|
Do Physician-Owned Cardiac Hospitals Increase Utilization?
This paper looks at whether physicians’ investment in heart hospitals during 1997–2001 was followed by an increase in the number of relatively profitable cardiac surgeries paid for by Medicare or in a shift toward operating on healthier (more profitable) Medicare patients. Although markets with physician-owned hospitals had slightly above-average growth rates in profitable cardiac surgeries during this period, the magnitude of the increase was small and statistically significant only for bypass surgery. There was no increase in the proportion of surgeries performed on healthier patients. These findings contrast with earlier studies of less-invasive services such as diagnostic imaging.
IS PHYSICIANS' OWNERSHIP OF A HEALTH CARE ENTITY associated with greater use of services? This question, which has been debated for many years, has become even more relevant with the emergence of physician-owned hospitals that specialize in cardiac, surgical, or orthopedic services. According to the U.S. Government Accountability Office (GAO), the number of specialty hospitals grew from twenty-nine in 1990 to ninety-two in 2003, and an additional twenty-six hospitals are under development.1 Although these hospitals accounted for only 1 percent of Medicare inpatient spending in 2000, they have attracted attention from general hospitals, physicians, researchers, and policymakers because they compete with general hospitals for highly profitable patients and because they give physicians a financial stake in hospital profitability.2 This paper reviews the debate between supporters and opponents of physician-owned specialty hospitals, examines the literature on physician-induced demand, and describes our study on the impact of physician-ownership on cardiac surgery volume.
Supporters of physician-owned specialty hospitals argue that such facilities develop focused processes of care that are tailored to a limited set of conditions, which allows them to provide higher-quality care at a lower cost.3 Proponents also contend that such facilities increase competition for hospital services and thus spur innovation in the marketplace. Supporters claim that ownership by physicians is critical because it gives physicians the freedom and motivation to design the hospital layout and operations in a way that maximizes efficiency and quality.4 Opponents of these hospitals argue that they select the most profitable patients: those who are healthiest (and hence less costly to treat), who have desirable insurance coverage, and who require profitable services (such as cardiac surgery). Critics claim that this behavior makes it more difficult for general hospitals to cross-subsidize less profitable services (such as burn units and trauma care) and care for poor and uninsured patients.5 Opponents also contend that physician-ownership creates a potential conflict of interest between physicians’ financial incentives and their patients’ clinical needs. This conflict of interest might influence which hospital a patient is referred to or might lead physicians to refer patients for unnecessary services, thus increasing overall volume. Because of differences in views among members of Congress, the Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003 placed an eighteen-month moratorium on the development of new physician-owned specialty hospitals and restricted the expansion of existing facilities.6 That moratorium expired 8 June 2005.
There are two ways in which the presence of physician-owned specialty hospitals could lead to more services in a market. First, if such hospitals increase overall hospital capacity, this might lead to additional admissions and procedures. A study of regional variations in Medicare spending by Dartmouth researchers found that high-spending areas have more hospital beds and physicians than other areas.7 These additional resources are associated with more so-called discretionary services, such as inpatient days, physician visits, and diagnostic tests. However, there was little impact on the rate of less discretionary services, such as surgical procedures (for example, coronary angioplasty). Another possibility is that physicians who invest in a hospital have a financial incentive to admit more patients who are likely to be profitable. There is a theory that physicians have at least some ability to induce demand for medical care among their patients. The theory of induced demand is based on the information asymmetry that exists between physicians and patients: Physicians generally have more information than patients about their condition and the relative benefits and costs of available treatments. Physicians might take advantage of this asymmetry by recommending "the provision of medical services that differ from what the patient would choose if he or she had available the same information and knowledge as the physician."8 Physician-induced demand might occur because physicians fail to understand or accept their patients’ preferences for health care services or because of their financial self-interest.9 It is very difficult to test this proposition directly because we usually do not know patients’ preferences and knowledge. However, we can test whether increases in physicians’ financial incentives to induce demand are associated with an increase in the volume of services provided. Researchers have discussed several factors that constrain physicians’ ability to induce demand.10 Some of these factors are physician-oriented: physicians’ ethical standards (professionalism); the limits of physicians’ time (a