Physician groups that have been aggressive in their pursuit of value-based care are making progress toward the Institute for Healthcare Improvement’s Triple Aim of improved patient experience, lower total costs and enhanced population health.
Progressive physician groups, such as Crystal Run Healthcare, the Carle Foundation and the Marshfield Clinic, among others, say several key competencies aided by IT, ranging from care management to data analytics, must be honed to survive and flourish as fee-for-service payment shifts to fee-for-value.
With the introduction of the Quality Payment Program in 2017, Medicare is tying a portion of physicians’ fees to their performance on quality and cost metrics. QPP also encourages physicians to move into Advanced Alternative Payment Models, which require physicians to take on downside financial risk, or cover some or all costs that exceed predetermined spending targets for a patient population.
Preparing for risk-based payment takes a combination of conviction and data, says Jeff James, CEO of Wilmington Health. “Our strongest point is that we have people with different talents who are really committed to what we’re doing, and we use the data as the main tool to do it.”
A commitment to value needs to begin at the top, says Scott Hines, MD, chief quality officer and chief medical officer, Crystal Run Healthcare, Middletown, N.Y. “To do this well, you have to really reconfigure the way that you do things in the practice, to focus even more than before on quality, experience of care and cost of care. You can’t just form an ACO committee. You have to go down to the grassroots level, convince the physicians that this is the best way to deliver care.”
Leaders at Crystal Run, an independent, multispecialty group, are continually selling their vision to the practice’s 400-plus physicians. In 2010, Hines and another senior leader met with small groups of physicians to explain why the practice needed to transition to value-based care. “Since then, we continue to beat the drum,” Hines says. Quarterly meetings update physicians on the practice’s progress; division and department leaders attend a Leadership Academy to learn key performance improvement competencies; and new physicians go through an orientation on the practice’s value-based approach.
Physicians also need to see that the practice is investing in key resources that will help them manage patient populations. A recent American Medical Group Association survey found that “hiring care coordinators” was the most-cited investment among AMGA members preparing for risk-based payment.
Other key investments revolve around using data analytics. “Business intelligence is really important,” says Marshfield Clinic’s Kori Krueger, MD, medical director, Institute for Quality, Innovation and Patient Safety. “It’s hard to be strategic about where you’re going, understand what your clinical outcomes are, and achieve financial targets if you don’t understand the current state and how that differs from the state you’d like to achieve.”
Data analytic needs
A lack of data to inform population health and improvement efforts is not an issue among leading-edge group practices, which have had EHRs for years. The opposite is actually the problem.
“Our biggest challenge is taking all the massive amount of data we have and trying to boil it down into actionable information,” says Rick Rinehart, CIO and vice president, information technology, The Carle Foundation, Urbana, Ill.
Two types of actionable information are critical: risk stratifying populations and identifying performance improvement opportunities.
Segmenting a patient population into high- vs. low-risk groups helps physician groups prioritize those patients who most need care management. “You’ve probably seen the Medicare numbers where 20 percent of patients are responsible for 80 percent of the costs—the 80/20 rule really works in healthcare,” says John Cuddeback, MD, the AMGA’s chief medical informatics officer.
Sophisticated risk stratification models identify patients when they are beginning to deteriorate and are likely to be admitted to the hospital or visit an emergency department in the next six months, Cuddeback says. By contrast, first-generation approaches, which have been used by insurers for decades, tend to identify high-risk patients after they were admitted to the hospital or visited an ED. These older models often rely on claims data or after-the-fact billing codes submitted by providers to describe healthcare services provided to patients.
Physician groups with EHRs containing up-to-date biometric information, including blood pressure and glucose readings, can develop predictive models capable of identifying patients who are on the cusp of a downward spiral. “The big advantage of including clinical data in your analytics is you can see patients as their clinical parameters are beginning to deteriorate, and you can predict not just a readmission but an initial hospital admission,” Cuddeback says. “This is a great time to intervene.”
The Carle Foundation has created disease registries within its Epic EHR for asthma, diabetes and other conditions. The vertically integrated network, which includes Carle Physician Group and Health Alliance Health Plans, developed its own risk score to help segment these disease populations.
“We have a means of stratifying that subset of a population to understand who’s at most risk for disease advancement because of their latest lab values or because they haven’t followed up on some of their health and wellness maintenance,” says Suzanne Sampson, system vice president for information management and analytics, as well as project management. “Through that functionality, we can do direct outreach to those patients very efficiently.”
Other group practices are investing in risk stratification software products. For instance, Crystal Run Healthcare uses Crimson Population Risk Management to identify those patients most in need of care management.
Stratifying a population by patients who did or did not receive critical screenings or lab tests is different from identifying those patients likely to be admitted to the hospital in the next six months, Cuddeback says. EHRs on the market can be programmed to send alerts about so-called gaps in care, like missed screenings, but fall short of predicting outcomes. “When you’re talking about predicting risk for poor outcomes, you need more sophisticated analytics,” he says.
Marshfield Clinic’s population health dashboard tracks 160 quality and cost measures, pulling data from the organization’s robust EHR and data warehouse. Physicians can see how well their patient panels or individual patients are doing on specific metrics and compare their scores against those of other physicians, while senior leaders can compare performance across different locations and over time.
“That dashboard is 100 percent transparent to everybody in clinical care delivery throughout our organization,” says Krueger. “Using a drill-down application, they can immediately see all the way down to individual provider-level performance on any measure.”
The dashboard has made it easier to engage physicians around organizationwide efforts, such as the blood pressure improvement initiative that began in 2015. “We used our medical record and our data warehouse to understand the population of patients who have elevated blood pressure and hypertension, and we put together a strong response for managing that population,” he says.
The initiative has enlisted all physician specialists, from dermatologists to plastic surgeons, in tracking blood pressure rates and ensuring any patient with high blood pressure is referred to a primary care physician for followup. “We developed some other dashboards for our medical subspecialties where they can see things like how often patients show up in their office with a blood pressure outside the desired range and how often staff appropriately take a second blood pressure in those cases,” Krueger says. As a result, Marshfield Clinic has increased the percentage of people with controlled blood pressure of 140/90 or lower to the high 80 percent range, from the low 70 percent range in 2015.
Crystal Run Healthcare’s business intelligence team has also been enlisted to help guide performance improvement. One tool it built helps pinpoint variation in physician treatment approaches, which informs best practice discussions. “If you had two people, same age, same sex, same genetics, who both had hyperthyroidism, and they saw two different doctors in the same organization, you would think they’d be treated in relatively the same way,” says Hines. “But we’ve found that’s not the case.”
The variation tool compares the overall charges per patient per year for specialists treating the same disease. For example, when looking at diabetes, the tool shows each endocrinologist’s average charges per patient for the past year.
“No matter what diagnosis, no matter what specialty, we usually find a three- to four-fold variation between the providers on the left and right side of the graph,” says Hines. “The reason is not that some providers have sicker patients or that others have better quality. The reason is a lack of awareness of, and adherence to, evidence-based practice guidelines.”
Register or login for access to this item and much more
All Health Data Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access