CORKSCREW 2013 CORK study of children's realistic estimation of weight

A Comparison of the Performance of Different Age-Based Paediatric Weight Estimation Formulae in Hong Kong Children

Objective

To evaluate the performance of five existing age-based weight estimation formulae - the original and updated Advanced Paediatric Life Support (APLS) formulae, Luscombe formula, Best Guess formula, Chinese Age Weight Rule (CAWR) - as well as a new two-part weight estimation formula, the Chinese Age Weight Rule-3 (CAWR-3), in Hong Kong children presenting to the emergency department (ED).

Methods

Data based on children aged 1-12 who presented to the ED of a tertiary referral centre in Hong Kong over a six-month period. Actual weight was compared against estimated weight from the weight estimation formulae. Level of agreement was assessed by Bland-Altman analysis using mean percentage difference (MPD) and 95% limits of agreement (LOA). Root mean squared error (RMSE) and proportions of weight estimates within 10%, 15% and 20% of actual weight were calculated.

Results

A total of 4600 children were included. The CAWR-3 outperformed the five existing weight estimation formulae. The CAWR-3 had the least MPD in age 4-12 (MPD +3.2%), as well as the second least MPD in age 1-12 (MPD -0.7%). The CAWR-3 had narrowest 95% LOA in age 1-3 (95% LOA -32.6% to 21.9%) and the second narrowest 95% LOA in age 1-12 (95% LOA -37.5% to 36.1%). The CAWR-3 had the smallest RMSE of 6.33 kg in age 4-12 and the smallest RMSE of 4.90 kg in age 1-12. Furthermore, the CAWR-3 had the highest proportion of weight estimates within 10%, 15% and 20% of actual weight.

Conclusion

The CAWR-3 outperforms the five existing age-based weight estimation formulae in Hong Kong children presenting to the emergency department.

The accuracy of emergency weight estimation systems in children-a systematic review and meta-analysis

The safe and effective administration of fluids and medications during the management of medical emergencies in children depends on an appropriately determined dose, based on body weight. Weight can often not be measured in these circumstances and a convenient, quick and accurate method of weight estimation is required. Most methods in current use are not accurate enough, but the newer length-based, habitus-modified (two-dimensional) systems have shown significantly higher accuracy. This meta-analysis evaluated the accuracy of weight estimation systems in children. Articles were screened for inclusion into two study arms: to determine an appropriate accuracy target for weight estimation systems; and to evaluate the accuracy of existing systems using standard meta-analysis techniques. There was no evidence found to support any specific goal of accuracy. Based on the findings of this study, a proposed minimum accuracy of 70% of estimations within 10% of actual weight (PW10 > 70%), and 95% within 20% of actual weight (PW20 > 95%) should be demonstrated by a weight estimation system before being considered to be accurate. In the meta-analysis, the two-dimensional systems performed best. The Mercy method (PW10 70.9%, PW20 95.3%), the PAWPER tape (PW10 78.0%, PW20 96.6%) and parental estimates (PW10 69.8%, PW20 87.1%) were the most accurate systems investigated, with the Broselow tape (PW10 55.6%, PW20 81.2%) achieving a lesser accuracy. Age-based estimates achieved a very low accuracy. Age- and length-based systems had a substantial difference in over- and underestimation of weight in high-income and low- and middle-income populations. A benchmark for minimum accuracy is recommended for weight estimation studies and a PW10 > 70% with PW20 > 95% is suggested. The Mercy method, the PAWPER tape and parental estimates were the most accurate weight estimation systems followed by length-based and age-based systems. The use of age-based formulas should be abandoned because of their poor accuracy.

It is time to abandon age-based emergency weight estimation in children! A failed validation of 20 different age-based formulas

AIM

When children's weight cannot be measured during medical emergencies, it must be estimated, as weight is required for many interventions. Age-based formulas are the oldest weight estimation systems in children, but have been shown to be very inaccurate in many studies. This aim of this study was to evaluate the accuracy of age- and length-based formulas and to see if a measure of body habitus could be used to improve the performance of these formulas.

