It’s a curious paradox of health care’s current quest for fiscal salvation through “innovation” that a tight focus on new ideas may actually thwart the process of finding new methods for lowering costs, improving patient outcomes and changing ineffective clinical practices.

In some ways, new ideas can be distracting. At a recent gathering of health policy researchers, anecdotes made clear that hospital executives who announce their new assignments to manage some aspect of “innovation” are routinely bombarded with ideas at every turn. “I’ve come to view them like baby pictures,” said one. “Everyone has them and tends to push them at you.”

David Asch, executive director of the new Penn Medicine Center for Innovation, concurs that innovation has become a “faddish and very ‘Oh, Wow'” concept whose mechanics are not well understood.

“Many believe the process of health care innovation is the process of ideation — of coming up with often-wacky ideas that seem incredibly exciting,” said Asch. “But it’s really very different from that; innovation in a health care setting is a highly disciplined enterprise involving scientific rigor, testing and implementation. It can seem boring if you think innovation is just about creating trendy smartphone apps.”

Asch, the former head of the University of Pennsylvania’s Leonard Davis Institute of Health Economics, was one of four Penn experts who participated in a recent roundtable discussion about the management of health care-related innovation programs. Other participants included Roy Rosin, Penn Medicine’s chief innovation officer and former vice president for innovation at Intuit; Christian Terwiesch, Wharton professor of operations and information management, and co-author of the book, Innovation Tournaments: Creating and Selecting Exceptional Opportunities; and Raina Merchant, an emergency physician and assistant professor of emergency medicine at Penn’s Perelman School of Medicine.

An article on the roundtable discussion recently appeared in LDI Health Economist. Below are excerpts of that discussion. To see the full article, click here.

So exactly what is innovation, and how do you begin to do it inside something as large and complex as the $4.3 billion-a-year Penn Medicine network of health care facilities?

“People are thinking about ‘innovation’ so many different ways,” said Terwiesch. “It’s important to define it so we’re all clear about what we’re trying to accomplish. Innovation is a match between a solution and a need that creates better value than what currently exists. The idea that you begin with is a hypothesis; it has to be scientifically proven.”

Terwiesch noted that you can’t manage what you don’t measure, pointing out that health care institutions with innovation programs have to measure the overall innovation process as well as targeted innovation projects. He said few institutions were currently able to do this.

“These places have [many] potential innovations flowing through the system,” Terwiesch said. “They should start measuring that overall process just like they would measure a production process: What are the defect rates, the number of widgets in the system? Where do the ideas come from? How many did we try out, and how many then moved to the next step of refinement? Why?”

At the national level, according to Terwiesch, “nobody is thinking of this as a structured innovation process — it’s just all a big mess where, at best, they measure individual innovations but ignore the rigor of defining the process structure. Somebody needs to own the process of innovation management.”

Roy Rosin noted the importance of understanding “that ‘innovation’ is not about executing that first big idea you fell in love with, but rather about establishing a support infrastructure that enables you to systematically figure out what really works. Most innovation fails due to premature scaling — it’s rolled out and pushed big before its managers really know whether or not it works.”

A well-controlled innovation process feeds on lots of new ideas, but discards almost all of them as failed elements of an ongoing, evidence-based experiment, Rosin added. “In other areas of business in recent years, most of the innovations that became successful started out in the wrong direction,” Rosin said. “The best venture capitalists tell you that most business plans they invest in — even those involving entrepreneurs with the best track records — initially fail and are saved by the right innovation technique: That means the ability of managers to learn, pivot and adjust quickly enough to enable success to rise from the ashes of those initial failures. The same thing is true for health care innovation.”

Increasingly in both government and industry, the game-like format of “innovation tournaments” is being used to generate the initial wave of ideas that are fed into the innovation evaluation process – a trend that was accelerated by Innovation Tournaments, co-authored (along with Terwiesch) by Wharton vice dean of innovation Karl Ulrich. These tournaments are contests that use crowdsourcing methods to rapidly compile large numbers of ideas for improving business policies, practices or products. A tournament committee reviews the thousands of incoming ideas and selects a few that are granted a reward or developmental funding.

