Alternative Payment Methodologies
In response to the charge to assess alternative payment methodologies, the committee identified a range of options for paying for different types of medical services, using the existing literature and surveys and interviews with selected payers and providers.1 This broad list of options was clearly too extensive for analysis within the time and resources available. The committee, therefore, narrowed its focus to those methods likely to be most suitable to the particular characteristics of clinical laboratory services. Discussion of alternative methodologies in this chapter is organized around the structure and elements of payment systems considered initially in Chapter 4 and the goals that the committee articulated in Chapter 1.2
In the next section, the committee analyzes the major elements of a payment system, discusses the alternatives for structuring each element, and assesses the
See Appendix C for a description of the methodology used in the payment survey conducted for the Institute of Medicine (IOM) by CHPS Consulting. Katie Merrell (see footnote 2) collected information during January and February 2000 through interviews with several key informants from the laboratory and insurance industries. Informants were identified through specialty societies and other knowledgeable sources and were assured anonymity.
Very little literature exists on the issues surrounding payment methodologies for clinical laboratory services. This chapter, therefore, draws heavily from the literature on paying for other health services providers, particularly physicians, and on general health services research. Much of the chapter is drawn from a paper commissioned by the IOM for this study by Katie Merrell, Center for Health Administration Studies, University of Chicago.
strengths and weaknesses of these options. The third section discusses other key design issues, such as updating the payment schedule, payment adjustments, and cost sharing by beneficiaries. Finally, the chapter concludes with a discussion of implementation considerations, including the legislative and administrative steps necessary for implementation and the paperwork and financial costs of introducing and using the various options.
The current payment system and its various elements were among the options considered by the committee. Since elements of the current payment methodology are discussed extensively in Chapter 4, they are mentioned only briefly in this chapter. Also, the committee’s assessment of the current system, discussed in Chapter 5, is not repeated here, but it did influence the discussion of the advantages and disadvantages of the various alternatives.
ELEMENTS OF A PAYMENT SYSTEM
Type of Payment
The committee recognizes the general advantages of prospective payment and did not examine in depth any retrospective payment methods for clinical laboratory services.
Payment amounts can be determined on a retrospective or prospective basis.3 Retrospective payment means that the amount paid is determined by (or based on) what the provider charged or said it cost to provide the service after tests or services had been rendered to beneficiaries. Here providers have no incentive to hold down their charges or costs, and the payer has few mechanisms for controlling expenditures (Sing et al, 1998). In a prospective payment system (PPS), prices are set in advance and are known (or knowable) by all parties before care is provided.
Prior to 1984, Medicare paid clinical laboratories on a complicated retrospective system, similar to that used to pay physicians. The reasonableness of the charges was judged based on customary, prevailing, and reasonable criteria. Since the mid-1980s Medicare, as well as virtually all private payers, has moved to using prospective payment systems. The Health Care Financing Administration (HCFA) has paid for inpatient hospital services on a prospective basis since 1983. It recently implemented a prospective payment system for outpatient hospital care and is currently designing a PPS for the remaining specialized hospi-
tals not yet covered by a prospective system. Medicare pays for physicians’ services on a prospective basis and for laboratory services, defacto, on a prospective basis.
A retrospective payment system gives providers greater influence over payment rates than they generally have under PPS. In turn, it reduces the payers’ ability to constrain expenditure growth. The use of prospective payment leads to more predictable payment levels for payers, patients, and providers and is typically associated with simpler administrative systems. Prospective payment systems also provide an opportunity for the payer to exercise some control over total spending through such mechanisms as constraints on updates and volume adjustments.
Unit of Payment
The committee concludes that the risks of a capitated payment for laboratories outweigh its advantages of administrative simplicity and efficiency and that payment per service or test is preferable.
The unit of payment for laboratory services can be defined in several ways. It can reflect a single test or service, a group of services, or all potential services used by a beneficiary in a specified period (capitation). Under the current payment per test or service, tests are identified and classified through specific HCFA Common Procedural Coding System (HCPCS) codes, which encompass the American Medical Association’s (AMA’s) Current Procedural Terminology (CPT). A dollar amount is then set for each coded service. Under a method that groups services for payment purposes, related laboratory services are bundled. For example, panels of automated tests are bundled into groups and a payment amount is set for each group of tests. Under capitated payment, the provider is paid a fixed amount per beneficiary for a list of covered tests that may be medically necessary during a given time period. The payment is provided whether or not beneficiaries use any services.
Payment per Test. Each test or service has an assigned CPT code or HCPCS code in the case of Medicare. For most tests and services, it is clear to the laboratory which HCPCS code to use. Multiple testing methodologies for a single analyte or similar methodologies testing different analytes may be represented by a single code number, so assigning an appropriate payment amount can sometimes be problematic.
Payment per Group of Tests. For tests that are routinely and frequently ordered together, bundling them and using panel-level payments can create administrative savings to physicians and the program. If the test can be done on a multichannel analyzer, the laboratory’s cost savings from performing all of the tests at once should be reflected in the payment. The payment for automated chemistry panels should be less than the sum of payments for individual tests in the panel because of economies of scale. The appropriate use of panel payments is complicated, in part, because the definitions of panels as well as the rules governing their payment have changed frequently over the years. Relatively few tests are included in panel fees, but they tend to be among those most frequently ordered. Grouped or panel tests could be a component of a system using payment per test or service, but could not be a separate option, because bundling is not appropriate for all tests.
Capitation. The use of capitation in the health sector grew during the 1990s. Increasingly, private health plans that receive capitation payments from Medicare under Medicare+Choice or from private purchasers, use capitation payments to buy laboratory services. Two basic approaches are used to capitate laboratories: (1) a managed care organization pays the laboratory an agreed-upon amount per member per month to provide all, or a defined list of, ordered laboratory services; or (2) the managed care organization includes laboratory services in the capitated payment to the physician.4 The latter approach is rarely used.
Capitated payments for laboratory services under fee-for-service Medicare are conceivable, although the committee is not aware of this method being used. This approach would require Medicare to contract selectively with certain laboratories to provide services to identifiable subgroups of beneficiaries, possibly on the basis of geography. It would be necessary to determine an appropriate per-person capitation payment rate. One or more laboratories would be paid a capitated amount to provide required laboratory services to the covered group of beneficiaries. By determining the size of the covered groups, Medicare would effectively be deciding how many laboratories in each area (or nationwide) would get Medicare business, which in turn could have tremendous implications for the numbers and types of laboratories that survive in future years. Physicians
treating beneficiaries and ordering outpatient laboratory tests would have to send patients or specimens to the appropriate selected laboratory, much as they would for private patients in managed care.
