Supervisor behavior emerged as a measurable predictor of physician turnover
In a large academic medical center cohort, physicians who rated their immediate supervisors more favorably were less likely to leave within 2 years. The association was not soft or anecdotal; it was anchored to human resources records and tested against occupational well-being measures that are often used to explain departures. Each 1-point higher score on the Mayo Clinic Leadership Impact Index was associated with a lower risk of objective turnover, with a hazard ratio of 0.89 and a 95% confidence interval of 0.85 to 0.93.
The finding matters because physician turnover is usually discussed as a retention problem, but the underlying mechanism is often behavioral before it is structural. Supervisors shape workload tolerance, psychological safety, recognition, and the day-to-day sense that a physician can remain effective without constant friction. When those conditions erode, burnout may rise, professional fulfillment may fall, and departure becomes more likely. This cohort study connects that lived experience to a hard endpoint: actual exit from the organization.
A cohort built from repeated surveys and employment records
The study followed practicing physicians who completed organizational surveys at a large academic medical center between October 3, 2016, and May 1, 2024. Investigators paired individual survey responses with human resources records to determine objective turnover during the following 2 years. Five survey time points were used, and participation remained above 50% in every survey year, with rates of 53.6% in 2016, 66.8% in 2019, 77.5% in 2020, 69.0% in 2022, and 66.2% in 2024.
Altogether, 7351 survey responses from 3124 unique physicians were analyzed, including 1528 women, who represented 48.9% of the cohort. The repeated-measures structure gives the work a practical advantage: it reflects how physicians’ perceptions and job experiences can change over time rather than freezing them into a single snapshot. That design also makes the turnover finding harder to dismiss as a transient mood effect.
Several validated instruments were used. Physicians completed the 9-item Mayo Clinic Leadership Impact Index, the Stanford Values Alignment Scale, the Stanford Professional Fulfillment Index, and a measure of self-reported intent to leave within the next 2 years. Objective turnover was the main outcome, and multivariable Cox proportional hazards regression models adjusted for age, sex, survey year, burnout, and professional fulfillment. By adjusting for those occupational well-being variables, the analysis asked a sharper question: does supervisor behavior still matter once burnout and fulfillment are already in the model? The answer was yes.
Why leadership behavior can influence departure risk
The clinical significance of the result is best understood at the organizational level. Physicians do not leave only because of pay, schedule strain, or workload, although those factors remain important. Departure often follows a cumulative loss of trust in the immediate environment, especially when day-to-day leadership feels inconsistent, unsupportive, or disconnected from the values clinicians expected when they joined the institution. A physician who sees poor leadership repeatedly may interpret the workplace as one in which professional judgment is discounted and future growth is limited.
That pathway is biologically and psychologically plausible. Chronic workplace strain activates the same stress systems that contribute to burnout, sleep disruption, emotional depletion, and detachment from the work itself. Leadership behavior can either amplify that burden or buffer it. A supervisor who communicates clearly, resolves conflict, and protects time for clinical work may not eliminate institutional pressure, but can reduce the sense that the pressure is personal and inescapable.
Values alignment operates through a related but distinct mechanism. When a physician’s personal values and the organization’s values are aligned, routine hardships may feel more tolerable because the work still feels coherent. When the mismatch grows, even a manageable workload can become psychologically expensive. The study was designed to examine both supervisor behavior and values alignment, which makes sense clinically: turnover rarely follows a single cause.
Objective turnover offers a harder endpoint than intent to leave
The use of human resources records strengthens the study in a way that self-reported intent alone cannot. Physicians may say they plan to leave and then stay, or they may understate their intentions while quietly preparing to depart. Objective turnover avoids that ambiguity by capturing the actual employment outcome. In workforce research, that difference is consequential because institutions make hiring, staffing, and coverage decisions based on real departures, not survey sentiment.
The study also reflects a point that health systems often miss. Retention is not only a matter of recruiting the right people; it is also a matter of whether the current environment makes good physicians want to remain. A favorable score on a leadership scale is not just a satisfaction metric. In this cohort, it tracked with a lower probability of exit over the next 2 years.
What the findings can and cannot support
This is a single-center cohort study, so the results cannot be generalized automatically to every practice setting, specialty mix, or health system culture. Academic medical centers have particular hierarchies, promotion pressures, and administrative structures that may intensify or modify the effects of leadership and values alignment. The study also relies on survey participation, which can leave room for response bias even when participation rates are healthy.
Even with those caveats, the direction of effect is difficult to ignore. Physicians who viewed their immediate supervisors more positively had lower turnover risk, and that association persisted after accounting for burnout and professional fulfillment. For leaders trying to reduce attrition, the message is concrete: the daily behavior of a named supervisor is not a soft variable sitting at the edges of workforce management. It is a measurable feature of retention risk.
The abstract does not provide all of the values-alignment estimates in the text available here, so those effects should not be inferred beyond what was explicitly reported. What is already clear is enough to change the frame of the discussion. Retention strategies that focus only on hiring bonuses, recruitment pipelines, or downstream exit interviews miss a proximal driver operating much earlier in the employment relationship: the physician’s immediate leader.