deficiencies are two possible major reasons why Indian children have a lower body mass and weight compared to the NCHS reference. In 1998 and 1999, 47% of Indian children younger than 3 years were underweight or severely underweight, and an additional 26% were mildly underweight. In total, three-quarters of Indian children were underweight.26 The World Health Organization’s 2006 growth standards suggest that genetic influences on childhood weight for length are minimal and, indeed, that breastfed children growing around the world in healthy communities can be expected to achieve similar heights and weights for given ages.27 This assumption implies that the same anthropometric methods used to predict children’s weights in the United States, like the Broselow system, would be useful in similar settings internationally. Confirming this, studies in the United States, Europe, and other developed countries have found no significant variation in weight prediction by race or ethnicity.12,18,28 Recent studies in India have also shown that the growth performance of younger children (<15 kg) in affluent areas is similar to developed countries, resulting in a high correlation coefficient for the Broselow-predicted weight with measured weight in these communities.20,29
Our study raises important questions about the internationalization and validation of the Broselow-Luten pediatric emergency weight estimation tape in developing country populations other than India. In addition, once the correction factor is prospectively validated for clinical utility, our study methodology can be followed to test the tape before introducing it to other parts of the developing world.
Sufficient evidence exists from previous studies to show that most children growing up in affluent communities will be well represented by the current Broselow tape (Pacific Islander/Maori children is one proven exception28). There are also significant challenges to the widespread use of the tape in India and other developing countries beyond validation and correction. Reducing the price of the tape in developing countries and allowing international printing of corrected versions will allow the Broselow-Luten color-coded system to be utilized throughout the world. This would pave the way for a simple, population-appropriate, transportable emergency tool to assist emergency care providers across the developing world.
The Broselow tape reduces estimations, memorization, and complex calculations. In Indian children, performing a simple calculation (reducing predicted weights by 10% for children over 10 kg) during resuscitation will improve accuracy and preserve the tape’s utility. Prospective validation of the 10% correction factor in the Indian population is required before its implementation.
The study was conducted at a single hospital. However, the ICH is a large, quaternary referral center that attracts a wide array of urban, rural, poor, and middle-class patients from all over South India. The less affluent population cared for at a public hospital such as ICH is representative of India, where 75% of children are underweight. A second potential limitation is the lack of reliable age data for these children. Accurate age data would have allowed us to make more direct comparisons to American children, since data from previous studies were analyzed in terms of age groups (infants, toddlers, etc.) and not weight groups (<10 kg, 10–18 kg, etc.). Finally, our correction factor of 10% for children >10 kg requires prospective validation in an emergency practice setting.
Our limitations are also our strengths. Our sample largely comes from poorer, more underweight populations of children in India seen in a general hospital that comprises more than 75% of Indian children. Although an affluent private hospital population might have shown better correlation with the Broselow tape, they would have been representative of only the 25% well-nourished or overnourished Indian children. In addition, our analysis of the data on the basis of Broselow-predicted weight groups grounds our recommendations on the only factor available to physicians in an emergency setting: the Broselow weight. Physicians using our data do not need to know the child’s age or measured weight to determine whether the tape is accurate or not. If the child’s Broselow-predicted weight is more than 10 kg, a correction of 10% can be applied.
The Broselow pediatric emergency weight estimation tape accurately predicted weights in Indian children <10 kg, but overestimated weights in Indian children >10 kg by more than 10%. Application of a 10% correction factor to the Broselow-predicted weights for Indian children over 10 kg more accurately predicts their true weights and may reduce medical errors; however, this practice requires prospective validation. Emergency care providers in India and other developing nations should be aware of the potential benefits and limitations of this commonly used pediatric resuscitation tool.
- 25WHO Working Group on Infant Growth. An evaluation of infant growth: the use and interpretation of anthropometry in infants. Bull World Health Organ. 1995; 73:165–74.
- 26India’s Undernourished Children: A Call for Reform & Action. World Bank Report, May 2006., , , , .
- 27The WHO Multicentre Growth Reference Study (MGRS): rationale, planning, and implementation. Food and Nutrition Bulletin 2004; 25(Suppl. 1):S3–S84., , , , .