top of page

Dangerous topics in healthcare: Quality in the Medicare Advantage program

Updated: Jun 22, 2023

Now is the time to get grounded on this inevitable topic


Last week, I made an educational workshop about Medicare Advantage (MA) for our Strategy Boot Camp, which helps people who are newer to the conversation (e.g., new board members, out-of-industry executives, etc.) get conversant on essential topics. In general, we define essential topics as the ones with that special one-two punch of being (1) most required and (2) most dangerous, meaning that it's really easy to get yourself in trouble if you don't "speak" that topic well enough to get around.


MA was top of my list for this because MA takes the cake on both fronts.


The general dangers of speaking MA


On the 'danger' front, MA can be technical and complicated. Its situation is changing all the time, making it easy to get out of date. It's very sensitive (more on this in a moment). And the industry information environment is riddled with ‘selectively chosen’ information, because different stakeholders are competing hard to shape the overall narrative.


Together, all that means if you don't have enough context, it's easy to say something way off base—including accidentally quoting a specious data point or assertion—which undermines the soundness of your advice and your personal credibility.


But speaking MA is required


Anyone with any kind of thought partner or information-giving role has speak at least some MA, because so many different kinds of organizations, products and services rely on MA revenue that you can’t talk about those organizations' economics without talking about MA. In addition, the program is now heating up to full boil as a controversial pan-industry conversation topic. This makes it virtually certain that if you haven't already been pulled into conversations or giving presentations that touch on MA, at some point soon, you will.


For those not familiar, current hot topics include:

  • Cost: MA spending represents a major policy problem at the federal level. While MA is less costly than traditional Medicare to many enrollees, MA is viewed as being overall about 4% more costly to the federal government (1). Four percent may not sound like much until you think how many billions that is in absolute dollars each year—plus Medicare's overall high growth rate, the increasing share of the program that is MA, and so on.

  • 'Innovation catalyst'/public-private partnership/population-health support: The flip side of federal spending is also the revenue that flows to different organizations in healthcare—with all the different commercial benefits that entails—employment, innovation, etc. Health insurers, including their diversified services, are the highest-profile beneficiaries here, but providers, technology companies, services companies--the whole healthcare ecosystem pretty much—is involved. MA's risk incentives (a lot more on that shortly) are generally credited with creating a more friendly environment for 'population health'-type investments and activities.

  • Prior authorizations: Policymakers are raising concerns over MA organizations inappropriately delaying or denying needed services.

  • Over-coding: Widespread accusations of insurers, and to a lesser extent providers, 'gaming the system' of risk-adjustment by inappropriately adding unsupported diagnostic codes to increase payment.

  • Valuations of MA enablement companies: For those in the digital health/investment space, discussions of MA economics are tied to the fact that 'MA enablement' is a whole category of health tech companies. Some industry observers argue that these companies are over-valued (and there has certainly been a shakeout in recent months); other investors say this is still an area of growth and in fact a bright spot in digital health investment because the actual market need is so solid.

Today's focus: Getting savvy on MA quality


The hazards of talking MA are too big to fit into one blog post, so the focus for today is 'MA quality'. I picked quality even though that doesn't appear on the list above, because at some point any conversation about any MA hot topic, especially cost/spending, is likely to touch on the domain of quality, because of the necessity of talking about value for dollar. Plus, quality is always at risk of getting lost in the noise—and we should all fight to keep this from happening. So let's take a moment and get ready to watch out for these five danger areas when it comes to talking about MA and quality, and make sure you're sticking to the information high ground in each.

Recommendation #1: Be skeptical about broad assertions regarding quality—in MA and beyond. Always remember that healthcare quality is not monolithic. Quality spans diverse domains—experience, safety, access, avoidable utilization, mortality, etc.—and it is very difficult to generalize across them. It is highly unlikely, in the real world, for any program, plan, provider organization, intervention, or even nation to have a ‘straight-A's’ (or 'all-Fs') holistic quality scorecard. Instead, there will be nuanced quality highs and lows across different dimensions. So when you see a general claim about overall quality, look twice and check to see how ‘quality’ is being defined.


