Michael E. Chernew, PhD

 To measure variation in spending and inpatient prices associated with the primary care physician (PCP) practice to which patients are attributed.

 Cross-sectional analysis of claims data.

 We used random effect models to estimate case mix–adjusted spending across large PCP practices within 3-digit zip codes. We compare inpatient prices for patients in high-spending practices with those in low-spending practices.

 The physician practice to which a patient was attributed is associated with significant differences in spending after controlling for patient comorbidities and geography. Patients attributed to practices in the top quartile of total medical expenses have about 30% higher spending than patients attributed to practices in the bottom quartile of adjusted spending in their 3-digit zip code. If patients attributed to practices in the top 2 quartiles had spending equivalent to those in the median practice, total spending would drop by 8%. Price variation accounts for a meaningful amount of the variation, with inpatient prices 17% higher in top-quartile vs bottom-quartile practices. We cannot disaggregate the large variation in utilization into practice patterns and unmeasured case mix (including unmeasured differences in patients’ socioeconomic status) vs random health shocks, but correlation in spending patterns across years suggests that some persistent differences in spending patterns exist.

There are meaningful opportunities to reduce spending by changing patient PCP selection, encouraging patients to use lower-priced specialists and hospitals, and eliminating wasteful care. Attention must be paid to the best ways to reap these savings.

Physician practices account for a significant amount of variation in spending.

There is widespread variation in spending associated with physician practices.

This likely reflects variation in utilization and in the prices of care delivered by the practice and the sites they refer to.

Significant savings can be obtained if patients shift from practices associated with high spending to those with low spending.

Although the country is making preliminary steps toward recovery, the coronavirus disease 2019 (COVID-19) pandemic has sent the United States into its most significant economic downturn since the Great Depression, which will significantly exacerbate pressure for employers to control health care spending. One strategy available to employers is to encourage the individuals and families they cover to use more efficient health care providers.

A large body of literature documents the significant variation in prices within markets, from which it can be inferred that considerable cost-saving opportunities are associated with price shopping.

 Importantly, physician practice prices have not been found to be associated with quality or efficiency metrics.

 The literature on practice style variation suggests that there may also be savings opportunities from identifying physicians with less intensive practice styles.

 For example, physician beliefs about treatment regimens contribute to significant regional differences in physician practice spending.

 Similarly, physician experience is associated with spending, as less experienced primary care physicians (PCPs) and specialists have higher overall costs.

Much of the practice variation literature focuses on market-level analysis, which masks widespread variation within geography.

 For this reason, the National Academy of Medicine has called for greater understanding of variation attributable to physician practice styles.

We extend the nascent research on this topic using unique data on commercial prices and spending. We focus on variation in total medical and prescription drug spending attributable to PCP groups, as well as variation in inpatient prices. Specifically, PCPs influence primary care and have some sway over the frequency with which their patients visit specialists, as well as which specialists and facilities their patients visit. There is considerable price variation by specialty practice and site of service.

It is important to note, however, that the extent to which PCPs can influence specialty care and overall utilization varies and can be limited. Although PCPs can influence how care is utilized by serving as gatekeepers to other services, their impact is limited by a variety of factors. For example, patient preferences regarding specialist visits are associated with higher specialist utilization, suggesting that other factors besides PCP referrals can play significant roles in care usage.

 Moreover, research has shown that patients are often unwilling to sacrifice time or money to maintain continuity of care with PCPs; such continuity is associated with better outcomes and more cost-effective care.

 However, given that PCPs serve as a first contact and main provider for most individuals, identifying variation in practice spending and prices still provides valuable insight into how best to utilize primary care efficiently.

Using medical and pharmacy claims for commercially insured adults from a large, national health insurer, we analyzed the variation in per-enrollee total health care spending associated with large primary care practices in 2017 and 2018. Using an approach similar to CMS’ method of attribution of Medicare beneficiaries to accountable care organizations, we attributed each individual to a primary care practice.

 Our baseline analysis was restricted to continuously enrolled adults aged 18 to 64 years with at least 1 primary care visit during the sample period and excludes pregnant women. To attenuate concerns related to sampling error, we further restricted the analysis to large practices with a final sample size of at least 50 enrollees in both 2017 and 2018 and pooled the data over both years. This yielded a sample of 2 million commercial enrollees with both medical and pharmacy coverage. We calculated total annual medical and pharmacy spending for each enrollee, including both the insurer-paid amounts and the enrollee cost-sharing amounts.

To measure variation in case mix–adjusted spending, we used random effect models to estimate the spending associated with each primary care practice, controlling for their patients’ characteristics, age, sex, and comorbidities, measured using both the Elixhauser Comorbidity Index (Healthcare Cost and Utilization Project Elixhauser Comorbidity Software) and Agency for Healthcare Research and Quality comorbidities. We also controlled for region using 3-digit zip codes. Importantly, this risk adjustment accounts for clinical conditions but not variation in use within those conditions. For example, the risk adjustment will account for presence of cardiovascular disease but not for which of the patients with cardiovascular disease experience serious events (eg, heart attacks). The random effects models, using 2 years of pooled data, generate a robust estimate of relative spending per practice adjusted for patient covariates and for randomness associated with smaller practice size. The estimates reflect variation in practice styles and referral patterns, as well as variation in unobserved case mix or variation in partial-year enrollment. We then classify each practice into quartiles. Because we controlled for 3-digit zip code, the quartiles should be interpreted as relative to other practices in their 3-digit zip code.

To assess the role of variation in inpatient prices in driving variation in spending, we focused on inpatient prices because they are easier to measure and because inpatient care drives a substantial portion of the overall spending variation. Specifically, for each admission in the 20 most common diagnosis-related groups (DRGs), we computed inpatient price as the total allowed amount for the facility claim. We regressed prices on the quartile of spending for the patient’s PCP practice, DRG fixed effects, Elixhauser comorbidities, 3-digit zip code, age, and sex. To reduce measurement error associated with coding, we limited this analysis to admissions where the patient had only 1 admission in the month.

We found significant differences in adjusted annual per-enrollee health care expenditure across primary care practices within their respective 3-digit zip code. Patients in practices at the 75th percentile have 31% higher annual expenditure than patients in practices at the 25th percentile, even after adjusting for covariates. If patients in above-median practices had spending levels of comparable patients in median spending practices in their area, mean per-enrollee spending would drop by approximately 8%, or about $562 per year, as shown in 

These spending differences are primarily driven by differences in facility spending (both inpatient and outpatient) and prescription drug spending. As reported in 

, per-enrollee mean inpatient spending accounts for a smaller share of total mean annual spending (16%) than outpatient facility spending (35%). However, inpatient and outpatient settings account for comparable amounts of the excess spending among above-median practices (24% and 30%, respectively) (

To better understand the variation in inpatient spending (which lends itself better to price measurement than outpatient facility spending), we compared mean per-inpatient admission spending for the 20 most common DRGs in practices in the upper quartile of spending in their 3-digit zip code to that of practices in the lowest quartile. After controlling for DRG, patients’ characteristics, age, sex, and comorbidities, we found that patients at practices in the upper quartile of the spending distribution had case mix–adjusted inpatient prices that were, on average, approximately 17% higher than practices in the lowest quartile. These differences were still 11% and 8.5% when the upper quartile was compared with the middle lower and middle upper quartiles, respectively (

). The 17% figure suggests that variation in the price of hospitals where PCPs’ patients receive care is an important contributor to the variation in spending associated with PCP practices. Yet it is well below the 31% estimate of spending differences, thus suggesting that utilization differences may contribute to the variation. Some of this variation in utilization may reflect inefficient care, whereas some of it may reflect unmeasured case mix or random health events. For example, individuals who start the year with similar risk may experience different health events over the year. Although the random effects models will address some of the random variation, we do not speculate as to the extent to which the differences in utilization are due to inefficient practice patterns as opposed to unmeasured case mix. However, we know that randomness cannot account for all of the variation in spending. As a robustness check, we ran our main spending specification separately for 2017 and 2018. We found that 60% of the practices in the top quartile in 2017 are in the top quartile in 2018; if variation was just random, this share would be 25%.

