BOSTON -- Healthcare is full of great ideas and awesome innovations, said Leora Horwitz, MD, director of the Center for Healthcare Innovation and Delivery Science at New York University Langone Medical Center. But how do you make them stick?

For example, consider inadequate discharge communications between hospital physicians and primary care doctors. It's easy to improve, but better discharge summaries could be seen by some physicians as just more annoying documentation.

"There's really no incentive for this discharge doctor to make a nice summary for a primary care physician he's never met," said Horwitz at Tuesday’s Pop Health Forum in Boston.

  Read more of our event coverage.

To improve routinization of those interventions, the best thing to do is automate it, she said – which is just what NYU Langone did with its Epic discharge summaries.

Complex, low-value interventions are even trickier. They hard to do, hard to automate and not of obvious value to clinical staff.

For example, Horwitz discussed efforts to combat overcrowding in the emergency department, working to encourage staff to ensure patients are discharged by noon wherever possible.

If the policy is simply dictated from above, most staff will comply with it, she said. But compliance is quite different from committed use.

Her advice here is to "turn low-value interventions into high-value." Reframe the new way of doing things to show its intrinsic worth and make its overall benefit explicit and visible.

The question, then, should not be what percentage of patients are discharged at noon, but instead how much has emergency department wait time decreased?

"Try to make it clear to people why it is you want people to discharge patients by noon, so they can at least understand the value to others," said Horwitz. "The point is not noon. The point is how to make beds available to patients."

Horwitz offered these five guidelines for making innovation stick:

1. Choose and implement an intervention. The most successful interventions to improve care delivery share some hallmarks, she said. They're user-centered, incorporating input from the frontline end-users. They're designed to fit workflow. They're also iterative, able to be changed based on feedback from staff: Effective healthcare innovations incorporate evidence to try to solve a given problem but are rarely gotten right on the first go-around

2. Find a champion. Implementation requires enthusiasm, and that means clinical champions, said Horwitz. Committed champions can encourage adoption among potentially skeptical staff, helping facilitate system-wide change.

3. Keep the champion. An advocate who quickly moves on to the next project du jour is not going to be as effective, however. Innovation champions need to stay committed, said Horwitz, and help hold new projects together – perhaps for as long as a year or more – before they get routinized. "If you don't give up, it eventually sticks," she said. Eventually, that newfangled innovation becomes just the way of doing business: "It's not something we talk about every month when we meet anymore. It's just kind of what we do."

4. Understand the value-add of intervention. Not every new innovation is created equal, however. Some are easy to do, and make obvious sense – "high-value interventions," Horwitz called them. Those that provide immediate and visible rewards "will become sticky very fast and be sustained over the years." Others, however, require a more concerted and creative effort to gain traction.

5. Tailor your approach accordingly. If the simple, high-value changes are easy make stick, Horwitz focused on two others that require much more work – simple and complex low-value interventions. The first may be easy to implement, but at first blush would seem to offer little incentive for staff to change their behaviors. The second is similarly "low-value," but also requires much more in the way of change management.

  Related articles from our Pop Health Forum Coverage:

• Healthcare policy 'a mess,' so IT execs need to stay prepared
• Don't expect politics to derail population health, value-based care
• To break down telehealth silos, connect senior execs with the doctors on the ground
• Should IT or doctors lead population health programs? It's not so clear cut
• 4 population health must-haves: goals, leadership, organization, analytics
• You don’t need 30 products for population health
• Why healthcare innovation needs to become routine
• To fix broken US healthcare system, deliver value, VA director says

Twitter: @MikeMiliardHITN