|
Specialty Versus Community Hospitals: Referrals, Quality, And Community Benefits
In this paper we compare physician referral patterns, quality, patient satisfaction, and community benefits of physician-owned specialty versus peer competitor hospitals. Our results are based on evidence gathered from site visits to six markets, 2003 Medicare claims, patient focus groups, and Internal Revenue Service data. Although physician-owners are more likely than others to refer to their own facilities and treat a healthier population, there are rationales for these patterns aside from motives for profit. Specialty hospitals provide generally high-quality care to satisfied patients. Uncompensated care plus specialty hospitals’ taxes represent a greater burden, in percentage terms, than community benefits provided by nonprofit providers.
AS PART OF THE MEDICARE PRESCRIPTION DRUG, Improvement, and Modernization Act (MMA) of 2003, Congress established an eighteen-month moratorium on the development and expansion of new physician-owned specialty hospitals. The central concern among policymakers is whether these hospitals enjoy an unfair competitive advantage relative to other community hospitals. During the moratorium, Congress required the Medicare Payment Advisory Commission (MedPAC) and the Centers for Medicare and Medicaid Services (CMS) to report on two different aspects of this issue. At issue is whether specialty hospitals’ physician-owners are able to control the referral of patients, choosing between their own facilities and other hospitals in the community, in a way that results in favorable selection. Other related issues are whether specialty hospitals provide high-quality care, how their patients perceive care, and what types of community benefits they contribute in their markets. Although the congressional moratorium focuses largely on whether or not physician-ownership contributes to an unfair competitive advantage for specialty hospitals, at stake is the level of competition that will be allowed in hospital markets, or whether greater regulation of hospital markets is a likely policy direction of the future. We conducted a study that focused on these issues, under contract to the CMS to support the secretary of the Department of Health and Human Services’ (HHS’) required report to Congress. In this paper we apply our findings to four policy questions: (1) Do specialty hospitals enjoy an "unfair" competitive advantage in their markets, driven by the incentives of physician-ownership? (2) Does physician-ownership result in favorable referral patterns to specialty hospitals? (3) Do specialty hospitals provide lower quality of care to less satisfied patients than local community hospitals? (4) Do specialty hospitals fail to bear an equal burden in providing community benefits compared with community hospitals?
A critical first task in our study was to determine which facilities we should define as specialty hospitals. Section 507(a) of MMA defines them as "primarily or exclusively engaged in the care and treatment of one of the following categories: (i) patients with a cardiac condition; (ii) patients with an orthopedic condition; (iii) patients receiving a surgical procedure," and any other category deemed to a be a "specialty" and subject to the eighteen-month hospital-building moratorium. Unfortunately, a Medicare designation of "specialty hospitals" (other than certain specialized hospitals, such as children’s or psychiatric facilities) does not exist. Recent studies have investigated the effects of specialty hospitals on other providers, each using a somewhat different definition and different numbers of "specialty hospitals" in their analyses.1 Study sample. Given the lack of a well-defined list of specialty hospital facilities, we generated a list of physician-owned specialty hospitals based on Medicare data. We adhered generally to the MedPAC definition of specialty hospital in which at least 45 percent of discharges were in one of three clinical groups: diseases and disorders of the circulatory system (major diagnostic category, or MDC, 5), diseases and disorders of the musculoskeletal system and connective tissue (MDC 8), or surgical diagnosis-related groups (DRGs). Our goal was to identify all such hospitals in existence in mid-2004 using the most recently available data, because, as the earlier studies have noted, the number of specialty hospitals has increased dramatically during the past few years. To this end, we identified physician-owned specialty hospitals in two stages. In the first stage we identified a set of physician-owned hospitals that were "potential" specialty hospitals using an Internet search and listings provided by relevant associations. In the second stage we used Medicare claims data from January–June 2004 to determine the specialization (if any) of the hospitals identified in the first stage. This method yielded ninety-two specialty facilities—a much larger and more current study sample than previous studies have used. Once we identified specialty hospitals using this approach, we identified competitor hospitals in the same markets (defined as located within twenty miles). The Medicare Provider of Services (POS) File, issues of the American Hospital Association (AHA) Guide, and full calendar year 2003 Medicare claims were used for many of the analyses.2
Site-visit data.
