Skip to main content

Data Analytics Helped Mercy Slash $4.3M in Nursing Costs

Analysis  |  By Alexandra Wilson Pecci  
   August 28, 2017

Combining data from the hospital's existing SaaS scheduling application with predictive analytics is helping schedulers identify and stop nurse leakage.

At Saint Louis-based health system Mercy, like most other hospitals and health systems, nursing labor is one of the top expenses.

According to Curtis Dudley, vice president of Mercy's integrated performance solutions, that number tops $800 million annually.

Executives realized that gaining control of the cost was imperative. It had full-time equivalent (FTE) leakage among nurses across its 630 units in 44 hospitals that reached 80,000 hours per year and forced an overreliance on agency nurses to fill the gaps.

The solution was a tool that combines data from the hospital's existing software as a service (SaaS) scheduling application with predictive analytics to help schedulers identify leakage and fill those gaps with their own employees.

This tool has saved $4.3 million since it was deployed in September 2016, Mercy reports.

Mike Gillen, vice president of Mercy's system of operations, says using technology made sense here because of the great variation in scheduling practices across the organization, not to mention the "sheer magnitude" of the work involved with existing templates and paper tools.

Mercy knew it needed to not only improve its scheduling processes, but to do it in a uniform way.

It also knew that it needed to start where the nurses already were.

"We know that integration into the workflow is the key," Dudley said.

Starting At Home

Mercy built the solution in-house on top of its existing scheduling platform.

It aggregated data from the nurse scheduling tool, then used predictive analytics to project how many FTE hours would be needed based on predicted patient volumes and historical data such as hospital activity during certain days or times of year.

Nurse schedulers can now access a dashboard showing them how their schedule compares to predicted needs and where leakages are likely to occur.

According to Dudley, the tool allows users open a number of different looks to drill into data specific to leakage and staff cancellations. Users also can easily view when there's a leakage on a unit.

"It all comes back to having that data, having that analytics right in front of those nurse leaders," says Gillen.

Looking at cancellations allows schedulers to see which individuals among the core staff are leaking at any given time, and allows them to be redeployed in schedule before a pay period ends.

This ensures that all staff is being used to their full-time commitment and allows nurse leaders to see what gaps remain afterwards.

Working With Nurses for Solutions

In designing the tool, Dudley says nurse leaders met regularly with the IT staff to identify which metrics were most meaningful, and how they should be calculated. They would build, test, and tweak versions based on feedback in "30-day sprints."

"There were many things we tried that we thought were good ways to look at things that nurses reacted negatively to," Dudley says. "You design your delivery to the audience, and the nursing audience is different than the lab audience, or the physician audience."

Gillen says nurse schedulers were using the tool to check for leakage and make modifications for staff, but because of the way the data feeds were updated—just twice a day—nurse leaders would fix a schedule only to have to redo it later.

Duplicate work was required just because of timing of the data feeds. So they worked with the vendor of scheduling application, to change the data feed from twice a day to five times a day.

"It was a lesson learned for us," Gillen says.

In September 2016, a couple of years after it was first conceptualized, the project was officially deployed. It grew so quickly, that it was eventually moved from Mercy's general database to an in-memory platform that is faster and more efficient.

In addition to the cost savings, there's anecdotal evidence that nurses are more satisfied because their schedules are more thoughtfully created. Unit turnover has improved and will nurse satisfaction will be tracked more formally soon, Gillen says.

Staff Education Critical

Gillen says that they learned that leakage is somewhat of a "hidden metric," and Dudley agrees.

"I think a lot of nursing leadership would have said there's no opportunity here," Dudley says, believing that managers were already doing a great job with scheduling.

Showing them the data, though, surprised them and showed that "there's a bigger opportunity here than you might think."

Both Dudley and Gillen also agree, though, that if they had to do it over again, they would have had more education about leakage upfront, and consider more deeply how scheduling fits into users' workflow.

Once the users saw how the tool not only improved scheduling, but also freed them up to do other things, there was more significant buy-in.

"Don't underestimate the change-management requirements to make this work effectively," Gillen advises. Nurse managers are incredibly busy, so "anytime we start messing with the scheduling and the staffing… we have to know the impacts upfront."

Overall, the tool has been a huge money and time saver, and Gillen and Dudley agree that there are even more opportunities waiting to be uncovered.

"We [have just begun] to scratch the surface," Dudley says.

Alexandra Wilson Pecci is an editor for HealthLeaders.

Tagged Under:


Get the latest on healthcare leadership in your inbox.