Health Care Affordability: Iron Triangle Or Iron Curtain?

“The high cost of health insurance and health care services now affects not only the uninsured but also middle-class Americans with employment-based health insurance (ESI), enrollees in the Affordable Care Act exchanges, and Medicare beneficiaries. A popular concept in health policy discussions is the “iron triangle”: herehereherehere, and here. These authors posit that it is impossible to increase access and quality of care while simultaneously reducing spending. That idea warrants further scrutiny.

[..] evidence suggests that failure to improve affordability is primarily due not to a mathematical “iron triangle” constraint, but an “iron curtain” of stakeholders who are aware of promising alternatives but oppose their implementation. [..]

We suggest an alternative approach—“discriminant” coverage designs—and offer real-world examples of this approach. We explore why discriminant designs have not been more widely adopted and why various health care stakeholders have not taken the lead in advocating for them. [..]

As acknowledged by David N. van der Goes, one of the iron triangle proponents, “major gains in efficiency” would shift the frontier outwards [..] thus breaking the iron triangle and allowing coverage expansions with no increase in total spending. Examples include the development of a new technology or passage of a state law that allows nurse practitioners to work “at the top of their license” without physician oversight.

But there is another approach to increased efficiency that may be even more effective and easier to attain than technological or legislative changes. A review of the literature on medical errors, waste, and low-value and avoidable health care (hereherehereherehere, and here) demonstrates that many health care providers lie inside the frontier [..]. Moving providers horizontally from the interior of the frontier out to the sloping left-hand portion of the frontier [..] represents lower cost with the same or better level of health.

Among providers on the sloping left-hand portion of the frontier, the difficult tradeoff between better health and lower cost—for either public insurance programs or individuals—is inescapable, and the iron triangle is one way to characterize that choice. But so long as patients are being treated by providers who are inside the frontier, discussion of the iron triangle is premature.

Instead, the focus should be on ways to facilitate movement of patients, particularly low-income patients, from provider A to providers on the frontier. Loss of patients by inefficient providers will give them an incentive to become more efficient and move closer to the frontier, thus benefiting all patients. How can this be accomplished?

The answer is to:

  1. Collect data on the prices, practice styles, and health outcomes, including structural and process measures of quality of care, for health care providers and their patients.
  2. Give the data to consumers in easily understandable formats.
  3. Share the savings with consumers who choose more efficient providers. In other words, structure the health insurance benefits so that consumers who choose more efficient providers face lower out-of-pocket costs.

We refer to health insurance benefit designs that embody those three principles as “discriminant” designs because coverage levels discriminate between high- and low-cost providers. These three steps address the affordability problem regardless of the size of the employer—giving consumers the information in understandable form will let them find lower-cost providers without depending on employer leverage or sophistication. An added benefit is that if enough employers adopt discriminant health insurance benefit designs, then high-cost providers will face pressure to become more efficient. This approach, originally introduced by Walter McClure in the early 1980s (later known as “Buy Right”), spurred early efforts at developing and sharing cost and quality measures, which have improved dramatically over the past 40 years.

We discuss two discriminant health insurance benefit designs: reference-based pricing (RBP) as implemented in the California Public Employees’ Retirement System (CalPERS); and tiered cost sharing as implemented by the state of Minnesota’s State Employee Group Insurance Program (SEGIP).

Both systems use past fee-for-service claims data to evaluate providers’ performance. Neither system requires negotiations between health plans and providers. The provider’s past performance, evaluated by anyone who has access to claims data, is taken to be the provider’s initial “best offer.” Of course, providers are free to improve their performance over time.

CalPERS RBP generally is applied to the inpatient hospital costs for a specific procedure such as total hip or knee replacement. Past claims are reviewed, and CalPERS sets the reference price at a specific dollar amount or percentile in the price distribution—in CALPERS’s case, approximately the 60th percentile for most procedures. If consumers choose a provider whose price falls below that point, the procedure’s cost is covered in full. If the consumer chooses a provider whose price falls above that point, the consumer pays the full difference out of pocket, and that expense does not contribute to the consumer’s deductible. [..]

