Improving Nurse Staffing Schedules Using Responsible and Ethical Machine Learning
In our presentation, Indiana University Health (IUH), the largest hospital system in Indiana, has endeavored to partner with an Artificial Intelligence (AI) first, digital engineering company to develop responsible and ethical machine learning models that combine algorithms to predict the number and type of patients that will be admitted on a daily basis and combine that understanding with associated work drivers post admission that will be used to accurately estimate nursing staffing needs with no compromise in patient care. We will demonstrate how we leveraged organizational change management best practices to drive ownership and buy-in from nursing and how the partnership between nursing and the IUH Digital Transformation Leadership was critical in gaining adoption and commitment to improving the models over time. Lastly, we will share our approach, feature sets, results, and lessons learned as we tune the models and optimize nurse demand staffing and how this was coupled with nursing supply models to optimize results in near real-time.
- Design a demand-based system that takes into consideration actual work drivers to adjust supply-side nurse scheduling systems using machine learning techniques and associated models
- Maximize the use of best-practices organizational change management practices to drive maximum support and end-user satisfaction and support
- Discuss lessons-learned in an attempt to improve effectiveness, efficiency, and impact over time to incorporate into new initiatives in the future