Because most of the policy issues involve physicians’ behavior and motivations, we felt that it was important to observe and interview staff in specialty and competitor hospitals. Therefore, to complement the Medicare claims data analyses and to help interpret these empirical findings, project staff conducted site visits at hospitals in six cities. Cities were selected based on the number and types of specialty hospitals in operation (cardiac, orthopedic, and surgical) (Exhibit 1
Limitations of study design. Despite the advantages of a large national sample of specialty hospital Medicare claims and direct data on physician ownership share, this study has a number of limitations. First, it is limited to Medicare patients and hospital inpatient claims. Second, we did not analyze Medicare claims for hospital outpatient departments or ambulatory surgery centers (ASCs). This is work that we are performing as a follow-up to our original analysis. Use of services in these facilities would likely be necessary to present a complete picture of competition in specialty hospital markets. Third, project time and financial resources limited our analysis to one year of Medicare claims data (2003), six case-study markets, and six patient focus groups in three markets. Finally, our findings (and those in the MedPAC study also mandated under MMA) consider the impact of specialty hospitals during a limited period. The dynamic nature of specialty hospital entrants into markets suggests that longer-range impacts might be different from the short-term analyses presented here.
Hospital competition. Because physicians generally drive the decision of which facilities their patients are referred to, physician-owners might have a financial incentive to direct the most profitable patients to specialty hospitals in which physicians have a financial interest. This is particularly true for Medicare beneficiaries, given the current case-mix adjustment limitations of the DRG-based prospective payment system (PPS) that has been shown to lead to differences in the relative profitability of various DRGs.3
From our analysis, we found that ownership by physicians is positively related to the likelihood of referring patients to a specialty hospital. However, the relationship between ownership and referrals varied by specialty hospital type and (nonlinearly) the size of the ownership stake (Exhibit 2
First, although we found that physician-owners do tend to favor their own specialty hospitals, they also refer patients to competitor hospitals; the size of the ownership share appears to be an important factor, not the fact of ownership per se. We also found that most physician-owners have very small shares in their specialty hospital and, possibly as a consequence, make few referrals to the facility. For example, only one of ten cardiac facility owners with shares less than 0.5 percent (one-third of all owners) referred more than half of their cases to their own specialty hospital. By contrast, one of two physician-owners with greater than a 1 percent share referred more than half of their patients to their own hospital. The relationship was similar for owners of surgical specialty hospitals but somewhat stronger for owners of orthopedic specialty hospitals. Second, case-study interviews revealed that many local physicians invested in the specialty hospital either out of a personal relationship with the major physician owners or to ensure that they could refer patients to the facility if need be. These reasons might explain the relatively small ownership shares of many physicians; they can accomplish these goals without more sizable financial stakes. That physician-owners sometimes refer sizable numbers of patients to their hospitals is not inconsistent with physicians’ behavior more generally. For example, an inspection of admission patterns of sixty-four competitor cardiologists in one of our site-visit cities (Oklahoma City) showed that all but nine admitted more than 90 percent of their patients to a single facility. Of sixty-one orthopedic surgeons, all but twelve admitted to a single facility. Detailed discussions with physician-owners uncovered a range of factors—other than ownership share—that affect physicians’ referrals. An appreciation of these factors will help one interpret our quantitative findings. Insurance participation. We found that insurance was a strong determinant in physicians’ referral decisions. In several markets, community hospitals had entered into exclusive contracts with major insurers. We found no evidence that specialty hospitals were holding exclusive insurance contracts that would draw much business away from the community hospitals. In fact, a few specialty hospitals were lobbying for "any-willing-provider" legislation in their states, to be able to refer insured patients to their facilities. Emergency "call" in competing community hospitals. We found that it was fairly common for physician-owners of specialty hospitals to take emergency department (ED) "call" in community hospitals; this practice also affected referrals. This occurred for a number of reasons. Physician-owners said that they needed to see patients in the larger EDs of community hospitals to serve the community and to make a living. Community hospitals reported that specialized expertise was needed to properly care for emergency patients entering their doors.4 Thus, not only do most physician-owners see large numbers of patients at competitor hospitals, they also tend to admit sicker patients coming in through the ED where they are taking call. Patients’ preferences and service needs. Patients’ preferences were factored into referral decisions. Some patients might prefer a specialty over a general hospital because of its single rooms and hotel-like amenities or, conversely, might prefer to return to a local hospital where they had been treated in