Clinical Rationale/Problem Solved
Aim 1: Wrong weight data entered into the EHR is very hard to detect and leads to potentially catastrophic dosing errors.
Aim 2: Weight data errors are amenable to machine learning, but it can be very time consuming to build the training sets needed to employ ML.
Potential Impact
Aim 1: Prevents overdosing because of bad weight data.
Aim 2: Cutting edge use of ML, more robust detection capabilities.
DSAW Investigators
S. Andy Spooner (co-PI), Danny Wu (co-PI, UC), Eric Kirkendall (co-I), Kevin Dufendach, Josh Courter
Collaborators
Yizhao Ni, Pieter-Jan VanCamp
Grants
Mary K. Logan Research Award, Association for the Advancement of Medical Instrumentation (AAMI)
Publications
Hagedorn PA1, Kirkendall ES2, Kouril M3, Dexheimer JW4, Courter J5, Minich T5, Spooner SA6. Assessing Frequency and Risk of Weight Entry Errors in Pediatrics. JAMA Pediatr. 2017 Apr 1;171(4):392-393.