Development and validation of a method to estimate body weight in critically ill children using length and midarm circumference measurements: The PAWPER XLMAC system

• Main Authors: M Wells, L Goldstein, A Bentley
• Format: Article
• Language: English
• Published: Health and Medical Publishing Group 2017-11-01
• Series: South African Medical Journal
• Online Access: http://www.samj.org.za/index.php/samj/article/download/12106/8278
• ISSN: 0256-9574; 2078-5135

Background. Erroneous weight estimation during the management of emergency presentations in children may contribute to patient harm and poor outcomes. The PAWPER (Paediatric Advanced Weight Prediction in the Emergency Room) XL tape is an accurate length-based, habitus-modified weight estimation device, but is vulnerable to errors if subjective visual assessments of children’s body habitus are incorrect or erratic.Objective. Mid-arm circumference (MAC) has previously been used as a surrogate indicator of habitus, and the objective of this study was to determine whether MAC cut-off values could be used to predict habitus scores (HSs) to create an objective and standardised weight estimation methodology, the PAWPER XL-MAC method.Methods. The PAWPER XL-MAC model was developed by creating MAC ranges for each HS in each weight segment of the tape. This model was validated against two samples, the National Health and Nutrition Examination Survey datasets and data from two previous PAWPER tape studies. The primary outcome measure was to achieve >70% of estimations within 10% of measured weight (PW10 >70%) and >95% within 20% of measured weight (PW20 >95%) for children aged 0 - 18 years.Results. The PAWPER XL-MAC model achieved very high accuracy in the three validation datasets (PW10 79.2%, 79.0% and 81.9%) and a very low critical error rate (PW20 98.5%, 96.0% and 98.0%). This accuracy was maintained across all ages and in all habitus types, except for the severely obese.Conclusions. The PAWPER XL-MAC model proved to be a very accurate, fully objective, standardised system in this study. It has the potential to be accurate across a wide variety of populations, even when used by those not experienced in visual assessment of habitus.