As an additional robustness check, we conducted a permutation test in which we randomly assigned each patient to our practice identifiers and reestimated our models with the “fake” practice identifiers (

). In doing so, we found only a 16% spending “difference” between the 75th percentile and 25th percentile and savings potential of 5% if patients at above-median practices had spending levels comparable with those of patients at median spending practices in their area. Because our estimated effects are roughly twice the estimates obtained by randomly assigning patients to practices, we are more confident that our estimates are not purely spurious.

Our analysis identifies significant within-market variation in risk-adjusted spending associated with the primary care practice to which patients are attributed. A large part of this variation is likely due to the prices of the hospitals (and likely specialists) to which the practice refers. Some may also be due to practice styles, such as propensity to refer or to order various diagnostic tests.

These estimates have several caveats. First, these estimates capture only associations and are inherently imprecise. Although we pooled data over 2 years and required that practices have at least 50 patients per year, unmeasured health status or socioeconomic status characteristics of attributed enrollees, as well as the skewness of health care spending, may present a significant potential for confounding. For example, it could be the case that the patients at “high-spending” practices have health conditions that are costlier in ways that we do not account for. A primary care practice that excels at treating patients with multiple chronic conditions may draw patients with disproportionately complicated conditions; even if such a practice were identified to be in the top quartile of the spending distribution, it would not necessarily be the case that their patients’ care would be less costly at any other practice in their area. Our analysis of prices is less affected by this concern. Second, and most importantly, we do not assess whether the outcomes at higher-spending practices are better; thus, additional research is needed to assess the relationship, at the practice level, between spending and outcomes.

The economic downturn associated with COVID-19 is exacerbating pressure on employers to control health care costs. The wide variation in spending within markets, including that driven by price variation, demonstrates significant opportunity for savings if patients are directed to PCPs with more conservative practice styles, resulting in more efficient utilization patterns or referral patterns that steer patients to lower-priced sites of service.

There is likely no magic bullet for reaping these savings. Benefit design, tiered networks, and other instruments can play a role by encouraging patients to use high-value physicians who deliver affordable, high-quality care and by encouraging high-spending practices to examine their behavior, such as by shifting referrals to lower-cost sites of service. Similarly, innovative payment models can encourage efficient utilization and may promote lower-cost referral patterns. Finally, some policy response to address exceptionally high prices may be warranted. For example, some proposed policies to address surprise billing include provisions that effectively place soft or hard caps on provider prices, which could offer a model for broader policy efforts to address imbalances in market power. Regardless of one’s preferred solution, the documentation of wide variation in total medical spending among PCPs demonstrates that practice style variation is an important factor contributing to unnecessary and potentially avoidable health care costs. Additional research should explore the specific practice characteristics (eg, size, ownership) related to cost variation, utilization, and referral patterns.

 Harvard University (SM), Cambridge, MA; UnitedHealth Group (EK, EDP, HT, LGS), Minnetonka, MN; Healthcare Markets and Regulation Lab and Department of Health Care Policy, Harvard Medical School (MEC), Boston, MA.

 Ms Koep, Dr Parker, and Dr Sandy are employees of UnitedHealth Group and own UnitedHealth Group stock. Dr Trenz is an employee of UnitedHealth Group. Dr Chernew is an unpaid board member of Health Care Cost Institute; is an unpaid advisory board member of Blue Cross Blue Shield Association and Blue Health Intelligence; is a paid advisory board member of National Institute for Health Care Management and Virta Health; and has received speaking honoraria from America’s Health Insurance Plans, Blue Cross Blue Shield of Florida, HealthEdge, Humana, American College of Cardiology, and American Medical Association. Mr Moy reports no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

 Concept and design (SM, EK, EDP, LGS, MEC); acquisition of data (EK, EDP, LGS, MEC); analysis and interpretation of data (SM, EK, EDP, HT, MEC); drafting of the manuscript (SM, EK, EDP); critical revision of the manuscript for important intellectual content (SM, EK, EDP, HT, MEC); statistical analysis (SM); administrative, technical, or logistic support (EK, EDP, HT); and supervision (HT, LGS).

 Michael E. Chernew, PhD, Department of Health Care Policy, Harvard Medical School, 180 Longwood Ave, Ste 207, Boston, MA 02115. Email: chernew@hcp.med.harvard.edu.

1. Cooper Z, Craig SV, Gaynor M, Van Reenen J. The price ain’t right? hospital prices and health spending on the privately insured. 

. 2019;134(1):51-107. doi:10.1093/qje/qjy020

2. Roberts ET, Mehrotra A, McWilliams JM. High-price and low-price physician practices do not differ significantly on care quality or efficiency. 

. 2017;36(5):855-864. doi:10.1377/hlthaff.2016.1266

3. Van Parys J, Skinner J. Physician practice style variation—implications for policy. 

. 2016;176(10):1549-1550. doi:10.1001/jamainternmed.2016.4433

4. Cutler D, Skinner JS, Stern AD, Wennberg D. Physician beliefs and patient preferences: a new look at regional variation in health care spending. 

. 2019;11(1):192-221. doi:10.1257/pol.20150421

5. Mehrotra A, Reid RO, Adams JL, Friedberg MW, McGlynn EA, Hussey PS. Physicians with the least experience have higher cost profiles than do physicians with the most experience. 

. 2012;31(11):2453-2463. doi:10.1377/hlthaff.2011.0252

6. Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending. part 1: the content, quality, and accessibility of care. 

. 2003;138(4):273-287. doi:10.7326/0003-4819-138-4-200302180-00006

7. Fisher ES, Wennberg DE, Stukel TA, Gottlieb DJ, Lucas FL, Pinder EL. The implications of regional variations in Medicare spending. part 2: health outcomes and satisfaction with care. 

. 2003;138(4):288-298. doi:10.7326/0003-4819-138-4-200302180-00007

8. Leavitt MO, McClellan M, DeVore SD, et al. Competencies and tools to shift payments from volume to value: a vital direction for health and health care. National Academy of Medicine. September 19, 2016. doi:10.31478/201609o

9. Anthony DL, Herndon MB, Gallagher PM, et al. How much do patients’ preferences contribute to resource use? 

. 2009;28(3):864-873. doi:10.1377/hlthaff.28.3.864

10. Pereira AG, Pearson SD. Patient attitudes toward continuity of care. 

. 2003;163(8):909-912. doi:10.1001/archinte.163.8.909

11. CMS, HHS. Medicare program; Medicare Shared Savings Program: accountable care organizations. 

Articles

Variation in Spending Associated With Primary Care Practices

 As you both know, this is the last interview in 2020 that the journal will run as part of a monthly series. We didn’t expect 2020 to turn out the way it did, and 2021 will start off with a new president taking office in January. What would be your priorities for President-elect Joe Biden in Medicare, Medicaid, and managed health care overall?

The first thing, whether they want to or not, there’s going to have to be some attention to the Part A trust fund. It’s scheduled to be depleted in 2024, and some strategies are going to have to be undertaken to figure out how to maintain the solvency of the trust fund. The second thing that I think is going to be particularly important for the new administration is to chart a path for promoting alternative payment models. There’s been a lot of movement in the area of alternative payment models; some things have worked and some things have not worked. I’d like to see a strategy for beyond [this administration] to be developed as we begin to transform the health care system. The third thing that I think you will see from this administration is that it would be a priority to strengthen where the Affordable Care Act (ACA) is in its implementation. There are a number of actions taken largely by executive order in the current administration that I think the new administration will revisit, and I think it’s important to lay out a path for the ACA for the next 4 years and beyond.