the past. Referring physicians also considered the service needs for a particular patient and where he or she would receive the highest-quality care. Physician-owners and nonowners using the specialty hospital felt that the overall quality of care was better because of higher procedure volumes, lower patient-to-nurse ratios, and more patient amenities (such as private rooms). In sharp contrast, physicians in community hospitals were concerned that patients in specialty hospitals lacked the necessary intensive medical care backup on site that was available in their facilities. Physicians’ preferences and convenience. Physicians working in specialty hospitals reported that they find it more convenient, on occasion, to admit a patient to a general hospital that is closer to their offices or that offers other conveniences. Physician-owners and nonowners at all sites we visited argued that specialty hospitals are better able to schedule patients in a way that is convenient to the physician and to complete procedures or operations on time. Favorable selection. Patients with greater severity of illness require more-intensive and -expensive care; yet, under the current Medicare payment system, reimbursement is generally the same, regardless of severity. Therefore, profitable favorable selection occurs when one facility systematically admits Medicare patients with lower severity levels. Our results for specific MDC and DRG analyses in the six site-visit markets found that, consistent with findings from the GAO and MedPAC, physician-owned specialty hospitals treat less severely ill patients than their competitors, although not necessarily across every DRG and not systematically across all specialty hospital types.5 These results suggest that physicians are referring less severe and therefore more profitable patients to specialty hospitals that they own. We found that a range of rationales explain the existence of the clearly defined, narrow focus of specialty hospitals that go beyond physician-ownership and a motive for profit making. Our findings suggest that although profit might well be a motive for referral behavior in some hospitals and by some physician-owners, other explanations might apply in other hospitals and markets. It seems logical to ascribe their narrower service offerings, in fair part, to the dominant specialty of the major physician-owners—usually cardiologists and orthopedic surgeons.
If physicians’ ownership and incentives for profit were the primary driving factors in referring healthier patients to specialty hospitals, we would expect to see different patterns of referrals among owners and nonowners. Nonowner referring physicians do not share physician-owners’ profit incentives for specialty hospital referrals (Exhibit 3
Within each of the six site-visit markets, we observed little difference in referral patterns between owners and nonowners, which suggests that specialization of the hospital is potentially a primary issue, not ownership alone. The results for cardiac specialty hospitals in three cities (Fresno Heart Hospital was in operation only two months in calendar year 2003 and was excluded) show that in Dayton, both physician-owners and nonowners referred a higher percentage of severely ill patients to the cardiac specialty hospital. In Oklahoma City and Tucson, the trend was reversed, but only differences in Oklahoma City were significant. The Oklahoma Heart Hospital is majority-owned by a full-service hospital located next door that treats many heart patients with serious comorbid illness. Findings for orthopedic specialty hospitals show a more consistent pattern of less severely ill patients referred to specialty hospitals, but, again, the pattern is consistent for both owners and nonowners. Too few observations were available to test for percent-of-ownership effects on severity of referrals. Quality of care and patient satisfaction. Our study also included a number of analyses to assess the quality of care and patient satisfaction for Medicare beneficiaries in specialty relative to community hospitals.6 Specialty hospitals put forward the argument that by focusing on a limited range of diagnoses and procedures, they have the potential to increase the quality of care provided to their patients. This argument centers on the notion that focus, practice, and repetition are known to improve outcomes.7 Competitor hospitals, however, argue that specialty hospitals, by offering a limited range of services, lack the ability to deal appropriately with complications and other complex problems.
In addition, specialty hospitals feature an all–registered nurse (RN) staff; low patient-to-nurse ratios; high procedure volumes; electronic physician ordering; single rooms; and the latest equipment, structure, and process measures that are associated with quality. We examined three measures of quality using Medicare claims data: mortality rates (Exhibit 4
Mortality rates. Risk-adjusted thirty-day mortality rates were significantly lower for specialty hospitals than for community hospitals. Across both cardiac and orthopedic specialty hospital types (the small numbers of deaths made analysis of surgical specialty hospital unreliable), for both moderate (All Patient Refined, APR-DRG 1 or 2) and severe patients (APR-DRG 3 or 4), the proportion of patients who died while hospitalized or within thirty days of discharge was significantly less for specialty hospitals than for community hospitals for all DRG groupings. Although we controlled for admission type and severity, we did not stratify by specialty and community hospital volume, which has been found to account for differences in mortality.8 Although mortality is a very rare event for the types of orthopedic conditions treated at specialty hospitals, other measures such as return to functioning and improved mobility are not available in the claims data.
