How Price Variation Impacts Employers

Alexander Love: Hello and welcome to Insights Uncovered, a real world data podcast Series presented by MarketScan, a Truven data solution and brought to you by Global Data. My name's Alexander Love. I'm a senior editor at Global Data and I'm going to provide you with an introduction for this episode. This is the first episode in our, uh, second season exploring the evolving landscape of real world data and its role in shaping health care decisions. Today I am joined by Caitlin Quinn, who is vice president of data and analytics at uh Lantern, a specialty care platform for surgery infusions and cancer care navigation available through employers. And my second guest is John Linehan, who is Vice president of data strategy and operations at UH, MarketScan. If you've ever wondered why the same knee replacement costs dramatically different amounts depending where you are in the country, whether you're in New York City, Minneapolis or Chicago, and whether moving care to an ambulatory surgery center actually saves money, today's conversation will provide you with real answers. Not estimates, not posted prices, actual commercial allowed amounts from one of the largest claims databases in the world. Let's start with the introductions. Hi everyone. Welcome to the podcast.

John Linehan: Thanks, Alexander. Uh, it's great to be here. Uh, my name is John Linehan, Vice President of Operations and Data Strategy at Truven. Uh, for those who don't know, Truven is a platform offering real world data, healthcare analytics and member engagement for the employer health plan and life sciences community. Our flagship real world data set at Truvan is Marketscan. And Marketscan is gonna be the key component of what we talked through today. Although as we'll learn, um, the rest of the Truvin business really impacts the value of the Marketscan dataset. So today I'm really excited to be joined by Caitlin Quinn, Vice President of Data and analytics at Lantern Care. Caitlin, why don't you tell us what Lantern does in your role there.

Caitlin Quinn: Thanks, John. I appreciate it. Hello everyone. Um, yeah, so Lantern is a specialty care navigation company. Uh, we help self insured employers guide their members to high quality, effective providers for surgical and specialty procedures. Um, as VP of Data and Analytics, I lead the team responsible for the data and benchmarks that power how we build our network and advise our employer clients on those benchmarks as well as on the savings offered.

John Linehan: Excellent. And segueing off of that. Um, one of the main reasons we're recording this podcast today is that in March we actually co released a white paper, um, between Truven and Lantern Care, uh, using the market scan data to look at how the price of components of common elective Surgeries in the US varied across geographies. And so we'll spend a good amount of time at the beginning here of the podcast talking through what we learned. So Caitlin, I'll just dive right in. So, um, you know, elaborate as you wish on Lantern, but really, I'd really love to understand why did you want to publish such a deep and data driven analysis looking at elective surgery prices?

Caitlin Quinn: Yeah, great question. So Lantern's own price benchmarks are seen as self serving. We build on that data so clients and consultants push back. Naturally, uh, we needed an independent third party validation from a credible external source like MarketScan, uh, which is the gold standard. Employers naturally can't do this analysis themselves. Uh, because of sample size limitations. Even the largest employers don't have enough claims for procedure, acuity, site geography, ah, all simultaneously. And therefore plus the expertise, um, around that problem. And claims data is so messy that without that, those rigorous rules in place, you get a lot of nonsense. And so as you know, you know, we've even seen claims on a spreadsheet and from various sources that have a negative amount for a knee replacement. Right. So you do need to have some standardization rules in place in order to handle the amalgamation of various data sets and structures that you would expect to see. So this certainly allows us to do that and allows us to validate some of those benchmarks, obviously using such a gold, uh, standard source like MarketScan.

