Why The Best Hospitals Are Managed by Doctors

The nation's top doctor-managed hospitals' quality scores are 25 percent higher than manager-run hospitals.

by James K. Stoller,Amanda Goodall, and Agnes Baker

Healthcare has become extraordinarily complex — the balance of quality against cost, and of technology against humanity, are placing ever-increasing demands on clinicians.  These challenges require extraordinary leaders. Doctors were once viewed as ill-prepared for leadership roles because their selection and training led them to become “heroic lone healers.”  But this is changing.  The emphasis on patient-centered care and efficiency in the delivery of clinical outcomes means that physicians are now being prepared for leadership.

The Best Hospitals

The Mayo Clinic is America’s best hospital, according to the 2016 US News and World Report (USNWR) ranking. Cleveland Clinic comes in second. The CEOs of both — John Noseworthy and Delos “Toby” Cosgrove — are highly skilled physicians.  In fact, both institutions have been physician-led since their inception around a century ago.  Might there be a general message here?

A study published in 2011 examined CEOs in the top-100 best hospitals in USNWR in three key medical specialties: cancer, digestive disorders, and cardiovascular care. A simple question was asked: are hospitals ranked more highly when they are led by medically trained doctors or non-MD professional managers?  The analysis showed that hospital quality scores are approximately 25% higher in physician-run hospitals than in manager-run hospitals.

The findings of course do not prove that doctors make better leaders, though the results are surely consistent with that claim.  Other studies also find this correlation. Research by Nick Bloom, Raffaella Sadun, and John Van Reenen revealed how important good management practices are to hospital performance.  But they also found that it is the proportion of managers with a clinical degree that had the largest positive effect; in other words, the separation of clinical and managerial knowledge inside hospitals was associated with worse management.

Support for the idea that physician-leaders are advantaged in healthcare is consistent with observations from multiple other sectors.  Domain experts – “expert leaders” (like physicians in hospitals) — have been linked with better organizational performance in settings as diverse as universities, where scholar-leaders enhance the research output of their organizations, to  basketball teams, where former All Star players turned coaches are disproportionately linked to NBA success, and in Formula One racing where former drivers excel as team leaders.

Why doctors make good managers…

What are the attributes of physician-leaders that might account for this association with enhanced organizational performance?  As leaders, do physicians create a more sympathetic and productive work environment for other clinicians, because they are “one of them”? Does being a physician inform leadership through a shared understanding about the motivations and incentives of other clinicians?  When asked this question, Dr. Toby Cosgrove, CEO of Cleveland Clinic, responded without hesitation, “credibility … peer-to-peer credibility.”  In other words, when an outstanding physician heads a major hospital, it signals that they have “walked the walk,” and thus have earned credibility and insights into the needs of their fellow physicians.  But we would argue that credibility may also be signaled to important external stakeholders — future employees, patients, the pharmaceutical industry, donors, and so on.

The Mayo website notes that it is physician-led because, “This helps ensure a continued focus on our primary value, the needs of the patient come first.”   Having spent their careers looking through a patient-focused lens, physicians moving into executive positions might be expected to bring a patient-focused strategy.

In a recent study that matched random samples of U.S. and UK employees with employers, we found that having a boss who is an expert in the core business is associated with high levels of employee job satisfaction and low intentions of quitting.  Similarly, physician-leaders may know how to raise the job satisfaction of other clinicians, thereby contributing to enhanced organizational performance.

Our research suggests that if a manager understands, through their own experience, what is needed to complete a job to the highest standard, then they may be more likely to create the right work environment, set appropriate goals and accurately evaluate others’ contributions.  Having an expert leader at the helm, such as an exemplary physician, may also send a signal to external stakeholders, such as new hires or patients, about organizational priorities.  These factors are revealed in new work soon to be released.

