Model Predicts Which Pediatric ER Patients Likely to Be Admitted
TUESDAY, April 25, 2017 -- A new model can accurately predict pediatric patient hospitalization early in the emergency department encounter by using data commonly available in electronic medical records, according to a study published online April 25 in Pediatrics.
Yuval Barak-Corren, M.D., from Boston Children's Hospital, and colleagues retrospectively analyzed all visits to the Boston Children's Hospital emergency department from July 2014 through June 2015. Half of the data were used to derive a model for early prediction of hospitalizations, and half of the data were used for model validation.
Based on analysis of 59,033 patient visits (11,975 hospitalized cases and 47,058 discharged controls), the researchers found that using data available within the first 30 minutes after presentation, the model identified 73.4 percent of the hospitalizations with 90 percent specificity and 35.4 percent of hospitalizations with 99.5 percent specificity (area under the curve, 0.91). The emergency department could potentially save 5,917 hours per year, or 30 minutes per hospitalization, by applying this model in a real-time setting.
"Such early identification can be used to advance patient placement processes and shorten ED boarding times," conclude the authors.
© 2019 HealthDay. All rights reserved.
Posted: April 2017