Sources: Survey of ASHA membership, Medicare HCRIS Cost reports, Area Resource File, and Bureau of Labor Statistics; see section 5.1.1 for description. Notes: *Significant at p ≤ 0.10 (t-test); **Significant at p ≤ 0.05 (t-test) Multivariable Profit Function Regression Models, Dependent Variable is Sources: Survey of ASHA membership, Medicare HCRIS Cost reports, Area Resource File, and Bureau of Labor Statistics; see section 5.1.1 for description. Notes: *Significant at p ≤ 0.10 (t-test); **Significant at p 5≤ 0.05 (t-test) 5.2 Physician Self Referral The remaining policy issue is the potential effects of physician self-referral. The costs and benefits of physician self-referral has been debated for many years, mainly because the dominant physician payment mechanism in the U.S. has been and continues to be fee-for-service, which creates financial incentives for self-referral. In the case of specialty hospitals, the general argument against physician self-referral is that physician ownership may result in financial incentives to admit patients to the facilities in which they have an ownership stake. These arguments are to some extent based on research that has found that utilization of ancillary services is higher when an ownership relationship exists between referring physicians and ancillary services (Mitchell and Sass 1995; Lynk and Longley 2002; Kouri, Parsons, and Alpert 2002; Zientek 2003; O'Sullivan 2004). However, there are at least four important limitations to applying these arguments to acute care hospitals. First, the vast majority of studies of higher utilization resulting from self-referral are based on physician ownership of ancillary services, rather than acute care hospitals. Mitchell and Sass (1995), in their frequently cited study of physician referral, failed to find higher utilization rates associated with self-referral to acute care hospitals. This lack of association has been one of the main reasons that the two phases of Stark anti-kickback legislation have exempted physician ownership of acute care hospitals (Stout and Warner 2003; Rohack 2004; O'Sullivan 2004). In addition, there is no direct evidence that the observed higher utilization rates resulting from selfreferral to ancillary services represent inappropriate or unnecessary care (Kouri, Parsons, and Alpert 2002; Zientek 2003). Second, there is no direct evidence that physician self-referral is motivated disproportionately by financial incentives. Physician self-referral is motivated by four factors: appropriateness, quality, efficiency, and financial returns. The relative magnitude of each of these incentives has been the subject of debate, but there is no direct evidence to suggest how, on average, physicians assign weights to each factor. Consistent with the empirical findings, anecdotal evidence suggests that physicians may disproportionately weight financial incentives when the referral is for standardized products or services (e.g., lab or pharmacy), and disproportionately weight appropriateness and quality when the referral is for more intensive procedures, such as surgery (Moore 2003). Third, there is no evidence that self-referrals result in worse outcomes than other types of referral (Kouri, Parsons, and Alpert 2002; Zientek 2003). A likely reason for these findings is the endogeneity of three factors: physician quality, the likelihood of self-referral, and the quality of patient outcomes. In the case of specialty hospitals, site visits and trade press literature indicate that physician investors in specialty hospitals tend to be those who highly value efficiency in quality and cost dimensions. Thus, for many physician investors, self-referral is likely to represent the most optimal referral in terms of quality and cost. Fourth, in the case of physician ownership of acute care facilities, it is likely that the magnitude of financial incentives is limited. The General Accounting Office (2003a) found that 30 percent of specialty hospitals surveyed had no physician investors. For half of the facilities with physician investors, the average individual physician ownership share was less than two percent. In the ASHA survey, virtually all physician investors owned only five percent or less (Table 2). Moreover, the entrepreneurial returns (i.e., the fraction of the facility fee considered operating margin) for any single case are likely to be substantially less than the professional fee charged by physicians. Given the order of magnitude difference between these two revenue streams, physician incentives are likely to be driven more by professional fees, which do not vary significantly by practice setting.26 Indeed, in this context the potential for a surgeon to enhance his or her own productivity is a more likely source of financial incentive for self-referral to a specialty hospital. In other words, the primary financial motivation may be to enhance the return on investment for the surgeon's investment in "human capital" (associated with the number of procedures performed) 27 rather than any effort to assure a return on investment in the form of financial assets (associated with the overall financial performance of the hospital). In terms of policy options, even if we were to assume that these limitations were not important, a more central question is whether creating barriers to market entry are the most appropriate means of addressing the issue. The net social welfare losses associated with barriers to market entry are likely to be greater than those 26 It should also be noted that high variation in utilization and referral patterns exist without respect to physician ownership. For example, Weinstein et al. (2004) recently observed significant variation in utilization patterns for major surgery for degenerative diseases of the hip, knee, and spine in several South Florida hospital referral regions where there are no physicianowned specialty hospitals. 27 Refer to section 3.4 attributable to physician referral incentives, particularly in light of the weakness of these incentives. 