|
Getting The Price Right: Medicare Payment Rates For Cardiovascular Services
Specialized, physician-owned cardiac hospitals have grown rapidly. Physicians have also expanded their capability to provide cardiovascular diagnostic services in their offices. In this paper we consider evidence of errors in Medicare’s prices for hospital care and physician services and discuss ways to improve the accuracy of those prices. We find that recent proposals to change the inpatient prospective payment system would help dampen hospitals’ financial incentives to favor some kinds of patients and related investments. For the physician fee schedule, we suggest that the Centers for Medicare and Medicaid Services (CMS) review the accuracy of prices for high-growth diagnostic services.
A GOAL OF MEDICARE PAYMENT POLICY is to get good value for what the program spends. Achieving this goal means maintaining beneficiaries’ access to high-quality services while ensuring wise use of program resources. One way to increase value in Medicare is to improve the accuracy of the prices paid for individual services. Medicare uses administered prices; if these prices are inaccurate, there are consequences for the program. If services are underpriced, beneficiaries could have problems with access to care. If services are overpriced, providers might furnish more services than beneficiaries need. In addition to the wasteful spending that accompanies overuse of services, there is the potential for harm through medical errors or iatrogenic disease.1 Inaccuracies in Medicare prices can also have consequences beyond the program, including capital investment and workforce decisions. For instance, physicians might make specialty choices based on perceived underpayment for some services relative to others.2 If private payers model their payment systems on Medicare’s, errors in Medicare pricing can lead to errors in those payers’ pricing.3 Heart disease has long been the leading cause of death among people age sixty-five and older.4 Treatments for cardiovascular conditions also have undergone much technological innovation.5 Medicare’s prices might not be accurate now even if they were in the past. In this paper we discuss ways to improve the accuracy of Medicare’s payments for hospital inpatient care and physician services, focusing on cardiovascular services. Available evidence suggests that Medicare beneficiaries have no widespread access problems.6 However, there are signs that some of Medicare’s prices might not be accurate. For instance, questions have arisen about the rapid growth of small, physician-owned hospitals specializing in cardiac services, which are thought to have above-average profitability. The Medicare Payment Advisory Commission (MedPAC) found that these hospitals had a profitable mix of patients for two reasons: Cardiac surgery cases tended to have higher payment-to-cost ratios than the average Medicare case, and physician-owned specialty hospitals tended to serve lower-severity patients than their community hospital competitors.7 In this paper we show that cardiac surgery cases in 2002 were more profitable than the average case. We also discuss how the relative profitability of cardiac surgery changed under the policies the CMS adopted for fiscal year 2007 and how profitability might be affected by further policy changes that are under consideration for FY 2008. Growth has also occurred in the cardiovascular services furnished in physician offices. Although this growth is no doubt driven in part by technological advances, its rapidity raises questions about payment accuracy under Medicare’s physician fee schedule. Our approach to evaluating prices for physician services, however, is different from the one we take for inpatient hospital care. In the absence of cost data similar to those for hospitals, we cannot observe physicians’ resource costs directly. Instead, we must infer the accuracy of their payments. Inferences about payment accuracy can help prompt the CMS to undertake more detailed assessments. By law, the agency reviews prices in the physician fee schedule once every five years during a process known as the five-year review. For the review, the CMS relies heavily on physician specialty societies to identify specific services whose prices might be inaccurate and to collect data, including surveys of physicians about their resource costs. During the three five-year reviews that have occurred since the fee schedule was implemented in 1992, the process has led to many more price increases than decreases. To strengthen the process, MedPAC has recommended that the CMS establish a standing panel of experts with knowledge in health economics and physician payment in addition to clinical expertise. This panel would recommend services for assessment of payment accuracy. Our findings suggest that the panel should consider cardiovascular diagnostic services experiencing rapid volume growth. The CMS should also revisit assumptions about the cost of operating costly medical equipment.
From 1996 to 2004 the number of cardiac surgeries per thousand Medicare beneficiaries grew 22 percent.8 Growth is expected to continue. In a 2006 Banc of America survey of nonprofit hospitals, 30 percent of executives stated that they planned to add cardiac surgery, and 15 percent, to add cardiac catheterization laboratories.9 No other new service was mentioned as often. Community hospital executives have repeatedly asserted that their hospitals make money on cardiac and other surgical procedures and that they use those profits to subsidize medical admissions. If this claim were true, we would expect to find that (1) Medicare patients in diagnosis-related groups (DRGs) for cardiac surgery are more profitable than patients in other DRGs; and (2) hospitals that specialize in cardiac surgery are more profitable than the average hospital. We used cost-accounting data to test whether cardiac surgery cases are more profitable than the average Medicare admission. We examined hospitals’ financial statements to test whether specialty heart hospitals tend to be more profitable than general community hospitals. Assessing accuracy of payments. To assess whether community hospitals have an incentive to favor cardiac surgery, we evaluated the relative profitability of cases in each DRG. To measure relative profitability, we estimated costs and payments for hospitals’ Medicare claims. Estimates were derived from detailed claim charges by type of service and each hospital’s departmental cost-to-charge ratios. Payments were derived from MedPAC’s 2002 inpatient payment model. We also modeled claims payments for several payment alternatives. We used these estimates to calculate relative profitability ratios for each DRG in use in 2002 and for each all-patient refined diagnosis-related group (APR-DRG, version 15). APR-DRGs do a better job than DRGs in distinguishing patients with different severity levels.