Finally, we might expect a highly talented physician to know what “good” looks like when hiring other physicians. Cosgrove suggests that physician-leaders are also more likely to “tolerate crazy ideas” (innovative ideas like the first coronary artery bypass, performed by René Favaloro at the Cleveland Clinic in the late ‘60s).  Cosgrove believes that the Cleveland Clinic unlocks talent by giving safe space to people with extraordinary ideas and importantly, that leadership tolerates appropriate failure, which is a natural part of scientific endeavor and progress.

…and how training can make them even better ones.

Physician-leaders appear to be the most effective leaders precisely because they are physicians. Yet, great leadership also takes social skills. Medical care is one of the few sectors where lack of teamwork might actually cost lives, yet physicians are not trained to be team players. Nor is there evidence that it is the team players who select into medicine. Indeed, the favored nature of physician leadership of hospitals is even more remarkable for the leadership and followership handicaps that physicians must overcome in becoming doctors.  In view of this handicap, Dr. Victor Dzau, President of the National Academy of Medicine, considers those successful physician-leaders (who largely lack formal leadership training) as “accidental leaders.”

Physicians have traditionally been trained in “command and control” environments as “heroic lone healers” who are collaboratively challenged.  In the context of this paradox, that medical training on the whole conspires against great leadership, there is a clear need to train physicians more systematically.

One model has been pioneered by Paul Taheri, CEO of Yale Medicine, who has been engaging doctors in management training for some time.   He has focused on a two-tier approach: the first introduces physicians to the fundamental principles of business in the delivery of healthcare, and personal leadership development, through a day a month programme spread over a year.  Taheri sends around 40 medical faculty annually.  For those physicians who stand out as emergent leaders, the next step is an MBA.  Taheri insists that in the executive programs physicians are always trained with other physicians, but by design they are taken away from their hospital environment into the safe learning environment of the business school.

The Cleveland Clinic has also been training physicians to lead for many years.  For example, a cohort-based annual course, “Leading in Health Care,” began in the early 1990s and has invited nominated, high-potential physicians (and more recently nurses and administrators) to engage in 10 days of offsite training in leadership competencies which fall outside the domain of traditional medical training. Core to the curriculum is emotional intelligence (with 360-degree feedback and executive coaching), teambuilding, conflict resolution, and situational leadership. The course culminates in a team-based innovation project presented to hospital leadership. 61% of the proposed innovation projects have had a positive institutional impact. Moreover, in ten years of follow-up after the initial course, 43% of the physician participants have been promoted to leadership positions at Cleveland Clinic.

In-house programs have been developed in many healthcare institutions (including Virginia Mason, Hartford Healthcare, the University of Kentucky, etc.), by medical societies like the American Association of Physician Leadership, and by business schools (including Wharton, Harvard Business School, the Weatherhead School of Management, and soon at Cass Business School in London).  There seems to be a widening consensus that training physicians for leadership matters.  Such training promises to enhance the pipeline of physician-leaders so that the benefits of physician leadership can be more broadly realized.

James K. Stoller (M.D., M.S.) is a pulmonary/critical care physician at the Cleveland Clinic, where he also serves as chairman of the Education Institute.

Amanda Goodall, Ph.D., is a Senior Lecturer in Management at Cass Business School.

Agnes Baker is an assistant professor at the University of Zurich.


Thanks for the interesting article although I am not sure the title of the articles or the conclusions are supported by its contents. The evidence cited seems to support the insight that most doctors are poor healthcare managers by virtue of their training and some doctors, having received extensive leadership training are excellent managers.In my experience so-called professional managers have received little by way of the training cited so it would be interesting to see if non-medical professionals who have received the same training cited would get the same excellent results? Separately, Nicholas Bloom et al has written extensively about the generic problem that most organisations / managers (in all sectors) fail to adopt known best practices so arguably poor management is a universal problem and can be solved by the kind of extensive training cited in is article .

