Trilliant Health | June 16, 2023

Becker's Hospital Review | 2023 Trilliant Health SimilarityIndex™ | Hospitals


Trilliant Health’s CEO, Hal Andrews, spoke with Molly Gamble on the Becker's Hospital Review podcast about the 2023 SimilarityIndex™ | Hospitals


Molly Gamble: Welcome to the Becker's Healthcare Podcast. I'm Molly Gamble, VP, Editorial. If you are listening to this, you likely work in healthcare, and if you work in healthcare, you are likely familiar with the numerous ranking, rating and grading systems that have proliferated over the past decade. There are dozens of frameworks out there for how we look at healthcare providers and how we think of them. It has really come to resemble something of a frame shop.

Today I'm speaking with Hal Andrews; Hal is President and CEO of Trilliant Health. Last year, Trilliant Health released its inaugural Similarityndex | Hospitals, which is a fresh and compelling take on traditional rankings that you may be used to. It uses machine learning models to identify the 50 most similar peer hospitals for approximately 2,000 US hospitals based on quality, financial, competitive and price data. This year, Trilliant Health is releasing its second edition of that index. This time, the index also includes rates data from the health plan price transparency machine-readable files. This shows previously inaccessible insights about commercial reimbursement across the US.

Hal, thank you so much for joining me today. This is a big week for you and the Trilliant Health team in releasing the Similarityndex | Hospitals. So thanks for taking some time to connect with Becker's listeners.

Hal Andrews: Molly, it's my pleasure, thanks for having me on.

Molly Gamble: I shared a brief overview and description of Similarityndex | Hospitals, Hal, but I'm, I'm curious, can you share more about the origins of this? Where did the idea for this outlook of hospitals come from?

Hal Andrews: Well, as often is the case with good ideas, it came from a couple of different directions. We had just finished working on a new demand forecast model back at the end of 2021, and the data scientist who had led that effort approached me and said that he would like to explore uses of similarity in healthcare analytics. At the same time, a few months later, Sanjula Jain, who leads research for us and is familiar to many of your listeners, came to me and said, you know, we should build a benchmark product for the health systems and hospitals that we serve.

We combined those ideas and decided to create an index using similarity, using math to create new rankings, indexes and benchmarks for hospitals with the hope that we would give people something that's a little more actionable than most of the historical rankings. As my friend Paul Keckley says, they're 812 top 100 hospitals, which is an interesting thing from a marketing perspective, but not really relevant to the day-to-day operations of a hospital.

Molly Gamble: Paul Keckley is always good for a quote, Hal. I appreciate that perspective. Something you mentioned is actionable rankings and actionable benchmarks. You've also talked about aspirational rankings, ratings, the like. Can you talk about the difference between the two?

Hal Andrews: Well, the aspirational comment is the one that helps me coalesce the idea for what our data scientist has suggested and applied what Sanjula had suggested back in 2007. I was with a hospital that's a community hospital in the Midwest, a good hospital in their market, but certainly nothing more than a regional player.

When I had gone to talk to them about their clinical benchmarking, I asked them who their peer group was, and they said, why our peers are the Cleveland Clinic and Johns Hopkins. I smiled and thought, well, no, not exactly, but I appreciate the sentiment and that's the idea of aspirational benchmarking, that there's a lot of effort based on things that are never going to be relevant. So, you know, the Intermountain is a leader in clinical quality. Most people are not going to ever able to get to Intermountain's clinical quality, and there are market considerations that impact that.

Nobody competes with the Cleveland Clinic. As the new similarity index reveals, the top hospitals are not really even peers of each other. When you look at the data, the Cleveland Clinic's not similar to Cedar Sinai. Cedar Sinai is not similar to John Hopkins. John Hopkins is not similar to Mayo. They're each unique, because of a number of characteristics that separate them from each other. For a community hospital, or a traditional urban, suburban community hospital, of which they're about 2,000 in the US. Having an aspirational goal to do something like Mayo does, or like New York Pres. does, or like Cedar Sinai does, is not really a useful exercise. It's really just a waste of time. I was surprised the first time I used the tool last year, and I was again surprised this year.

Molly Gamble: You mentioned the top hospitals are not even peers of each other. The US News Hospitals ranking is about three years younger than me. Since working at Becker's, it's really been a framework for how I've come to see hospitals, especially when I was a writer, cutting my teeth. To look at tools like the one from Trilliant Health, you begin to see things a bit differently. You mentioned, I think of the honor roll hospitals, and even US News has acknowledged this themselves. They'll say those hospitals, the top 20, don't typically change the number they receive, or their precise ranking may differ a little bit year over year, but it's usually the same 20 hospitals. In looking at the Trilliant Health tool, I wanted to learn what surprised you, Hal? What assumptions did it challenge for you? Or what feedback have you gotten about what surprised people the most?

Hal Andrews: Well, we'll start getting feedback on Thursday of this week, I'll tell you a little background. When we started this last year, we first released markets, and then later in the year, we did hospitals. Those two things were separate. Frankly, a lot of the response to the release of Markets was confusion. People said, 'Well, I don't understand what that's about.' I've been fortunate in my career, I'm a lot older than you, but I have worked with Charlie Martin and with Hud Connery and Joel Gordon, who recently passed and I learned the importance of markets early on and have traveled all around the country. I've been in the halls of probably 700 hospitals.