The technical side of quality data and frameworks also makes it easy to fall for misleading or flawed information. That means you must choose your information sources with extreme care and always read the fine print, including both study methodology and, frankly, authors. Online healthcare journals usually have “author affiliations” you can click to read. When reading about MA, click that “affiliations” plus-sign icon right out of the gate. It’s not usually necessary to reject published research based on the source, but it is helpful to know what kind of bias to check for.


Recommendation #2: Know that there are so many metric and framework issues in measuring MA quality that MedPAC has thrown in the towel, and does not assess MA quality at all. The list of barriers that Medicare Congressional advisory group MedPAC cites is long—but the short version includes three key points. First, there is no robust common source of data between MA and traditional Medicare. Second, MA quality is rated at the plan level, creating problems like the fact that the parent organization can combine ‘plans’ in ways that artificially boost the appearance of quality performance. And third, quality performance in general tends to differ by geography in important ways—and plan-level quality ratings obscure market-level differences. There is no good way today to compare program-to-program (MA vs. traditional Medicare) in a given market—or to assess market-specific quality within a given market, as a prospective beneficiary would want to do (2,3). And anyone who claims otherwise has an agenda you should probably figure out.


If you’re familiar with MedPAC, you know it has access to reams of data—and is highly motivated to be able to report on value for money in Medicare. So, when it says it simply cannot assess MA quality or compare across programs, I would take that seriously.


Recommendation #3: If you’re going to put aside MedPAC’s warning, and try to compare quality across programs, give credit where credit is due.


There are two obvious ways to go wrong characterizing MA quality--giving it too much credit, or not enough. Let's look at 'not enough credit' first, and tackle 'too much credit' next (in recommendation #4)


The dominant narrative about MA and quality at present is, “roughly the same quality as traditional Medicare, while being much more costly.” After reviewing the supporting information, I think it is more correct to say that MA overall has so far demonstrated a small overall quality edge, generally on the strength of its population-health/avoidable utilization performance over traditional/FFS Medicare (5,6). At the same time, it has shown problematic performance in some other quality dimensions.


Caveats: The WHY is important—spoiler alert, it may have more to do with the strengths of risk models generally, which is not an inherent attribute unique to MA --more on that in recommendation 4. And again, MA does not demonstrate better or even the same level of quality as traditional Medicare in EVERY study; there’s a troubling subset that talks about MA steerage patterns having negative impacts on quality, including adverse impacts on mortality after certain surgeries (4).


But, when someone argues that MA is 'same quality for much more cost," I think the real burning element is outrage about cost—and not a considered judgement on quality in itself. Separating the issues is difficult; cost is an aspect of quality, and to the extent that MA is over-spent upon, those are dollars that are being diverted from more necessary priorities. But if you're going to talk about quality, it's worth trying to talk about cost and quality separately, and tracking the quality highs and lows enough to present the nuanced and holistic story. Such as one can (remember, MedPAC refuses to do this at all, on the grounds of bad data.... and more caveats below).


Recommendation #4: Remember that the ‘why’ of improved quality performance—in MA or traditional Medicare—can be risk-based payment, which is neither a stable nor an exclusive aspect of either program's identity.


Here comes the part where MA, as a program, sometimes gets too much credit, and how you can avoid leaning on questionable talking points when talking about quality and MA. In short: What looks like better quality performance in the MA program may not be about MA itself at all, so much as it is about superior results from any risk or value-based-care (VBC) model—and of course, traditional Medicare contains those models too.


On the whole, risk/VBC models tend to achieve higher quality than pure fee-for-service (FFS) models, especially if quality is being measured as preventing avoidable low-value utilization (e.g., cutting down on avoidable hospitalizations), and especially if they contain two-sided risk (upside and downside). This assertion has support within traditional Medicare (7) and within MA (8). VBC's tendency to boost quality is great news except for the fact that it acts as a water-muddying confounding variable when we try to compare quality across programs (MA vs. traditional Medicare). That is: It may be that the different preponderance of risk programs are causing any differences in quality performance. Again, different preponderance of risk is neither a stable nor an inherent attribute of either program.