First off, it’s really great for Mike and I to be the finishing touches of the 25th anniversary celebration for 

. I think there are a lot of us who actually want to see the year 2020 come to an end for a variety of reasons. But as we turn to a new year and a new administration, I think my points would be to continue to grow on our bipartisan work, emphasizing that there is more than enough money in the US health care system. We need levers, like Mike just mentioned, like alternative payment models, that will hopefully shift some of the money more to the services that make Americans healthier and away from those that don’t. More specifically, as we first published in 

 in 2001, not quite 25 years ago, but getting close to 20 years, the idea of value-based insurance design (VBID), which aligns out-of-pocket costs with the value of services as opposed to setting consumer cost sharing at the price of the service, was included in the Biden platform. It has been used in both Republican and Democratic administrations. So it’s our hope, particularly with a split government with no major health care policy changes on the horizon, that there will be continued attention to aligning health care costs with the essential nature of clinical services, that not only will barriers be removed for coronavirus disease 2019 (COVID-19) vaccines and COVID-19 treatments, but hopefully this would extend to other high-value services that are particularly used in the area of chronic diseases, and that we could pay for the added generosity of that coverage with the identification and reduction of low-value care, as put forth by Mike and I in a template called VBID X, a benefit design that lowers cost sharing on certain high-value services and is paid dollar for dollar by increasing cost sharing on either specific services or line items of services like high-cost imaging and nonpreferred branded drugs.

Dr Chernew, do you want to talk more about alternative payment models, particularly some of the efforts that the current administration has made in chronic diseases?

 My broad thought is that we need some coherence in the alternative payment models, that we’re running too many of them, that they’re often conflicting, and that we haven’t projected where they are going. So, although I think there are some particular clinical areas of interest, I would rather see some statement about the role that, for example, population-based payment models like [accountable care organizations] will play relative to episode-based models that you see a whole series of. I think it’s been very challenging to be successful with the number of new models introduced, and I actually think we’ve gotten to the point where we’re introducing too many new models. And the new models, even if a model itself would be good standalone in this environment, are actually being counterproductive, because they’re distracting people given the range of choices of models they might be in or not be in and how they run. I would encourage the administration, with the Center for Medicare and Medicaid Innovation (CMMI), to just slow down in terms of the number of models instead of finding new and interesting ones.

 Dr Fendrick, when you talked about VBID and VBID X, does that require any sort of political will to get consumers and patients to accept it?

That’s a really good question. I think one of the reasons Mike and I would say we’ve had the amount of success we’ve had with [VBID] is its intuitiveness and its bipartisan and multistakeholder support. I mean, on the surface, who’s not going to support more of the good stuff and less of the bad stuff? So what we feel is that with the COVID-19 pandemic, lots of pain, lots of hardship, but as colleagues of mine from Tufts and I published in the September issue of 

 there might be a silver lining in the fact that the dramatic reduction in both high- and low-value care that Mike and his colleagues from Harvard have continuously updated us on allowed us this ability to really inform all stakeholders that we could realign payment, benefit design, and other levers to increase efficiency, in that, while I’ve not been overly sympathetic to the provider community over my career, I’ve written in multiple venues that it’s time to start paying clinicians and delivery systems more for some of these high-tech services that provide great value and hold them accountable for the services that might be very expensive and profitable but that don’t make Americans any healthier. We have seen some movement, like all things VBID, modest to this effect. But I’m hopeful that given the dramatic decreases that we saw in high-value care and sadly are likely to see again with a second COVID-19 surge, we’ll prioritize not just access but actually be clinically driven in how we spend our dollars moving forward.

Some people we’ve talked to this year have talked about the amount of waste in the health care system, and that if we cut out waste, that would be enough to create some of the solutions that we’re looking for. Would you agree with that, looking at waste in staffing, procedures, inefficient processes, and those sorts of things?

 Yeah, I’ll jump in here, because Mike has been very consistent over the 20-plus years that I’ve known him, and the 15-plus years we’ve worked on 

, to remind me that almost all the services that we’ve recommended to increase access to are cost-effective and not cost-saving, and it means that if you continue to push depression management, blood pressure management, diabetes management, better cancer care, and even rare disease [treatments] that are cost-effective, your total cost of care will go up and not go down, nd particularly as we’ve been pushing to have lower consumer cost sharing, the purchaser had to actually pick up a greater portion of that total spend. And I’ve finally succumbed to Mike’s recommendation 5 years ago, with great chagrin, to become a low-value care guy. The good news is that for select high-value services, not the whole gamut, there’s enough low-value care, if we were to have the courage to get rid of it, that we would be able to provide increased affordability. But as Mike reminds me, the pipeline of high-value but not cost-saving services is so great that we really need to not only focus on “increasing volume, good” and “decreasing volume, bad,” but we have to start paying attention to prices. When all the waste is gone, then I hope to be retired, because that’s going to turn over entirely to Mike, because that’s when it gets really hard.

 I want to echo everything that Mark said. I do believe that reducing waste is important. I believe that there is a lot of waste in the health care system. I also believe that we are currently able to get our hands on only a small amount of that waste. Because waste, as Mark will tell you, is quite nuanced. It’s not as if you could cut a particular service for everybody and get rid of the waste. It’s really who the service is delivered to that matters, and that is complicated to get at. In fact, my affinity for alternative payment models is because they set the financial preconditions for efforts for organizations to actually begin to go after waste. So if you look at what they’re doing in Virginia, for example, with Beth Bortz and Smarter Care Virginia, you will see delivery systems really working on getting rid of waste, but at least right now, there is not enough money there to fund all the things that we would broadly want to fund in terms of high-value services. And more to the point, independent of that, prices are an increasing problem, particularly in the commercial sector, with industry consolidation, and there’s growing work that suggests that the variation in prices is enormous, with relatively little evidence that the high end is providing high quality. There are a lot of specific examples and things one might do, but it will take a lot of effort for the government to go after those prices. Frankly, I don’t think much of that’s going to happen. At the federal level, you asked my priority to the administration in terms of Medicare and Medicaid, and managed care in general, but I will add, there are other priorities, like getting rid of surprise billing, for example, that would be very high on my list of things to do that might not fit into the first question and addressing really excessive pricing by consolidated organizations. I think those are important things, but they transcend the government’s role as a purchaser more into the government’s role as a regulator.

One thing that hasn’t been mentioned by either of you yet is drug prices, which started out with a big splash early on in the Trump administration but have not come to any sort of solution as of this moment.

 Most Americans don’t care about drug prices; they care about what it costs them. And I’m not sure how many of our listeners are familiar with the Hundred Acre Wood in Winnie the Pooh, but I’ll give the more Tigger message before Mike follows up with the Eeyore one. I’ll say that there is bipartisan support; there is multistakeholder support around more marginal initiatives, and I agree with Mike that they’re marginal, that would continue to reduce out-of-pocket costs for high-value services beyond prevention. As I mentioned earlier, Americans are thrilled about the prospect of no cost for FDA-approved COVID-19 vaccines and the Medicare announcement that certain antibody therapies will be made available at no cost to those who are clinically indicated. But our view is, that should be the same for high-value services like those we’ve listed in our VBID X template, in our high-deductible health care plan reform work, as well as our early writings about the Medicare Advantage Value-Based Insurance Design demonstration. I think that we were very fortunate at the V-BID Center to host CMS Administrator Seema Verma, just 2 weeks ago, when she announced a new rule requiring increased transparency for patients be able to see not what their health plan pays or their employer pays, but actually what they will need to pay out of their own pockets in real time. And while I think there will be a lot of changes and philosophies between the Trump administration and the Biden administration, I do believe that removal of surprise billing, lowering of out-of-pocket costs, and increased transparency will persist, and I support those things. Net prices of drugs are above my paygrade, so I’m more than happy to have my co-editor-in-chief take that on.