physician can only treat a limited number of patients per day); physicians’ interest in limiting their malpractice exposure by avoiding unnecessary interventions; and whether physicians have convenient access to facilities, such as operating rooms. Other factors relate to the patient: the patient’s diagnostic skills and knowledge (a better-informed patient is more likely to challenge a physician’s recommendation); and policies set by the patient’s insurance plan to ensure medical necessity, such as preauthorization. A third set of factors relates to the service itself: its cost to the patient, its riskiness, and its potential benefits. As price and risk increase, inducing demand becomes more difficult. To review the evidence on induced demand, we divided "inducement" into two types: direct inducement, in which the physician convinces the patient to use more services directly provided by the physician; and indirect inducement, in which the physician persuades the patient to use services not directly provided by the physician.11 In the case of direct inducement, physicians receive a professional fee for the service and, for surgical procedures, also receive a facility fee if the procedure is performed in a setting they own. Physicians who engage in direct inducement are constrained by the limits of their time and workload. In the case of indirect inducement, physicians generate passive income that is not constrained by the number of hours they are able to work each day. Physicians’ investment in specialty hospitals might represent direct (such as a surgeon-investor who refers patients to the hospital for surgery) or indirect inducement (such as a nonsurgeon-investor who refers patients to the hospital for surgery). Direct-inducement literature. Several studies have tested the theory of induced demand by examining how an increase in the number of physicians in a market affects the volume and price of services. In a competitive market, we would expect physicians to lower their fees in response to more competition, which would lead to greater use of physician services. Under the theory of induced demand, however, physicians should be able to increase their prices and recommend that their patients use more services, to maintain a target level of income. Gail Wilensky and Louis Rossiter found that higher physician-to-population ratios were associated with slightly greater use of some types of care (physician-initiated ambulatory visits) but not others (surgery).12 Jerry Cromwell and Janet Mitchell, however, found that areas with higher surgeon-to-population ratios had higher surgical fees and more surgeries, particularly elective surgeries for non-life-threatening conditions.13 These types of studies have been criticized on several grounds. Their econometric models are faulted for not specifying how physicians set their target incomes or what factors constrain them from inducing demand.14 Critics point out that the models often lack important variables that are correlated with demand, such as insurance coverage.15 There are also problems with how key variables such as physician supply and health care use are measured. Indirect-inducement literature. The literature on indirect inducement offers stronger support for the theory of induced demand. These studies have focused on whether physicians who own equipment or facilities providing ancillary services (such as diagnostic imaging and physical therapy) refer patients for more of those services, presumably for financial reasons. A study by the GAO found that physicians in Florida who were investors in diagnostic imaging centers referred their Medicare patients more frequently for tests such as magnetic resonance imaging (MRI), computed tomography (CT), and nuclear medicine than physician-nonowners did.16 Some of the differences were dramatic: Imaging center owners ordered twice as many MRI scans and 29 percent more CT scans for their patients than nonowners ordered. The GAO also found that physicians with imaging equipment in their offices ordered tests more frequently than physicians who referred patients to outside facilities. Another study, which adjusted for differences in patients’ conditions, found that physicians who performed imaging services in their offices used up to eight times as many tests as physicians who referred their patients to a radiologist.17 Other studies have found that physicians’ ownership of clinical laboratories and physical therapy facilities is associated with greater use of these services.18 Cardiac surgeon and cardiologist literature. There is some evidence that incentives can affect cardiac surgeons’ clinical decisions. For example, some cardiac surgeons tend to avoid the most complex and risky patients when patient outcomes are publicly reported.19 In one survey, 63 percent of cardiac surgeons said they were less willing to operate on severely ill patients when risk-adjusted mortality was reported.20 In a study of physician-owned specialty hospitals, the Centers for Medicare and Medicaid Services (CMS) found that physician-investors with larger ownership shares tended to admit more of their patients to a physician-owned hospital. Only 10 percent of physician-owners who owned less than a 0.5 percent interest admitted most of their patients to the cardiac hospital, whereas 46 percent of physicians who owned more than a 1 percent interest admitted most of their patients to that hospital.21 However, we are not aware of any studies that have tested whether hospital ownership can induce cardiologists and cardiac surgeons to increase the volume of invasive cardiac procedures.
