Navigation

Recent Publications

More...

Home

Improve healthcare operations using simulation analytics

oyhy's picture

oyhy - Posted on 30 October 2017

Project Description: 

We propose a simulation analytics method to evaluate the health care operations in a more detailed way: with this method, we would be able to predict the transient system performance based on the given system state at a given time. The idea is to identify whether the healthcare system would perform worse than some specified rules in the near future, and if it would what procedures will best get the system out of trouble. Take the ED blocking problem as an example, we predict the probability of blocking in ED for the next few hours, and if the probability is higher than some threshold we will suggest some immediate actions that would decrease the blocking probability under the specified threshold.

Research Project Details
Project Duration: 
10/2017 to 09/2019
Project Significance: 
With the research work, we could develop an online decision tool for healthcare operators. The healthcare operators may have some daily operation objectives that they would like to achieve. Previous operations research studies may have suggested some optimal or good policy based on the aggregated performance measures. Our research aims to predict the system performance from big data generated from simulation, without interrupting the system operations. We could provide conditional transient prediction on system performance real time quickly. Besides, for any given system state, we would be able to detect system risk in advance and suggest best solutions based on our analytics models. This simulation analytics method would be able to improve the health care operations in many circumstances, such as reducing ED blocking probability, improving ED to impatient unit transfer efficiency, predicting intensive care unit patient LOS based on current situation, etc. Furthermore, simulation analytics, proposed by Barry Nelson, is a new methodology in the operations research area that explores information generated during simulation for prediction and control. This is a complete new area and this research will provide significant theoretic contribution to the development of concepts, models, and applications of simulation analytics.