John Linehan: Yeah, I think you've come to the right place in that regard. So one of the things I love about MarketScan is really how we source the data. And so as I mentioned in the intro, Truvin is really a foundational, um, data warehousing data analytics platform. We work with, uh, healthcare benefits teams across large employers in the United States and have been doing so for over 30 years. And what that's allowed us to do is that as we collect, um, data of all types, claims, enrollment point solutions information, workers compensation, disability, etc. Um, we take advantage of the client base and the data that we have to create the de identified market scan data set which we use to establish benchmarks to help our clients, um, understand how their, their population looks relative to cost trends and of similar types of employers, uh, across the country. And one of the benefits that has for, for third party research, like what we did with you is just bring number one size and scale. Uh, there's 300 million cumulative patients in MarketScan at this point sourced by over 350 providers. So when you think about, uh, making sure that you have a view, not just within your client base but, but really across the market it's a, it's a perfect place to come. Um, the credibility is incredible. Uh, with 30 years on the market there's over 4,500 research publications and peer reviewed journals um, by organizations across academia, life sciences provider payer, um, uh and it's always helpful when you embark upon analytics as you know to do so with the data set that uh, stands up to peer review and then finally um, and I think we'll talk a lot more about this as we get into the findings. There are hard to find data elements that don't exist in all data sets and what we're really excited about with MarketScan is that things like actual costs of what the plan actually pays and not just the plan but what the patient actually pays are populated for over 90% of the claims in our data set. And so we get really, really granular view of um, the true cost exposure of care. And really that's the foundational point of the analysis that we did um, looking at elective surgeries. And so uh, with that my next question is really to dive into the findings. Uh and just curious when you saw the results of the analysis, what were two or three things that stood out most?

Caitlin Quinn: Yeah, I mean the paper was really comprehensive in the robustness of what an episode includes as well as the variation that we see in the amount in percentage to Medicare those costs um, really result in and that's across facility, physician uh, and surgeon fees and anesthesiologists. Um so you know it's expensive. We see around 175 to 290% above Medicare nationally. That's a useful heuristic for any employer doing commercial analysis. Um and it's geographically very variable. And so you know I think one thing that was really like really stood out was around NYC and Chicago are not playing the same ball game here. Right. Chicago is looked at from their very reasonable rates, um, especially lowest in the market. NYC is the opposite of that. It's all based on where you live in your current market to then analyze what should the benchmark be. And then the last thing is around ASC assumption is wrong. The ASC assumption in general is wrong. Pivoting site of care to just ASC is not the mechanism for lower and like smaller cost by episode. There's a lot of other factors that go into that including con law requirements and limitations by city and msa. Um, and that instance of ASE and the um, presence of those as well as the utilization of ASCs. Really change the cost, especially when you factor in physician fees. They are not one in the same for every asc. And so that was really interesting to see the ratio and how that relationship exists between the presence of an ASC relative to physician fees at that asc. Um, especially when you take that by MSA and by geographic location. Um, those are some of the things that we had had, uh, a lot of, of course, hypotheses about. But then seeing that in the data itself and with very specific examples was really, really helpful and eye opening.

John Linehan: Yeah, absolutely. I mean, I'm glad you mentioned the ASC topic. Um, you know, one of the things that, that we see, we hear a lot about across healthcare really is site of care and the desire to move care to, to quote, unquote, lower cost settings. Um, many employers assume that moving surgeries to ambulatory surgery centers will automatically save them money. But the reality is, as you pointed out that we found in the analysis was that that's not always the case. Uh, and there's significant implication based on not just geography but whether there are certain dominant practices that exist in those geographies, um, that really drive costs. And so it's a nuanced, it's a nuanced consideration that I'm sure you're focused on significantly with your employers. One of the things that um, we saw uh, in the paper was around anesthesia costs. Um, it was a striking finding throughout. Tell, um, us a little bit more about what you saw there.

Caitlin Quinn: So anesthesia was expectedly, um, and maybe in some cases unexpectedly elevated across all settings and all geographies. So there was really no exception to that rule. It is, in every facility it is elevated very significantly above Medicare. The national average was between 290 to 482% above Medicare depending on the procedure. A lot of times these can be the hidden and invisible components to episode costs, but they are, they make a massive difference to the total allowed amount and the end result of those of those procedures. And so NYC is an example. Outpatient anesthesia for knee replacement was 879% above Medicare. But even Chicago, as I mentioned, with the lowest cost market was at 427% of Medicare for knee replacement anesthesia. So there isn't any exceptions to the rule by geographic location. Um, it really is, um, one of those hidden costs that drives up the total, um, and it really was not dependent on site of care, um, whether it be asc, outpatient or inpatient. Um, so we found that to be really interesting also a bit of a hunch and hypothesis on our end, but actually having the benchmark data to prove that was very helpful. Helpful context.

