I was recently chastised by a colleague for being too negative in one of my pieces on hospital care. His is a remarkable story of what happens when things go well, and it has made me think a lot about why, in some places, things seem to work while in others, not so much.

He told me how a few months ago, soon after returning to Boston from a trip to China, he had started feeling short of breath. When his cardiologist convinced him to be evaluated, he found himself at the Beth Israel Deaconess Medical Center (BIDMC), arriving in the ER late one evening.  He was triaged within minutes, had an EKG within 15 minutes, at which time comparisons were made to previous EKGs.

After ruling out a heart attack, his ER physicians quickly ordered a CT Angiogram.  That test, completed within an hour of his initial arrival to the ER, revealed the reason for his shortness of breath:  he had a large saddle pulmonary embolus.  He was started immediately on IV heparin and sent quickly to the ICU, experiencing essentially no delay in care.  He spent three days there and reports receiving care that was attentive, expert, and consistently of the highest quality.  Even after discharge, he received two nursing visits at home to ensure he was doing OK.  In discussing his experience, he repeatedly emphasized the fantastic communication and teamwork that he witnessed.  Weeks after discharge, he continues to get better and feels the benefits of the excellent care he received.

This is the story we all hope for.  And when I heard it, I have to say that I wasn’t surprised.  There’s something about the BIDMC that’s unusual.  Of the 4,500 hospitals that report their mortality rates to Medicare’s Hospital Compare website, only 22 (less than 0.5%) have better than predicted  mortality rates for all three reported conditions:  heart attack, congestive heart failure, and pneumonia.  And, we know that the combined performance on these three conditions is remarkably good at predicting hospital-wide outcomes, including outcomes for pulmonary embolism.  If you are a patient and care deeply about good outcomes, BIDMC seems to be a good place for you.

So what’s so special about them?  What do they do that’s different?  I don’t know, specifically, all of their tactics, but I have some guesses about what seems to differentiate high performing institutions from the rest.  And in a word, it’s leadership.  BIDMC has had two CEOs over the past few years, and both of them have been unusually committed to achieving high quality care.  That commitment translates into real activities that make a big difference.  Let me divert us with a story of what this might actually mean.

A few years ago, I was working on a strategy for improving the quality and safety of VA healthcare.  As part of this effort, I called up senior quality leaders of major healthcare organizations across the nation.  One call is particularly memorable.  Because I promised anonymity, I will not name names but this clinical leader was very clear about his responsibility: every month, he met with his CEO, who began the meetings with three simple questions: “How many patients did we hurt last month? How many patients did we fail to help? And did we do better than the month before?”

The CEO and the entire hospital took responsibility for every preventable injury and death that occurred and the culture of the place was focused on one thing: getting better.  When I looked them up on Hospital Compare, they too had excellent outcomes and they regularly get “A” ratings for patient safety from the Leapfrog Group.

How do the BIDMCs and these other super-high performers pull it off?  How do they build a culture of quality when so many organizations seem to struggle?  High performance is complex, of course, and I won’t try to be overly simplistic.  But a few things seem common among many high performing institutions. They seem to be focused on three things:  timely, clinically relevant outcomes data; transparency within (and usually outside) the organization; and a constant focus on getting better.  You can see the kinds of data that BIDMC posts on its website — it’s not just the standard Hospital Compare stuff (which everyone has to do) but other data on a series of outcomes which are not required.   When I hear Kevin Tabb, their current CEO talk about quality — it’s obvious that quality is not a platitude.  He is genuinely focused on getting better.

So what’s the lesson from BIDMC, Mayo and other high performing institutions? There is no substitute for great leadership.  Each of them seems to have been blessed with leaders who, despite all the wrong incentives in the healthcare system, prioritize patient care and drive their organizations to great performance.  They are internally motivated and do all the things I describe above, despite the fact that our primary payment systems incentivize them to do more, not better.  They are extraordinary leaders- with not only great vision but also the ability to execute that vision.

But here’s the risk:  too many policymakers believe that all we need to do is figure out what BIDMC or Mayo or Kaiser does and just get everyone else to do it. Such an approach, while seemingly perfectly good on paper, fails to account for the human element.  The strategies that they have used have been executed by individuals unusually focused on improving care.  Barring substantial improvements in cloning technology, we can’t expect that each hospital will have a great leader.  We don’t expect that every technology company will have a Steve Jobs.  In every industry, there are a few visionary leaders, but the rest of the organizations?  They are run by mortals — and mortals respond to incentives.

And here lies the problem:  the incentives in the system are not motivating the typical CEO to improve care.  Whatever strategy we employ around timely data, transparency, etc. won’t work until the leadership is properly motivated and focused on quality.  And while that happens in pockets, it’s not happening across the entire healthcare system.  And this is where we will pick up in my next blog: how to get the rest of the organizations to make quality a real priority.

Ashish Jha is an associate professor of health policy and management, Harvard School of Public Health.  He blogs at An Ounce of Evidence and can be found on Twitter @ashishkjha.