 Yes, so first of all, Americans may care about out-of-pocket prices more than they care about total prices, but they should care about total prices, because they’re paying the total price as well. They’re just paying for it differently in a way that they don’t see it, but it gets spread across premiums in a bunch of ways that they should care about. Of course, they should also care about the introduction of new innovative medicines, and nothing like the COVID-19 pandemic has illustrated how important innovation is, so it’s a difficult decision how to deal with large prices and affordability. But let me add, there are areas that are not about such complex, controversial tradeoffs where I think progress can be made, particularly around things like the role that rebates are playing, and individuals paying out of pocket based on gross price instead of net. So sometimes they might actually be paying more than the actual net price of the drug, because their co-pay is based on the gross price. There are some other nuances about programs like 340B and things that I would consider quite weedy for this discussion that I think one can think about reforming in the prescription drug space. I think the biggest challenge is going to be how to maintain the environment for innovation that we want and still meet the goals we are going to face around affordability. My recommendation for the new administration would be that while it is reasonable to go after some of the bigger-picture affordability issues, it’s useful to start with trying to get rid of some of the obvious areas where there are market failures or inequities, and they might be able to do that. The previous administration tried as well, and I do think it is a complicated political and technical issue about how to do that, but I think some of the things are getting to the point where failure to act is just not going to be tolerated, given what people are paying out of pocket. I don’t think the solution is more government money to fund the out-of-pocket holes, but instead trying to find ways to make the system transition to a more efficient market mechanism.

Are there any takeaway lessons from the experiences that we’ve had in trying to get a COVID-19 vaccine to market that can be used in the future to balance innovation and affordability?

 Well, my view of the COVID-19 vaccine has been that this is a life-altering, economy-altering cataclysmic action. We basically shut down the economy. And so I think the core issue in many ways is that we were willing to spend a lot of money to get innovation in this targeted area, way more than we would spend in other areas. I think we could do a lot if we throw all the resources we threw toward this particular challenge; I’m not sure that’s what we want to do in all medical areas.

 Just to briefly add to that, most people do not have a lot of confidence in government, whether it be the executive branch or the legislative branch. But the fact that there was bipartisan political support this spring to include in the Coronavirus Aid, Relief, and Economic Security Act to mandate 100% coverage of the COVID-19 vaccine, using the section of the ACA that implemented value-based insurance design principles as the guide, suggests that there is an openness to not only making the investment to create a very important innovative medical intervention, but also thinking about how to both distribute it, which the military is taking over, and make sure there are minimal barriers to individuals who are interested in receiving the vaccine.

You’ve both talked about strengthening the ACA in the future. Do either one of you want to comment on the actions that you would like to see happen there?

 Strengthening the ACA could be looked at 2 different ways, one of which is looking at the law that was passed in 2010 and how to improve upon that. I think in the short term, particularly with a mixed government that we’ll mostly see at least early on in the Biden administration, not legislative fixes, but reversals of some executive orders that were put in place or other decisions like administration. As Yogi Berra said, predictions are dangerous, particularly those about the future, but some of my predictions would be that there will be a restoration of funding regarding communication and other educational efforts to inform Americans about the individual marketplace and the open enrollment periods such as the one we’re going through now. I think there will be some consideration about the removal of the work requirements and other waivers that were put in place for the Medicaid program. I do also think Mike could speak about other innovations that will occur directly through CMS or CMMI to fortify the Medicare program, both from a clinical perspective, as well as a financial one, but I do think that, not ACA-based, a restoration of trust in and the need for science [is necessary]. And this is where Mike and I feel particularly proud to be involved with 

 because we feel that, with the help of our staff at MJH Life Sciences

, we have continued to produce high-quality, relevant information for clinicians, for managers, for policy makers, and for other health care stakeholders. Because in the end, while there will be changes from Washington, DC, and state capitals, I believe most of the innovation will continue to come from organizations outside of government.

 That was a broad overarching answer. I will focus more on the ACA—Mark was making predictions, I think many were right—but I will have a little bit more of a wish list. I think you’ll see changes to things like the marketing rules that Mark mentioned, at least more emphasis on them. I suspect, although I’m not sure, you would see changes to the executive orders around short-term duration plans, for example, and association plans; I suspect they’ll be looking at reversing the current administration’s rule on cost sharing with the silver loading and a bunch of things of that nature. I suspect you’ll see some attention to the public charge rule, which discouraged participation. So I think there’ll be a lot of things to encourage participation.

One of the more interesting things will be how many states decide to expand Medicaid. It’s a good deal financially for the states that expand Medicaid, and a small set now have not. I think given the financial pressure that providers are under, there may be renewed interest for states to expand their Medicaid programs, because that federal money will be all the more attractive given the fiscal situation that states find themselves in and the fiscal situation that their providers find themselves in. So that’s the type of thing that I think you will see going on in the ACA. I don’t think, given the anticipated makeup of Congress, you will see some of the more aggressive aspects of the Biden platform implemented; I don’t expect, for example, to see a public option. That would surprise me. I certainly wouldn’t expect to see Medicare for all, but that was never part of the Biden platform in the first place. Nevertheless, I think that you will see a series of analyses and efforts to bolster the insurance market writ large, and I think they will use the existing ACA framework to do that.

I don’t know if it would happen, but I could see an effort, for example, to increase the generosity of the subsidies so that you get something that’s better than an existing silver plan. I think there’s going to be a concern that people are moving to lower-generosity plans, plans with more cost sharing, bronze plans, etc. And I think there’ll be some effort to address issues of underinsurance, even within the context of the ACA, but there will be limits as to what you’re going to be able to do in that spirit.

Also, the economy will still be struggling to recover during as the pandemic continues, and people have lost their insurance through work. So maybe there will be political pressure.

 Yes, that’s right. Again, all of that is true, although it turns out political pressure is much stronger when you’re about to come into an election. So, it depends who’s feeling political pressure to do what. I think it is important to at least be explicit that Mark and I, when we started looking at VBID, were motivated by desire to make sure that people got the care that they needed, the high-value care. And I continue to believe that, although I think it’s important to understand that a lot of what’s going to happen in the health care space over the next several years is going to be driven by fiscal concerns. The federal government is running a large deficit; there’s a lot of stimulus that was put in place; there may be more stimulus put in place, so there’s not going to be a lot of money there. The states are facing a very difficult financial situation. Private employers are facing a generally difficult financial situation. And so, while we want to make sure we can improve access, I think going forward, we have to make sure that our approaches to meeting our fiscal challenges don’t diminish access and quality. Of course, this brings us back to the theme of low-value care. Although it is hard to get our hands on all of that waste and, frankly, it’s hard to find ways to lower prices, even when we know they’re too high because of anticompetitive reasons. those are the 2 ways to find resources to allow us to preserve access and quality without spending more money than we have.

Do you want to talk about Medicare and what’s going to need to happen there?