I knew markets were important, and I knew that markets really determine the future of a hospital. As somebody buying hospitals in the nineties, our saying was, there aren't bad hospitals, there are only bad markets. So, we started with Markets and then we did Hospitals this year we merged those concepts. What we'll release on Thursday is a combination of the market characteristics, the competitive factors, the financial performance and the quality with the market considerations. What surprised me, even though I'm biased toward the importance of markets, was how much that jumped out in the identification of the peer cohorts.

One of the things we did differently this year was instead of just sort of laying out the Euclidian distance between different facilities, we grouped them into cohorts of 50 peers, as you suggested. What's really interesting is the distance that, the mathematical distance using Euclidian principles, between hospitals in the different cohorts. There are some hospitals where if you pick them again, like a Cleveland Clinic or Mayo or Cedar Sinai, or a New York Pres., their closest peers are not very close at all. There are other markets like Dallas, where you pick any hospital in the Dallas Metroplex, and what comes back are 20 hospitals that are virtually on top of each other from a Euclid distance perspective.

It really reinforced a couple of things; one is that markets really do matter. The peers are most likely to be peers of similar geographies, even some of the academic medical centers. You'll see the peers and you look at the markets they're in, and you see that this is an academic medical center in a state capital, and you start to see the similarities of, an Ohio State to a Duke or a UNC. When I first saw the data, I couldn't believe how much markets influenced your peer cohorts. Even with my bias toward that being the answer, the data suggests it's even more of an important factor for almost every hospital in the US than even I thought.

Molly Gamble: Yeah. I like the idea. There are stories to be told about hospitals and from hospitals, but then when you zoom out to a market level, boy, do things get interesting. There are a lot more dynamics in play. Over the course of your career, is there one market that stands out to you as highly interesting that has held your attention over the years?

Hal Andrews: Well, I think, there are a couple of markets. The best market I ever saw was New Albany, Ohio, where some doctors had gotten together and built a surgical hospital. I don't know, based on your age, which since I don't really know it, whether you are a fan of Shrek, but in Shrek, there's a scene when they show up in Duloc, they're going to see the emperor, they walk in and it's a shiny city, and everything's clean, it's like a fantasy land. New Albany was always that fantasy land market to me, but fantasies are not real.

So I think the market that is most dynamic is probably Houston. If I think about my background in Nashville, every new idea that we've had in Nashville about an alternate site for a hospital JV with a new entity or a JV with physicians, you always went to Houston. Part of that was because it was Texas and it was a big part of that's probably because there weren't really any zoning regulations in Houston or CON, so you could sort of do whatever you wanted to do.

I think Houston is probably the most dynamic in terms of site selection and alternate site services. I'm not really sure what the market is in this era of technology, which one's going to be most interesting from that standpoint.

Molly Gamble: It has been a tumultuous year for the legacy ranking systems, frameworks that have worked and been in health for decades, and how we assess the best are being challenged or questioned, but what good can come from this, Hal? At the same time, I imagine it also presents some challenges. What are those?

Hal Andrews: Well, from my standpoint, it would be helpful to reset what we're trying to do. So let's use the hospitals that have traditionally been highest ranked in US News as an example. If you look at the data and you realize that others aren't really like you, then your focus shouldn't be on trying to be like people who are excellent but are very different from you. The most prestigious hospitals in the US would probably be better off just focusing on being better themselves, on making their clinical quality better, on making their operational efficiencies better and not being distracted by the rankings. There's just no real value to thinking about somebody who's completely different from you.

I think for the average community hospital, it's a slightly different thing, which is A: don't think about people that you could never be like, but B: you think about who your true peers are and how you can improve against your true peer group. There was an example of this 15 years ago when HCAHPS came out and nobody knew what HCAHPS was. There was a two-year, or so, period where people were improving on it, and they were seeing how they improved, but they didn't, they didn't have any benchmarks to compare against.

A lot of people got lulled to sleep by comparing themselves to themselves and then finding out, well, everybody else had improved in that two-year period as well. Real benchmarking with your real peers has led to a tight clustering of health systems around HCAHPS. The same thing has happened with quality and so the data shows that there is some very tight clustering around readmissions that 1,500 or 2,000 hospitals are within decimal points of each other on readmissions. If that's the case, if you have figured out where you are with your peers on quality and patient satisfaction, well then maybe it's time to use that same mentality to think about who your peers are from financial performance and really from a value performance.

I think we're at a point in the US healthcare system where we have to start focusing on value; value from money as the Brits like to say, not necessarily value-based care as a program or an initiative, but delivering real value. I think by understanding who your peers are and comparing yourself to people that are like you, you can actually set some realistic goals. Back to markets; there are certain markets like New Albany that have an advantage that no other market's going to have. If I take a market like Brockton, Massachusetts, there are certain things that the people in Brockton, Massachusetts have to do. There are a lot harder than what you have to do in New Albany because of the market characteristics, whether that's poverty, whether it's inequities and provision of health, whether it's food deserts, whatever those things are, the market's really determined what you can do.

