Written by: SA Kushinka

Health Catalyst, a Salt Lake City based data warehousing and analytics company, drew over 800 attendees and analytics pioneers from leading health systems to its first annual Healthcare Analytics Summit September 24 and 25. A key theme throughout the two-day summit, echoed by many speakers, was using analytics to drive value in healthcare, a concept defined as service and outcomes divided by cost.

Billy Beane, the General Manager of the Oakland Athletics, revolutionized the way statistics and analytics were used in the selection of players. At the summit he spoke on why the On Base Percentage was more valuable than traditional metrics such as home runs or Runs Batted In. By fielding a team of low cost, undervalued players Beane provided a striking analogous example to the revolutions in healthcare. He drove home the point that you need to be disciplined about examining and acting on the data, and over time you’ll get the results you want. Beane, portrayed by Brad Pitt in the movie “Moneyball,” explained why he doesn’t actually watch the games: baseball (and sports) is emotional, often driven by gut feel or intuition on the field. By not watching the games, he maintains discipline, guided by statistics and analytics, and has consistently fielded the highest performing team for the least amount of money.

Glenn Steele, MD of the Geisinger Health System, asserted that “50% of what we do in medicine is too little, too much or wrong.” At Geisinger, they asked how they could use data to extract which costs don’t bring benefit to health. He asserts that there is an inverse relationship between quality and cost in healthcare, a belief echoed by other speakers throughout the summit. Lower costs should be part of the definition of quality, but this is a difficult point to get providers to rally around. “No one is energized by cutting costs,” Steele said, “but improved outcomes do motivate.” If you can use data and analytics to show that the two are related, providers become very engaged. Geisinger’s physicians came to consensus and adopted 67 clinical guidelines, translating those into 120 evidence-based practices. Individual variation in practice is regularly exposed to the judgment of colleagues. If you allow unjustified variation you do not get the highest quality and lowest cost. Near real time data feedback is critical for this process. Dr. Steele asserted that availability of data has never been a problem for them but usability and timeliness of that data is key.

James Merlino, MD, Chief Experience Officer of the Cleveland Clinic, spoke about using data to transform patient experience. Although we claim to be “all about the patients,” in reality we often fall short of these ambitions. Merlino described a survey asking providers what they thought patients wanted (in a hospital setting); they answered with things like new facilities, private rooms, food on demand, interactive bedside computers, a quieter room. What patients actually want respect, good communication between staff and happy people. Patients want us to treat them like people, even if this can go against objective med school training.

Further, because they are unsophisticated in assessing healthcare, patients judge the quality of care they get from providers based on proxy measures, like good or bad communication among staff. Patients are looking at visual and verbal cues – if the doctor rushes into the room, the patient is less likely to ask questions. Patients are looking and listening to get a sense of what is going on in this unfamiliar and stressful environment.

Patient experience is like the elephant parable, in that it is very difficult to get the full picture. How do you fix this? Merlino argued it’s data — and looking for critical pain points with focus groups and quantitative analysis. Organizations should drive the data down vertically: all tactics need to have data to support them. Look for best practices that truly make a difference, and put a process metric in front and an outcome measure at the end.

Dr. Penny Wheeler, CEO of Alina Health System, spoke about the unsustainable, spiraling healthcare costs in the US, noting that if food prices rose as fast as healthcare, a pound of bacon would cost $137. The solution is to find a way to move from fee-for-service to a value-based way of delivering healthcare. Wheeler offered an analytic system that allowed for three “looks” at cost data: retrospectively (was an event preventable?), current look (what’s the risk of a given action?), and predictive (how can we focus scarce resources on patients that need it?).

Dale Sanders, Vice President of Strategy for Health Catalyst, talked about predictive analytics in a provocatively titled presentation: “There’s a 90% chance your son is pregnant: predicting the future of predictive analytics in healthcare.” Sanders, who once worked on missile defense systems, noted gaps and challenges in today’s HIT systems. First, no one is really doing a good job of collecting data on outcomes. Without data on outcomes, we can’t possible get accuracy about value. He suggests that patients need to be in engaged in supplying that data and even incentivized to provide that kind of data. But there is no vendor on the market that does that.

Second, Sanders argues that we are asking our doctors and nurses to use unwieldy digital systems with no thought to their tolerance threshold for clickwork. A study found that it took 400 clicks from log on to completion of a clinical note. We’re asking providers to climb a hill they can’t conquer. EHR engineers should be challenged to build user interfaces that can be navigated with only 50 clicks.

Takeaways:

  1. There is an inverse relationship between quality and cost. Cost is inherent in the definition of quality.
  2. Cost per Unit of Health is healthcare’s On Base Percentage—this is the metric that we should use to measure quality.

                          

                           

Find this useful or interesting? We’re constantly sharing stuff like this. Sign up to receive our newsletter to stay in the loop.