Use predictive data analysis to reduce ER overcrowding, study suggests


Hospitals looking to reduce overcrowding in the emergency department should use current emergency department demand data to predict future trends, a study finds.

The research (.pdf), which comes from the Columbia Business School, included an algorithm that can be used to predict future ED trends. Using a simulation, the researchers found that the proposal could reduce delays by as much as 15 percent and may reduce the need to divert patients.

Using data to reduce wait times can lead to cost savings and improved patient satisfaction as well, according to the study, which was published in the journal Manufacturing & Service Operations Management.

“Patients on their way to the emergency room want to know that their emergency is going to be handled as expeditiously as possible,” Carri Chan, Ph.D., associate professor of business at Columbia and one of the study’s authors, said in an announcement of the findings. “Using predictive analytics is a step towards eliminating the overcrowding and long wait times that plague many of today’s emergency rooms, ensuring patients receive the care they need when they need it.”

Overcrowded emergency departments may rely on ambulance diversion to avoid compounding the problem, but that may be potentially dangerous for patients, FierceHealthcare has previously reported.

The study encourages healthcare leaders to be more proactive in making choices that may impact wait times in the ER. For example, begin diverting patients before the emergency department gets backed up, using previous ER use data to predict when those bubbles of use might form.

The predictive analysis used in the Columbia study may be applicable in other areas of healthcare as well, according to the research. Data for discharge and admissions, for instance, is predictable, according to the study, so hospitals may also be able to use similar tactics to improve care quality and efficiency in those areas.