As Covid taught us, people really rely on hospitals when things are tough. If you're in a tough market and you've got a lot of challenges, it doesn't do any good to worry about what academic medical centers are doing or what hospitals and growth markets like Austin are doing, because that's not your reality. My hope is that by focusing on true peers, who your real benchmark should be, people won't waste as much time trying to be things they can never be, and instead will focus on how they can be their best for the communities they serve.

Molly Gamble: I imagine Hal too, in a time when hospitals are more closely managing and scrutinizing costs, I imagine being able to right-size expenditures to make more meaningful improvements based on pure data and peer comparisons is a lot wiser than perhaps spending in a lot of different areas to try to catapult to be of the level of Cleveland Clinic, to use your example.

Hal Andrews: I think that's right. I think we have to be real about demand. The demand for inpatient hospital services in the United States has been declining since 2008, which suggests that we are probably not in a state where everybody can be all things to all people. We like to talk about having multiple centers of excellence when the data shows from either a quality perspective or a reimbursement perspective, that people are good at delivering value on two or three things. In an era that is as capital constrained as this one, focusing on the two or three things that you really do well as compared to your peers, that you do efficiently as compared to your peers, is probably a better use of time and capital than having multiple bets in multiple service lines.

There are some markets where due to migration patterns from the pandemic and demographic patterns that were already in place, we're probably not going to need as many OB doctors and LDPs over the next few years. It's starting to be rational about where future demand will be, where you can meet that demand best for your community and pulling back on things that are declining in terms of demand, or things where you're just not as good at it, as the folks across the street.

Molly Gamble: Before we wind down, Hal, I mentioned in my opening remarks too, but this year the index includes rates data from health plan price transparency files and I wanted to check in on that as well. What can the index this year tell us about the quality of some of the most well-known, prestigious hospitals compared to the cost of the services they offer?

Hal Andrews: Well, Molly, I would distinguish a little bit between true cost versus negotiated rate. I think what the data shows back to the patient satisfaction and the quality example in every market that we've looked at, you can pick a single DRG or a handful of DRGs, and what you see is a tight clustering of quality and a wide variety of negotiated price. There's a lot of talk in Washington about whether that's due to market dominance and we're diving into that. I think the early returns on that, is reimbursement is something that's a relic of longstanding patterns. It's not, there are certain monopoly markets where they're getting a lot lower rate than a market with a number of urban competitors. I think the real thing is the focus in Washington on the cost curve and the fact that it doesn't bend, and the health plan price transparency is going to reveal that there's a wide range.

It can be 2X, 3X in some markets, we've even seen 8X difference between the negotiated rate for a DRG, whereas the quality is, is almost indistinguishable. I think the health systems have to start thinking about who the real customer is, which is the employer. The employer is 65% of the commercial matured book in the US, self-funded employers. Starting to understand what's a relevant market price, there are people who are doing a lot of volume at high quality for a much lower rate. There are, there are people doing slightly better quality at a much higher rate. It's different in every market; it's different for every DRG, but I know that has the attention of the folks in Washington because the call I had just before this morning was with staffers from one of the committees that are focused on price transparency.

I think it's something everybody in the health economy has to think about going forward.

Molly Gamble: You're right, I used cost there, but you're right, Hal, negotiated prices. You mentioned there are differences in negotiated prices where quality is not easily distinguishable. In that situation, payers are paying more for the same outcome, essentially. Is the reverse true where there are the same negotiated prices and big gaps in quality? Or is that a less likely scenario?

Hal Andrews: No, we haven't seen that pattern emerge yet. I think back to the negotiated price, part of that is if you think about the Affordable Care Act and you think about the MLR rates being established at 85% in most cases, the one thing is to say, we set what the minimum amount that payers would spend on healthcare. The flip side of that is we codified what their profit margin would be. There are actually not that many incentives for the payers, 65% of whose business is just managing claims for them to control costs. What you see in the data is there hasn't been any real focus on controlling the costs from the payer side. It's truly the only thing that can explain the disparity of a single-payer paying a 4X difference within the same market for the same service at equivalent quality.

I think the questions are really there for everybody to answer. There are some health systems that, maybe they have over the years leveraged a quality brand or market share into a certain rate, but pretty random. It's not obvious to me until I look at the data to see whose lowest paid for a thing versus highest paid for a thing for which they have equivalent quality. I think we're starting a new era of starting to understand quality and patient satisfaction, safety and negotiated rate. When you add all those things up, describe them as value.

Molly Gamble: Hal, thank you for being my guest today, Hal Andrews, president and CEO of Trilliant Health, listeners. Hal's made mention of this, Trilliant Health will release its SimilarityIndex™ | Hospitals Thursday, June 15th. So I encourage you to check it out and use the tool. It might challenge some assumptions or long-held ways that you look at hospitals and health systems, as it did for me. Hal, is there anything else you'd like to share with listeners in closing here? No, I think, Molly, you've covered it all and I really appreciate being your guest today.

  • Podcast
Sign up to receive Free weekly studies