Abstract: Infants undergoing a palliative surgery series for Single Ventricle Disease are at risk of sudden, potentially fatal health deterioration. Remote monitoring using the CHAMP App has been shown to effectively eliminate the likelihood of fatality from these events, however the potential for related morbidity persists. This work presents two machine learning approaches to predict such events via resulting unplanned hospital readmissions in an effort to enhance patient review, and in so doing to improve outcomes.
Describe the new knowledge and additional skills the participant will gain after attending your presentation.: This work presents a practical example of how machine learning can be used to enhance existing workflows rather than to replace them, as well as how such enhancement can lead to better patient outcomes. Moreover, by contrasting two different machine learning approaches, it demonstrates how multiple techniques might be attempted in tandem, both to better understand the final results and to approach an optimal solution.
Christine Allen (Presenter)
University of Washington
Vikas Kumar (Presenter)
Carly Eckert, KenSci
Ankur Teredesai, KenSci
Amy Ricketts, Children's Mercy Hospital
Lori Erickson, Children's Mercy Hospital
Peter Churchill, Children's Mercy Hospital
Jennifer Marshall, Children's Mercy Hospital
Girish Shirali, Children's Mercy Hospital