John Linehan: Yeah, absolutely. I mean I think it just arms you so well with, with a perspective that you can bring to your, to your clients and kind of help them understand the various considerations and what one needs to consider in order to, you know, drive down cost trend and really maximize outcomes for your patients or your employee employees in this case. So I'd like to transition a bit. So marketscan is, as we've talked about and certainly was indicated in the findings is commercial claims data. Um, there's additional elements around uh, disability and productivity that are incorporated, social determinants of health, things like that. But fundamentally, ah, it's a longitudinal claims data set. So I'm curious, what do you wish you could see in analysis like that that claims just can't provide?

Caitlin Quinn: Yeah, it's a really good question of course from an analyst perspective. My question myself is what other data do I want to see in order to really drive a point home? And one of the biggest gaps that we see in this data, implant costs. So for joint replacements and spinal procedures, implants like titanium joints and spinal cages are expensive but rarely flow through the claims pipe. They come through as invoices or get bundled into facility fees. It's not unexpected, but it's just uh, a symptom of the architecture of our invoicing process. So nationally only about 170 joint implant billing encounters and three years of data, again not a data flaw, it's a system wide infrastructure gap. But it's one where it's a significant aspect and can have a really high price tag for those specific procedures. Um, and even some of our competing data warehouses in other respects really have missing implant data and 30% or above procedure types. Um, so it's a massive gap not just for us, not just for the broader landscape. It certainly is sort of an unknown. So the way that we really act as lantern as a specialty care platform are data takers. Uh, so whatever fields are present in the claims, EHR overlays, no machine readable file access, that's, that's a big gap for us as well. And so we have to do a good amount of assumption based and really identifying like how we fill in those gaps. And the third gap is incomplete episode bundles. So high cost surgical claims often have items not adjudicated within the claims lag window. Those again are deliberate reasons why we chose CPT code analysis for this, for this white paper versus episodes. So we did not have that lag or those gaps in the episode itself. Um, and so that that was, those are the few things that I really noticed is okay, what, what else would I, would I have wanted in order to continue to elaborate on this? And those are certainly ones that are not just specific to this white paper, but the landscape overall.

John Linehan: Yeah, thank you for that number one. And number two, I agree. I mean the nature of the data sets that exist and especially across claims just, you know, has inherent limitations. I think as you pointed out from a surgery perspective, um, the bundling, the bundling issue is significant. You know, it's very difficult to tell what happens within a bundle. And when you're trying to identify the component cost and the drivers within those component costs, it certainly provides challenges. That's as we think about how we build and continue to develop the market scan data set at Truvin, that continues to be a significant portion of our focus. Specifically how we can close gaps for our clients so that we can be as, as so for our clients who are data takers and your terminology, how we can be a one stop shop to be able to address all research questions and fill as many gaps as possible. Um, one of the things we're doing is, and it may even be relevant to a phase two of the work that we did is taking advantage of tokenization, uh, to allow us to link our claims data to third party data sets to open the aperture for what's possible in terms of analytics. Uh, later this month, uh, toward the end of May, we're launching a uh, partner ecosystem actually in which we have set up partnerships with organizations, uh, who have healthcare data, really across the spectrum, um, from electronic health records to labs to chargemaster data to really allow us to augment what's inherent in marketscan with other valuable data points that can open up uh, new research questions or answer more research questions. And I think, as I think about what your answer, there's really a couple places that might provide some interesting insight from an implant perspective. Uh, um, EHR data, we have a partnership that we announced with Varadigm um, about a year and a half ago which we have our market scan data linked to Varadigm data, um, which is electronic medical records in which there's some information around implants. But one of the areas that I'm excited to be launching as part of our ecosystem is bringing in charge master data where we can actually break through the bundle, so to speak, um, and be able to look at component pieces of uh, surgical procedures, uh, in an inpatient setting or surgery center setting. And who knows, that could be an interesting phase two analytics that we could do. Um, specifically to be able to capture that implant data that you won't find in a claim.

Caitlin Quinn: Yeah, really exciting stuff. Glad to hear there's so many moves being made on rounding out that data set.