Very glad to see my observations confirmed by your research data. As a Physician, from a family of Business leaders, my interests has always been in both areas. Working in 3 countries for around 40 years gives some oversight into this area. There is a lot to say, but to start somewhere I would like to point out that A) Doctors learn a lot about leadership through their training. A Doctor is essentially leading the health care process of their patient, taking a lot of decisions (for others to follow) day in and day out, often relying only on himself in the final analysis. Most doctors are privileged to work with many other doctors in leadership roles, picking up from them good and bad (also important to know what kind of leadership to avoid) types of management/leadership styles. Since doctors move around a lot during their training they get an overview, which is often lacking in the education of other leaders. B) To run a business where you do not understand "the product", which is essentially healing human beings, is always difficult, especially when your employees are the experts, who are trained to take their own decisions, and where you have very little to offer. In this situation leaders often have little to contribute aside from the budget, and may end up focusing only on the budget, usually keeping a tight control, since it is the only thing that they are praised for by their own superiors. It is inherent in what is said here, that a non-medical leader is unlikely to be able to have future visions for the organisation, when it comes to product development, which in this case is the development of Medicine. Leaders who are in this difficult situation often try to overcompensate to have the status they are aspiring for. Really, it takes a special leader personality to do well in this situation, alas, many have been more of a burden on the organisation, and are usually quick to move on to another job. C) Doctors need to understand the "leadership culture" and learn the "lingo" to be able to feel confident in leadership roles, and in dealing with non-medical leaders. Therefore education is necessary, preferably starting in the Medical Schools. D) Increased Health Care costs strangely coincide with decrease in Medical leadership in the Nordic countries, Iceland, Norway and Sweden. Again more to be said, and again thank you for an inspiring article.

Appreciate the article, and it supports what I suspected for a long time. Additional research would be helpful for confirmation, and building awareness that hospitals frequently go off the wrong track when lead by non-physicians. A recent trend in the US is the rapid growth of companies contracting with hospitals to employ and manage groups of physicians on behalf of their client hospitals. These companies are managed primarily by non-physician leaders, and so the effect of non-clinically trained leadership extends beyond thspital borders. This interest his should focus not only on the hospital directly, but one the contracted clinical services industry supporting them.

This reminds me of a true story from a leadership workshop at a hospital with a business leader CEO:
Participant feedback to CEO visiting the workshop = "We never see you, you lead from behind a closed door."
Action = CEO cleared his calendar, stayed at workshop and listened to physicians and administrators. He started to build trust and relational currency. This potent feedback was a catalyst for change. . . . . . .but would it last?
Result = HR director commentary 3 months later = "he is a changed man. He's walking the corridors, engaging, smiling and breaking down our old silos between clinical and administration. "

The other story worth sharing is about Mayo's very intentional journey to build a more coach-like and collaboration skills within their senior physician and administrator population. They use a 'Leader As Coach' workshop to develop leadership coaching skills. This is a 2-day practice and feedback rich development experience. Post workshop they are tasked to coach 3-4 others within Mayo so the impact spreads and the coach-like behaviors take root. 

"...For those physicians who stand out as emergent leaders, the next step is an MBA.." Not true because a more focused graduate program such as a Master's In Health Administration is far more apropos to the setting.

When I moved to Colorado many years ago, to be on the medical school faculty, all hospitals in the city were run by physicians (except for a nun at St. Joseph's, who was excellent). That gradually gave way to MBAs and eventually just anyone...burnt out nurses, anyone. It became the corporate control of medicine. This article does not address the unethical behavior of these hospital heads...they report false statistics, and, worse for patients, attack doctors who "rock the boat" or are whistleblowers who point out bad care simply to save their patients. But in many instances, administrators are not interested in making things better. They are just interested in covering up the bad care, keeping their careers going, and often will ruin the career of a physician as they take steps to get rid of him/her and his suggestions and revelations.
This is an important article as it gives body to the claim that hospitals should be run by physicians.

Add new comment