4 population health must-haves: goals, leadership, organization, analytics

At Kaiser Permanente Colorado, “personalized health management" relies on IT savvy, clear operational goals and data that paints the entire health picture.

By Mike Miliard

April 04, 2017

11:03 AM

While many health systems are still getting their sea legs when it comes to population health, Kaiser Permanente Colorado already has two decades worth of experience.

That has taught the health system a thing or two about effective population health management, including one crucial tactic of replacing "population" with "personalized" to make gains by focusing intently on individuals.

But that's easier said than done at a health system that serves more than 650,000 patients.

"Personalized care doesn't scale well," said Bill Hoberecht, who until recently was senior director of population health management at Kaiser Permanente Colorado. "That's the big challenge: How do we scale personalized population health management?"

But over the years, Kaiser has worked to do just that.

Read more of our read our event coverage

"We have made it an iterative process over a couple decades," said Ann Wells, MD, an internist specializing in prevention and chronic care solutions at Kaiser Permanente Colorado. "We've refined our process, learned a ton, and generated a population health program that is personalized and is scalable."

Wells and Hoberecht spoke Tuesday at the HIMSS Pop Health Forum 2017 in Boston. They said what started as a relatively simple chronic disease management program has evolved over the years into a tightly coordinated, team-based, and heavily IT-led strategy that's able to communicate individual patient's specific needs to care teams who make broad use of EHR and other data to ensure optimal care for each patient.

For organizations looking to take the first steps doing something similar, he had an initial piece of advice: "Go out and read any materials on organizational change management," said Hoberecht. "Anything those books say will be useful for you. It's just not as simple as saying, 'I have an answer, we're all going to do it this way.'"

Beyond that, Hoberecht outlined four key must-dos for those providers looking to set up or optimize a pop health management program.

Have a proper launch. A specific population health management strategy is critical, he said – that includes having defined executive and leadership roles and defined quality goals.

"That's foundational," said Hoberecht. "You need to have folks who understand what the strategy is and they agree with that strategy. IT staff need to be on board, but that's not who really matters here. Who really matter are the doctors who are going to be championing this and talking among themselves about how the practices are really going to run."

Keep a focus on operational responsibilities. At Kaiser Colorado, they established clinical governance councils and deployed centralized care management teams to ensure the program had both rigorous oversight and committed staff to carry it out.

"Ensure that you have the right operational responsibilities in place," he said. "Ensure you have all your bases covered."

Incorporate IT staff from the get-go. While physicians are important to the equation, population health management needs to be heavily supported by technology. A smart pop health program gets the IT staff involved from the very start.

Personalize analytics. The fourth key to success is to use all available data – and expand data sources whenever possible – to enable a more personalized look at analytics

"There's a great deal of personalization we've done with our program from the start," especially when it comes to tailoring messages that go out to individuals, said Hoberecht.

"You'll probably get a message that's different from me for conditions that seem to be similar," he said. "Your screening rate might be different from mine for a particular prevention because of other criteria.The discussions we have that are shared decision-making mean my course of treatment for a particular chronic disease might be different, even though we look similar. There's a tremendous amount of personalization that does happen in the exam room."

For personalized analytics, it's key to use "all available data that you have, so you have that full picture view of the individual and you're not just seeing that last A1C as the sole piece of information," he said. "And continually expand your data sources. In the last two years, we've added financial data, which has been very useful for us, allowing us to do a number of additional analytics based on that one new data source."

Related articles from our Pop Health Forum Coverage: