The Pricing of U.S. Hospital Services: Chaos Behind A Veil of Secrecy

Uwe E. Reinhardt

Disclosures

Health Affairs. 2006;25(1):57-69. 

In This Article

Hospital Pricing and Consumer-Directed Health Care

Until now, the U.S. health care "market" has been analogous to an imaginary world in which, say, employers offered to reimburse their employees 80 percent of the "reasonable cost" of all attire deemed "necessary" and "appropriate" on the job but, under the contracts negotiated with department stores by the fiscal intermediaries administering this "Clothes Benefit Program," employees had to enter department stores blindfolded. Only months after a shopping trip would the employee receive from the fiscal intermediary a so-called Explanation of Benefits (EOB) statement, explaining how much the employee had to pay for whatever he or she had stuffed, blindfolded, into the shopping cart on that shopping trip. Framed in bright red on that EOB would be the statement: "Pay X amount." X would represent 20 percent of what the intermediary would have judged, ex post, to be "reasonable prices" for those garments in the shopping cart deemed by that intermediary, ex post, to have been "appropriate" attire for the particular employee's circumstances. It also would include 100 percent of the prices charged by the stores for items in the cart that were deemed by the intermediary, ex post, as "not necessary" or "inappropriate" and that were therefore not covered by the Clothes Benefit Program.

Ridiculous though it sounds, such an arrangement closely resembles the current payment system for U.S. health care. It is difficult to reconcile this picture with increasing demands by employers, insurers, and policymakers that patients be forced to act as more responsible "consumers" of health care, a movement now gathering force under the banner of consumer-directed health care.

This term has come to describe health insurance policies with annual deductibles ranging anywhere from $2,000 to more than $10,000 per family and, often, coinsurance in addition. The costs borne by the insured can be defrayed out of health savings accounts (HSAs) into which both employers and families can make annual deposits that are not taxable income to the employee. Amounts in the HSA not spent in a given year can be rolled over to the following year, which means that for chronically healthy families, the HSA can become a major, tax-preferred savings account earmarked for future out-of-pocket payments on health care.

This construct can be offered by employers, in lieu of the traditional, more comprehensive employer-sponsored insurance. It also can be procured by households in the market for individual insurance.[32] A major advantage to both is that the premiums for the construct are lower than those for more generous insurance coverage.

The central idea of consumer-directed care is that the high degree of cost sharing will force patients to take a more active interest than they hitherto have had in the cost-effectiveness of their care. This "consumer empowerment," as it is sometimes called, can only occur, however, if prospective patients actually have easy access to user-friendly, reliable information on at least three dimensions of their care: the prices charged by competing providers of health care; the costliness of practice styles adopted by these various providers—that is, the prices times the quantities of services and supplies they package into the treatments they render; and the quality of these providers' services. If such a transparent information infrastructure now exists anywhere in the United States, it would be the rare exception.

In connection with hospital care, of course, it could be argued that prospective patients ("consumers") require only better information on "quality," because the cost of even moderately expensive hospital stays typically will exceed the patient's annual deductible, so that "price" and "costliness" are no longer important to the patient. That argument overlooks the fact that a growing fraction (now more than a third) of all hospital revenue comes from outpatient services, whose individual costs might be below the deductible. Furthermore, patients increasingly face coinsurance payments for inpatient care, too, which gives them an interest in the actual prices of those services. Thus, the question remains how the prices charged by hospitals could be revealed to prospective patients in ways they could digest and act upon as "consumers."

In June 2004, Michael Porter and Elizabeth Teisberg published a paper titled "Redefining Competition in Health Care."[33] Much of that paper repackages familiar ideas that have long been espoused by policy analysts such as Ellwood, Enthoven, and Etheredge; Pauly, Danzon, Feldstein, and Hoff; Herzlinger; and numerous other authors who have wrestled with the idea of forcing genuine price and quality competition onto the U.S. health system.[34] But Porter and Teisberg do offersome radical ideas on the pricing of health care services, which makes them relevant to the present inquiry.