Using these data, we found that the relative payment-to-cost ratios under 2002 payment policies were much higher for Medicare cardiac surgery patients than for both cardiac medical patients and the average Medicare discharge. On average, cases in twenty-two of the twenty-four cardiac surgery DRGs were relatively more profitable than the average discharge and accounted for 98.6 percent of all Medicare cardiac surgery patients. In contrast, nineteen of the twenty-four cardiac medical DRGs were relatively less profitable than the average discharge and accounted for 91.2 percent of all Medicare patients admitted for medical treatment of a cardiovascular disorder. Illustrative relative payment-to-cost ratios for four cardiovascular DRGs are shown in Exhibit 1
Severity of illness. By looking at the four severity classes within each APR-DRG, we could also see the impact that differences in severity of illness among patients have on profitability (Exhibit 2  ). Under policies in effect in FY 2002, our estimates suggest that profitability varied widely (and in the opposite direction) with increases in severity of illness. For example, low-severity coronary bypass patients (Level 1) had a relative payment-to-cost ratio of 1.47, or almost 50 percent more profitability than the average patient. In contrast, the most severely ill bypass patients were only 80 percent as profitable as the average patient.
Growth in physician-owned heart hospitals. Because cardiac surgery DRGs tended to be more profitable than other cardiac DRGs, hospitals with a focus on cardiac surgery are expected to have relatively high profitability and attract private capital. The number of heart hospitals with cardiologist investors more than doubled from twelve in 2002 to twenty-five in 2004, which suggests that many cardiologists believed that heart hospitals could provide a reasonable return on investment.10
During 2004, the median specialty heart hospital was able to generate more revenue per dollar of assets and more pre-interest income per dollar of revenue than the average community hospital (Exhibit 3
Higher returns on assets can be attributable to having either a mix of patients that have favorable payment rates or greater efficiency for the same mix of patients. On average, physician-owned cardiac hospitals and community hospitals have similar costs per discharge, after factors such as patients’ severity of illness and interest expenses are adjusted for.11 Therefore, it appears that the relative success of the average cardiac hospital is primarily attributable to the type of services it provides and not the efficiency with which it provides them. As cardiac surgery payment rates are reduced, only the most efficient cardiac hospitals will still exhibit above-average profitability.
MedPAC recommendations. To improve the match between Medicare’s payment rates and hospitals’ costs of furnishing care, MedPAC recommended in 2005 that the CMS and Congress make four changes in payment policy under the IPPS: (1) refine the current DRGs to more fully capture differences in severity of illness among patients; (2) base DRG relative weights on the estimated cost of providing care rather than on charges; (3) base the weights on the national average of hospitals’ relative values in each DRG; and (4) adjust the weights to account for differences in the prevalence of high-cost outlier cases in each DRG.
If the CMS were to adopt all of MedPAC’s recommendations, we estimate that the relative payment-to-cost ratios in each APR-DRG category would be much closer to 1.0 than it was under the 2002 policies (Exhibit 2 CMS policy changes for physician-owned heart hospitals. In its final rule for the inpatient prospective payment system (IPPS) for FY 2007, the CMS adopted one of MedPAC’s four recommended policy changes (cost-based weights) and stated that it will formally evaluate the other three during the coming year. Cost-based weights. From 1986 through 2006, the CMS based its payment weights on hospital charges. Starting in 2007, it will use cost-based weights, where charges will be converted to costs using hospitals’ cost-to-charge ratios for thirteen individual departments. The effect of the change will be to increase payments for admissions that primarily use routine services (for example, mental health) and decrease payments for admissions with large ancillary charges (for example, surgical admissions). In general, hospitals that have a balance of medical and surgical admissions will experience little change in overall Medicare payments. However, hospitals that focus on surgical admissions might see a decline in payments. For example, payments to cardiac specialty hospitals will be approximately 6 percent lower over a three-year period than they would have been under previous policies.13 The payment change is to be phased in over three years, resulting in 2 percent lower payments to cardiac specialty hospitals in 2007. By implementing cost-based weights, the CMS will take one step toward reducing investors’ incentives to form cardiac hospitals and, more importantly, decrease full-service hospitals’ incentives to allocate their limited capital toward expanding surgical rather than medical services. Cost-based weights, however, are only one step toward removing imbalances in the payment system.