Selection of the unit of payment should include consideration of administrative feasibility, efficiency, incentives for appropriate use of services, and adaptability to new technology. The incorporation of new technology into the system is affected by how finely the unit is defined. If the unit were defined by very specific test characteristics, then new service codes and payments would be needed every time a new, slightly different test is approved for Medicare coverage. Conversely, if the unit were defined by broad categories of tests or some basic equivalent service, then new codes and amounts would be needed only for breakthrough technologies that create a new class of test. The gain in administrative simplicity with broad code categories could limit access to new technologies that are not substantially less expensive than existing tests. That situation would arise because manufacturers might be reluctant to develop a new, relatively expensive technology of higher quality that is similar to an existing test and would be included in a current code and payment amount, if they felt the existing fee was insufficient. On the other hand, such a deterrent might be useful in the case of new, more expensive technologies that are not a significant improvement over current methods.
Continued use of payment per test or per service would be most consistent with other Medicare Part B, fee-for-service policies. The current system of payment per service is familiar to all stakeholders, would require no major system change, and could provide a fairer basis for payment than capitation. There would be no disruption to the structure of the laboratory industry with this option. It would be difficult to set an appropriate formula for calculating capitation rates, given the current level of knowledge about clinical laboratory utilization rates by the Medicare population. With test claims data, HCFA could more readily track patterns of practice and costs per episode if it so chose. Because of shortcomings of the current coding system, it would be worthwhile to explore alternatives for future use.
Capitation generally is thought to promote more efficient, appropriate, simple, and financially sustainable care. It also carries the risk of providers’ inappropriately using lower-quality tests that are less costly and limiting the number of tests ordered so as not to exceed the constraints of the per capita payment. Because several hundred million outpatient clinical laboratory tests are processed annually by Medicare, using a monthly capitation payment for each participating laboratory would mean a significant reduction in HCFA’s administrative burden and costs. It would be an administrative challenge for physicians as well as HCFA, however, to ensure that physicians and patients use the appropriate, capitated laboratories.
With a capitated Medicare laboratory payment, as with capitated laboratory services offered to managed care patients, the laboratory would have no control over utilization. In this situation, the purchaser, Medicare, would also have little control over a physician’s test-ordering behavior. Care would not be managed, merely capitated.
The potential effects on the industry of selective contracting for laboratory services would be a drawback of capitation. The possible disruption of the entire structure of the industry under a capitated laboratory payment with very limited numbers of contractors could have a negative impact on beneficiary access, quality, and value. Efforts to mitigate negative effects on the clinical laboratory industry’s composition, such as increasing the number of contracted laboratories or frequently revisiting contracts to encourage losers to re-compete and winners to provide good service, would likely increase the administrative complexity of this approach and reduce some of the expected gains.
Basis of Payment
The committee recognizes the advantages of a single, national fee schedule based on resource costs for fee-for-service payment.
Prospective fee-for-service payment, almost by definition, uses a fee schedule. The question is whether the fee schedule should reflect charges, competitive market prices, favorable pricing, or resource costs. Each option implies a different choice of payment methodology: charge-based payment, competitive bidding, a “most-favored-nation” (MFN) approach, or a fee schedule based on the National Limitation Amounts (NLAs) or resource costs. Some options more naturally lead to multiple fee schedules, while others result in a single national fee schedule. Multiple fee schedules may better reflect local market conditions. A single national fee schedule, however, has the potential to be simpler to administer, more understandable to stakeholders, and more equitable among laboratory service providers.
The following sections review two market-oriented approaches for developing fee schedules—competitive bidding and MFN—and two administrative approaches—NLAs and the establishment of a resource-based relative value scale (RBRVS). It is beyond the scope of this study to examine possible barriers to market entry, collusion, or other potential sources of market power in the laboratory testing industry. In addition, the limited availability of information about the current structure, competitiveness, and quality of the laboratory industry makes it impossible for the committee to assess the likely market implica-
tions of particular payment approaches. In the following discussion of alternative approaches, there are no precise estimates of how alternative approaches would affect the number and types of laboratories in a particular market. Because current Medicare payment for laboratory services derives from charges and Chapter 5 has assessed the merits of this charge-based payment, using charges as the basis of payment is not reviewed in this chapter outside the discussion of the NLA.
Competitive Bidding. This is a method by which buyers and sellers come to agree on prices. Competition among sellers to win business encourages each to reveal the minimum price at which a sale is acceptable. The goal of competitive bidding is to secure for Medicare a set of prices that reflect the cost of efficient production, including a normal profit. Bidding is supposed to discover these prices without the need for intrusive data collection or any explicit decision by Medicare about the amount of profit to be earned by individual firms.
The key to competitive bidding as a strategy for developing appropriate rates is in how the auction is actually structured. There are several design issues discussed below that could affect the operation of competitive bidding as well as the results.5 For purchasing clinical laboratory services, HCFA has developed various competitive bidding models that typically have the following features:
Multiple contracts would be awarded to avoid the risk of bad performance. This would allow physicians and their patients to move their business to laboratories that give satisfactory quality and service.
Laboratories would bid a price for each of a set of tests. These bids would typically be weighted by volume and evaluated. Winners would be selected based on a stated protocol. They would receive no guarantee of business, only a right to market their services and be paid at the winning prices. The number of winners would affect the number of fee schedules in use.
Bidders whose prices were too high either would be excluded from participating in the demonstration area during the period of the demonstration or would be allowed to participate at a discounted rate. Exclusion of losers would enhance the value of being a winner by increasing the prospect of business.
In most designs, bids would be required from all commercial laboratories and from hospitals that market outpatient laboratory services to the general population of physicians. Physician office laboratories (POLs) would not be required to bid and would automatically be paid at the winning prices. If there were federal budget constraints on Medicare, an upper limit on the bidding process might be necessary.
Prices determined by competitive bidding would be exempt from federal budget adjustments or rate cuts during the period of the procurement. Price updates would occur automatically with each round of competitive bidding procurement.
Strengths of Competitive Bidding. Competitive bidding should produce prices that are close to bidders’ actual costs. Conventional economic analysis, if applied to this issue, would suggest that competitive bidding could encourage lower costs and innovation by limiting existing firms’ market power. The government expects that these prices would, in most cases, be lower than those it pays under the current laboratory fee schedule and thus produce a net savings in total expenditures for Medicare, but this is not certain. There would be minimal disclosure of business data since laboratories would not have to reveal the basis for their bid, the details of their cost structure, or the amount of their profit.
HCFA became interested in competitive bidding for outpatient clinical laboratory services in the mid-1980s and issued contracts for both sizable studies of the laboratory industry and design options for a demonstration. Since 1995, HCFA has invested more than $350,000 and has created a significant body of literature analyzing design alternatives for competitive bidding (Hoerger and Waters, 1993; Hoerger et al., 1997; Mennemeyer et al., 1986). HCFA has an operational demonstration in one county for competitive bidding on durable medical equipment and supplies and it resulted in first-year aggregate savings of 17 percent, without a reduction in access. (DeParle and Berenson, 2000).
Weaknesses of Competitive Bidding. Under an exclusive bidding model, or selective contracting, where only firms submitting winning bids would be allowed to participate in Medicare, losing firms would not be able to participate in the Medicare program in the designated area during the time of the procurement. This could have a significant effect on the financial health of excluded laboratories and the structure of the entire industry. Even without the exclusion of losers, the impact of competitive bidding could disproportionately disadvantage certain segments of the laboratory industry. The committee had insufficient data to conclude whether the present number and mix of laboratories is appropriate or not to meet Medicare beneficiaries’ needs most efficiently. Hence it makes no recommendation whether the current size and structure of the laboratory industry should be maintained or changed. Nevertheless, the committee recognizes that policy changes that do significantly change the structure of the industry would likely have more short-run implications for Medicare beneficiaries and physicians than would policies that maintain the status quo. Such changes may be appropriate and desirable, but it is important that they be anticipated and steps taken to minimize the dislocations experienced by beneficiaries as they occur. The committee thought there might be efficiency gains from competitive bidding in terms of Medicare payments for services in some markets, but it is unclear how they would compare with the added administrative
costs of running competitive bidding processes. Hence, a demonstration would provide useful experience and data.
Competitive bidding would likely result in multiple fee schedules across the country and possibly even within separate bidding areas. This could add administrative complexities to the program and has the potential for laboratories in different areas to be treated unequally.
The committee believes that developing a feasible competitive bidding process requires testing through a demonstration in order to gather information about the needs for administrative resources, HCFA management at the local level, how beneficiaries and physicians would respond, how to educate them, and how their access to services would be affected. Although much developmental work has already been done, it would nevertheless take an extended time for a demonstration to become operational and evaluation findings to be available. At that point, it should be clearer whether the approach would be practical and advantageous on a national or local level.
There has been much opposition from the laboratory industry to competitive bidding. HCFA’s work was halted in 1987 when Congress imposed a moratorium in response to industry opposition. Provider opposition has prevented implementation of all Medicare competitive bidding demonstrations other than the one for durable medical equipment, not just the ones for laboratories. A former HCFA administrator concluded recently that the failure of competitive pricing demonstrations four times, in part due to the lack of broad political support, did not bode well for HCFA to develop new market-based approaches for Medicare (DeParle and Berenson, 2000).
The committee finds that the disadvantages of competitive bidding outweigh its advantages for use as the basis of payment. Nevertheless, it considers this method again as a possible means of collecting data that could inform calculation of the level of payment.
Most-Favored-Nation Approach. This title is borrowed from the language of international trade and refers to a system in which laboratories would provide services to Medicare beneficiaries for the lowest rate they accept from any other payer. In effect, each laboratory would create its own Medicare fee schedule. This approach would tend to eliminate laboratory discounts to selected private payers. In theory, the MFN approach could lead to economically efficient pricing across payers, minimizing the risk of Medicare subsidizing the discount that laboratories offer to other payers. It is uncertain how laboratories would translate private capitation rates into reasonably comparable amounts or fees per individual tests. Even comparisons of Medicare test fees with those of private payers can be difficult if the private payer includes different bundles of tests in its panels. MFN would result in variation among laboratories in relative and actual payment rates. It is not clear what enforcement policies, such as audits and reporting requirements, would be necessary to ensure that providers charge Medicare no more than their lowest commercial rate.
A concern of some laboratorians is that, although the test may be the same, services provided to Medicare are not comparable to the services that laboratories offer other payers because the laboratory must provide substantially more documentation for Medicare. Thus, the costs of doing business with Medicare may exceed the costs of serving private patients. Some laboratories feel justified in charging Medicare more to cover these costs. The committee has found no cost data to support or refute this claim.
Strengths of MFN. In theory, the MFN approach would allow Medicare to pay no more than private payers and could create a more equitable arrangement among all payers. It would not initially require extensive and possibly intrusive data collection. To the extent that Medicare has been cross-subsidizing lower rates to physicians or other private payers, this approach could produce lower payments for Medicare.
Weaknesses of MFN. The biggest drawback of MFN is the difficulty in ensuring that laboratories charge Medicare according to the rules. Detection and prevention of fraud could present major administrative problems. Other administrative functions, such as changes in computer systems to accommodate laboratory-specific fees that change continually could also cause difficulties. The MFN system could become costly if laboratories were less willing to bargain for discounted rates with managed care companies because the same rates apply to Medicare. If Medicaid was the lowest payer in a market, however, the laboratories probably would not have the same flexibility to negotiate rates with the state as they do with private payers. If Medicaid was included in the MFN approach, it could significantly lower Medicare rates for laboratories in these states or create an incentive for laboratories to drop out of the Medicaid program. Substantial developmental work would be necessary to design an MFN system, and as with competitive bidding, demonstrations would be desirable before national implementation is considered.
National Limitation Amounts. Initially, HCFA could consider establishing a single national fee schedule based on NLAs. This would look very much like current payments since most of the fees and even more of the payments (98 percent) are now constrained by the NLAs (Appendix B).
Strengths of NLAs. NLAs are available, known to all stakeholders, would cause minimal disruption to the industry, and would likely have a negligible effect on beneficiary access in the short-term. Since these fees are within the range of private payer fees, NLAs have face validity. A single, national fee schedule would be simpler to manage and explain to stakeholders than multiple fee schedules such as the current 56 carrier fee schedules or laboratory-specific schedules.
Weaknesses of NLAs. NLAs are based on historical charges, which may bear little logical relationship to the costs of providing specific laboratory services now. If used indefinitely, they could distort incentives for laboratories to offer certain tests.
Resource-Based Relative Value Scale. HCFA spent millions of dollars during the late 1980s and 1990s to develop an RBRVS for physician services. Much of the theoretical and practical lessons from that experience could inform the development of a resource-based fee schedule for clinical laboratory tests or services. Such a relative value scale (RVS) requires data on actual or estimated resources used for individual laboratory tests or services and shows the relative relationship of these resources from one test to another. Methods for collecting the necessary data on resources, including costs for equipment, supplies, labor, and other direct and indirect costs, used to produce laboratory services and their cost are discussed in the next section. The relationship of payments across services affects providers’ relative willingness to provide each service and in turn can affect efficiency, appropriateness, technological innovation, access, and the composition of the industry.
Policymakers have to consider whether such a relative value scale should be based on the costs or resources of an efficient laboratory (and how such a laboratory would be defined and identified), on the mean or median costs or resources of all laboratories (or a sample), or on some other measure. These design decisions will be important to stakeholders and will affect the way relative values are developed and the cost of their development, but it is not clear how much impact such a decision would have on relative, rather than actual, levels of payment. For example, the actual dollar cost of a test in an efficient laboratory might be significantly lower than the mean or median cost of all laboratories. Nonetheless, the cost of Test A might be twice as much as that of Test B, whether it was conducted in an efficient laboratory or a laboratory with median costs. The relative values would be the same in the different laboratories, although the level of costs would differ.
Strengths of an RBRVS. Relative values based on costs or resources would likely minimize financial distortions or incentives to provide some services rather than others (PPRC, 1989). In any case, the creation of some cost or resource-based data on laboratory tests is essential to assessing the adequacy of current and future payment levels. It would not be necessary to have precise data on each of the 1,100 codes in order to establish a laboratory RBRVS. Data on some tests within groups of closely related codes could be used to estimate values for those codes lacking data. Such a scale could readily be converted to dollar amounts for a single national fee schedule by the use of a dollar conversion factor and adjusted as necessary.