Questions. We tested whether physicians’ investment in heart hospitals was followed by either (1) an increase in the number of relatively profitable cardiac surgeries performed on Medicare beneficiaries in the market; or (2) a shift toward operating on healthier (more profitable) patients. During the time frame of our study (1996–2002), there was a large increase in cardiac surgeries across the nation. We controlled for overall industry trends by comparing utilization changes in markets with physician-owned heart hospitals with those in markets lacking such hospitals. We did not evaluate whether increased utilization equates with better care or better outcomes. An above-average increase in the rate of surgeries per Medicare beneficiary, in markets with physician-owned hospitals, could indicate that hospitals are meeting an unmet community need or that physician-owners are reacting to the economic incentives associated with hospital ownership by inducing demand for additional services. Such inducement could be direct or indirect. Investors in specialty hospitals could argue that their community needs the increased capacity provided by their hospital. If this were true, we would expect to see an increase in both highly profitable and less profitable patients. However, if the opening of physician-owned heart hospitals were followed by an increase in only the more profitable cases, this could indicate that financial incentives influenced at least some physicians’ behavior. Profitable types of cardiac surgery. Investors in heart hospitals have a stronger incentive to increase use of services with high marginal profits than services with low marginal profits. To test the relationship between marginal profits and volume growth, we first identified surgeries with high, moderate, and low marginal profit. Our example of a high-margin service is coronary artery bypass graft (CABG). In 2002, CABG surgery with cardiac catheterization had a Medicare base payment amount of roughly $24,000 in a typical small urban market. A review of cost-report data, the literature, and discussions with owners of cardiac hospitals confirmed that $6,000–$12,000 of the $24,000 represents marginal profit on an average patient.22 This profit represents a sizable share of total revenue because a large share of CABG costs are fixed and do not vary with patient volume. Our example of a moderate-margin surgery is angioplasty. Angioplasty with implantation of a stent for a patient without acute myocardial infarction (AMI) had a Medicare base payment of roughly $10,000. According to the literature and hospital financial officers, the marginal profit from angioplasty was sizable but lower than that from CABG.23 In contrast, implantation of implantable cardioverter-defibrillators (ICDs) appeared to yield little or no profit in 2002. Although implantation of an ICD without cardiac catheterization had a fairly high payment rate of $22,000 in 2002, there was a perception in the hospital field that the cost of purchasing an ICD was equal to a majority (or all) of the payment.24 Because this was a relatively low-profit or unprofitable service, investors in heart hospitals had little incentive to increase ICD implantation. Patient-severity categories. We also divided cases by level of patient severity. Physician-owned hospitals tend to treat lower-severity patients, who cost less and are thus more profitable.25 In contrast, the sickest patients require more resources and are less profitable. We tested whether physicians’ ownership of heart hospitals leads to a marketwide shift in surgical volumes toward lower-severity (and higher-profit) patients. The null hypothesis is that the overall ratio of low-severity to high-severity patients in the market stays the same. We are not suggesting that cardiac surgeons operate on perfectly healthy patients. The question is more subtle: Are some physicians who invest in cardiac hospitals more likely to perform surgery on less severely ill patients when their hospitals profit from those surgeries? Size of physicians’ incentives. A hospital’s marginal profit is equal to the payment received by the hospital minus the variable costs associated with having one additional patient (for example, nursing time). Fixed costs do not affect the marginal profit of treating an additional patient. A moderately efficient hospital could increase its profit by roughly $10,000 by filling an empty bed with a CABG patient. A physician who owns a 1 percent interest would receive a profit of $100. However, each physician-investor receives profits not only from his or her admissions but also a share of the profits from other physician-investors’ patients. For example, if twenty cardiologists each have a 1 percent interest in a hospital (20 percent in total), they will each receive 1 percent of the marginal profits from twenty patients ($2,000) if each cardiologist produced one additional admission for their hospital. Data and methods. We examined utilization in the market rather than each hospital’s admissions, to control for the shift of market shares from community hospitals to specialty hospitals. We defined markets as the 306 Hospital Referral Regions (HRRs) in the Dartmouth Atlas of Health Care. These regions were created by examining where Medicare patients received major cardiovascular surgery.26 We identified eleven physician-owned cardiac hospitals that opened between 1997 and 2001 in ten different markets. We compared changes in the volume of surgeries per 1,000 fee-for-service (FFS) Medicare beneficiaries in these ten markets with changes in the volume of surgeries in 295 markets without physician-owned cardiac hospitals. This is known as a "difference-in difference" approach. We computed a rate of change by examining two years of data, 1996 and 2002. The starting point, 1996, was before the opening of all eleven facilities. The ending point, 2002, was more than one full year after all of these facilities had opened. To test for significant differences in growth rates, we used a two-tailed t-test. We calculated growth rates for all surgical cardiac discharges and for three subsets of cardiac surgeries: CABG, angioplasty, and ICD implantation. We identified surgical and CABG discharges using diagnosis-related groups (DRGs). ICD implantation and angioplasty were identified using procedure codes from claims data because they did not have their own DRGs in 1996. We categorized patients into two severity categories using All Patient Refined DRGs (APR-DRGs, 3M version 15). Patients ranked 1 or 2 on the APR-DRG severity scale were deemed less severely ill; patients ranked 3 or 4 were deemed more severely ill. Based on work by the Medicare Payment Advisory Commission (MedPAC), we assumed that the less severely ill are more profitable and the more severely ill are less profitable.27 We used claims for all Medicare beneficiaries in 1996 and 2002 in our analysis. The number of FFS Medicare beneficiaries in the ten markets with heart hospitals ranged from 64,000 to 343,000 in 1996, and the number of Medicare cardiac surgeries ranged from 2,221 to 10,669. In these markets, physician-owned heart hospitals’ share of total cardiac surgeries increased from 0 percent in 1996 to 20–45 percent in 2002. On average, the physician-owned hospitals accounted for 27 percent of the angioplasties and CABG surgeries in their markets. Our difference-in-difference approach rests on the assumption that the need for cardiovascular surgery per Medicare beneficiary did not change at an unusual rate in markets with specialty hospitals. To test this, we examined the rate of AMI (heart attack) diagnoses per 1,000 Medicare beneficiaries. There was essentially no change in the reported rate between 1996 and 2002 in markets with and without specialty hospitals. Thus, the underlying need for heart surgery did not appear to change.
Changes in heart surgery rates. Exhibit 1  shows that the differences between heart-hospital markets and other markets were small relative to practice-pattern changes that were occurring in all markets over time. On average, from 1996 to 2002, heart surgeries per Medicare beneficiary grew 21 percent in markets with physician-owned heart hospitals and 16 percent in markets without such hospitals. This difference in growth is not statistically significant.
Bypass surgeries declined in most markets because of substitution of angioplasty for bypass surgery.28 Consistent with the financial incentives to provide additional bypass surgeries, the decline was slower in heart-hospital markets than in other markets. This difference is statistically significant. Rates of angioplasty rose rapidly in all markets. Consistent with the financial incentives to provide more angioplasties, the growth rate was slightly higher in heart-hospital markets. However, the difference in growth rate is not statistically significant. Rates of ICD implantation also grew rapidly. Although the growth rate was slightly lower in heart-hospital markets, the difference was not statistically significant.
Growth in low- versus high-severity patients.
We also tested whether markets with physician-owned heart hospitals had faster growth in low-severity (higher-profit) patients relative to their growth in high-severity (lower-profit) patients, compared with other markets. We did not find any statistically significant differences in growth rates by level of severity. Markets with physician-owned hospitals had slightly above-average growth in both low- and high-severity surgeries (Exhibit 2
Study limitations. Our difference-in-difference approach controlled for changes in the numbers of Medicare beneficiaries in markets by examining rates of surgery per 1,000 beneficiaries, but it did not control for changes in the type of Medicare beneficiaries living in different markets. Although we found that the incidence of AMI among Medicare beneficiaries did not change during this time frame, we cannot be sure that there is not some unknown factor that could have caused Medicare beneficiaries in cardiac hospital markets to experience an unusual change in their need for cardiac surgery.
The differences in surgical growth rates between markets with and without physician-owned cardiac hospitals were all in the direction predicted by the financial incentives but were significant only for CABG surgery. The differences in growth rates between high- and low-profit surgeries were usually not large enough for us to conclude that hospital ownership by physicians is associated with higher use. It is still possible that most physicians are slightly influenced by the financial incentives associated with hospital investment (or a small minority of physicians is strongly influenced), but the impact of these changes at the market level was too small to be detected by our study. It is also possible that physicians in both types of markets have been inducing demand for cardiac surgery to collect professional fees, but physician-owners’ ability to induce additional use is constrained by the factors discussed earlier (for example, professionalism and malpractice liability). Because the growth in the number of cardiac surgeries in markets with physician-owned hospitals was similar to growth in other markets, we believe that specialty hospitals primarily obtained their market share (27 percent of CABG procedures and angioplasties) by capturing market share from community hospitals. Although the literature suggests that physician-owned hospitals tend to treat less severe cases, our research indicates this is attributable to a shift of low-severity patients from community hospitals to cardiac hospitals—rather than a large increase in the number of low-severity patients served in the overall market. Taken together, our findings suggest that physicians’ ownership of cardiac hospitals has primarily affected where people get surgery, not who gets surgery. These results are consistent with other researchers’ finding that geographic areas with greater hospital and physician supply are not associated with significantly higher use of many major surgical procedures.29 The finding of at most a small change in utilization associated with physician-owned hospitals is in contrast to other studies that have found that physician-ownership leads to greater use of ancillary services, such as imaging studies and laboratory tests. It indicates that physicians’ surgical recommendations might be less affected by financial incentives related to facility ownership than their recommendations regarding ancillary services. This could be because surgery is less discretionary and carries more risk for the patient. However, physician-owned cardiac hospitals are a relatively new phenomenon, and this paper looks only at the first chapter in the cardiac hospital story. It is possible that the effect of physician-ownership on utilization could change over time. In the first few years of their existence, cardiac hospitals have apparently filled their beds by taking patients from community hospitals. Anecdotal evidence suggests that some community hospitals have responded by aggressively recruiting physicians and adding new cardiac catheterization labs to attract physicians from other markets.30 As the second chapter of the story evolves, we could see an increase in the total number of cardiologists and cardiac surgeons in these markets and, as a consequence, faster utilization growth.
Jeffrey Stensland (JStensland{at}medpac.gov) and Ariel Winter are senior analysts at 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. The authors thank Mark Miller for his valuable support and guidance.
This article has been cited by other articles:
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
The Debate Over Specialty...
The Literature On Physician...