Readmission rates.
The readmission analysis in Exhibit 5 Patient satisfaction. Patient satisfaction is another dimension of quality. To better understand patients’ perspectives on specialty and local competitor hospitals, we conducted a limited number of focus groups with Medicare patients in three of the six sites. Although focus groups are useful in providing uniquely detailed feedback on patients’ perspectives, they are inherently limited, and their findings should be generalized only with caution. Focus groups do not provide statistically robust findings. Still, focus groups are useful in identifying strong or prevalent patient perspectives when a survey is not feasible. With these limitations in mind, we noted that Medicare patients’ satisfaction was very high in specialty hospitals, as evidenced by numerous positive comments from patients who had been treated in specialty hospitals. Patients in our focus groups evaluated the limited clinical focus of specialty hospitals highly, particularly as this related to the perceived expertise of hospital staff. They thought that the level of knowledge and specialized skills of the nursing staff differed materially between specialty and community hospitals: I felt like the nurses were trained in that specific area and therefore we didn’t have to do as much explaining to them about what we felt was going on with our bodies. Those who had been hospitalized previously for a serious condition compared their experience to being in an intensive care unit (ICU) at a community hospital. One focus-group member remarked that the nurses’ demonstrated confidence and knowledge helped relieve many of the fears he had when going into surgery: They made me feel so comfortable.... They talked to me.... They explained the procedure, and this one nurse told me they had 98 percent of people coming out okay. Remarks about the specialized knowledge of nursing staff were not offered by Medicare patients treated at a community hospital, except in the ICU context. When asked in our specialty-hospital focus groups if people knew that the hospital was partially owned by physicians, most stated that they had known prior to hospitalization. Most agreed that physician-ownership was a positive factor that probably contributed to how well they felt the hospital was run. One commented: I think they care more because their name’s on it.... They own it.... It’s just normal that they would put more into it. Those receiving care at a specialty hospital had positive experiences with the hospital environment and commented on the private rooms, space, lower noise level, and treatment of family members, including pleasant waiting areas. Private rooms offered a quiet environment conducive to sleep and recovery: If you have a heart condition, it’s extremely anxiety producing. If you are in a setting where there’s a lot of ruckus or you are concerned about whether you are going to get the kind of attention you need or if there’s just generally a sense of disorganization or noise...it just adds to that anxiety level. In contrast, many Medicare patients who went to a community hospital expected the inconvenience associated with a shared room, a certain level of noise, fewer ways to accommodate family (including less plush waiting areas), and occasional teaching rounds of residents and interns. They also reported more delays after being admitted to the hospital when being transferred for tests, and so forth. This was generally considered part of the hospital experience. Community benefits. Concerns have been raised that physician-owned specialty hospitals exist primarily to generate profits for their physician-owners. To the extent that this is true, specialty hospitals might contribute little to the overall community in which they exist. But based on the ten specialty and twenty-one community hospitals in the six cities we studied, we found that specialty hospitals incurred a greater net community benefit burden than their not-for-profit competitors did.9 We estimated the sum of uncompensated care costs and taxes paid by these hospitals. We also computed the difference between uncompensated care costs and the value of the tax exemption received by nonprofits (this definition might better account for the value of unprofitable activities, since lower margins result in lower values of tax exemption). Equating uncompensated care cost with community benefit, for comparing to taxes or tax exemptions, is a standard approach in this literature.10 Under both definitions, the specialty hospitals we studied provided more net community benefits than their not-for-profit competitors as a share of total revenues: 5.5 percent versus 2.5 percent under the first definition, and 1.0 percent versus –0.4 percent under the second. We did find particularly low uncompensated care percentages for not-for-profit hospitals in cities with a publicly owned hospital. On average, the low community benefit burden of not-for-profits did not justify the value of their tax exemption. The higher net community benefits generated by specialty hospitals were attributable almost entirely to the taxes they paid as for-profit entities. Additionally, the cardiac hospitals in this study provided evidence of a nontrivial level of uncompensated care, but they were generally less profitable (and hence paid fewer taxes). Our results are also generally consistent with findings that uncompensated care in not-for-profits costs somewhat less than the value of their tax exemptions.11 We also found that a large proportion of patients in cardiac specialty hospitals have Medicare coverage. Most specialty hospitals also treat Medicaid and self-pay patients, although as a smaller percentage of total patient revenues than their competitors—especially orthopedic and surgical specialty hospitals. Orthopedic specialty hospitals also treated patients with other forms of public insurance (for example, workers’ compensation).