METHODS

This was an analysis of data from a sample of 1085 children aged from 1 month to 16 years, collected from four Emergency Departments in Johannesburg, South Africa. Basic demographic and anthropometric data was collected and each child had a visual assessment of body habitus, quantified as a habitus score. Weight estimates from 20 existing age-based formulas and two length-based formulas were then compared against measured weight to determine their accuracy. Age- and length-based, habitus-modified models were developed and similarly evaluated.

Finger counting method is more accurate than age-based weight estimation formulae in estimating the weight of Hong Kong children presenting to the emergency department

OBJECTIVE

The aim of the present study was to evaluate the finger counting method and compare its performance with four commonly used age-based weight estimation formulae in children aged 1-9 years presenting to the ED in Hong Kong.

METHODS

A cross-sectional, observational study of children aged 1-9 years who presented to the ED of a tertiary referral hospital in Hong Kong over a 6 month period was conducted. Actual weight was compared with estimated weight using the finger counting method and four commonly used age-based weight estimation formulae. Bland-Altman analysis was performed to evaluate the degree of agreement in which the mean percentage difference (MPD) and 95% limits of agreement (LOA) were calculated. Root mean squared error (RMSE) and proportions of weight estimates within 10%, 15% and 20% of actual weight were determined.

RESULTS

A total of 4178 children were included. The finger counting method was the most accurate method (MPD 0.1%; 95% LOA -34.0% to 34.2%). The original Advanced Paediatric Life Support (APLS) formula (MPD -7.0%; 95% LOA -38.4% to 24.3%) and the updated APLS formula (MPD -0.4%; 95% LOA -38.5% to 37.8%) underestimated weight whereas the Luscombe formula (MPD 7.2%; 95% LOA -31.8% to 46.2%) and the Best Guess formula (MPD 10.6%; 95% LOA -27.3% to 48.4%) overestimated weight. The finger counting method had smallest RMSE of 4.06 kg and estimated the largest proportion of children within 10%, 15% and 20% of actual weight.

CONCLUSION

The finger counting method outperforms the commonly used age-based weight estimation formulae in children aged 1-9 years presenting to the ED in Hong Kong.

Comparison of the finger counting method, the Broselow tape and common weight estimation formulae in Filipino children after Typhoon Haiyan

OBJECTIVES

We sought to evaluate commonly used paediatric weight estimation techniques in a sample of children in the Philippines.

METHODS

We prospectively collected age, height and weight data for a sample of 207 children aged 1-9 years seen during a medical aid trip. Weights were estimated using the finger counting method, the Broselow method and four formulae. Bland-Altman analysis was performed to evaluate agreement with measured weight.

RESULTS

Mean difference and range of agreement in kilograms were as follows: 0.6 (95% CI 0.1-1.1) and 14.9 (95% CI 13.1-16.7) for the Broselow method; 1.1 (95% CI 0.5-1.7) and 17.3 (95% CI 15.2-19.3) for the traditional APLS formula; 3.1 (95% CI 2.4-3.7) and 18.6 (95% CI 16.4-20.8) for the finger counting method; 4.0 (95% CI 3.2-4.8) and 23.1 (95% CI 20.3-25.8) for the updated APLS method; 5.1 (95% CI 4.4-5.8) and 21.1 (95% CI 18.6-23.6) for the Luscombe formula; and 5.3 (95% CI 4.5-6.1) and 22.5 (95% CI 20-25.2) for the Best Guess formulae.

CONCLUSIONS

The Broselow tape and the traditional APLS formula performed best in our sample. The finger counting method also outperformed newer weight estimation formulae. 'Updated' age-based formulae created recently in developed countries should not be used in disaster relief efforts in the Philippines. Caution should be used when applying these formulae to other developing countries and in disaster response.