A predictive staffing model developed at Manning Family Children's has boosted the hospital's nursing workforce by 66%, essentially eliminating the use of agency nurses. The model is the result of a unique collaboration among nursing leadership and data scientists.
“The common ground for us was the data,” said Lisa Labat, RN, BSN, MBA, NEA-BC, the hospital’s assistant vice president of patient care services. “Having that shared language enabled us to get on the same page to address our needs.”
The hospital began building the tool in the wake of nursing shortages exacerbated by COVID-19 to anticipate workforce gaps and hire nurses before they occur. To do so, the team analyzes data across these standard metrics:
- Census. A 12-month rolling patient census by unit provides a snapshot of recent trends. Users have the option of including actual or budgeted census figures in the calculation.
- WHPUOS. The calculation of worked hours per unit of service — basically, total staff hours divided by the services they provide over a certain period — accounts for the different staffing needs across various units.
- Turnover. This unit-level measurement of nurses who’ve left the organization provides insight into retention rate trends.
An additional metric is key to the predictive tool’s value: internal transfer data. Using its HR vendor platform, the hospital assesses staff movement throughout the organization — information not represented in turnover figures but integral to understanding staffing trends from unit to unit.
“It's really important because our primary goal is to keep our nurses at our hospital,” Labat said. “We want them to grow and advance as nurses, so we are pretty open about transfers, and we know we're going to fill those positions because they’re captured in the data.”
The advantage of a manual model
Hospital leaders didn’t know the extent to which the predictive staffing tool would be used. So, they simply developed it as a series of Microsoft Excel spreadsheets rolling up to an executive dashboard. Over time, the team has added layers of iterations and adjustments to the model.
“It’s a beast of a tool,” said Camille Messa, MHA, senior practice administrator. “It’s got a million formulas and is very manual.” Messa acknowledged that migrating the model to a software platform where it can be automated is “a lofty goal,” but the current arrangement has its advantages.
Unusual situations — such as a nurse who splits time across multiple units — are more easily recorded in the current tool. The team’s familiarity with its complexities means they can more easily spot inconsistencies in the data. “One time, our HR platform reported turnover in triplicate,” Labat said. “I do the exit interviews, so I knew that was way higher than it should have been.”
Streamlining the hiring process
When hospital’s nursing leaders, HR representatives, and executives meet monthly to discuss workforce planning, the predictive staffing model guides their decisions. It’s also integral in matching students from the hospital’s revamped nurse tech program with anticipated openings. Because the data is regularly refreshed and analyzed, the model’s accuracy has earned the trust of hospital leadership. Though it’s not the only factor the organization uses to manage its nursing workforce, that confidence has led to faster hiring decisions and fewer vacancies.
“It's one data point in the process,” said Lindsey Casey, MSN, RN, NEA-BC, senior vice president and chief nursing officer. “But it's given us a really extraordinary place to start. It’s our source of truth.”