Preponderance of risk in MA is currently higher, and this could explain the “quality edge” that different researchers have found in program-to-program comparisons (to the extent that such a thing is feasible at all — etc., etc.). It's particularly likely that risk/VBC is the active ingredient if the 'quality performance' being cited is really a set of process metrics based on population health principles. Example measures include: rate of enrollees reporting that they received population-health-type supports (such as a prescription drug review), a usual source of care, providers coordinating care and responding to concerns, and receiving written discharge information after a hospitalization (9)


Thus a more sophisticated quality argument for MA would be, 'it's managing to pull more providers into risk arrangements, and risk arrangements are helpful in boosting quality'. But instead of that argument, what I see instead are problematic apples-to-oranges comparisons showing up in industry literature that just claim the risk-quality mantle for MA to a degree that isn't correct. Specifically, I found at least one journal-published study that claimed to compare MA and Medicare quality overall, when a close read shows that the authors were actually comparing quality outcomes of one program’s risk models (apples) versus the other program’s pure-FFS approach (oranges). Big mistake if we’re giving these authors the benefit of the doubt—or, actively misleading if not.


Recommendation #5: Remember that MA is not synonymous with risk/VBC – both MA and Traditional Medicare contain a mix of risk and FFS. Here’s a related point. It’s common in the industry for people to talk about MA as if the whole thing is a risk-based or VBC model—and to make comparisons between it and 'FFS Medicare,' as if all of Medicare is FFS. But both are major oversimplifications.


In fact, both MA and traditional Medicare have a mix of risk/VBC vs. FFS program participation. MA has a higher preponderance of (at least two-sided) risk within it, which is especially important to keep in mind if it helps explain differences in quality performance at the overall program level. But both programs are a long way off from being 100% VBC/risk (or 100% FFS, or 100% anything) (10-15). Spending time with this point caused me personally to stop saying "FFS Medicare" to make comparisons with MA and switch to saying "traditional Medicare" instead.





Conclusion: proceed with caution—and context


The bottom line on talking quality in any healthcare strategy context, MA and beyond, is always the same: It’s a challenge to talk about quality in a way that is both accessible and correct, but quality is a necessary and important part of the conversation. So go ahead and talk quality, including in conversations about MA--just make sure you're savvy on the basic ways to go wrong, and how to stay on solid information ground.


Read the fine print.... remember metric flaws, give credit where credit is due... consider whether risk is the reason ...and remember the risk picture is mixed. There you go -- easy!



Let’s Talk Some More


At Union we have a passion for ensuring everyone in healthcare can participate in strategy conversations. Reach out to learn more about our Strategy Bootcamp curriculum, which is perfect for new board members, out of industry execs, new-to-general-management-role clinical leaders, and client-facing teams. Don’t hesitate to reach out to us at info@unionhealthcareinsight.com or contact me directly at amanda@unionhealthcareinsight.com.



Notes and Sources


(1) Estimated overpayment has varied by year; currently (as of 2022) pegged at about 4%. See figure 12-3, p. 431, MedPAC March 2022 Report to Congress. Available at: https://www.medpac.gov/wp-content/uploads/2022/03/Mar22_MedPAC_ReportToCongress_Ch12_SEC.pdf


(2) MedPAC March 2019 Report to Congress. See Chapter 13, ‘Medicare Advantage Status Report’, subsection “Quality in Medicare Advantage is Difficult to Evaluate’ . Available at: https://www.medpac.gov/wp-content/uploads/import_data/scrape_files/docs/default-source/reports/mar19_medpac_ch13_sec.pdf