 Well, again, I think the challenges with Medicare are going to be how to maintain solvency of the trust fund. I think there are some places where Congress will look: I think they’ll look to Medicare Advantage for some savings; I think they’ll look to prescription drugs for some savings. I think there are other aspects of the fee schedule where there may be opportunities to save some money. I don’t think they will be able to solve all of the fiscal problems with Medicare through those types of fee cuts, and I think that’s for 2 reasons. The first one is that the core problem Medicare faces is the ratio of the number of beneficiaries to the number of workers; it’s a demographic problem that is changing, and so we have to think about how we’re going to manage that portion of the problem. And the particular trust fund problem, at least 2 years of the trust fund problem is because of a loss of revenue due to the economic collapse because of COVID-19. So it doesn’t strike me that the right policy response to a revenue reduction is to lower the fees to all the providers; now it might be we want to lower fees for other reasons, more efficient payment, get the efficiencies we can if we were overpaying in places, but I do think we’re going to have to think about some of the revenue side for parts of Medicare. That’s a very complicated thing to say; I am not an expert in how one might begin to solve Medicare’s revenue problems. But I think that in the long run with the demographic trends that we’re facing, we have to be as efficient as we possibly can. And if those efficiencies don’t hit the fiscal needs that we have, and again, I’m not sure how much we will actually be able to harvest on the spending side, then we’re going to have to think about revenues, or we’re going to have to have a much broader debate about how much debt we want to have in this country, which transcends my area of expertise.

As we talk about inefficiencies and emphasizing higher-value care, is there anything else that you spend most of your time thinking about when you think about COVID-19 extending into next year and the impact it’s going to have beyond that?

 There are real challenges in terms of choosing high-performing providers and understanding the risk-benefit tradeoffs, whether you should see your clinician in person or virtually or not at all. But one of the things that keeps me up at night, and Mike knows this, is that there was this belief that in some type of global budget or capitated model, low-value care would go away, because if the money was coming out of your own pocket, you’d be unlikely to do that. It turns out that there have been lots of studies in situations where clinicians are not paid fee for service, and they just behave badly. And I think a lot of that has to do with the culture in which American physicians were trained, which is they’re afraid to be right, as opposed to being afraid to be wrong. So they order every test, understanding they might glean something clinically valuable, and completely underestimate the negative implications from a clinical standpoint, as well as a financial standpoint, about the utilization of services that do not have a clinical indication. One thing that even makes it more complicated is that even with artificial intelligence (AI) and big data, and really good health services researchers like Mike, we find that the clinicians who tend to behave badly on low-value services are very frequently those clinicians who perform very, very well on high-value services. So it’s not like there are people who are only doing bad stuff and only doing good stuff. And to understand the levers we need to get to where you want to be is going to be challenging, which is why in 

® we wrote that one is we’ve got to pay people well for doing things that are going to make individuals and populations healthier and no longer pay for those situations that don’t. Another area where I think 

® has been a leader is discussion of innovative technologies like electronic medical records and more recently telemedicine and about how we could use these innovative technologies to get us to a post–COVID-19 new normal. But I think in the end, we now have the ability to use clinically driven approaches to both pay providers and encourage or discourage consumers to get the right mix of care. I’ve called this, as Mike knows, peanut butter on the provider side and jelly on the demand side or the patient side, suggesting that I think the only path forward is that we make it easy for clinicians and patients to do the things that make them healthier, and harder to [do the things that do] not. In the situations that we pay doctors like myself to get their patients with diabetes to the eye doctor, we increasingly enroll diabetics in plans with high deductibles that don’t allow them to go to the eye doctor. I use the peanut butter and jelly analogy because I believe it’s the best example of the sum being greater than the individual parts. I believe the academicians and the behavioral psychologists and economists would support the idea of aligning incentives for both provider-facing and consumer-facing initiatives. So I could only hope that as we think about coming out of the COVID-19 pandemic and putting new processes in place, they will at least slightly lean on better payment, increased technologies, and encouraging consumers to do more of the right stuff and less of the wrong stuff.

My addition to that is that we spend a lot of time thinking about financing and prices and what services are delivered, all of which is important. I think the other area that I’m increasingly pondering is what I would call broadly the production of care and the extent to which we can produce the care that we’re getting more efficiently and thereby spend less, so there’s a lot of discussion on telemedicine. We haven’t talked about it much, but I think one of the enduring changes from the pandemic may be the role of telemedicine. Certainly there are likely to be changes in the regulations around the use of telemedicine. I don’t think it will take over the way in which some have predicted. I think if you look, you see a huge increase in telemedicine when no one was allowed to go to the doctor; that is actually pulled back some, so I do think there are some natural limits of telemedicine. But I think there are other broad potential changes in the delivery of care—things like digital clinics, which aren’t quite the same as telemedicine, or other sites of care that may be used more efficiently. We’re going to have to think through how we can build a delivery system that can deliver high-quality care that’s responsive to consumers’ needs for less per-unit cost, and one of the advantages of different payment reform models, of course, is that they incent that type of innovation. I think you’ll see more [artificial intelligence] targeting of people so we can focus the services to get used on the people that will benefit most; I think you’ll see a broader set of alternative providers, there might be scope of practice changes that might enable care to be given better. There are certain clinical areas I worry about a lot. I’m very worried about mental health, for example. We have real problems in the workforce from mental health and a lot of mental health problems, even prepandemic. So I think we have a lot to do with thinking about an efficient mechanism for delivering care at a price that people can afford, and we’re not quite there yet, but I do think you’ll see work in those areas, and I think that will only proceed if we can create incentives for organizations to find those efficiencies.

 Do you think we’ll still be having this discussion in 25 years?

I probably won’t be on the other end of Zoom, but I think that almost every country in the world, no matter what their health care system is like, has a version of this discussion. I’m involved in some work in the United Kingdom, for example, which is having different versions of this discussion. I think it’s part of the human condition that we age and we face illness and death, and so I think we will continue to spend time as a country as individuals trying to figure out how we can get the longest life possible and the healthiest life possible. Some of that will be through medical care; a lot of it will be through social determinants of health and a broader set of services that extend beyond those traditionally thought of as health care. They will be brought into the discussion; they are being brought into the discussion. And I think, as long as we are interested in our health and in some related medical services to improve our health, we will continue to be having discussions about how those services are financed and who pays for them. So in that sense, I think all of that is true. When 

 was founded, managed care was a new thing, and so there was sort of traditional care and managed care. Now everything is managed care, and it’s just a question of how within that big field of managed care we do the best job of financing and delivery.

And 25 years ago, we weren’t talking about social determinants of health as broadly as we are now. Dr Fendrick, do you think 25 years from now those will become even more integrated into what we do?

 I think it’s very interesting. I wasn’t on the editorial board at the journal’s inception but came on very early on. And I want to reiterate this point that 

, we hope people know it as much as the acronym as “

,” because our managed care that we’re talking about is small-m small-c managed care, because I completely agree with Mike that we’re not talking about the managed care of 25 years ago. I think it’s not only interesting and worthwhile to scroll back at some of the highlighted 

, but I also want to give a shout-out to the amazing 

 editorial staff in putting together the 12 Q&As for this year: a remarkable combination of clinical leaders like Eric Topol and Mandy Cohen, who is the secretary of health in the state of North Carolina; people who are currently in government or previously had high-ranking positions in both Republican administrations like Gail Wilensky and Democratic administrations like Don Berwick and Sherry Glied; and really interesting insights from policy insiders like Larry Levitt and Kavita Patel. And I do want to especially note, I don’t know how they did it, a really fascinating Q&A with now everyone’s favorite American, Dr Anthony Fauci. It’s been really a pleasure to work with Mike and the 

 team as they celebrate 25 years. I hope that I’m passing the baton, so to speak, to others to carry on the research activities and the policy activities, and trying our best to provide relevant, timely information to decision makers who are on this, as Mike says, timeless journey to figure out how to keep individuals and populations healthy and not bankrupt their countries or their private industries.

1. Kim DD, Ollendorf DA, Neumann PJ, Fendrick AM. Crisis into opportunity: can COVID-19 help set a path to improved health care efficiency? 

. 2020;26(9):369-370. doi:10.37765/ajmc.2020.88412

To mark the 25th anniversary of the journal, each issue in 2020 includes an interview with a health care thought leader. The December issue features a conversation with the journal’s co-editors-in-chief, Michael E. Chernew, PhD, and A. Mark Fendrick, MD.