6. CONCLUDING REMARKS In this study we have reviewed the theory and evidence on some of the key characteristics of specialty hospitals, including efficiency, demand, case mix, and quality. These findings were supported by observations from five specialty hospital site visits. We also conducted statistical analyses of the effects of specialty hospitals on the profit margins of general hospitals. The main findings of the study can be briefly summarized in the following three points. First, there are economic advantages associated with specialization, due mainly to process redesign, learning, avoidance of diseconomies of scope, and focus on core competencies. Specialty hospitals appear to have equal or better patient outcomes compared to their general hospital counterparts. Hence, there is no evidence to suggest that specialty hospitals should be barred from entering acute inpatient care markets on the basis of efficiency or quality of care. Second, there is no evidence, other than anecdotal, to suggest that general hospitals have been financially harmed by competition from specialty hospitals, or that such competition is undesirable from a societal perspective. Specialty hospitals compete with general hospitals in the same manner in which general hospitals compete with each other. Based on a longitudinal study of general hospital profit margins in markets with and without specialty hospitals, we find that profit margins of general hospitals have not been affected by the entry of specialty hospitals. Consistent with economic theory, the models consistently showed that the most important predictor of general hospital profitability was the extent of competition from other general hospitals in the same market area. General hospitals in less competitive markets (i.e., those with fewer competitors) had higher profits than general hospitals in less competitive markets. Contrary to the conjecture that entry by specialty hospitals erodes the overall operating profits of general hospitals, general hospitals residing in markets with at least one specialty hospital have higher profit margins than those that do not compete with specialty hospitals. These findings are also consistent with economic theory, which suggests that firms will enter markets in which extant profit margins are comparatively higher. Third, though often cited as a significant policy concern, there is no evidence that physician self-referral is a problem in specialty hospitals. Physician self-referral is likely to play a relatively minor role in specialty hospitals, for four reasons: (1) the vast majority of studies of higher utilization resulting from self-referral are based on physician ownership of ancillary services, rather than acute care hospitals; (2) there is no direct evidence that physician self-referral is motivated disproportionately by financial incentives; (3) there is no evidence that self-referrals result in worse outcomes than other types of referral; and (4) in the case of physician ownership of acute care facilities, it is likely that the magnitude of financial incentives is limited. APPENDIX A 2004 Survey of Specialty Hospital Instruction: 1. These results will be kept strictly confidential. Under no circumstances will the data leave the control of ASHA and its principal contracted researcher, John Schneider. Only aggregate data will be presented publicly (e.g., means and standard errors). 2. All responses, unless otherwise noted, should refer to your previous full fiscal year. If your facility has not been open for an entire fiscal year, indicate so at the beginning of the questionnaire. Also, unless otherwise specified, responses should refer to the main patient care facility. 3. Please answer each question as accurately as possible. In the event that it is not possible to answer a question, use the following codes: Unknown DK, Refused NA. Before resorting to these codes try to at least provide = RF, Not applicable a reasonable estimate. = = 4. For technical questions contact John Schneider at john-schneider@uiowa.edu or 319-331-2122. Question 1. Name of facility: 2. Zip code (main patient care facility) 3. Has your facility been open for at least one whole fiscal year? (1=Yes; 0=No) 4. Beginning date of most recent full fiscal year (MM/DD/YYYY) Licensing & Accreditation 5. Is your facility licensed in your state as an inpatient hospital? (1=Yes; 0=No) 6. Accredited by Accreditation Association for Ambulatory Health Care? (1=Yes; 0=No) 7. Accredited by Joint Commission on Accreditation of Health Care Organizations (JCAHO)? (1=Yes; 0=No) 8. Other accrediting organizations (1=Yes; 0=No) Specify: History 9. First calendar year in which facility was licensed as inpatient hospital 10. First calendar year in which beds were added, if different from Q9 Beds and Capacity 11. Total bed capacity 12. Number of staffed inpatient beds 13. Number of operating rooms 14. Number of intensive care beds 15. Number of recovery beds (all stages) 16. Do you maintain & staff an urgent/emergent care center? (1=Yes; 0=No) 17. If Q16 staffed? = yes, how many hours per day is the care center Ownership Structure (Q21-Q24 sum to Q20) 18. Total number of owners 19. Total number of physician owners 20. Total number of physician owners who admit 28 at least 5 patients per year 21. Number of physicians in Q20 with 0-1% ownership stake 22. Number of physicians in Q20 with 2-5% ownership stake 23. Number of physicians in Q20 with 6-9% ownership stake Response Question 24. Number of physicians in Q20 with 10% or more ownership stake Volume and Case Load 25. Number of inpatient discharges 26. Number of inpatient days (overnight stay) 27. Number of inpatient days (observation days) 28. Number of surgeries (overnight stay) 29. Number of outpatient surgeries (no overnight stay) Patient Care Revenue 30. Total gross patient care revenue (inpatient + outpatient) 31. Outpatient revenue as percent of total gross patient revenue (Q30) Sources of Patient Revenue (Q32-Q35 sum to 100%) 32. Medicare revenue as percent of gross patient revenue % 33. Medicaid revenue as percent of gross patient revenue % 34. Commercial (private health plan) insurance revenue as percent of gross patient revenue % 35. Other revenue as percent of gross patient revenue % Charity Care 36. If your state has a charity care risk pool, do you pay into it? (1=Yes; 0=No) 37. If the answer to Q29 was yes, indicate annual amount paid $ into risk pool 45. Average patient to RN ratio (e.g., for 3:1 write "3;" for 5:1 write "5") 30 % |