Integrated Delivery Networks: In Search of Benefits and Market Effects

By Jeff Goldsmith et al.
For the National Academy of Social Insurance, February 2015

Advocates believe that IDNs [Integrated Delivery Networks] should be structured like large prepaid group practices such as Kaiser [Permanente]…. Those advocates believe that as IDNs assume more economic risk, either delegated risk through capitated payments from health plans or actual insurance risk through “captive” health plans, IDNs will evolve into Kaiser-like entities, with compelling incentives to control costs. [p. 5]

Because there are so few organizations like Kaiser, there is naturally little comparative data on its performance. There are occasional books and articles extolling the virtues of the Kaiser model, often written by Kaiser-affiliated researchers (Enthoven and Tollen, 2004; Crosson, 2005). There are also some consulting firm reports that suggest the superiority of the Kaiser model (Aon-Hewitt, 2011). Otherwise, there is no evidence that we know of that documents the competitive efficiency of the Kaiser model.

Indeed, during the 1980s and 1990s, Kaiser had difficulty exporting its own model to other parts of the U.S. beyond its native Pacific Coast market (Gitterman, Weiner, Domino et al., 2003). [p. 16]

The latest growth spurt in IDN formation has been stimulated in major part by the quasi-risk contracting model embodied in ACOs. If the intended end state for regular Medicare payment is full-risk contracting with IDNs, and present day IDNs do not display either increased operating efficiency or lower total cost of care compared to community-based alternatives, policymakers need to find another payment approach. [p. 29]

For nearly a half century now, the holy grail of the managed care movement has been the consolidation of the nation’s insurance companies, hospitals and clinics into large “Kaiser-like” entities. The name used by managed care proponents for these entities has changed over the years. First it was “HMO,” then it was “integrated delivery system” (IDS), and now it’s “accountable care organization”(ACO), but Kaiser Permanente always remained the movement’s poster child. Now comes a study that says there was never any evidence for the claim that Kaiser, HMOs, and IDSs cut costs. The authors predict the same will turn out to be true for ACOs.

The study was commissioned by the National Academy of Social Insurance, an organization supported by much of the health policy establishment, including AARP, United HealthCare, the Centers for Medicare and Medicaid Services, the Robert Wood Johnson Foundation, and Kaiser itself.  The lead authors, Jeff Goldsmith and Lawton Burns, are prolific writers who are well known in the health policy world. The study consists of a literature review and an examination of the finances of 15 IDSs.

The authors make it clear they understand they are criticizing the most fundamental belief in managed care theology. In a blog essay about the study, Goldsmith and Burns state:

For the past four decades, there has been one dominant theme in healthcare delivery-system reform: Hospitals and physicians must transform themselves into comprehensive-care enterprises to be paid a population-based global budget. In the vision of pioneering health policy researchers Paul Ellwood Jr. and Alain Enthoven, consumers should choose among multiple Kaiser-like entities competing based on premium (e.g., total cost of care).

Ellwood, Enthoven and their disciples claimed, without evidence, that subjecting insurance-hospital-clinic conglomerates to a “global budget” – which means shifting all or much of the insurance risk traditionally borne by insurance companies onto hospitals and clinics – would create the best of all possible worlds. It would lead hospitals and clinics, with helpful instructions from the insurance component of the conglomerate, to reduce costs and improve quality. Exactly how the Kaiser-like entities were supposed to do this has never been clear.

The study found that evidence supporting the claims made by Ellwood, Enthoven and others for HMOs/IDSs/Kaiser-like entities has never existed:

What we found was, frankly, disappointing. We reviewed more than 30 years of academic literature on vertical integration and diversification in healthcare, and found virtually no measurable benefits … of putting health insurance, hospitals and physician services under the same structure.

The study concludes, “there is no evidence that we know of that documents the competitive efficiency of the Kaiser model.” (p. 16)

The study in fact presents some evidence indicating “integrated delivery networks” (IDNs, Goldsmith et al.’s label for IDSs) are driving prices up. Goldsmith and Burns summarized this portion of the study as follows:

[W]e found that IDNs’ flagship hospitals … were more expensive than their direct in-market competitors…. Further, the size of the IDN (measured either by hospital bed count or total revenue) did not correlate with profitability…. Neither scale nor scope economies could be detected.

Perhaps most importantly, the authors found evidence indicating that IDN’s that subject their flagship hospital to risk-sharing make those hospitals more expensive:

What we found was that flagship hospitals within IDNs that have no revenue at risk are on average 6.8 percent less expensive than their in-market competitors, while flagships within IDNs that have some revenue at risk are on average 20 percent more expensive than their competitors. This finding is similar to one found in the literature review. If there is a cost of care advantage conferred on IDN hospitals by their owner operating a health plan, it was not apparent from this analysis. (p. 24)

The authors suggest profits for the provider components of IDNs fall. They state, “From the provider perspective, the available evidence suggests that the more providers invest in IDN development, the lower their operating margins and return on capital.” But what about non-provider investors in IDNs? Do they ultimately suffer lower returns too, or do they enjoy higher profits because the insurer component of the IDN sucks more money out of clinics and hospitals than it returns? The authors do not attempt to offer an answer. They report that IDNs are so “inscrutable” it is impossible to tell how money moves among the three components of IDNs.

Finally, the study found no differences in quality between the IDN flagship hospitals and their nearest competitor.