After 25 Years, AJMC® Looks to the Future: A Q&A With Michael E. Chernew, PhD, and A. Mark Fendrick, MD

 Understanding variation in spending across organizations, rather than across geographic areas, is important because care is delivered by organizations and interventions increasingly focus on organizations. Accountable care organizations (ACOs) are particularly important to study given their incentives to reduce spending. Analyzing spending differences across ACOs may help identify cost savings opportunities.

 Cross-sectional analysis of Medicare claims.

 We stratified ACOs into quartiles based on the deviation between each ACO’s risk-adjusted spending and average risk-adjusted fee-for-service spending in the same market (hospital referral region). We compared spending between top- and bottom-quartile ACOs on each of 7 major service categories and 10 clinical condition groups to identify areas of potential savings. We simulated spending reductions if ACOs with high adjusted spending reduced spending to the levels of lower-spending ACOs.

 In 2016, geographically adjusted and risk-adjusted total per-beneficiary spending for the highest-spending quartile of ACOs was 14% higher than for ACOs in the lowest quartile. Variation between high- and low-spending ACOs was greatest, at 27%, in the use of skilled nursing facilities—a service category in which ACOs have reduced spending by the greatest percentage. Inpatient care was the largest driver of absolute dollar differences in spending, however, accounting for 37% of the total spread. If spending in ACOs above median adjusted spending were brought down to the median, savings would be 3% to 4%.

 By extending the variations literature to focus on ACOs, we illustrated that meaningful further savings opportunities exist both within and across markets.

There was a 14% difference in risk-adjusted total spending between accountable care organizations (ACOs) in the highest and lowest quartiles of spending relative to local fee-for-service Medicare spending, suggesting meaningful opportunities for savings even after 4 years of the ACO program.

If spending in ACOs with above-median adjusted spending were brought down to the median, savings would be 3% to 4%.

Spending variation by service category suggests savings opportunities in inpatient, postacute, and home health care. Variation in spending for patients by select condition categories indicates substantial savings opportunities in aggregate for prevalent chronic conditions, but the magnitude of variation in dollars per patient is small, suggesting that interventions must be low-cost on a per-beneficiary basis to achieve net savings.

Variation in healthcare spending and utilization exists at every level of the healthcare system, ranging from states to individual physicians within a given practice.

 Yet most of the research characterizing variation in healthcare spending has been focused at the geographic level, as exemplified by the Dartmouth Atlas. Results of geographic analyses argue convincingly that a great deal of spending variation across areas is not connected to patient morbidity or outcomes.

 However, considerable differences in spending exist even within geographic areas,

 and geographic areas are not a natural unit for actionable interventions to reduce spending variation.

 The National Academy of Medicine’s 2013 report on spending variation emphasized the importance of extending analyses of practice pattern variation to “the unit of decision-making,” such as individual providers or provider systems.

Evidence characterizing organizational variation remains relatively limited. Studies of organizational variation in Medicare spending typically focus on a specific clinical context, like outpatient cardiology or oncology practices, or examinations of specific services, like low-value care.

 Evidence from the commercial sector focuses on variation in price, rather than total spending, and generally looks at variation across hospitals rather than organizations such as accountable care organizations (ACOs) that are responsible for population-level spending.

 As in the geographic variation literature, research examining practice pattern variation in provider organizations reveals large and unexplained differences, but none of that work examines variation in total spending at the population level.

ACOs are provider groups that agree to take accountability for spending and quality of care for patients in the ACO contract. ACOs represent an important unit of observation because they are CMS’ most important alternative payment model and the foundation of many current efforts to reduce wasteful medical spending. Little research exists on spending variation in ACOs: One early study reported considerable variation in ACO spending, even after controlling for regional variation and case mix, but did not delve into sources of variation.

 Breaking down sources of spending variation among ACOs could inform where organizations might intervene to reduce avoidable spending.

We examined the deviation between each 2016 ACO’s risk-adjusted spending and average risk-adjusted spending in its market, then stratified ACOs into quartiles based on these deviations. We defined top-quartile ACOs—those with the largest positive deviations, in dollars, from adjusted local market fee-for-service (FFS) spending&mdash;as high spending (quartile 4 [Q4]). We defined bottom-quartile ACOs&mdash;those with the largest negative deviations&mdash;as low spending (quartile 1 [Q1]). We compared spending between Q1 and Q4 ACOs on each of 7 major service categories and 10 clinical condition groups to identify areas of potential savings. Finally, we simulated spending reductions if ACOs with high adjusted spending (within their market or nationally) reduced spending to the levels of lower-spending ACOs.

We analyzed a 100% sample of FFS Medicare beneficiaries, refined to include only beneficiaries continuously enrolled in parts A and B for the full year (or until death) and continuously enrolled in the prior year. We further limited the sample to beneficiaries potentially eligible for enrollment in an ACO: individuals with at least 1 qualifying ambulatory evaluation and management visit in the year of interest.

We analyzed ACOs participating in the Medicare Shared Savings Program (MSSP), the largest of Medicare’s ACO initiatives. We assigned beneficiaries to ACOs using the provider identifiers (taxpayer identification numbers [TINs] and CMS Certification Numbers [CCNs]) defining each ACO in the ACO SSP Provider-Level Research Identifiable File for that year. We assigned beneficiaries to an ACO if it accounted for a larger share of qualifying office visits with a primary care physician than any other ACO or non-ACO TIN/CCN during the year (see 

]). Patients assigned to non-ACO TINs served as FFS controls. As in the MSSP, we used retrospective assignment, assigning patients to ACOs based on utilization in the year of interest.

Our sample included 432 ACOs in 2016. (There were 220 in 2013, 333 in 2014, and 387 in 2015, and 145 ACOs were present all 4 years.) For ACOs present in multiple years, the composition of TINs or CCNs included in the ACO may have changed over time, which we addressed in sensitivity analyses.

Our outcomes of interest were total annual per-beneficiary Part A and Part B Medicare spending and annual per-beneficiary Medicare spending by service subcategory.

In all analyses, we controlled for beneficiary age, sex, race/ethnicity, Medicaid status, disability as the reason for enrollment, and end-stage renal disease. We controlled for comorbidities using prospective hierarchical condition category (HCC) diagnoses and risk scores using CMS 2014 version 12 software for all 4 years, using CMS’ 

International Classification of Diseases, Tenth Revision, Clinical Modification

International Classification of Diseases, Ninth Revision, Clinical Modification

. We identified whether beneficiaries had ever had 1 of 27 chronic conditions identified in the Chronic Conditions Warehouse (CCW). We assigned beneficiaries to a hospital referral region (HRR) based on their residential zip code.

Ranking ACOs by Adjusted Spending Deviations From Their Markets

To rank ACOs by total adjusted spending, we estimated a linear mixed-effects regression model predicting total per-beneficiary spending (eAppendix) as a function of the covariates described previously: HRR fixed effects and ACO random effects. Estimated random effects represent an adjusted mean spending deviation from local market FFS spending for each ACO. In contrast to fixed-effects models, the random-effects model accounts for within-ACO variation in spending in the estimation of the between-ACO variance; without this correction (eg, in a fixed-effects model), the ACO-level variance would be overestimated. We defined an ACO’s market as the HRRs that it serves. ACOs can serve multiple HRRs, so an ACO beneficiary’s spending relative to other FFS beneficiaries in its market is implicitly estimated as a weighted average of the beneficiaries in each HRR served by the ACO. We were thus able to measure each ACO’s spending deviation relative to other ACOs and non-ACO FFS beneficiaries in the ACO’s market.