Weaknesses of an RBRVS. A relative value approach could be very difficult to design because the resource costs are difficult to establish. Furthermore,
some frequently performed tests or services are parts of panels that use the same specimen and equipment for all tests. The costs associated with each test are almost impossible to separate accurately. Also, there are numerous challenges to collecting cost data from laboratories.
Level of Payment
If a resource-based relative value scale were used to determine payments for clinical laboratory services, multiple approaches could contribute useful information on resource costs for developing the relative value scale. Issues related to setting the initial conversion factor also must be addressed.
A fee schedule shows the payment amount set for each test or service—that is, the level of payment. The previous discussion on the basis of payment noted that certain payment methods would automatically result in a dollar payment level: competitive bidding, MFN, and the NLA. The RBRVS would result in a list of ratios showing the relationship of each test to a standard test. A conversion factor would have to be used to turn ratios reflecting the relative value of each test into dollars.6
This subsection reviews four approaches that the committee considered for collecting the data necessary to establish resource-based relative values. If it were possible to collect consistent, accurate cost data readily from a sufficiently large sample of laboratories for each test on the fee schedule, this discussion would be simpler, but obtaining objective and reliable data is difficult. Each approach to developing a laboratory RBRVS has its own strengths and limitations. The approaches differ with regard to cost and likely acceptance by stakeholders. Some of these approaches may be more easily adapted than others to setting national relative values. They are not mutually exclusive, and by using an appropriate combination of approaches, perhaps the weaknesses of each can be minimized. The options include micro-costing studies, a consensus approach, charges, and demonstrations using competitive bidding or negotiations.
Regardless of the approach adopted to establish relative values, a resource-based fee schedule also needs a dollar conversion factor to translate the RVS into payment amounts. This combination of the RVS and the conversion factor determines the fee schedule’s level of payments. Setting the initial conversion
factor for a newly established fee schedule is thus a critical policy decision. Considerations in setting the conversion factor are discussed at the close of this subsection.
Micro-Costing Studies. Micro-costing studies using standard accounting practices involve collecting detailed data on direct costs and developing an appropriate basis for allocating indirect costs to specific services to create service-level cost estimates. Direct costs are those that are incurred in providing a specific service to a particular patient; indirect costs are those that would accrue regardless of the number or mix of services provided in a particular day. The lack of standard cost accounting systems for the laboratory’s costs, distinct and separate from the hospital or physician’s practice makes any data collection difficult. In addition, there are no consensus documents or conventional methods for micro-cost accounting that are agreed on by laboratory experts and used throughout the industry. To reduce problems of comparability of data from laboratory to laboratory, any cost study would have to establish definitions that are meaningful to survey respondents, regardless of their own accounting practices. There also has to be a standardized method to allocate indirect costs that is used consistently from laboratory to laboratory.
Collecting data on direct costs is a serious challenge in micro-costing studies. A number of important methodological issues have to be addressed that will affect the quality and collection costs of the data and stakeholder acceptance of both the process and the outcome. Decisions must be made about the type of sample: Should it be representative of all laboratories, of laboratories weighted by Medicare service volume or revenue, of “efficient” laboratories, or of some other group of laboratories? Similarly, the question of how to administer the survey and validate results—by phone, by mail, or in person—also has important implications for data costs and validity.
Strengths of Micro-Costing Studies. The cost data acquisition method not only is important for establishing initial relative values, but also could be used on a periodic basis for contributing to reviews of the relative values. It also could provide insight into whether Medicare was paying the “right price” for a particular test. The size of the sample, number of tests, and specific research methods could be scaled according to the research funds available.
Weaknesses of Micro-Costing Studies. Such an approach has been used in limited ways to date, but it has not yet been used successfully on a national basis to establish national Medicare payment rates for any providers. HCFA incorporated a modest, reasonably simple, cost survey administered by mail, with phone follow-up, as part of its efforts to develop new practice expense relative values for the Medicare Fee Schedule for physicians. HCFA discontinued the survey because of the low response rate of less than 30 percent to the first round of the study.
The committee used a very small resource-costing study to gain insight into how current payment rates relate to service-level costs and to understand the process that would be required to conduct such a study on a broader scale.7 The cost survey, including site visits, encountered serious difficulties. Researchers found no uniformity of existing accounting systems from laboratory to laboratory and a reluctance of laboratories to participate in a time-consuming data collection process without compensation. Given the current sensitivity of the laboratory industry to divulging cost data, the potential response to a HCFA cost survey could be poor if there were no direct incentives or penalties associated with responding.
Consensus Approach. A consensus approach relies on interdisciplinary groups of experts and aims to get some estimate of relative costs or resource use at the service or test level. It can be structured to provide opportunities for public input. Because this approach is used to develop relative values, not to set prices, it can minimize the potential for anti-competitive regulatory capture. Data-driven consensus approaches can have more credibility and can lead to more accurate results than micro-costing studies alone. Data drawn from various sources— including audits, competitive bidding and negotiations, and actual costs and charges—can inform, challenge, and validate the decisions of a consensus panel.
This approach has been used to establish physician work and practice expense values under the physician fee schedule and is the basis for periodic review of the entire fee schedule. As a result, HCFA has experience in establishing this type of process, working with providers, and using the results to develop (and later refine) relative values.
Strengths of the Consensus Approach. This approach could be much less expensive than an extensive micro-costing effort. It could be readily used in conjunction with other approaches to integrate data from various sources. If well structured, it can enhance the credibility of the resulting relative values among stakeholders.
Weaknesses of the Consensus Approach. Considerable attention must be given to structuring the process because the acceptability of the results to stakeholders will likely depend largely on their view of the fairness of the process used to select participants, the structure of the meetings, and how rigorously
information is used to establish relative values. Care must be taken to avoid the possibility or appearance of regulatory capture.
Charges. Relative charges could be accepted as a reasonable indicator of relative costs, could be compared with cost data to establish the relationship initially, or could provide the starting point for a consensus process. Since charge data are continuously updated and available to HCFA, they could also be used for future revisions of the national fee schedule.
Strengths of Charges. Developing charge-based relative values would be relatively inexpensive and administratively simple since HCFA has charges recorded on all laboratory claims submitted for payment. In addition, a charge-based system provides an automatic and timely methodology for accommodating the need to set payments for new laboratory tests and for updating the relative values of the national fee schedule.
Weaknesses of Charges. This approach is based on the assumption that charges are highly correlated with costs. That assumption may be spurious, and any correlation may vary for different types of providers and different markets, and be affected by payment policies of major payers. Without reasonable evidence that charges accurately reflect costs, stakeholders unhappy with the approach could present data showing low correlation between charges and costs, undermining the credibility of the new fee schedule.
Competitive Bidding and Negotiation Demonstrations. Information on relative prices resulting from a competitive bidding demonstration could provide the basis for setting relative values. The demonstration could be structured to provide data for establishing initial relative values and would not necessarily be repeated annually or implemented in a manner that ultimately excluded laboratories in the demonstration sites. Although the resulting prices from this competitive bidding demonstration would be used only to establish relative values, not actual payment rates, they could provide information for assessing whether Medicare is paying too little for laboratory services or is a relatively generous payer.
