Yale researchers claimed that they have created a machine-learning model that could predict which physicians will leave the job and what contributes to their departure in light of the recent rise in physician burnout and their eventual departure in healthcare facilities.
In the study, which was released on February 1 in the journal PLOS One, electronic health records (EHRs) from more than 300 doctors representing 26 medical specialties were examined over the course of 34 months. Physician usage patterns, clinical productivity indicators like patient volume, and demographic information like age and job history were all included in the data. Nearly 14% of physicians left their positions throughout the study period.
Which doctors would leave within six months could be predicted by the machine-learning algorithm with 97% accuracy. The AI-based model also determined which variables changed as a doctor’s probability of leaving increased from low to high and how strongly these variables affected turnover risk.
According to the researcher, the length of time a physician had been working, their age, the complexity of their cases, and the demand for services were the top four indicators that indicated departure risk.
Another study that offered light on the issue of physician leaving and was released in October 2021 found that less time spent on the EHR, particularly inbox management, was related to physician departure, enabled only evaluations of linear connections, and several other factors were ignored.
According to Yale researchers, this machine learning technique enabled researchers to more nuancedly observe the difficulties faced by clinicians. For instance, the risk of leaving was higher for doctors who had been employed more recently and for those with longer tenures, but it was lower for those with a more average length of tenure. The probability of leaving was also higher for physicians under 44, lower for those between 45 and 64, and again higher for those 65 and older.
Variables interacted with one another as well. For instance, for doctors who had been employed for less than ten years, spending more time on EHR-related tasks reduced their likelihood of leaving. But the chance of leaving increased for doctors who had been working there for a longer period of time.
Robert McLean, MD, medical director of Northeast Medical Group, stated in the report that “physician burnout is an increasingly recognised problem, and healthcare systems, hospitals, and large groups need to figure out what they need to do to ensure the emotional and physical health and well-being of the doctors and other clinicians who do the actual caring for patients.”
Many healthcare systems already have wellness officers and committees that might be tasked with gathering, analysing, and drawing conclusions from this data. These conclusions would then result in implementation plans for adjustments and, ideally, improvements, McLean continued.
The researchers reasoned that since physician turnover affects patients negatively and is expensive for healthcare institutions, understanding the factors that influence physician leave could enhance job satisfaction and prevent costly turnover.