We grouped ACOs into quartiles based on their risk-adjusted spending deviation from average FFS spending in their market. We used ACOs’ total spending quartile rank in all analyses, including condition or service subcategory analyses.

&#8203;&#8203;&#8203;&#8203;&#8203;&#8203;

We ran a linear regression model with Q1 and Q4 quartile dummies and covariates for beneficiary demographic characteristics, case mix (HCC), and market (HRR). We subtracted the Q4 coefficient (spending attributed to high-spending ACOs) from the Q1 coefficient (spending attributed to low-spending ACOs) to calculate the spread in total spending. We focused on this “spending gap” between the Q1 and Q4 means as a useful summary of the range of the spending deviations.

We repeated the same total spending per beneficiary model described previously for each service category, changing only the dependent variable to subcategory spending rather than total spending. Additionally, we calculated the percentage share of the total spending gap between quartiles attributable to each service category.

We estimated deviations for patient populations with specific conditions, using the 27 CCW conditions to create 10 CCW condition groups (eAppendix). We restricted some conditions, such as hip fracture, to diagnoses within the last year to better isolate spending associated with that diagnosis. We included chronic conditions if the patient had ever received a diagnosis. Using linear regression models with quartile dummies, condition dummies, and interactions between quartiles and conditions, we computed spread in total spending and service category spending between Q1 and Q4 ACOs. We determined the prevalence of conditions in Q4 ACOs to estimate potential savings in aggregate associated with each condition; prevalence was approximately evenly distributed across ACO quartiles and other FFS beneficiaries.

Simulation of Potential Spending Reductions

We simulated cumulative spending reductions in 3 scenarios to crudely measure ACO savings potential. We estimated change in total spending if all ACOs with above-median adjusted spending were able to reduce spending to the level of the median (50th percentile) ACO. We repeated this analysis for thresholds at the 25th and 10th percentiles.

Our primary analyses focused on 2016. We repeated the analyses for 2013 to 2015, as well as for ACOs present in all 4 years. We replicated the service and clinical area analyses (both quartile rankings and spending category variation models) for each year, with and without controlling for geography (HRR). We also simulated savings without HRR controls.

Our sample included approximately 26.6 million eligible beneficiaries in 2016, with 8.3 million beneficiaries attributed to ACOs. Mean ACO membership was 19,179. Adjusted mean total per-beneficiary spending in 2016 was $10,212 for low-spending ACOs and $10,850 for high-spending ACOs. Results for 2013 to 2015 are reported in the eAppendix.

In 2016, mean total risk-adjusted spending was 14% higher for ACOs in the highest quartile relative to those in the lowest (

). Although the relative percentage difference between high and low spending quartiles for inpatient care was only 15%, inpatient spending was the largest service category in dollars; consequently, it accounted for 37% of the absolute spread in total spending (ie, in dollars per beneficiary). Outpatient spending in the upper quartile was 22% greater than that in the lowest quartile, and, because of its large spending share, it was the second largest driver of the overall spending gap, accounting for 32% of the spread in total spending.

In all years, variation on a percentage basis between Q1 and Q4 ACOs was greatest in skilled nursing facility (SNF) and home health. However, because SNF and home health spending accounted for a smaller share of spending than inpatient and outpatient spending, they account for a smaller share of the overall gap in spending.

 summarizes spending and prevalence by condition group for high- and low-spending ACOs. Taking patients with chronic kidney disease (CKD) as an example, the mean risk-adjusted difference in spending for Q4 is $1372 higher than spending for Q1 ACOs. Prevalence of CKD in Q4 ACOs was 22%. If Q4 ACOs reduced incremental spending for patients with CKD to that of Q1 ACOs, Q4 ACOs could achieve cost savings of $302 per beneficiary across the entire ACO population (ie, including people with and without CKD).

We observed substantial spending variation across all condition groups. In general, acute or newly diagnosed CCW condition groups tended to have a high dollar difference, driven by higher mean spending, in per-beneficiary spending between Q1 and Q4 ACOs, but low prevalence. Chronic condition groups tended to have smaller per-beneficiary dollar differences in spending for patients with those diagnoses but high prevalence, and thus they were responsible for greater absolute spending differences between high- and low-performing ACOs when considered over the full ACO population.

If all beneficiaries in ACOs with geographically adjusted and risk-adjusted spending higher than the median ACO were brought down to that threshold, savings would be 3.6% (

Without adjusting for HRR, the between-quartile spread in total risk-adjusted spending deviations for the top and bottom quartiles of ACOs nationally was 10 percentage points larger than the models adjusting for HRR (eAppendix). Our primary analysis is thus conservative in that it assumes that ACOs are more readily able to reduce spending relative to their markets than relative to national benchmarks. On a national basis, if beneficiaries in ACOs above median national ACO spending (adjusted for case mix but not geography) were brought down to that threshold, savings would be 9.2%.

Our results were largely insensitive to the year of analysis (eAppendix): The gap between Q1 and Q4 ACOs in 2016 was $1409 (14% of mean per-beneficiary spending). The analogous figures for 2013, 2014, and 2015 were $1328 (14%), $1364 (14%), and $1475 (15%), respectively. The services contributing to this variation were largely unchanged. ACO quartile rankings were largely stable over time; only 6 ACOs had a year in the highest quartile and a year in the lowest.

This analysis suggests meaningful variation in ACO spending within markets and overall. Despite widespread recognition that high spending in the United States, particularly in the commercial sector, reflects high prices, our savings estimates largely measure variation in utilization because Medicare prices are set administratively and vary much less than commercial prices.

 Thus, our analysis can broadly identify areas of savings opportunity related to utilization. Adjusted within-market spread between Q1 and Q4 ACOs was 14% in 2016, suggesting meaningful opportunities for savings even after 4 years of ACO adoption.

Our analysis suggests that key savings opportunities exist in the inpatient, SNF, and home health care settings, consistent with ACO evaluations reporting spending reductions in these areas.

 ACOs may find savings by changing the site of care (eg, shifting from hospital outpatient department to office settings), which is consistent with observed evidence on ACO savings.

Research involving ACO executives suggests that they have adopted strategies including strengthened primary care, greater care management, and reductions in avoidable ED use—all with an emphasis on high-risk patients.

 However, care management strategies do not appear to be how ACOs, on average, have found savings. Prior research indicates that ACOs have not differentially reduced admissions for ambulatory care—sensitive conditions, nor have reductions in hospitalizations and spending been disproportionately concentrated among high-risk patients.

 ACOs increasingly report focusing on standardizing and systematizing clinical processes.

 And, as ACOs have discovered the savings opportunities in postacute care, they have initiated strategies such as preferred SNF networks.

Spending variation in clinical groups suggests considerable aggregate savings potential among patients with high-prevalence chronic conditions, but the magnitude of spending variation per beneficiary in dollars is small, meaning that interventions aiming for net savings would need to be highly efficient on a per-patient basis. Incremental spending differences between Q1 and Q4 ACOs for people with newly diagnosed and acute conditions are sizeable in dollars per beneficiary, but these conditions are low in prevalence, meaning that aggregate savings would be modest in condition-specific efforts to control costs.

Targeted interventions to improve chronic conditions may not provide a return on investment for ACOs in the short term because per-beneficiary cost savings are modest and the costs of many interventions, such as case management, are high. Further, returns to chronic disease management may be realized in future years, and beneficiary attribution to an ACO is not guaranteed from year to year. Many popular clinical interventions may be cost-effective but not cost-saving.

 The most proximate savings may lie in interventions to reduce utilization of low-value services; this requires developing effective interventions, as existing tools to measure low-value care capture only a small fraction of it.

 Because low-value care is not necessarily confined to the high-risk group, interventions almost certainly need to be broad.

