Cost Containment and the Convenient Tale of Care Coordination

Nobody likes waste. Nobody likes fragmentation. Evidence that both are hallmarks of the US health care system has therefore fueled vigorous debate over how to redesign payment and delivery systems to root out inefficiencies. With the broader imperatives of cost containment and quality improvement at play, a powerful narrative has emerged from this debate that is now widely held and dominates policy—care coordination not only improves outcomes but lowers costs, too.

While tidy and attractive, this notion is not evidence-based. Studies of programs or practice models designed to enhance coordination and management of care for patients with multiple conditions and multiple providers have demonstrated minimal, if any, consistent savings from such interventions.¹ There are several reasons for this. First, efforts to coordinate care and improve patient outcomes appropriately involve interventions to correct underuse and ensure timely access to care. In isolation, these efforts tend to increase use of care and thus act to negate, at least partially, any reductions in preventable or unnecessary care resulting from better coordinated care. Second, as with any therapeutic intervention, what an epidemiologist would call “the number needed to treat” is greater than one. That is, for every costly complication prevented per patient per year, a care coordination program needs to manage multiple patients, even when selectively targeting high-risk patients. Third, coordinating care is itself costly. Even if a program effectively reduces spending on unnecessary or preventable care through better care coordination, the costs of information technology, personnel, and other inputs to such programs partially offset or exceed those spending reductions. Thus, touted reductions in hospitalizations and readmissions constitute offsets to the costs of care coordination, not net savings.

Nevertheless, care coordination is widely considered to be not only a winning strategy for achieving savings under new payment models, such as the Medicare accountable care organization (ACO) and bundled payment programs, but the leading strategy. For example, although ACO and bundled payment models provide incentives to lower spending via any mechanism, the Centers for Medicare and Medicaid Services (CMS) defines these models in terms that focus almost exclusively on care coordination: “ACOs are groups of… providers, who come together voluntarily to give coordinated high quality care to their Medicare patients.”² The Pioneer ACO model is described as “a program designed for early adopters of coordinated care” and the Bundled Payments for Care Improvement initiative as another way Medicare is “encouraging coordinated care.”²

There is undoubtedly room for improvement in the design and implementation of care coordination programs, and good ones might break even. What such programs might look like is unclear, however, as the definition of care coordination has become increasingly muddled by rhetoric that has outpaced evidence. Typically, care coordination is defined as a set of patient-centered activities to integrate care across services, specialties, or settings, with patient-specific investments to encourage engagement in the treatment plan. Although such efforts could plausibly prevent duplicative testing, medical errors, and adverse consequences of rocky transitions or poorly managed chronic disease, coordination efforts intersect with care in ways that may be largely orthogonal to the major causes of wasteful care. More direct and less costly strategies to reduce overuse would instead focus on getting providers to do less. Limiting the provision of low-value services such as imaging for low back pain, for example, typically does not require coordinating care or modifying patient behavior. Savvy care coordination could potentially enhance efforts to encourage evidence-based decisions—for example by alerting primary care providers when patients are offered ineffective procedures—but wasteful care can still be well coordinated.

Why then have concepts such as efficiency, value, and bending the cost curve become nearly inseparable from care coordination in the vernacular of US health policy? And what are the dangers of promulgating the myth that care coordination is the path to savings?

The answer to the first question is political expediency. Nobody likes waste, but nobody likes rationing either. The idea that less is more is a tough sell. The idea of care coordination bending the cost curve, on the other hand, resonates with the romantic notion shared by physician and patients alike that if we could only do more for patients, the savings would materialize. As health services and health economics research has repeatedly shown, however, the only guaranteed outcome from doing more is that more is ultimately done.

As for the potential dangers of conflating cost containment and care coordination, there are many. First, it causes the merits of care coordination to be judged on the basis of savings, diminishing the importance of coordinated care itself as a worthy goal that can enhance patient experiences and improve outcomes even if it does not reduce utilization or produce net savings. Early evaluations of the Medicare ACO programs, for example, suggest that ACOs’ care management and coordination efforts have meaningfully improved patient experiences but with minimal effects on hospitalizations for ambulatory care-sensitive conditions.^3–5 That should be considered a success.

Second, it could prevent a meaningful science of waste reduction from emerging in health care. An overemphasis on care coordination as a cost-cutting strategy could divert attention and resources away from the development of approaches that eliminate the provision of low-value services more effectively. Innovation in this area is sorely needed but undercapitalized relative to the burgeoning industry of health analytics companies promising big returns from patient engagement and seamless care. Despite the rhetoric around care coordination, at least some ACOs seem to be pursuing more direct means of lowering spending. Much of the early savings in the ACO programs have been driven by lower use of post-acute care in skilled nursing facilities (SNFs), home health care, and high-priced hospital outpatient departments—all thought to be leading sources of wasteful spending.⁴

While the specific mechanisms mediating these savings have yet to be elucidated, these spending reductions do not require improved care coordination in the classic sense of facilitating smoother transitions and better communication between providers. Rather, they can be achieved more directly by steering patients to lower priced or more efficient providers or instructing clinicians to curb excessive use when it arises (e.g., leveraging “SNF rounders” to shorten inappropriately long stays), without requiring enhanced information exchange, smoother transitions, or additional patient-specific investments. While an unusually broad definition of care coordination might include more direct strategies, expansive labeling also contributes to the misdirection of cost-containment efforts toward activities more commonly associated with care coordination that do not lower spending. If the early savings by ACOs were driven primarily by better coordinated care in the standard sense, one would expect accompanying evidence of its presumed mechanisms—particularly fewer readmissions and preventable hospitalizations—but such evidence has been largely absent.^3,4

Thus, what modest savings ACOs have produced to date appear largely due to selective targeting of obvious sources of overuse. Inhibiting the evolution and diffusion of these early efforts could jeopardize the viability of new payment models that will ultimately be judged on their fiscal impact. Regulatory efforts to encourage participation will likely be stronger if the savings are greater. In addition, guiding ACOs to exclusively focus on care coordination could cause them to fail to qualify for shared-savings bonuses no matter how much they improve quality, and thus to withdraw from the ACO programs.

Third, characterizing care coordination as a primary mechanism for cost containment perpetuates the notion that there is a tradeoff between competition and coordination in the pursuit of efficiency when this may be a false dilemma. Not only has it never been convincingly demonstrated that care coordination lowers spending, it has also never been demonstrated that large integrated health systems are necessary to coordinate care or do it better than smaller provider groups. Exaggerating the efficiencies from care coordination thus lends unjustified credibility to provider arguments that establishing ownership over the continuum of care through mergers and acquisitions will produce efficiencies under new payment models that should allay concerns over price increases from market power. On the contrary, independent physician groups in the ACO programs have thus far achieved savings that are as great if not greater than savings achieved by larger hospital-integrated ACOs.^3,4 For all we know, provider competition may play a critical role in spawning more effective ways to coordinate care and reduce overuse in response to incentives to lower spending and improve quality.

Framing is important because it can affect resource allocation in the presence of uncertainty. A change in the national conversation over cost containment may be needed for payment and delivery system reforms to live up to their promise. Magical thinking that we can spend less by doing more is counterproductive. We should coordinate care not to save money but because coordinated care is better care. Patients deserve more from the health care system and should get more. But to spend less, we need to do less. Providers and payers do not have to convince patients that less is more with complex arguments about iatrogenesis or incidentalomas. Nobody likes waste. So let’s get rid of it.