We found that specialty hospitals treat Medicare patients with lower-severity illnesses, compared with the illness severity of patients treated in community hospitals. Furthermore, physician-owners do prefer to admit to their own facilities when possible. However, contrary to the notion that community hospitals are languishing passively on the perceived unfair playing field created by physician-owned specialty hospitals, we found in our site-visit discussions that community hospitals have responded vigorously to local competition and the entry of specialty hospitals in most markets by (1) purchasing "feeder" primary care practices committed to sending patients to their facilities; (2) providing valuable operating room time as an incentive for surgeon referrals; (3) negotiating exclusive managed care contracts with insurers; (4) providing lucrative "management" subcontracts with inpatient specialists in lieu of actual ownership stakes; (5) opening heart and orthopedic "centers of excellence" on campus for specialists; and (6) building physician offices on campus.12 Any disadvantage that not-for-profit hospitals face from being prohibited in offering their physician staff an "ownership stake," while real, has been diminished somewhat through alternative financial arrangements. Although the policy debate tends to focus on specialty hospitals’ possible "unfair" competitive advantage, we found that they actually stimulate a competitive environment in some markets, which could have positive effects on quality of care. Cardiac specialty hospitals in general, and orthopedic specialty hospitals in small markets in particular, heightened local competition for patients. In no case were we aware of a specialty hospital opening in a market without at least one local competitor. Given the short-run nature of our evaluation, we were not able to evaluate specialty hospitals’ long-run viability, but it was clear that not all of them were financially viable in the longer run. Patient satisfaction among Medicare beneficiaries treated in specialty hospitals was very high. Contrary to allegations made by competing hospitals, we found very little evidence of poor quality of care in specialty hospitals relative to community hospitals; instead, we found many instances of high-quality care that should be encouraged. Physicians’ commitment to and pride in their specialty hospitals are powerful positive forces that critics have underappreciated. Encouraging physicians’ involvement in community hospitals’ decision making could address many of the reasons why physicians choose to sponsor their own specialty hospitals. Although specialty hospitals provide less uncompensated care, they do contribute substantial tax revenues, contrary to the notion that these facilities are simply a drain on community resources. For-profit hospitals are legitimate, legal entities and are not required by the Internal Revenue Service to provide uncompensated care because they pay taxes instead. Policy concerns regarding "unsupported" uncompensated care in U.S. hospitals should not focus on specialty hospitals alone; rather, the problem of uninsurance deserves a more fundamental solution, including broader coverage and better enforcement of community benefits when tax exemptions are granted. Favorable selection in many facilities, including specialty hospitals, arises from the flawed Medicare payment system that overpays for healthy surgical cases. "Overpayments," not necessarily physician-ownership, encourage all types of investors to open specialized facilities and "unravel" care from full-service tertiary care hospitals. Physicians are able to take advantage of the profitability differences created by the current Medicare payment system because they have a strong influence over where patients receive care. However, these incentives, and the resulting effects on favorable selection, do not seem to be unique to specialty hospitals. Changing "self-referral" incentives for physicians could be addressed much more directly and effectively through review and modifications to the Medicare DRG-based payment methodology than through policies that limit only referrals to specialty hospitals.
Leslie Greenwald (lgreenwald{at}rti.org) is a senior scientist at RTI International in Research Triangle Park, North Carolina. Jerry Cromwell is a senior fellow there; Walter Adamache, a research economist; Shulamit Bernard, program director; Edward Drozd, a senior economist; and Elisabeth Root, an analyst. Kelly Devers is an associate professor in the Departments of Health Administration and Family Medicine, Virginia Commonwealth University, in Richmond. This project was funded by the Centers for Medicare and Medicaid Services (CMS) under Contract no. 500-00-0024, T.O. no. 12. The statements contained in this paper are solely those of the authors and do not necessarily reflect the views or policies of the CMS. The authors assume responsibility for the accuracy of the information contained herein and thank Fred Thomas, their CMS project officer, for many helpful comments and support in conducting this research.
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Methods: Definitions And...
Study Findings
