Better Supervisor Leadership Was Linked to Lower Physician Turnover
Physicians were less likely to leave their organizations when they rated their immediate supervisors more highly on leadership behavior. In a Stanford Medicine analysis published Oct. 14 in JAMA Network Open, a 1-point improvement in leadership behavior scores was associated with an 18% decrease in physician turnover over the following two years.
The finding fits a familiar clinical workforce reality: people rarely leave solely because of salary or schedule. Poor management, weak communication and inconsistent decision-making can erode trust faster than any formal retention program can repair it.
Researchers drew on 7,351 survey responses from 3,124 physicians across five survey periods spanning 2016 through 2024, then matched those responses with human resources records to capture actual departures rather than stated intentions alone. The design matters because voluntary turnover is often discussed in broad terms, yet real employee movement gives a much sharper measure of organizational stability.
Values Alignment Mattered Early, Then Lost Statistical Weight
A second exposure also tracked with lower turnover in simpler analyses: alignment between physicians’ personal values and the organization’s values. Each 1-point improvement in values alignment was associated with a 14% reduction in turnover over the following two years.
The association did not survive the fully adjusted model. After accounting for age, sex, survey year, burnout and professional fulfillment, higher leadership ratings still tracked with lower turnover, but the values-alignment signal no longer remained. In practical terms, the supervisor relationship appears to carry a more durable retention effect than abstract cultural affinity once other physician-level factors enter the model.
That pattern makes biological and behavioral sense in a workplace setting. Leadership behavior shapes the daily environment in which physicians practice: workload expectations, responsiveness to conflict, recognition of effort and whether concerns can be raised without penalty. Those forces influence morale continuously, and repeated friction can push a clinician toward departure long before a formal exit plan begins.
Voluntary Departures Were Uncommon, But Intent to Leave Still Predicted Risk
Across the cohort, the overall two-year voluntary turnover rate, excluding retirements, was 5.0%. Most physicians who voiced some level of intent to leave did not actually exit during the observation period, which shows how often frustration remains unconverted into action.
Still, intent was not meaningless. Physicians reporting moderate intent to leave were more likely to depart, but only 8% actually did so within two years. Among those who reported likely or definite intent to leave, 17% departed in that same window. The gradient is clinically useful for organizations because it identifies a group that deserves attention before the resignation lands on an administrator’s desk.
One sentence captures the operational risk: intent is an early warning, not a guarantee.
Why Leadership May Outlast Other Retention Levers
Leadership quality is not a soft metric. It can alter whether physicians feel heard after a complication, whether staffing concerns are dismissed, and whether professional effort is recognized in ways that reduce exhaustion. Those day-to-day exchanges shape professional fulfillment, and fulfillment in turn influences whether a physician sees a future inside the organization.
Burnout remained an important covariate in the adjusted analysis, which is consistent with the broader literature showing that organizational strain and interpersonal strain often travel together. A physician may tolerate heavy clinical volume for a time, but repeated exposure to poor leadership can make the same workload feel less bearable. In that sense, supervisor behavior does not replace structural reform; it determines whether structural stress becomes psychologically survivable.
The study also helps clarify why turnover programs sometimes underperform. Recruitment bonuses, retention incentives and exit interviews all have value, but they operate late in the chain. Leadership development acts earlier, at the level of the physician’s lived experience inside the workplace. If a department chair, division chief or practice lead improves in specific behaviors, the effect may accumulate quietly before HR ever sees a resignation risk.
Single-Institution Evidence With Clear Workforce Implications
The analysis came from one institution, so the results cannot be assumed to apply identically across every health system, specialty mix or regional labor market. A single academic center may have different cultural pressures than a community practice, a safety-net hospital or a national integrated delivery system. Even so, the linkage between supervisor behavior and actual turnover is strong enough to warrant attention from physician leaders who treat turnover as a staffing problem alone.
The study also used repeated survey periods, which strengthens the case that leadership ratings and values alignment were not one-time impressions. Matching those responses to HR records adds an important layer of credibility because the outcome was real turnover rather than self-reported dissatisfaction. That distinction gives the findings immediate relevance for organizations trying to decide where to invest limited retention resources.
For physicians, the message is blunt: bad management has a measurable cost. For health systems, the implication is equally direct. Improving immediate supervisors may be one of the few interventions that can influence physician retention without waiting for market forces, compensation redesign or broader cultural reform to arrive.
What the Data Do and Do Not Prove
The study does not establish causation. Physicians who rated leadership more favorably may differ in ways not fully captured by the adjustment model, and organizations with stronger leadership may also have other retention advantages. The analysis also cannot tell leaders exactly which behaviors matter most: communication style, fairness, accessibility, autonomy support or conflict resolution may each contribute differently.
Even with those limits, the central pattern is hard to ignore. Better-rated supervisors were linked to less physician turnover over the next two years, and the association persisted after adjustment for major physician-level factors. In a profession already strained by burnout and administrative burden, leadership quality is not an abstract management theme. It is a measurable retention variable, and the Stanford data show that physicians vote with their feet when it is missing.