Improved severity adjustment.
The expected profitability of services varies widely based on patient severity (Exhibit 2 Hospital-Specific Relative Values (HSRV). In the current DRG weight calculation method, differences in costs among hospitals that are not accounted for by the payment adjustments (for example, teaching status) affect DRG weights. As a result, weights are overstated for DRGs in which cases are disproportionately treated in hospitals that have high standardized charges or costs. Similarly, weights understate costs where cases are disproportionately treated in hospitals that have low standardized charges or costs. Instead of paying more for types of discharges that occur in more costly hospitals, MedPAC has suggested first computing the relative costliness of DRGs within each hospital and then taking the weighted average of those cost ratios at different hospitals to determine payment weights. Calculating relative values for all Medicare cases in each specific hospital (the HSRV method) cancels out any differences in costs (or charges) among hospitals. This prevents hospital-specific differences in costs from affecting the DRG relative weights. The HSRV method also has the advantage that the CMS would not have to standardize costs before calculating the weights. Thus, it would not be dependent on the accuracy of the wage index or other payment adjustments when correcting for the associated differences in input prices or outputs among hospitals. The CMS is evaluating the use of the HSRV method for implementation in FY 2008. Outlier financing. Even with cost-based weights and severity adjustment, some systematic differences in profitability would remain among severity-adjusted DRGs because of the current system of financing outlier payments. Outlier payments primarily go to high-cost cases, but the funds for those payments come from a uniform reduction of 5.1 percent in payment rates for all Medicare discharges. The consequences for profitability are clear. The costs for extremely high-cost cases that trigger outlier payments are included in the calculation of the DRG weights, increasing the weights for DRGs that have lots of outlier cases. However, hospitals pay a flat premium of 5.1 percent for outlier insurance in each DRG. Thus, the premium doesn’t match the prevalence of outlier cases and the distribution of outlier insurance benefits across DRGs. As a result, severity-adjusted DRGs that tend to receive a disproportionate share of outlier payments also tend to have higher payment-to-cost ratios. Conversely, severity-adjusted DRGs that tend to have few outlier payments tend to have lower payment-to-cost ratios. In its final rule for FY 2007, the CMS stated that it lacked statutory authority to adjust the DRG weights for differences in the prevalence of outlier payments. However, the CMS stated that it would evaluate this proposal during 2007. By adopting cost-based weights in 2007 and MedPAC’s other proposed policy changes in 2008, the CMS could reduce the substantial disparities in profitability among types of patients that have existed under IPPS payment policies. This would help dampen hospitals’ financial incentives to favor some kinds of patients and related investments over others.
We now consider the accuracy of Medicare payments for physician cardiovascular services and MedPAC’s recommendation that the CMS establish an expert panel for advice on the five-year review. About 6,700 services are priced in the physician fee schedule and eligible for the five-year review; however, only selected services are considered.14 Simply identifying which services to review is a critical first step for the panel. To identify services, MedPAC has recommended that the CMS conduct data analyses in consultation with the expert panel. These analyses would identify services that have experienced substantial changes in volume, site of service, and other factors that may indicate changes in resource requirements. The panel’s analyses would also be useful later in the review process when the CMS seeks advice from a group of physicians known as the Relative Value Scale Update Committee (RUC).15 Data analyses could inform the RUC’s deliberations.
Among cardiovascular services, imaging—especially nuclear medicine and echography of the heart—grew fastest from 1999 to 2004, at average annual rates per Medicare beneficiary of 16.1 percent and 10.5 percent, respectively (Exhibit 4
Physicians have a choice about where they furnish services, and increasingly they have chosen to furnish high-volume cardiovascular services in their offices. As we discuss below, doing so often requires the purchase of expensive equipment. It seems likely that payments for the services are at least adequate. Some have argued that part of the growth in the volume of office services represents a shift from other settings, such as outpatient departments. As we show below, the volume of services is growing most rapidly in physicians’ offices, but the data do not show a drop in volume in other settings. Because of differences in billing between imaging and other services, we discuss each separately. Cardiovascular imaging. As with imaging services generally, physicians can bill for one or both of two components of a cardiovascular imaging service. The technical component covers use of equipment, supplies, and nonphysician clinical staff. If Medicare pays for this component under the physician fee schedule, the study was usually performed in a physician’s office. The professional component consists of interpreting results and writing a report. When billed separately, this component is usually for a study performed in a facility.
