Trilliant Health looks to supplant 'black box' top hospital lists with evidence-based comparison tool

Trilliant Health has a straightforward pitch for hospital strategy leaders: ditch the “best of” lists for hospital comparisons.

These "top hospital" lists are often based on subjective measures. As a remedy, the analytics and market research firm developed benchmarking tools based on quantifiable data points most relevant to hospitals.

Following up on this summer’s release of a market comparison resource for healthcare organizations, Trilliant Health has cut the ribbon on a new machine learning application that determines a hospital’s most-similar peers across quality, operations, market share, finances and other factors.

Trilliant said its SimilarityIndex | Hospitals application includes data on more than 2,000 acute care hospitals across the country and pulls from Trilliant’s proprietary all-payer claims database representing more than 320 million Americans, alongside other objective hospital data sources such as Centers for Medicare and Medicaid Services’ Quality Measures.

“The 800-plus ‘Top 100 Hospitals’ lists may be useful for marketing campaigns, but they are useless for developing evidence-based strategies,” Trilliant Health CEO Hal Andrews said in a statement. “Evidence-based strategies depend on accurate benchmarks against relevant peers. The healthcare industry is both capital-intensive and capital constrained and applying mathematical rigor to benchmarking and strategic planning reduces the risk of suboptimal capital allocation for healthcare providers, suppliers and payers.”

The company has published a free, public version of the benchmarking visualizer app online that allows users to select a hospital and review its 10 most-similar organizations in regard to care quality alone or as an aggregate measure.

Meanwhile, customers will be able to dig deeper into the tool to customize their comparisons around specific measures of, for instance, patient mix or market share, Trilliant Health Chief Research Officer and SVP of Market Strategy Sanjula Jain told Fierce Healthcare.

Even with the public tool, Jain said that users accustomed to seeing the same collection of big names will be surprised by the matches Trilliant’s system returns.

Punching in the Cleveland Clinic for overall quality analysis, for instance, pulls up matches like the University of Michigan Health System, Tampa General Hospital and the Southern New Hampshire Medical Center. Adding additional features to the analysis for an aggregate comparison returns another slew of names such as Vanderbilt University Medical Center, Strong Memorial Hospital and the University of Utah Hospital.

A table of SimilarityIndex | Hospitals' five closest peers for several well-known hospitals

“There’s no other kind of ‘Top 20 [List]’ with these hospitals,” she said. “A lot of the so-called top hospitals have peers that are all across the country in places that some of us may not have heard of or thought about as being quality, right? There are a lot of great hospitals across the country that have not been getting the credit for being great, so to speak, because they haven’t shown up on some of these lists but, objectively, they’re actually performing on par with some of the leading brands.”

Jain said that the methodologies of other well-established hospital ranking lists (such as those from U.S. News & World Report or Healthgrades) rely at least in part on individuals giving subjective measures of quality that are influenced by the prestige of a hospital’s brand.

But even when comparing hospitals strictly from a consumer marketing perspective, Trilliant’s hard data suggests consumers’ perception of a hospital don’t always align with their care-seeking behaviors, she said.

“So I'm in the DC area—you can calculate a loyalty score for me based on all my medical claims. Thirty percent of my medical visit went to [Johns] Hopkins, another 40% went to Inova, another 30% went elsewhere,” she said. “That is a true measure of customer loyalty. I would argue that's significantly more meaningful to hospital executives to figure out how to better market to consumers.

“I see this day in and day out working with hospital systems. They have a tendency to overestimate their market share time and time again, and it isn’t until they see this hard data that they realize that they aren’t capturing as much of the market. … A lot of what they thought to be true was biased by a lot of these subjective benchmarks.”

Although Jain said none of Trilliant Health's clients were asking for an evidence-based benchmarking tool outright, she noted that some top hospitals have pushed back against the “black box” approach of other rankings. These providers have previously tapped Trilliant Health to measure their individual quality metrics alongside those of competitors who they suspected “may have been misreporting” their measures to ranking lists that accept self-reporting, she said.

SimilarityIndex | Hospitals taps into those hospitals' requests by openly displaying the data it uses to generate its comparisons, “helping to expose the inaccuracy and flaws of the existing benchmarks,” Jain said.

“All the data points I’m talking about—market share, quality—you can see that it’s all attributed publicly to every hospital. We are being very transparent about how things are calculated. It’s just a mathematical formula running in tandem, which is meant to get folks to realize what is comparable, how you actually stack up—which is very different than black box lists that you have no idea how they actually came to that conclusion or what underlying data support that ultimate ranking.”