Case Study: Regional hospital in Georgia

Published by Hiro Kawashima on September 01, 2018

Client Background

The emergency department of a regional hospital in Georgia serves a 20-mile radius.  In line with common industry practice, the department had long operated with a fixed schedule approach, primarily 12-hour shifts at 6 am and 6 pm, the same every day of the week.

In 2015 visit volumes shifted to earlier in the day, along with an increase in patient severity and behavioral health cases, requiring more nursing resources than were being scheduled at those times.  This caused delays in treatment, greater length of stay and increased patient ‘boarding’, starting late morning and extending through the late evening.

“Often, by the time the mid-day staff arrived demand had built too rapidly and high to possibly be handled by the team.  The queue for service that regularly developed usually could not be relieved until evening time.” – Emergency Department Director  

Nursing hours were not being scheduled at the right time, and ‘day-of’ staffing adjustments were not adequate to address the issue because they could not be made quickly enough to prevent a large back log.  This resulted in increased costs, decreased staff and patient satisfaction, and decreased patient care quality.

Adding to this was the difficult issue of scarcity in nursing resources across the U.S., especially in rural locations.  The trained and experienced Emergency nurses required in this location are no exception.  Vacancies could last months, and Agency staff of good caliber were hard to locate and expensive.  With limited ability to obtain new nursing staff, any scheduling changes had to limit the potential turnover effects of those changes.

Client urgently needed to improve the alignment of staffing levels to match the dynamic visit volumes, using current FTE’s, while maintaining productivity targets.


Prescience Health’s AgileCast™ Predictive Analytics engine, was able to detect and forecast the Time of Day demand patterns within each Day of Week.  Management was confidently able to adjust the scheduling and re-distribute staff by time of day and day of week and add a new 7-hour late-night shift.  This re-allocation brought team members in incrementally, ahead of the mid-day build to mitigate queuing, and in the early afternoon to resolve any residual queue. 

“I found I could move to a base schedule of staggered starts.  I was able to see how to make available adequate staff resources to meet daily upcoming and continuing patient volumes.” – Emergency Department Director

Additionally, medical care coverage was confidently adjusted to shift the hours of Non-Physician Providers, to provide more support for physician coverage in line with demand patterns. Coverage moved to an earlier start, and an afternoon Non-Physician Provider shift was added.


During the 18-month period following the implementation of scheduling changes suggested by Prescience Health’s AgileCast Predictive Analytics engine, there was an approximately 15% improvement in several departmental key performance indicators, including the rates of Left Without Being Seen and Left Against Medical Advice, and most importantly Average Length of Stay.

Next post: Achieving 91% forecasting accuracy

Take the guesswork out of staffing

Schedule a demo and learn more about how predictive analytics can help you staff better