A variant of competitive bidding would be a demonstration using a negotiation process to establish acceptable fees that could be the basis for calculating relative values. This approach would draw from experience in the private sector, since some health plans negotiate contracts directly with laboratories. For Medicare, the carrier could negotiate with laboratories in its area.
Strengths of Competitive Bidding and Negotiation Demonstrations. These approaches could reveal market prices and the relationship between payments for one test and another with minimal federal intrusion. HCFA has done considerable research on competitive bidding. On a demonstration basis, these approaches would not be as disruptive to the industry or as expensive to implement as they could be on a national basis.
Weaknesses of Competitive Bidding and Negotiation Demonstrations. There might be distortions from using data from a few demonstration sites to set national relative values. Laboratories might not be interested in bidding or negotiating if the outcome was to be used primarily for calculating national relative values, not their own payment rates. The negotiating capability of small laboratories, especially POLs, could bias the results, since POLs probably have had relatively little experience negotiating payment rates with health plans. Negotiations with the large number of POLs could also be an overwhelming administrative burden for the carrier.
Setting the Initial Conversion Factor. Once the RVS is established, the conversion factor that translates relative values into payment amounts must be set. The level of the conversion factor has implications for all stakeholders. From the Medicare program’s perspective, the lowest conversion factor that will not compromise beneficiary access both reflects prudent purchasing and contributes to the longer-term financial sustainability of the program. Because it is difficult to determine precisely what that conversion factor would be, there is a risk of setting payment rates too low, which would make serving Medicare beneficiaries less attractive and less feasible for the laboratories. The information on costs and prices collected while developing the relative value scale could provide some indication of the adequacy of payment rates and guidance in setting the conversion factor. At the same time, the parameters for setting the initial conversion factor are likely to be specified by the Congress. Under current budget and legislative requirements, Congress may mandate that implementation of a new fee schedule for clinical laboratory services not increase program spending. Maintaining budget neutrality—ensuring the same level of spending for laboratory services under the new system as would have occurred under the system it replaced—would impose constraints on the level of the initial conversion factor.
OTHER ISSUES IN PAYMENT SYSTEMS
The committee recognizes that both the mechanism used and the factors to be considered in keeping a laboratory fee schedule up to date will depend on the basis of payment chosen and the method used to determine payment levels.
How payment systems change over time depends on a number of factors. Retrospective payment systems, such as cost-based and charge-based systems, are self-updating: as costs or charges increase, so do payment rates. This is exactly the characteristic of prior payment systems that led payers, including Medicare, to adopt administratively set updates that adjust for expected or desirable changes over time while providing some control over spending growth. For some prospective payment systems, the method of determining payments also
automatically changes them over time. For example, MFN fee schedules for laboratory services would unilaterally be updated continuously for each laboratory, and payments under a competitive bidding methodology would be updated with each bidding cycle. In all of these cases, the payment changes that would naturally occur could be constrained by congressional action if the Congress were to choose to control spending. For example, the NLAs currently constrain the extent to which changes in charges will be reflected in Medicare payments. In a similar manner, Congress could set an upper bound for payments determined through competitive bidding. Prior to the Medicare Fee Schedule, updates in Medicare’s charge-based payments were constrained by the Medicare Economic Index.
Total program spending is determined by both payment levels and service volumes, and payers can consider the relation between the two when they set annual updates. These administratively determined updates, which are increasingly common among payers, take into account factors such as input price inflation and expected service volume and intensity changes due to changes in technology, the demographics of the enrolled population, and services covered, or shifts in services that are complements or substitutes for those in another service sector.
Medicare has adopted an approach to updating the conversion factor for the physician fee schedule that results in lower updates when volume growth exceeds expectations and higher updates when volume growth is less than expected. This method is designed to hold total spending to anticipated targets. The update mechanism includes laboratory service volume in the annual measure of volume used to set future physician payment updates. Thus, excessive growth in laboratory spending could negatively affect future physician fees. Laboratory services are provided on referral from physicians. For this reason, laboratory volume growth is, to a large extent, beyond the control of laboratories. Placing the responsibility for laboratory volume growth on physicians, by considering laboratory service volume in physician fee updates, is likely to be more effective than having a separate volume target for updating laboratory fees. Adjustments for input price growth for laboratory services, however, would still be necessary.
Keeping a payment system current depends on more than annual fee updates. It requires periodic review and revision of all key policy elements, because the way in which services are produced and used changes over time. Different payment methods will pose different administrative challenges related to the way they are structured. For example, the relative values in the Medicare Fee Schedule for physicians are now subject to review every five years by AMA’s RVS Update Committee (RUC) and HCFA.8 This provides an opportunity to
ensure that annual additions of new services to the RVS and other refinements to specific relative values do not lead to distortions in the relative value scale over time.
The committee has found no evidence to suggest that there are currently beneficiary access problems or other concerns that would need correcting through an adjustment to the laboratory fee schedule. If a new payment system were adopted, there could be a need for certain payment adjustments depending on how the method is structured. Examination of the appropriateness and implications of making adjustments for factors such as geographic cost differences, sole community hospital costs, access in rural areas, or STAT tests thus follows decisions about the payment method.
Because circumstances change over time, monitoring the impact of the payment system, and any modifications to it, on beneficiary access, laboratory participation, and program costs may identify areas in which future adjustments would be appropriate. The current system, based on 56 locally set fee schedules, conceptually includes a geographic adjustment, although geographic variations in payments are limited by the NLA. It also includes an adjustment for sole community hospitals. The appropriateness of these and other adjustments depends on whether the current system remains unchanged or the committee’s recommendations are implemented.
Some of the committee’s stated goals could be achieved through explicit adjustments made to payments, depending on the design of key elements of the payment system. For example, under a national fee schedule, payments could be adjusted for local differences in costs in order to promote provider equity. Such an adjustment would require decisions about the geographic units within which laboratory payments would be equal and about the most suitable price index or indices to use for adjusting payments. The Medicare Fee Schedule for physicians and the hospital inpatient PPS provide different models for approaching these decisions.9
Special adjusters could also explicitly promote the goal of patient access if instances of precarious or inappropriately limited access could be objectively identified. For example, Medicare physician payments are increased for care provided to beneficiaries who live in health professional shortage areas (HPSAs), in an effort to improve access to physician services for beneficiaries in these areas, Omnibus Budget Reconciliation Act (OBRA, 1989). Any evidence that labora-
the first five-year review process and pp. 50–51 for a discussion of other issues that may be relevant to periodic review.)
tory services are unavailable in particular communities could support some comparable adjustments to help promote the development of a local service supply.
Specific adjustments based on site of service have been proposed for laboratory payments. Underlying the use of adjusted payments in different settings is the notion that unit costs differ for different types or settings of laboratories and that the lack of such a site adjustment would make it difficult for settings with higher costs to continue to provide services. Opponents of such adjustments argue that there is no reason for Medicare to pay more for services in the higher-cost setting if they are available for less in other settings.