John Linehan: Our plan is just always to allow be able to answer as many research questions for our clients as possible. And um, we think that um, it certainly increases the value of what we can offer to our client base but also just increases and augment enhances or magnifies the value of MarketScan itself. What I'd like to do now is really get a bit forward looking. So um, we walked through the analysis that we did and the findings and certainly anyone interested who's listening to the podcast can take a look at the white paper that's published on the Truvan website. But I'm curious about how this enables Lantern's offering going forward. So in practically speaking how is Lantern actually using the findings from the white paper? Does it change the conversations you have with your employer clients and their consultants?

Caitlin Quinn: Yeah, no. Great question and I'm really happy to say that it has been put to significant use really in two veins. Um, one is around credibility. A uh, reputable independent source validating that commercial claims are in fact 250% of Medicare is more powerful than Lantern saying it. We've done extensive research and legwork to actually get to that conclusion. But of course third party validation is, is really helpful and certainly necessary. So partnering on that has been a nice um, hypothesis, uh, testing and validation process. Um, and so that's been really a big game, a game changer for us and we're always on the hunt to continue to refine that and get even further clarity into that hypothesis um, as we take it to clients and consultants. The second is around methodology validation. So specifically the rule that removes episode bundles below 100% of Medicare as a floor commercial payers just do not pay below Medicare rates. And so we got pushback on this um, you know, and we have over the years. But the white paper provides evidence based support of that hypothesis. There is a right skewed distribution argument. Right. Healthcare spend is not normal and symmetric outlier trimming is wrong for, for this data. Um, because commercial carriers don't contract below Medicare for high cost surgery. We need to take that into account when we're considering the, the bottom five or you know the bottom 5% or the like 95th percent quartile. We, we certainly doesn't follow a normal distribution. Um, and then Further, it validates the component level CPT approach. Right. Like we, as we mentioned, it's the cleanest way to get an apples to apples benchmark. Um, there is not, not, not full episode cost but reliable and defensible approach. Using CPT codes is really the way that we get to that ultimate clarity and as specific as possible while getting to clean and complete data. So that's, it's, it's really been applied to multiple use cases, especially in a lot of our client and consultant convers. Um, it's been a really fruitful process and certainly a great outcome.

John Linehan: Yeah, thanks for that. And so just as from a scenario, right. So let's assume I run total rewards at an employer and I have responsibility to um, consider both my own cost trend, but my own, which therefore means I also need to think about the outcomes of my employees and their beneficiaries. So with all the variation that exists and that we saw in the analysis and cost costs in site of care in geography, what should employers actually do? What are the concrete steps that Lantern would recommend in this case?

Caitlin Quinn: Yeah, it's a great question, thanks for asking it because it's one of those things that we don't typically get to voiceover around, uh, providing advisement to employers at this point, but it's one that we feel really strongly about. So step one is really stop using national benchmarks for a local decision. Your market is your benchmark. You need to a benchmark site of care, uh, different or fees within those sites of care. Right. Physician anesthesia and really do the legwork to understand what in your geographic location are the standards and those benchmarks that you need to then use to then identify savings opportunities. Um, the second one is what we've already talked about asc, it's don't assume it automatically saves money. Right? Evaluate site of care and physician rates in your specific market, not generically. And this is also very influenced by con laws as we already described. So some places that have a 12% penetration or um, optionality around ASCs is very different from like a Minneapolis, it's 53% ASCs. Right. And so really taking that into account is really important and building that benchmark. Third is look at the full surgical episode. So facility surgeon anesthesia implants are relevant. Anesthesia is the visible cost driver as we had mentioned. So really take that into consideration, ask for that data, look at it with a very, a fine tooth comb and understand what percentage of the total that really makes up, um, demand methodology, transparency from your vendors. If someone gives you a Benchmark ask how did you handle outliers? Did you use symmetric trimming or floor? Did you account for incomplete bundles? Those are really important questions because you have to address this. You have to understand the assumptions going into any methodology or approach used um, with any, with any review of a potential savings approach. So those are really important questions and we get them all the time. We find them to be due diligence, we appreciate them and if they're not asked we suggest that they be. And so that's certainly a very important step in the process. And then fifth and final is invest in specialty care navigation. Of course the variation documented here is the opportunity steering members to high value providers in your specific market is more powerful than plan design alone. Um, obviously that's certainly part of the lantern value prop but that is an, a massive opportunity for any employer right now um, is to find you know really high value specialty care with the right price that's certainly more palatable to an employer and an employee. Um, and it's something that insurance you know built on so that it's, it's a lot of um. I hope this employers can take this as guidance. Um, as well as with the third party validation mixed with our robust data set it really provides the. Those employers with the tools to then go to market and act as advocates for themselves.

