The world of technology has its “hype cycle” and analytics is by no means immune from exaggerated claims and premature projections into the future. What’s more helpful to healthcare executives is to view some “real world” examples of success in analytics. In this three-part series I’ll highlight some of what’s being done on the front lines of healthcare where analytics is being applied in ways that are making a demonstrable difference.
In this post, we’ll talk about how hospitals can reduce quality variation to lower costs.
Using Analytics to Improve Quality
This topic was discussed in a presentation given at the Institute for Healthcare Improvement (IHI) Annual Conference in December 2014. Both speakers were leaders in the Yale New Haven Health System (YNHHS). The Vice President of Financial Planning, Steve Allegretto, represented the administrative side of the house, and the Vice President of Performance and Risk Management, Dr. Ryan O’Connell, the clinical side. They gave the presentation as a team, which turned out to be an important aspect of the story.
Their premise was that you can actually improve quality and lower costs at the same time. And by the end of the presentation they made this seem like a “no-brainer”. The presentation summarized the approach – and the results – of work that YNHHS has been doing for several years to identify and eliminate sources of unnecessary variation. Managing variation is at the heart of formal quality improvement methodologies like LEAN and Six Sigma that have been used to transform manufacturing and other industries. A major goal of the ongoing YNHHS work is to apply these techniques to healthcare at a very practical level.
At its essence the team’s approach consisted of:
- Identifying meaningful quantitative measures on both the financial side (revenue and cost) and the quality side (readmissions, adverse events and various other avoidable variations.) The quality measures were also consolidated into a single composite measure called a “quality variation indicator” or QVI.
- Segmenting patient populations using available data. Examples included: patients with QVIs versus those without; disease categories, such as asthma and heart failure; and hospital settings, such as critical care units.
- Comparing patterns in the quantitative measures among the various segments to identify opportunities to reduce variation.
- Reworking, or in some cases totally redesigning, care processes to minimize or eliminate important variations.
Why Better Quality and Lower Costs Are Not at Odds
The idea that improving quality and reducing costs are compatible can seem a bit counterintuitive at first. However, the following example illustrates that it can actually be quite logical:
Consider patients on ventilators in an intensive care unit. Ventilator Acquired Pneumonia (VAP) is an example of a serious QVI event that can often be avoided. An analysis of patients with VAP showed significantly higher costs compared with patients without VAP. The costs are so high, in fact, that even a handful of VAP cases can have a significant impact on the average cost of caring for all intensive care patients. With this insight, YNHHS redesigned care processes with the goal of completely eliminating instances of VAP. Interestingly, their successes span several areas:
- Average cost to care for intensive care patients declined. Incidentally, so did revenue; however, some of that revenue would not have been reimbursed due to “pay for quality” programs that reject claims for care associated with avoidable adverse events. Bottom line: it’s not the kind of revenue you really want.
- Quality improved. This was both in terms of the VAP-related measures and from reduced length of stay, since the longer patients are in the hospital, the more prone they are to QVIs.
- Patient satisfaction increased. This was a somewhat unanticipated side benefit. Patients who don’t experience QVI’s – and get out of the hospital more quickly – are generally happier patients!
There were several other examples, but the message was consistent: improving quality by reducing variation pays dividends.
Analytics Leads to Performance Improvement for Impressive Results
Analytics played a prominent role in this initiative, particularly in terms of highlighting the opportunities to reduce variation. But the more powerful lessons were really about how people, information, and technology can come together in a performance improvement initiative to achieve impressive results.
Here’s a brief summary of these lessons:
- A cross-functional team focused on achieving and sustaining measureable, outcomes-based results is essential. The members of the YNHHS team clearly respected one another and had worked hard to understand points of view that may have initially been outside of their comfort zones.
- “Work with the data you have” is a strategy that often yields better than expected results. The YNHHS project primarily used financial claims information. Even though this may have been far from “clinically perfect”, it provided enough insight to drive significant improvements, both clinical and financial.
- Select measures that are relevant, outcomes-oriented, and within your control. The YNHHS team was able to “move the needle” on both their financial and clinical measures through their improvement work. They chose well!
- And finally, a good story is one of the best ways to persuasively deliver a message to a diverse audience. Using summarized information and visualization to illustrate the relationship between quality and cost – as well as the impact of improvement efforts over time – helped to shine a light on what might otherwise seem like an academic exercise to people without a background in data and analysis. And narratives, such as the “VAP scenario”, bring in the human element around how QVIs impact not just abstract measurements but actual people. That’s the thing that really resonates with people who care for other people in their professions!
Latest posts by George Dealy (see all)
- Exploratory Data Analysis Part 2: Helping You Make Better Decisions - October 11, 2019
- Practical Analysis: Understanding Visualization Concepts - September 19, 2019
- How Exploratory Data Analysis Helps You Generate Meaningful Insights - August 1, 2019