Thus far, the innovation tournament offering the largest prizes has been the CMS innovation center’s 2012 Health Care Innovation Awards program, which pushed out nearly a billion dollars in innovation grants to 106 various health care and academic groups.

Terwiesch wonders if grants that gamble so much money on a single concept represent the most effective strategy for initiating and managing broad scale innovation.

“When you’re thinking of the billions of dollars that are going out through health care innovation tournaments,” he said, “you have to ask, ‘which form of that process would work best?’ If you have a billion dollars, do you fund 10 $100 million projects or do you fund a thousand projects with much smaller amounts but require them to send in intermediate results of ideas that have tested out in a promising way? Do you go for a few big ideas, or do you use the whole process as a way to validate much larger numbers of promising ideas?

According to Rosin, one of the biggest changes in innovation management seen in other business fields is the emergence of “rapid validation” techniques that facilitate the scientifically-sound testing of an idea in a matter of days or weeks, rather than months or years.

“In health care, that ability to apply scientific method faster, to be able to test lots of concepts quickly and cheaply, is going to be very important for getting a good return on innovation investment,” Rosin said. “We need to constantly generate good evidence that tells us we’ve learned something new and should change direction.”

The drive to innovate is also forcing health care to rethink its traditional silos and professional boundaries. For instance, aside from the health policy researchers who make up its core, Penn Medicine’s innovation center has recruited faculty from the University’s schools of business, engineering and design.

“In many ways,” said Asch, “we’re trading on the idea of doing something a lot of academic medical centers ought to be able to do, but are really very ineffective at doing — which is connecting the academic enterprise with the clinical enterprise.”

Terwiesch cautioned that academic traditions themselves can get in the way of such efforts. “It’s naïve to think we could sit an interdisciplinary crew in an office and actually solve anything. Smart academics like to think big thoughts as opposed to going out on the front lines — and, in this case, ‘front lines’ means where the patients are. You cannot fully understand or improve the overall process unless you form empathy with the end customers. The more we are facing the patients, the more we will learn about their needs and how to meet them in some better way.”

Health care innovation projects aimed at those patients can take many forms. Some are as simple as designing an electronically tethered pill bottle to increase medication adherence. Others, like accountable care organizations or medical homes, represent a disrupting reinvention of the entire institutional business model. There are a seemingly infinite number of other innovation targets in between these extremes.

Asch and the other panelists agree that, ultimately, the holy grail of health care innovation will be changing the traditional behavior and daily culture of clinical communities — something that will be difficult to achieve.

“There are no easy fixes and no low-hanging fruit when you’re talking about that,” said Raina Merchant. “It’s really complicated, but can sound easy at first. Like, let’s reward everyone who washes their hands after they take care of a patient. But actually, there are so many different things that have to happen just to get to the point of trying to do that implementation. And then you have the issue of how to sustain it on the scale of an entire hospital…. In my work on projects that try to change smaller systems, I have been continually humbled by how tough the process of altering behavior can be,” Merchant said.

Also crucially important to the success of in-hospital innovation programs are the IT systems whose functions are targets for cost-saving and care-improving innovations. But throughout much of health care, IT operations are rigid and inflexible in ways that present major obstacles to innovation researchers and managers.

“Right now,” said Rosin. “We’re not architected to facilitate rapid, frequent and low-cost changes. Often, you want to create foundational sandboxes or other experimental capacities that will allow people to turn ideas into some kind of action that can be studied. So, in many places, there’s a whole enabling bucket needed for innovative IT use.”

A related and equally important agent of innovational change are the tablets, smartphones, wireless body monitoring devices, electronic records systems and other digital utilities that offer dramatic new ways to communicate and gather clinical data.

“The general tendency is to focus on the individual devices and see those items of hardware and software as the innovations,” said Merchant. “But from the 30,000-foot view, it’s the evolving matrix of connections linking all these devices that could be the ultimate innovation. Think of it as an electronic ‘nervous system’ whose digital tentacles are rapidly spreading in all directions in a way that could connect everything to everything else.”

Others have found that “there are a lot of current opportunities to link existing systems better — but few places are taking advantage of that,” Merchant added. “So in terms of innovation, we’re really at the very beginning of harnessing the full power of digital interconnectedness within a health care system setting.”