Analysis of spending differences among accountable care organizations (ACOs) may help identify cost savings opportunities. We examined the magnitude and sources of spending variation among ACOs over 4 years.

Spending Variation Among ACOs in the Medicare Shared Savings Program

To compare supply of and access to physicians and hospitals for low- and high-income communities in the United States following the Affordable Care Act.

Cross-sectional study using 2015 data on physicians and hospitals from CMS, SK&A, commercial claims, and the American Hospital Association.

We computed densities of primary care and specialty physicians in low- and high-income zip code tabulation areas (ZCTAs), further stratified by urban, suburban, and rural categories. We also estimated geodetic distances to the nearest general acute care, teaching, children’s, and psychiatric hospitals for each ZCTA.

Urban low-income ZCTAs were 0.5% more likely to have very low primary care physician density (<5 physicians per 10,000 population) relative to high-income ZCTAs. The disparity was great in rural areas (where the analogous figure was 1.7%) and greatest in suburban areas, where low-income ZCTAs were 7.4% more likely to have very low primary care density relative to high-income ZCTAs. The percentage of urban and suburban low-income ZCTAs with very low specialty physician density was 1.3 to 1.4 times larger than that of high-income ZCTAs. Although the average distance to a general acute care hospital was similar for low- and high-income ZCTAs (median difference range, 0.4-2.1 miles), residents of rural low-income areas had to travel substantially farther to teaching, children’s, and psychiatric hospitals (10.2-35.6 miles).

Our findings document geographic variation in provider supply related to income and suggest that improving access to care may require greater attention to local delivery systems.

The American Journal of Accountable Care. 2019;7(4):11-18

One of the chief goals of the Affordable Care Act (ACA) was to expand access to healthcare to all Americans, particularly through insurance coverage.

 However, while health insurance increases access to care, it does not guarantee such access.

 Specifically, the structural aspects of the local delivery system—supply, location, and organization of healthcare resources&mdash;all vary widely across markets, but they have received much less attention than coverage as a driver of disparate access. Although considerable literature relates availability of primary care to beneficial outcomes and lower costs,

 it does not focus on disparities in access related to income. Although many studies have shown that Medicaid expansions have reduced the number of uninsured Americans,

 few studies have looked at primary care supply after many states’ Medicaid expansions took effect in 2014. The effects of improved financial access for disadvantaged populations depend on the nature of their local delivery systems.

Barriers to access posed by lack of healthcare facilities and trained physicians has been well documented.

 Policies to address scarcity of healthcare resources date back to at least the 1946 Hill-Burton Act. Today, the Health Resources and Services Administration identifies Health Professional Shortage Areas (HPSAs) or Medically Underserved Areas/Populations based on shortage designation criteria that it develops in collaboration with state partners.

However, the HPSA definition of a shortage area is based on context and specific applications, not a broad assessment of community needs. While adequacy of the supply of providers for low-income individuals is largely unknown, dependence on local safety net providers is substantial and potentially precarious

: Today, 1 in 13 Americans obtain care from the 10,000 community health center sites funded by the ACA.

Much of the existing literature relating provider supply to outcomes focuses on rural areas. Prior research suggests that rural populations experience greater difficulties in accessing needed care compared with urban populations.

 Rural areas often encounter challenges in attracting and retaining providers because providers feel “overburdened and underpaid,” among other reasons, compared with those in urban areas.

Yet rural is not synonymous with low income, and quantitative analyses of delivery systems in low-income areas are lacking despite the focus on access expansion policies for low-income populations. Existing research on this topic is largely qualitative and in the form of case studies of high-functioning safety net organizations.

 The large literature on the safety net generally takes the provider as the unit of observation and hence cannot speak to differences among communities in access to care associated with the local availability of providers. There has been little research to characterize healthcare delivery systems in low-income communities nationally.

We use a unique relational database to examine potential barriers to access and to quantify variation in the supply of primary care and specialty physicians, as well as general acute care and specialty hospitals, across the United States. In this paper, we present a large-scale, national description of the local healthcare delivery systems found in low- versus high-income communities following the ACA.

We selected zip code tabulation areas (ZCTAs) as our primary unit of observation, excluding 667 ZCTAs with fewer than 10 residents and 517 ZCTAs with missing income data. Our final database included 32,322 ZCTAs with 316,196,416 residents (99.9% of the US population in 2015).

ZCTAs capture variation in income better than larger units such as primary care service areas (PCSAs). There is often substantial income heterogeneity among the constituent zip codes within a single PCSA, for example. Because individuals may seek care outside of their ZCTAs, we assign to each ZTCA measures of provider supply that encompass a wider area (described below).

 and using income data from the American Community Survey (ACS), we classified a ZCTA as low-income if the percentage of individuals with income below 200% of the federal poverty level (FPL) fell in the top decile of the national distribution (57.4%-100% population <200% FPL; distributions were computed using all ZCTAs with poverty data and after the population size restriction for metropolitan areas). Analogously, we classified a ZCTA as high-income if the percentage of individuals with income above 500% of the FPL fell in the top decile of the national distribution (43.0%-100% of population >500% FPL). Our results were robust to other thresholds, such as percentage of individuals with income below 100% or below 300% of the FPL.

In a sensitivity analysis, we classified low- and high-income ZCTAs based on their state income distributions and refer to these as relative low- and high-income ZCTAs. Specifically, we classified ZCTAs in each state as being a low- (or high-) income community if the percentage of individuals in the ZCTA with low (high) income fell in the top decile of the state distribution. Income distributions vary from state to state, and relative low-income status in a state may be as important for disparities as national low-income status. Implications for state- and federal-level policy may also differ.

 Our data on practicing physicians and their practice locations were derived from 2015 data from the CMS Provider Enrollment, Chain, and Ownership file; SK&A physician file; and CMS MAX Provider Characteristics (verified in the National Plan and Provider Enumeration System) file. The total number of physicians included in this study (n = 818,951) compared favorably with numbers in other sources. The American Medical Association Masterfile contains 829,962 physicians in 2013 and the 2014 SK&A data file contains 565,000 unique physician National Provider Identifiers. Nurse practitioners and physician assistants also deliver care to patients, in some states do not require physician supervision, and sometimes serve as patients’ primary care clinicians. However, they were excluded from this study because there was no comprehensive data source on these nonphysician clinicians.

Because some patients cross ZCTA lines for care, we assigned to each ZCTA the physician supply of its PCSA. The ZCTA unit of analysis captures low income more accurately than broader geographic areas, whereas PCSA-level physician supply considers border crossing and more accurately represents the service area in which an individual obtains care. Specifically, we computed primary care and specialist physician density (ie, number of physicians per 10,000 population) for each PCSA and assigned these measures to their corresponding ZCTAs. We counted the 236,575 physicians who practiced at more than 1 site by proportions; for example, a physician who practiced at 2 sites was assigned 0.5 effort at each.

We classified physicians as primary care or specialist. Physicians trained in family medicine, pediatrics, and internal medicine (with no subspecialty training) were classified as primary care physicians. We included only non—hospital-based specialists in our tabulations of specialist supply (ie, not emergency medicine, radiology, anesthesiology, or pathology). Of the 818,951 National Provider Identifiers included in this study, 274,449 (33.5%) were classified as primary care physicians and 396,588 (48.4%) as specialists.

This large-scale, national study shows geographic variation in provider supply and hospital access for low- and high-income communities following the Affordable Care Act.

Michael E. Chernew, PhD

Articles

Variation in Spending Associated With Primary Care Practices

After 25 Years, AJMC® Looks to the Future: A Q&A With Michael E. Chernew, PhD, and A. Mark Fendrick, MD

Spending Variation Among ACOs in the Medicare Shared Savings Program

Comparison of Healthcare Delivery Systems in Low- and High-Income Communities