John Linehan: Absolutely. Data and analytics enable smart decisions is how I distill it. And uh, I think the advice is terrific and, and in the spirit of thinking of data analytics I think we'll close with what is often a challenging question which is to go forward looking. So think the next three to five years out. Um, think about from the employer perspective like what do you think employers will expect from data and analytics that they don't get today as we think three to five years out?

Caitlin Quinn: Yeah, I mean this is one of my favorite topics because it's also an expectation of ourselves. Right. Obviously John, yourself and I and things that we are like, like constantly day in and day are striving for. One of those is real time benchmarks. Right. So we use 2021 and 2023 data and it's already historical markets move inflation um is here especially in a post GLP1 uh market. So the landscape is evolving constantly and it's evolving in a very, a far shorter time frame than it ever has before. And so we need to keep up with the most relevant data set in order to drive to the most relevant benchmarks and outcomes. Um, we need to solve implant gap. We already talked about that. The EHR data and non Claims integration will eventually enable this. It's a roadmap item for both Lantern and MarketScan. But getting there sooner than later as well as agreeing on assumptions in the near term without that solve in place will be a game changer. Um, another like on the third point is quality linked to cost. So employers will want to know not just who is cheaper but who produces better outcomes. Right. So readmissions, complications, patient reported measures, those are all factors in quality and those need to be built out in a more comprehensive and robust way. Um, second to last is full episode transparency. So facility and physician anesthesia implants in one integrated view. Um, so transparency across the board would make I think all of our data analysis lives a little bit easier. But striving to that just as a general um, you know, baseline is will, will certainly drive a lot clear a lot more clarity in the outputs, um, and then predictive guidance. So AI driven member routing to the best provider for their specific condition and as well as dependent on their geography and their plan. That's going to be a massive impact, a positive impact for care navigation and it's something that we're really excited about going forward.

John Linehan: Yeah, well I can see you, you know that I appreciate the sense of urgency and the need to get there quickly and the amount of rigor is, is clearly something that I'm hearing loud and clear that you're expecting in the space and especially to enable you to support employers. We, as you mentioned our roadmap for this year talked about the work we're doing to expand data access, to answer the deeper, more nuanced questions and we continue to invest in the core mission of empowering high quality gold standard analytics and research. And we're really pleased that MarketScan was able to, to support LANTERN care and the work that you were doing. Um, so I think in closing, um, Caitlin, it's really been a pleasure to have you um, and just to hear about how you take uh, raw data and uh, use it to empower insight that you can bring to customers. The um, rigor that you bring, the work you do to steer employers and their employees and beneficiaries toward quality, um, is terrific and on the market from speaking for the Marketscan and Truven teams. We're just happy to be, be part of it. So thanks very much.

Caitlin Quinn: Thank you John. It's been a wonderful partnership so far and looking forward to the road ahead.

Alexander Love: Thank you Caitlin. And John, what comes through in this conversation is something important. Price variation in elective surgery isn't a new problem, but having granular market specific component level data finally gives employers and health plans real numbers to act on, not just estimates. The takeaways are clear. Commercial rates are expensive relative to Medicare, that variation is geographically determined, and the assumption that an ambulatory surgery center is always the cost effective choice is simply not supported by the data. For more on the research discussed today, the full white paper is available at uh meritiv.com real-world-evidence to learn more about Lantern's specialty care platform, visit lanterncare.com subscribe on your favorite podcast platform to stay up to date with insights uncovered. Until next time, thank you for joining us.