Weight Estimation without Waiting: Design, Development and Testing of a Mobile Application to Measure the Length and Estimate the Weight of New Zealand Children for Advanced Paediatric Resuscitation
MetadataShow full metadata
Background Weight estimation is critical in paediatric resuscitation as the time taken to weigh a child could directly influence their survival and quality of life. Unfortunately, the weight estimation techniques used in New Zealand are not accurate which, impacts the complexity of prescribing medication doses and selecting equipment size used in treatment during paediatric resuscitation. Aim Mobile technology could streamline the process of weight estimation and paediatric resuscitation. Therefore, this research aimed to design, build, test and evaluate a mobile application that will estimate the weight of NZ children in varied environments using augmented reality on a mobile device. Weight estimates using the device aim to be within 10% of the child’s actual weight in 75% or more instances. Methods An adapted design science approach was utilised that included investigation of existing solutions, application design, development, software testing (functional testing with preliminary exploratory testing, user observation, user testing via processes using the “think aloud method”). Regression modelling/equations are developed, tested for fit and compared with existing weight estimation techniques endorsed in NZ. The accuracy was assessed using MPE, limits of agreement and the proportion of weights within a 10%, 20% and 30% of actual weight. The distribution of errors was examined and limitations and future work are specified. Findings The Weight Estimation without Waiting (WEWW) mobile application was designed, developed, tested and evaluated. Even though the WEWW application outperformed (MPE 1.1, SD 23.2) the New Zealand Resuscitation Council (NZRC) (MPE 21.6, SD 16.7) and St John (MPE 18.9, SD 16.9) the accuracy of the WEWW application could be improved. For example, by either transforming the data using distribution of error to improve regression or machine learning models. Users were positive about the application and believe that it is easy to use and would make their weight estimates more reliable. However work still needs to be completed around regulation of the application and accuracy may be improved further in the future by more in-depth analysis of workflows and medication doses to support weight estimation. Conclusion The WEWW application provides a novel method to estimate the weight of children during resuscitation. Even though the WEWW application currently outperforms weight estimation methods endorsed in NZ it has the potential to become even more accurate after further, post-doctorate work around the further development of machine learning and the user interface.