Under current policy, laboratories in qualified sole community hospitals are paid slightly more. The committee found no data with which to determine whether sole community hospitals, in fact, do have higher laboratory costs, the current adjustment is an appropriate amount, or there would be access problems for beneficiaries if the adjustment was eliminated. Likewise, the committee found neither cost data nor evidence on access to support the need for an adjustment for other types of laboratory settings, such as hospital-based laboratories or POLs.
There may be some service characteristics for which higher payments are justified. For example, some assert that hospital laboratories incur higher costs because of their need to have equipment available to provide services on an emergency, or STAT, basis.10 In this case, it may be appropriate to have a payment adjustment for STAT services, regardless of site, if providing quick turnaround is typically associated with higher costs and is viewed as clinically important in some circumstances.
The risk of a STAT adjuster is upcoding—creating an increase in the number of laboratory services performed on a STAT basis in order to bill at the higher amount if this would increase profits. The fact that the physician, not the laboratory, orders tests could limit opportunities for upcoding. Successful implementation of a STAT adjuster, however, would require carefully crafted descriptions of the circumstances under which STAT services are viewed as appropriate. Close monitoring of the medical necessity of these claims might be necessary.
The committee concludes that because of the administrative costs and burdens it would impose and the limited effect it would have on reducing excess testing, cost sharing for laboratory services is inconsistent with its goals for a laboratory payment system that ensures beneficiary access and maintains administrative simplicity.
Payers often use cost sharing to reduce overutilization of insured services. There is evidence of general overuse of health services in the United States, and laboratory services are not an exception to this pattern (Axt-Adam et al., 1993; Hindmarsh and Lyon, 1996; van Walraven and Naylor, 1998).
Currently there is no cost sharing associated with laboratory services used by Medicare beneficiaries, although there was under the charge-based payment system used in the early years of Medicare.11 Two different justifications have been offered for the present lack of beneficiary cost sharing for laboratory services.12 First, the referral nature of these services diminishes the potential for inappropriate patient overuse that is typically assumed in cases where services are fully paid by insurance. The lack of cost sharing for laboratory services presumably creates fewer additional program costs than it would for services that are fully covered by insurance and are sought directly by patients. Second, the relatively low levels of payment for many common laboratory services have led many observers to claim that the cost to laboratories of collecting the standard 20 percent copayment amount from beneficiaries would frequently outweigh the revenue generated. In fact, it is estimated that a 20 percent copayment would average less than $2.30 per test for the top 100 laboratory procedures, according to HCFA data.
Despite these arguments, the President’s originally submitted FY 2001 budget included a 20 percent copayment for clinical laboratory services meant to prevent overuse and reduce fraud (President’s FY 2001 Budget Proposal, 2000). However, HCFA actuaries did not expect the proposal to have a significant impact on utilization.13
Options to avoid the impracticalities of collecting copayments for relatively inexpensive services have been proposed but found lacking. One suggestion was the introduction of copayments only for services (individual or total ordered) with payments above some dollar threshold amount. Alternatively a copayment could be applied only for tests at sites where the beneficiary has direct contact with the laboratory and the provider could collect the copayment at the time of testing. These might include POLs, hospital-based laboratories, or independent laboratory specimen collection stations. Each of these options raises issues of administrative complexity, inequitable application of cost sharing, burden on the laboratories, and potential ineffectiveness in reducing unnecessary utilization.
Any expected revenue gains from new cost-sharing requirements therefore have to be weighed against the potential costs to laboratories of implementing them, the access barriers that they could present for some beneficiaries, and the potential equity issues raised at the beneficiary and provider levels.
Some cost-sharing alternatives would shift the collection burden from laboratories to other parties. For example, one way to reduce the administrative costs to laboratories of a cost-sharing requirement would be to have the referring physician collect it from the beneficiary along with the copayment for associated physician services. Physicians, however, are likely to resist taking on this role. It is unclear how they would know the appropriate amount to charge the beneficiary. Another way to shift some of the administrative burden associated with cost sharing away from laboratories would be through the introduction of a deductible that was administered by HCFA and the carriers.
Advantages of Beneficiary Cost Sharing. Beneficiary cost sharing for laboratory services will produce savings for Medicare. The administration’s budget estimates substantial savings to Medicare of $2.4 billion over the next five years, primarily from the reduction of Medicare spending from 100 percent to 80 percent of the fees. Under the alternative where the physician initially pays the copayment to the laboratory, introducing the beneficiary’s costs into the physician-beneficiary relationship might help make both more aware of the costs of laboratory services and less likely to use unnecessary services, restoring some of the educational role of cost sharing into the overall demand for insured services.
Disadvantages of Beneficiary Cost Sharing. Although the rationale given for the cost-sharing budget proposal is to prevent overuse of services, this is not the expectation. In fact, the expected savings would come mainly from shifting costs from Medicare to beneficiaries and laboratories, not from more appropriate use of laboratory services that would produce real savings for the health care system. Instituting any form of cost sharing could create access problems, at least for some beneficiaries (Solanki and Schauffler, 1999; Solanki et al., 2000). Many beneficiaries have some form of supplemental insurance that likely would cover the costs of their laboratory copayments. Those who do not have supplemental insurance, however, are among the most vulnerable elderly and would be the most adversely affected. Evidence suggests that low-income individuals react to cost sharing through decreased use of all services, not just those of limited health value (Brook et al., 1984). These same access problems are associated with all cost-sharing requirements in Medicare.
Cost sharing would also introduce new administrative burdens to the system, whether borne by laboratories, referring physicians, beneficiaries, or Medicare. It could also increase costs to laboratories or referring physicians due to bad debt and to beneficiaries who currently do not pay cost sharing for laboratory services. Having the physician collect the copayment would add significant complexity to the system. The expectation that beneficiaries would become
more aware of fraudulent practices because of their copayments seems unrealistic, since they generally are not familiar with the variety of tests available, the appropriate use of various tests, or even the laboratory to which the physician sends their specimen. Significant fraudulent practices by laboratories are more likely to be detected by whistle-blowers than by beneficiaries.
The copayment alternatives based on a dollar threshold or applied only to providers with direct contact with beneficiaries would raise issues of equity for beneficiaries. Beneficiaries who received testing services from independent laboratories that were exempted from collecting copayments would have a financial advantage over those who had to pay on the spot because they had direct contact with the laboratories that conducted their tests. Applying a dollar threshold for tests before the laboratory copayment could be charged would create an incentive to game the system. Beneficiaries who had multiple services simultaneously could be subjected to cost sharing, while those who were able to receive the same tests on separate occasions might have individual charges that stayed below the cost-sharing threshold.
MAKING IMPROVEMENTS A REALITY
Designing Payment Policies
Any change in payment methods can cause major concerns since the economic well-being of so many players is involved, but the apprehension and disruption accompanying such changes can be minimized. During HCFA’s implementation of the Medicare physician fee schedule, observers noted that there were a number of aspects of the design and implementation phase that contributed to an apparently smooth process (Oliver, 1993; Smith, 1992):
A single, comprehensive proposal for reform was able to provide some elements that were attractive to each constituency, along with less popular policies, by combining several key policy changes into one package.