In a nutshell, they would replace the current payment system with an all-payer system, albeit one centered on individual providers. Thus, a hospital could set its own prices, but it would have to post them for public view and apply them to all patients, without price discrimination. Where possible, prices for individual services and supply items would be bundled into lump-sum prices for major procedure categories, akin to Medicare's DRG and APC systems.

In 1993 I proposed a somewhat similar idea.[35] I proposed that the government should expand the DRG system to all hospital patients (which would now include the recently established APC for outpatient services). There would be only one national set of weights for the various DRGs and APCs, which every hospital would have to adopt. These weights would be developed and continuously kept up to date by an authoritative national body of experts—perhaps the current Medicare Payment Advisory Commission (MedPAC).

To avoid the potential pitfalls of centrally administered, uniform prices for the entire hospital system of as far-flung a nation as the United States, each hospital would be free to set its own monetary conversion factor for DRGs and APCs.That hospital-based conversion factor would then translate the national relative value scales into hospital-specific, case-based fee schedules. Each hospital, of course, would be required to make its monetary conversion factor publicly known. It would be a one-dimensional index of the hospital's absolute level of case-based prices and could be easily understood and used by patients.

Even streamlined systems such as these would still confront prospective patients with lists containing more than 1,000 prices for the various hospitals' DRGs and APCs. But these price lists could be supplemented by smaller lists, showing the total cost billed by a hospital for the far fewer cases that constitute, say, half of its total revenues. Furthermore, software could be written that would enable prospective patients to obtain, from a dedicated Web site, comparative averages or medians of the total prices actually billed by a hospital in the past year for a specific case not on the smaller list. Although never perfect and easy, any such system would be a step far ahead of the chaos that now reigns behind the opaque curtain of proprietary prices in the U.S. hospital system.

Such proposals are, of course, are more easily stated than implemented. To transit from the current payment system to the proposed system in a way that would not unfairly create winners and losers raises a host of conceptual and practical questions, even leaving aside for the moment the nettlesome problem of the uninsured and underinsured.

There is, for example, Porter and Teisberg's recommendation that hospitals must accept third-party payments as payment in full, without balance billing of patients (aside from regular coinsurance or deductibles that may be extracted from patients by the third-party payer). Is that imperative? An alternative approach might be reference pricing, which might also be called "defined-contribution pricing." Here third-party payers (perhaps even Medicare) would pay hospitals no more than a stipulated conversion factor, benchmarked perhaps on lower-cost hospitals in a market area, which would leave patients to pick up the entire difference between the conversion factor covered by the third party payer and the conversion factor the hospital has announced publicly and actually charges. It would be a version of the tiered pricing now being contemplated by some private insurers.

There is the other open question of whether a system of common relative value scales for inpatient and outpatient care would continue to allow individual hospitals to charge different payers different, negotiated monetary conversion factors. A case can be made for permitting it, as long as patients with high-deductible policies would be charged by a hospital only the conversion factor that was negotiated with that hospital by their catastrophic insurer. To shop around for cost-effective care among hospitals in a market area, patients then would have to know all of the conversion factors that their own insurer had negotiated with all of the competing hospitals in the relevant market area. That information, in turn, would reveal to all the world all of the negotiated conversion factors. In the end, the system would most likely lead to something approximating an all-payer system in a market area, without the need for added regulation.

Under any such novel payment system, separate provisions would have to be made for low-income households and self-paying patients, whether or not price discrimination among third-party payers were to be allowed. One approach would be to have hospitals post their means-tested conversion factors, which presumably would be income related and might be zero for genuine charity cases. As Porter and Teisberg also observe on this point, however, in the long run any U.S. hospital payment system will be seriously impaired by the presence of large numbers of uninsured Americans.

Comments

3090D553-9492-4563-8681-AD288FA52ACE

processing....