For nuclear medicine and echography of the heart and the carotid arteries, about three-quarters of total volume is in billing for the global service—professional and technical components—performed in a physician’s office (Exhibit 5
Cardiovascular stress tests. The two components are not distinguished in billing for stress tests. Instead, payment rates vary by setting, with higher payments for nonfacility settings. The pattern of where stress tests are furnished is the same as for imaging. About 83 percent of volume is in physician’s offices, and volume growth is much higher there than in facility settings (12.6 percent versus 1.1 percent). Changes in clinical indications. Before leaving the topic of volume growth, we note that the volume of some cardiovascular services—coronary artery bypass graft (CABG), aneurysm repair, and thromboendarterectomy—is falling. We do not interpret these decreases, however, as a sign that payments are inadequate. In the case of CABG, for instance, the decrease likely represents growing substitution of less invasive services, such as coronary angioplasty. Review of payments for physician work. Physician work includes the time, mental effort, technical skill, psychological stress, and risk of performing a service. For a given service, the amount of physician work required could change because of factors such as learning by doing, improvements in technology, personnel substitution, reengineering, changes in patient severity, and documentation requirements.17 In particular, when volume grows rapidly, we expect requirements for physician work to fall as proficiency improves through learning by doing and other factors. Review of the fee schedule’s RVUs every five years permits the CMS to account for such changes. It appears that review of work RVUs, however, does not do a good job of identifying services that may be overvalued.
We see this problem in cardiovascular services with high rates of volume growth (Exhibit 6
Pricing of equipment. Practice-expense RVUs are another potential source of inaccuracy in the payments for cardiovascular services furnished by physicians. Here, we focus on questions about the way the CMS accounts for the price of imaging equipment in practice-expense RVUs.18
Many cardiovascular services that physicians are choosing to provide in their offices require expensive equipment. For example, nuclear medicine and echocardiography together account for 43 percent of the volume of cardiovascular services (Exhibit 4 The way in which these equipment costs are factored into determining practice-expense RVUs gives rise to two concerns. First, the CMS assumes that procedure-specific equipment, such as the gamma camera, is used twenty-four hours per week.20 This assumption was first adopted in 1997 based on comments from specialty groups. Second, the CMS assumes that the interest rate paid to purchase such equipment is 11 percent; this also dates back to 1997. With a sensitivity analysis, MedPAC has shown that per service costs for equipment are sensitive to assumptions about equipment usage and interest rates. For instance, cost per service drops by 40 percent if the equipment usage assumption is increased from 50 percent to 75 percent and the interest rate falls to 6 percent.21 Whether a 75 percent usage rate is appropriate for cardiovascular services (it might be too high or too low) is unknown. The point, however, is that estimated equipment costs for these services are sensitive to these assumptions, and we can expect them to affect the accuracy of the practice-expense RVUs. The CMS could update the usage assumption with survey data.22 Or it could adopt a standard that represents efficient use of equipment. On the interest rate assumption, it could use more recent estimates from the Federal Reserve Board.
Medicare fills a critical role in our society. Along with that role, however, comes a responsibility to ensure that resources are used wisely. Medicare spending already has been growing faster than growth in the national economy. The new out-patient prescription drug program and enrollment of the baby-boom generation will soon lead to even higher spending growth. Improving the accuracy of Medicare’s prices is one response to these challenges. In the case of the IPPS, the adoption of cost-based weights in October 2006 is one step toward removing distortions in the payment system that can result in mis-allocation of hospital capital. The payment system can be further improved by adopting three policies (a more refined severity adjuster, hospital-specific relative values to set weights, and a revised method for financing hospital outlier payments) recommended by MedPAC that the CMS stated it will evaluate between now and when its proposed rule is issued in April 2007. Physician services present a different set of issues because of the difficulty of measuring resource costs. MedPAC has recommended that the CMS establish an expert panel that would identify services that might be mispriced. Data analyses of changes in the volume of services and the site of service, as we have shown, could inform the panel’s work. The CMS could also update its assumptions about the cost of operating medical equipment. Improving the accuracy of Medicare payment rates is a necessary but not sufficient step toward modernizing Medicare. Medicare and other payers face many problems, including a lack of coordination of care among providers, a lack of information and the tools to use it, and poorly targeted diffusion of technology. It will take a public-private partnership with concerted efforts in pay-for-performance, investment in information technology, and other initiatives to overcome these problems.
Kevin Hayes (khayes{at}medpac.gov) is a principal policy analyst, Julian Pettengill is a staff consultant, and Jeffrey Stensland is a senior analyst with the Medicare Payment Advisory Commission (MedPAC) in Washington, D.C. The views expressed here are those of the authors and do not necessarily reflect those of the Medicare Payment Advisory Commission, which funded the research. The authors appreciate comments from Glenn Hackbarth and three anonymous reviewers. They also thank Mark Miller for his leadership and input into the research that forms the basis for this paper.
This article has been cited by other articles:
Comments:Read all Comments
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Incentives To Expand Surgical...
More Accurate Inpatient Payments