Whereas RBP focuses on single procedures and thus the unit price of the procedure, the SEGIP system is based on total cost of care, including primary, specialty, and inpatient care, as well as pharmaceuticals. Each November, Minnesota’s state employees choose a primary care “gatekeeper” clinic that is responsible for their total cost of care in the coming calendar year. SEGIP places the clinic in one of four cost-sharing tiers based on the annual per capita risk-adjusted total cost of care for their SEGIP patients. Employees who choose higher-cost clinics face higher cost sharing. The tiers give employees clear information on the expected total cost of care for each clinic, and the cost-sharing differentials immediately share the savings with employees who choose lower-cost clinics. Data on providers’ quality of care and health outcomes are available from Minnesota Community Measurement.

The SEGIP model brings three types of pressure to bear on providers. First, if a clinic loses a patient to a competitor, the clinic and the specialists and hospitals to whom it refers patients risk losing all the revenue associated with the patient, not merely a percentage as in most provider payment initiatives (for example, accountable care organizations [ACOs], shared savings programs, and bundled payments). Second, when SEGIP moves a clinic to a higher tier, the clinic’s patients will produce less revenue due to the moral hazard effect of higher cost sharing. Third, publicly available tier information makes providers aware of their relative performance, creating peer pressure effects.

SEGIP clinics that would prefer to be in a lower-cost tier can agree to a price discount. About 20 percent of clinics currently are in price discount arrangements with SEGIP, and the discounts range from 10 percent to 20 percent.

Despite their clear incentives for provider-driven improvements in efficiency and demonstrated ability to identify lower-cost providers, neither RBP nor tiered cost sharing has been widely adopted in either ESI or Medicare. In its 2019 survey of ESI offerings (the last year in which the question was asked), the Henry J. Kaiser Family Foundation found that only 14 percent of employers with more than 50 employees were offering any type of high performance or tiered network design, similar to the percentage in 2018.

The most frequently cited reason for non-adoption (28 percent) was disruption of employee-provider relationships or employee backlashAnna Sinaiko also found that employers were concerned about RBP’s complex benefit design, with potentially large out-of-pocket costs that would be difficult to explain to employees. Dennis Scanlon identified some of the same employer concerns but added an important factor: Insurers and third-party administrators do not facilitate awareness of price variation or make reference pricing easy to implement. [..]

But which employees are most likely to be upset? It seems probable that the employees most likely to find reference pricing and tiered cost sharing disruptive are the highly paid employees who prefer to patronize expensive, prestigious providers. Why should they object to discriminant insurance benefit designs?

If any employee somehow finds and chooses lower-cost providers but receives no financial reward for lowering their expected total cost of care, then those employees provide an implicit subsidy to employees choosing higher-cost providers. Discriminant coverage could end that subsidy.

Similarly, when all employees pay the same premium and receive the same level of coverage, for example, a large deductible, the employees who cut back the most on consumption of health care services—likely the lower-income employees—provide another implicit premium subsidy to higher-paid employees who are less deterred by the high deductible. Keeping the high deductible for the latter employees while reducing the deductible or premiums for employees who patronize more efficient providers would result in higher out-of-pocket costs for the high spenders. The same regressive redistribution is true of mandated benefits if higher-income individuals are more likely to consume the services covered by the mandate. These conditions all lie atop the regressivity of the tax deductibility of health insurance premiums that allows employees in higher tax brackets to purchase the same health insurance policy at a cheaper price than employees in lower tax brackets. Victor Fuchs provides this succinct summary: “In the United States, the interests of high-income individuals dominate decisions about what medical care is offered and how it is financed.”

[..] we conclude that the problem with affordability in both publicly and privately financed health insurance is not primarily a mathematical constraint like the iron triangle, but an iron curtain of providers, health plans, and well-to-do, heavily subsidized beneficiaries, employers, and employees who like the system just the way it is.

[..] current efforts in public and private health plans should focus on facilitating patient movement from the interior of the frontier [..] to the frontier itself. Basing that approach on informed consumer choice does not require changing provider behavior, although provider behavior may well change when inefficient providers begin losing patients. Nor does it require negotiating with providers, engaging in regulatory guesses about the efficient market equilibrium price for provider’s services, or relying on non-clinicians’ pre-certification restrictions on providers’ use of resources.

Nor does encouraging patient movement require increasing taxes or the federal debt. It requires only telling consumers where to find more affordable providers and sharing the savings with consumers who choose them.”

Full article, B Dowd and T McDonald, Health Affairs Forefront, 2023.2.16