(3) MedPAC March 2022 Report


(4) Higher mortality after some cancer surgeries, seemingly because patients were likelier to go to a lower-volume, not NCI hospital: Raoof M, Ituarte PHG, Haye S, et al. Medicare Advantage: A disadvantage for complex cancer surgery patients. J Clin Oncol. Published online November 10, 2022. doi:10.1200/JCO.21.01359 available at: https://pubmed.ncbi.nlm.nih.gov/36356283/


(5) MedPAC March 2019 Report to Congress; Newhouse, J. and McGuire, T ‘How Successful Is Medicare Advantage?’ Milbank Quarterly, June, 2014. Available at: https://www.milbank.org/quarterly/articles/how-successful-is-medicare-advantage/


(6) Systematic review of peer-reviewed papers published between January 1, 2010, and May 1, 2020. Dugoff, E. et al. “Quality, health, and spending in Medicare Advantage and traditional Medicare”, American Journal of Managed Care, Sept. 2021; available at: https://pubmed.ncbi.nlm.nih.gov/34533909/. See also Agarwal, R. et al, “Comparing Medicare Advantage And Traditional Medicare: A Systematic Review,” Health Affairs, June, 2021. Available at: https://www.healthaffairs.org/doi/pdf/10.1377/hlthaff.2020.02149


(7) Lewis, C. et al. “The Impact of the Payment and Delivery System Reforms of the Affordable Care Act”, Commonweath Fund, APRIL 28, 2022. Available at: https://www.commonwealthfund.org/publications/2022/apr/impact-payment-and-delivery-system-reforms-affordable-care-act


(8) Gondi S, Li Y, Antol DD, Bourdreau E, Shrank WH, and Powers BW. Analysis of value-based payment and acute care use among medicare advantage beneficiaries. JAMA Netw Open.Published online March 17, 2022. doi:10.1001/jamanetworkopen.2022.2916 . Available at: https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2790209


(9) Jacobsen, G. et al. “Medicare Advantage vs. Traditional Medicare: How Do Beneficiaries’ Characteristics and Experiences Differ?’ Commonwealth Fund, October 2021. Available at: https://www.commonwealthfund.org/publications/issue-briefs/2021/oct/medicare-advantage-vs-traditional-medicare-beneficiaries-differ


(10) LAN’s survey of five states, 73 health plans and traditional Medicare found steady participation in alternative payment models overall. It found 40.5% of Medicare and plan spending in 2021 was devoted to providers in a fee-for-service model with no link to quality or value, slightly up from 39.3% in 2020. Health Care Payment Learning and Action Network (HCP-LAN). Available at: https://hcp-lan.org/workproducts/apm-infographic-2022.pdf ; King, R. “LAN report: Medicare Advantage plans increased spending on 2-sided risk models in 2021”, November, 2022. Available at: https://www.fiercehealthcare.com/payers/lan-report-medicare-advantage-plans-increased-spending-two-sided-risk-models-2021


(11) Gondi et al


(12) Includes discussion of HCP-LAN data from HCP-LANB’s APM Measurement Effort (2021 APM data) Raths, D. “Panelists Discuss Steady Growth of Two-Sided Risk Models,” Healthcare Innovation (formerly Healthcare Informatics), November, 2022. Available at: https://www.hcinnovationgroup.com/policy-value-based-care/alternative-payment-models/article/21286721/panelists-discuss-steady-progress-on-twosided-risk-models


(13) Humana. “Value-based Care Report: Highlighting physician progress and patient outcomes based on calendar year 2021 data,” see member data note in sidebar on (non-numbered). Page 7 ; available here: https://docushare-web.apps.external.pioneer.humana.com/Marketing/docushare-app?q=IQrAcqeO%2fyLLnx6jG3tuPw%3d%3d


(14) Gondi et al


(15) “489,796 beneficiaries, of whom 16.6% were in an FFS model, 32.4% in upside-only risk, and 51% in 2-sided risk” – Gondi et al


Comments


Join our mailing list to see future posts

Thanks for submitting!

bottom of page