Virtual Fall Summit Encore 2023-Predicting Turnover

Predicting Turnover - Dr. Hu

Pdf Summary

Dr. Xi (Sisi) Hu from Harvard University discusses the use of machine learning to predict turnover risk and clinician burnout. The current solution of engagement surveys provides an incomplete view of clinician retention and well-being issues, as only 30% of clinicians respond to these surveys. Healthcare organizations need reliable and representative data for effective planning. Dr. Hu presents a methodology that leverages existing organizational data to quantify the costs of clinician burnout, setting the foundation for investing in clinician well-being. The study also explores ethical implications and the ROI of well-being programs. It is found that 64% of clinicians are burnout nationally and 1 million plan to leave the healthcare industry in the next two years. Additionally, 114 health systems struggle to get ahead of turnover. A case study is presented, demonstrating the effectiveness of predictive analytics for hotspot prioritization. The study also estimates the comprehensive costs of clinician burnout and turnover at a health system, which helps in planning and justifying investments in well-being programs. Various economic models are used to calculate cost savings opportunities and ROI for specific interventions. The study also highlights ethics concerns related to machine learning, such as fairness, bias, privacy, and accountability. A framework is proposed to address these concerns, including transparency, robustness, fairness, privacy, accountability, human control, and safety. Dr. Hu concludes by offering her expertise in turnover risk modeling and calculating health system-specific burnout and turnover costs.

Keywords

machine learning

predict turnover risk

clinician burnout

engagement surveys

clinician retention

healthcare organizations

costs of clinician burnout

investing in clinician well-being

predictive analytics

ethics concerns