Nurse teams around the world are facing burnout and shortage — straining care delivery, and posing a meaningful challenge to health system margins. An often-ignored but pivotal driver of the problem is provider scheduling. Building shift schedules is extremely time-consuming, and often inefficiently or inequitably allocates already-stretched nurse teams.
Design partnership for scheduling technology
Assuta Medical Centers is a health system in Israel with eight facilities that perform a total of 100,000 procedures a year. As part of a series of operational initiatives, Assuta sought a way to systemize nurse scheduling to improve workforce productivity.
In 2023, Assuta, in an effort supported by nursing leadership, operational leadership, and its innovation group RISE, formed a design partnership with a new healthcare AI company, In-House Health, which was developing automations for clinical schedules and clinical workload prediction.
The goals of the partnership were to:
- Develop a novel algorithm to predict unit-level clinical workload based on clinical data
- Create a platform for managers to easily update schedules based on predictions
- Improve the accuracy of clinical schedules
After validating the algorithm and scheduling application through a series of sessions with nursing leadership and management at Assuta, In-House’s platform was rolled out to the network, beginning in four units at the main facility: Orthopedics, General Surgery, Urology and Cardiothoracic.
The path to implementation was quick: after an initial 2-week training and testing period, the four units went live on the platform and began using it to make and adjust schedules. After the second month, users reported an average 28% reduction in the difficulty of “building and adjusting schedules”, and an average 14% increase in “confidence that the schedule is accurate to patient care needs”.
Schedules that respond to workload
At the core of In-House’s tech is an algorithm that predicts clinical workload by unit to automatically tailor schedules to patient care needs. Across inpatient care, many units schedule to a “fixed staffing grid”, which follows an unchanging pattern between weeks.
In practice, most units experience variation in census and acuity, not to mention employee-driven changes such as callins. As a result, nursing teams spend significant time updating schedules in reaction to changes, by removing and adding staff, sometimes at the last minute.
The main engine of the In-House platform takes clinical data and uses machine learning to predict the needs on future shifts. This allows nursing teams to plan further in advance, with specific details on what will drive the workload and employee availability in the future. On top of this, In-House tracks and considers staff preferences, availability and reliability, so that shift fills are properly prioritized and equitably distributed.
Reducing complexity and effort in scheduling
Each unit has particular patient mix factors that drive scheduling complexity. These are already factors that are fairly intuitive, even second-nature, to the relevant managers, who try to incorporate as much as possible when making scheduling decisions. However, solving for so many factors at once is often unrealistic and challenging, particularly alongside their other responsibilities, which may also include as acting as charge nurse during the day shift.
For example, high volume days or weeks with specialized, complex procedures, such as bladder removal surgery in Urology, can result in significantly higher clinical workload per patient. In such days and weeks, Urology benefits from having a relatively more tenured team on the floor. As another example, Assuta’s Cardiothoracic unit is often responsible for several patients at an intermediate/PCU level of care, which can drive more demanding shifts.
As explained by the unit manager for Urology: “In-House has transformed our scheduling process. The platform's predictions are highly accurate, and updating schedules is much easier now. In just a few weeks, we’ve seen a noticeable difference in efficiency and team morale."
From unit-level improvement to a full operating system
While improving accuracy of clinical schedules helps relieve some of the strain on nursing teams, it is just the beginning. With Assuta and others, In-House has developed additional tools that build on workload prediction, such as facility-level float pool management (used by other roles, Director of Nursing and Float Pool Director).
The mounting problem of burnout and churn in the healthcare workforce, particularly inpatient nursing, demands increasingly thoughtful solutions, particularly to keep up with rising patient demand. This means continuing to evolve the models for employment, staffing and scheduling, incorporating alternative care delivery models, and enabling dynamic scheduling across region with models like inter-facility float.