A consensus approach, open meetings, constructive engagement with all interested stakeholders, and well-regarded staff analyses contributed to an acceptable reform proposal.
These observations suggest a number of considerations for designing and implementing new payment policies for laboratory services.
Anticipate Effects of New Policies. Before a new approach to paying for laboratory services under fee-for-service Medicare is proposed for legislation and implementation, its potential implications for beneficiaries, providers, and the Medicare program should be carefully assessed. Understanding the potential effects of policy change allows the development of mechanisms to identify important problems that arise as new payments are introduced and to respond promptly
to them. Clear, accurate information about likely effects of new rates may also help key actors; even those who may not benefit from them, accept them.
In some alternatives previously discussed, the short-run dislocations might be minimal for laboratory providers. To the extent that laboratories provide a broad array of services at volumes comparable to overall Medicare utilization rates, a decision by Congress to mandate a budget-neutral fee-for-service payment change should not have a significant short-term effect on a laboratory’s Medicare revenues unless the laboratory is in a geographic area that would see many of its local fees increased. Also, an increase in some fees and a decrease in others, which could happen with a move to resource-based relative values, would balance out in laboratories with average Medicare utilization of services. Nevertheless, any changes are likely to have some impact on the operations of individual laboratories.
Plan to Minimize Dislocations. Strategies for minimizing potential short-run dislocations caused by new policies could be worthwhile depending on the magnitude of anticipated changes. Payment changes could have more serious effects on individual laboratories or on segments of the industry. If their service mixes were different from the overall Medicare distribution, then some laboratories could realize dramatic revenue increases, while others would experience losses. For example, only laboratories certified as high complexity by the Clinical Laboratory Improvement Amendments (CLIA) can do the full range of tests, and some of those laboratories specialize in only a selected menu. Most POLs do only a very limited range of tests. If information about laboratory-specific service mixes suggests that laboratories with certain service mixes would be disproportionately affected by payment changes, it might be appropriate to phase in new payment rates over several years. This would give laboratories a chance to respond to payment changes through changes in staffing, purchasing, or other appropriate strategies. Identifying ways to reduce the shocks associated with a new payment system becomes more difficult the more the proposed changes differ from the current system. For example, in a payment system based on selective contracting with a limited number of providers, it would be impossible to eliminate the dramatic effect on the composition of the industry.
Plan to Monitor Continuing Effects of Changes. Besides finding ways to help ensure smooth implementation of a new system, plans should be made to monitor the effect of new rates on beneficiary access. The clear difficulty here is that changes in service use from current volumes are not necessarily evidence of problems since there is no way to confirm that the present distribution of laboratory services is optimal. As new policies are crafted and proposed, stakeholders will likely present analyses of how payments under the proposed policies will differ in allegedly harmful ways from the current distribution of Medicare laboratory payments. It is difficult to weigh such evidence, however, since it does not clarify the underlying issue of beneficiaries’ ability to get needed services on a timely basis under either the current or any proposed sys-
tern. Knowing that decreased payment for a particular service may reduce its use compared to current rates is not equivalent to knowing that the decreased payment has important implications for beneficiaries’ health. Evidence of changes in service use must be combined with other studies that link utilization rates to clinical appropriateness.
Plan for Continuing Operations. Plans should be developed for ongoing administration of the new system. Issues that are important at the time of initial implementation may be quite different from those that require attention once the new system is in place. The methods used to develop initial payment rates may not be the same ones that should be used to update or revise them. Two predictable issues for smooth operation of a new system are its ability (1) to incorporate new technologies and to (2) adjust to changes over time in the costs of providing existing services. Before new payment policies are implemented, attention must be paid to how these longer-run needs will be addressed. Otherwise, the new system would be changed constantly but unpredictably, undermining stakeholders’ confidence in it and possibly diminishing the chance that it works as desired or expected.
Implementing New Payment Policies
Implementation of any payment alternatives discussed would depend on many details and design questions to be answered during the planning and development phase for a new methodology. The committee’s intent in assessing various payment options and making its recommendations is that any changes made be consistent with the goals of administrative simplicity, efficiency, and transparency. The size and scope of any proposed solutions ought to be reasonably related to the size of the problems.
To discuss legislative amendments, administrative steps, paperwork requirements, and costs for implementation of a payment methodology in any detail, one would have to know very specific aspects of each of the method’s elements. Getting data for such a discussion would be sufficiently expensive that it should not be attempted for hypothetical options or general recommendations. More decisions would have to be made and details clarified in order to make realistic assumptions that are needed for cost projections. Hence, the committee decided to discuss implementation issues at a general level and compare options to each other and to the current system. Keep in mind, however, that any change is likely to entail at least some changeover costs, dislocations, and inconveniences, even with careful planning. The alternative of totally maintaining the status quo, with no changes to the current payment methodology, also has costs both for today and, even more, for the future (see Chapter 5 for an assessment of the status quo).
Legislative Changes. Essentially any alternative for payment that would change or eliminate the existing formula, which takes the median fee of the 56
fee schedules and reduces it by 26 percent to set the NLA, would require new legislation. Thus, a single national fee schedule or any of the alternatives discussed previously would require congressional action. Additional legislation would be required to extend the current competitive bidding demonstration authority and to permit the use of demonstrations for the purposes of developing data with which to set relative values.
Administrative Steps. Under any of the payment options raised earlier in this chapter, administrative changes would be necessary. The committee’s goal would be to take advantage of this opportunity to simplify, streamline, and open the administrative procedures whenever possible. Some methods, such as market-based alternatives, would require a very different approach from that required by use of the NLA. The extent of the administrative changes under any approach would depend on how elaborate the procedures were for setting the initial payments and later sets of fees or relative values. A system that began with the NLA would create the opportunity to simplify some administrative procedures by eliminating certain functions related to the 56 carrier fee schedules. Alternatives for refining relative values, for instance, could require policy decisions about research contracts and surveys, how to refine market-based demonstrations, how to structure stakeholders’ meetings, and how to solicit public comments. Additional administrative procedures could be required for updating the fee schedule under some alternatives. Incorporating new technologies under certain options could mean reduced administrative activity by the carriers and more work done centrally by HCFA, resulting in greater efficiency.
A change of payment methodology could provide an opportunity for HCFA to address related administrative issues needing attention, such as the clarification of instructions to carriers concerning claims denials, that are not necessarily related to the particular alternative selected. At a time of significant policy changes, it would be particularly important to have all of the stakeholders understand the proposed changes and participate in their refinement, to the extent possible.
Paperwork. Under the fee-for-service alternatives considered in this chapter, there is no reason to expect that the claim form and accompanying data currently required of physicians and laboratories to bill for laboratory services would be changed substantially. New paperwork burdens could result from the studies suggested for planning a new system and monitoring its impact. The extent of the burden would depend on the type of data collection required, study design, survey sample size, and other methodological requirements. As discussed earlier in this chapter, a survey of laboratories to determine the costs per test could be quite complex and demanding of the respondents. A competitive bidding or negotiation demonstration would be relatively less intrusive, but demanding of the laboratories as they calculated their risks and bids. Other research, evaluation, and administrative requirements could create paperwork burdens and costs associated with conducting a demonstration. The paperwork burden of the consensus approach would depend, in part, on the extent of the
information incorporated into the process and the method chosen to identify participants in the consensus process.
