Predicting patient discharges

To optimize hospital operations, we use a machine learning model to improve patient discharge predictions. This is critical to cost saving and efficient facility management in order to improve overall healthcare delivery. Doctors and nurses often have to estimate discharge times but these estimations are not always accurate, leading to a suboptimal utilization of beds and other patient space.  MODL, directed by Sauleh Siddiqui, is collaborating with other researchers at JHU and the Sean Barnes-led team at the University of Maryland Smith School of Business in developing this predictive tool.

Relevant publications

  • Barnes, S., Toerper, M., Hamrock, E., Siddiqui, S., Levin, S., 2015, “Real-time prediction of inpatient length of stay for discharge prioritization,” Journal of the American Medical Informatics Association, DOI: 10.1093/jamia/ocv106, August 2015