Goldsmith et al. assert that a “major reason” why the merger of insurers, hospitals and clinics into one entity raises costs and fails to improve quality is that the new entity is more difficult and expensive to manage than the individual components were by themselves. The authors observe that this explanation is consistent with the larger economic literature on mergers of firms selling different products or services. “[T]he unrelatedness of the new businesses” (p. 15) now under the control of one set of managers makes the managers’ job far more difficult and expensive, the authors argue.

To translate into plain language that will resonate with critics of managed care, Goldsmith et al. are saying:

• insurance companies don’t know how to practice medicine,
• doctors and hospitals don’t know how to run insurance companies, and,
• when insurers, hospitals and clinics are thrown into the same entity, the management of the new entity is forced to spend more on administration than the components of the entity did prior to the merger because the new entity is so much more complex than any of the three components were operating alone.

This explanation will not surprise single-payer proponents. Since the birth of the American single-payer movement in the late 1980s, the movement has been highly critical of the cost of administering America’s increasingly consolidated, increasingly micro-managed, multiple-payer system. PNHP founders David Himmelstein and Steffie Woolhandler greatly enriched the study of administrative costs, and arguably invented the field, with their 1991 paper documenting the high administrative costs of the US system. That paper and subsequent research has shown that US administrative costs have risen from about one-fifth of total spending in the 1980s to one-third today. The period since the 1980s is, of course, the period in which managed care ideology and IDSs became widespread.

If it’s true that Kaiser-like entities incur higher administrative costs and ultimately save no money for anyone, why have mergers among insurers, hospitals and clinics become so common? The answer hinted at by the authors is the that economic forces set loose by the rise of the managed care movement gave first movers – the early consolidators bent on “managing care” – an advantage over the later movers. As the myth that insurance companies should micro-manage clinics and hospitals spread among the insurance industry’s most powerful customers (large employers and the politicians who control Medicare and Medicaid), insurers rushed to consolidate with each other and with providers to enhance their power to control clinics and hospitals and to extract discounts from them.

Anecdotal evidence indicates the advantage gained by the early consolidators was enormous. In an interview published in Health Affairs, Thomas Pyle, the former CEO of Harvard Community Health Plan (an HMO) observed that the most important “lever” the HMO used to control its costs during his tenure was to cut a deal with Brigham and Women’s Hospital in which the hospital agreed to give the HMO a 42 percent discount. “When we made that deal, we closed our own hospital down,” said Pyle, “and we gave the Brigham almost all of our business, in return for which we got a 42 percent discount off rate card. It was the first time that a Harvard hospital had really broken on rate cards.”  This ability to destroy hospitals and build up others by “giving all their business” to one hospital at the expense of others gave the early HMOs a significant advantage over insurers that did not have that ability.

Similarly, the discounts HMOs extracted from Twin Cities hospitals soared during the 1980s, from 9 percent in 1981 to 38 percent by 1990. (The 1981 figure is from Allan N. Johnson, “Cost-shifting: The discount dilemma,” Journal of Health Politics, Policy and Law, 1984;2:251-260. The 1990 figure is from Katherine Hiduchenko, “Do Health Maintenance Organizations control costs or shift costs?” New England Journal of Medicine, 1993;328:971.)

But that advantage acquired by the early consolidators was reduced, and for some eliminated, as the provider sector responded in kind. As the insurance industry’s game plan became clear to providers, providers rushed to merge with each other to create countervailing bargaining power. As Goldsmith et al. put it, “The rise of managed care, and the perceived threat posed by the rise of capitated contracting, created anxiety among providers that fueled their efforts to form IDNs.” (p. 10)

This dynamic resembles the arms race among nations. No nation can afford to stop the race by itself and so the race goes on. Similarly, because American politicians and the intellectuals who influence them encourage merger mania within the healthcare system with their celebration of “integrated care,” and because state and federal anti-trust authorities don’t have the resources to bring the race to an end, no insurance company, hospital or clinic can afford to “disarm” unilaterally – to stay out of the race to get big and complex. Thus, even though the race to bigness and complexity ultimately benefits no one because it drives up administrative costs by more than it reduces medical costs, the race goes on. And it will go on as long as the managed care ideology which sparked the race in the first place holds a firm grip on the minds of large employers, policy-makers, and other members of the health policy elite.

If market forces unleashed by the rise of managed care are responsible for the race to gigantism and complexity, and the race is generating more expense than it is saving, then the most important question before us is: How did managed care ideology become so powerful and what can be done to reduce that power? That, ultimately, is the most important question raised by Goldsmith et al.’s study. Goldsmith et al. give a nod to the problem. They observe that “IDNs have … operated under a halo of presumed societal benefits (quality, efficiency, care integration, etc.) for the better part of four decades with remarkably little evidence that these benefits in fact exist.” (p. 29) Where did this “halo” come from?

Goldsmith et al. make no attempt to answer that question. That’s understandable. The topic they bit off was large enough for one paper, and they analyzed that topic well. By demonstrating the gaping chasm between the claims made by Ellwood et al. and reality, Goldsmith et al. have performed a valuable service. But the research must not stop there. We badly need studies of the etiology of groupthink within the American health policy establishment.

Kip Sullivan, J.D., is a member of the board of Minnesota Physicians for a National Health Program. His articles have appeared in The New York Times, The Nation, The New England Journal of Medicine, Health Affairs, the Journal of Health Politics, Policy and Law, and the Los Angeles Times.