Financial Costs of Introduction and Use. The initial financial costs of the NLA approach would be minimal because no new data collection would be necessary for calculating the basic fee schedule. Software changes would likely be relatively minor since this would be more a matter of dropping the 56 local fee schedules than of creating totally new systems. The MFN option could require extensive software changes to prepare carriers to pay the laboratories based on their individual charges. The cost of a competitive bidding demonstration would depend on the number of geographic locations and laboratories included. Determinations of what data to use for payment updates and any geographic adjustments, how they should be calculated, and to what costs they should be applied, could draw heavily on existing information. HCFA has had extensive experience in creating such updates and adjustments for other types of providers. The costs of implementing adjustments would depend on how much precision was needed and how much original data collection and modeling would be necessary.
Significant expenses could be incurred for initial calculations of the costs or resource base for alternatives using relative values. There would be costs for calculating certain types of adjustments as well. Depending on how initial relative values are calculated, sufficient data could be produced by the process to contribute to the determination of whether some adjustments should be made. The cost of acquiring the necessary data depends largely on several key design decisions, such as the data collection approach (which affects both data costs and validity) and the type of measure (mean, median, or efficient model) upon which relative values will be based. Ultimately, the costs of developing an RVS would depend on the process chosen for establishing relative values and the amounts and type of data it requires.
The choice of a payment alternative does not have to be “either/or”; it could include various options. Incurring the major start-up costs and longer-term redistributive costs of some of the more complex alternatives is presumably reasonable if concerns about inefficiencies, inequities, or access problems with the current system merit the implementation costs and potential disruption associated with them. In the absence of evidence of such problems, it is likely that the industry has accommodated whatever payment distortions there are relative to costs under current policy. In this case, the less expensive, more pragmatic approach of a national fee schedule based on the NLAs may be justifiable as a first step toward a more coherent Medicare laboratory payment policy with minimal short-run implications for the industry or beneficiaries. However, the NLAs could present difficulties if used indefinitely since it is unclear how these fees relate to costs in aggregate or to costs of specific tests. A resource-based fee schedule, set at a reasonable level, would eliminate incentives for providers to limit access to
Medicare beneficiaries. Once the NLA fee schedule is in place, it could be gradually improved so that the fees more closely reflected the relative resource use of each test. This refinement could occur through a consensus process informed by data gathered through the types of approaches described earlier.
There would inevitably be costs related to any policy reform and changes needed in the administration of the program. There would also be costs and potential access problems associated with making no changes, and maintaining the status quo indefinitely into the future. Given current concerns about distortions in payments and constraints on efficiently updating the system, as well as demands on the current payment methodology that will result from new laboratory technologies under development, it is important that HCFA and Congress consider the committee’s recommendations in the following chapter. Changes cannot be effected well if done hastily in a crisis situation, so it would be advantageous to begin planning, collecting the needed data, and analyzing them so that a properly designed payment methodology can be developed.
Axt-Adam, P., J.C.van der Wouden, and E.van der Does. 1993. Influencing behavior of physicians ordering laboratory tests: A literature study. Med Care 31, No. 9:784– 794.
Brook, R.H., J.E.Ware, Jr., W.H.Rogers, et. al. 1984. The Effect of Coinsurance on the Health of Adults: Results from the RAND Health Insurance Experiment. R-3055-HHS. Santa Monica, CA: RAND.
Committee on Ways and Means, U.S. House of Representatives. 1998. 1998 Green Book: Overview of Entitlement Programs. Washington, DC: Government Printing Office.
Congressional Budget Office (CBO). 2000. Budget Options. Web page, accessed April 2000. Available at http://www.cbo.gov/.
DeParle, N.M., and R.A.Berenson. 2000. The need for demonstrations to test new ideas. Health Affairs 19, No. 5:57–59.
Hindmarsh, J.T., and A.W.Lyon. 1996. Strategies to promote rational clinical chemistry test utilization [see comments]. Clin Biochem 29, No. 4:291–299. Comment in Clin Biochem 1997; 30, No. 4: 361, 363.
Hoerger, T.J., and T.M.Waters. 1993. Competitive bidding for Medicare services. Med Care 31, No. 10:879–897.
Hoerger, T.J., J.L.Eggleston, E.Basker, and R.C.Lindrooth. 1997. Background Report on the Clinical Laboratory Industry, Final Report. Research Triangle Park, NC: Research Triangle Institute.
McAfee, R.P., and J.McMillan. 1987. Auctions and bidding. Journal of Economic Literature 25, No. 2:699–738.
Mennemeyer, S. 1989. Competitive bidding for Medicare outpatient laboratory tests. Advances in Health Economics and Health Services Research 10:313–333.
Mennemeyer, S.T., J.B.Christianson, R.Englbrecht-Wiggins, and P.Chemawat 1986. Competitive Bidding for Durable Medical Equipment and Clinical Laboratory Services: A Review of Related Literature. Prepared under contract No. HCFA 500– 85–0052. Cambridge, MA: ABT Associates, Inc.
Oliver, T. 1993. Analysis, advice, and congressional leadership: The Physician Payment Review Commission and the politics of Medicare. Journal of Health Politics, Policy and Law 18, No. 1:114–174.
Physician Payment Review Commission (PPRC). 1989. Annual Report to Congress, 1989. Washington, D.C.: PPRC.
PPRC. 1995. Annual Report to Congress 1995. Washington, D.C.: PPRC.
The President’s FY 2001 Budget Proposal. Web page, accessed April 2000. Available at http://whitehouse.gov/WH/New/00Budget.
Raslavicus, PA. January 20, 2000. Statement to the IOM Committee on Medicare Payment Methodology for Clinical Laboratory Services. Washington, DC.
Root, C. January 20, 2000. Testimony before the IOM Committee on Medicare Payment Methodology for Clinical Laboratory Services. Washington, DC.
Sing, M., R.Brown, and S.C.Hill. 1998. The consequences of paying Medicare managed care plans their costs. Inquiry 35, No. 2:210–222.
Smith, D.G. 1992. Paying for Medicare: The Politics of Reform. New York: de Gruyter.
Solanki, G., and H.H.Schauffler. 1999. Cost-sharing and the utilization of clinical preventive services. Am J Prev Med 17, No. 2:127–133.
Solanki, G., H.H.Schauffler, and L.S.Miller. 2000. The direct and indirect effects of cost-sharing on the use of preventive services. Health Serv Res 34, No. 6:1331– 1350.
van Walraven, C., and C.D.Naylor. 1998. Do we know what inappropriate laboratory utilization is? A systematic review of laboratory clinical audits [see comments]. JAMA 280, No. 6:550–558. Comment in JAMA 1998; 280, No. 6:565–566.