Challenge:

Like many care facilities in Norway, Munkvoll Health and Welfare Center in Trondheim faces the perennial issue of high short-term sick leave, especially during flu season and around holidays. These unpredictable absences often lead to frantic calls and messages to off-duty staff, creating stress for both managers and employees, and risking understaffing during critical times.

Traditional scheduling tools lacked the foresight needed to proactively address staffing shortfalls, leaving unit leaders with limited options beyond reactive planning.

Foto: Ole Martin Wold, Enhetsleder Jorun Slettahjell får hjelp fra programvare basert på kunstig intelligens når neste års turnus skal planlegges. 

Solution:

To tackle this, Munkvoll began piloting SynPlan, an AI-based planning tool developed by NTNU-based entrepreneur Hai Thanh Nguyen in collaboration with Trondheim Municipality. The software uses machine learning to analyze up to five years of anonymized absence data, identifying patterns related to short-term leave such as self-certified illness and sick children.

Instead of simply reacting to gaps, SynPlan forecasts absentee trends and flags vulnerable shifts up to a year in advance, offering guidance on where and when to consider increasing staffing.

Results:

80% Accuracy in Absence Forecasting

  • According to unit leader Jorun Slettahjell, SynPlan’s predictions are correct about 80% of the time.

  • The AI system often flags shifts at risk of understaffing that wouldn’t have been obvious through traditional planning.

Proactive Resource Allocation

  • Instead of rigidly “topping” every flagged shift, team leaders evaluate flagged slots with SynPlan’s help to prioritize where additional staff are truly needed.

  • About half of the predicted high-risk shifts are proactively bolstered with extra staff—shifting budget from overtime to planned labor.

Reduced Administrative Load

  • Michele Silva, head nurse and department manager, notes a significant drop in last-minute phone calls and messages to cover for sick-leave:

“I spend less time calling around trying to fill shifts. It’s a change for the better.”

  • Slettahjell adds:

“Now I can spend more time on leadership and strategic tasks. Staff also appreciate less hassle during their time off.”

Improved Planning and Staff Morale

  • By narrowing the gap between planned and actual staffing, more shifts are now covered within regular working hours, reducing strain on staff and boosting morale.

  • The system encourages evidence-based planning discussions, where leaders decide collaboratively which shifts to reinforce based on data.

Positive Trends in Sick Leave

  • Early indications show a reversal of the upward sick leave trend, thanks to better foresight and improved planning.

Quote:

“We don’t aim to save money directly—we aim to create better plans. And we’ve seen a reduction in the difference between planned and actual staffing.”
— Gørild Brekke, Workforce Planning Advisor, Trondheim Municipality

“It’s a relief for staff not to be constantly contacted for extra shifts during their time off. This brings more predictability and less stress.”
— Jorun Slettahjell, Unit Leader, Munkvoll Health and Welfare Center

https://sykepleien.no/2023/10/bruker-kunstig-intelligens-til-forutse-sykefravaer