Physician-Owned Specialty Hospitals and Coronary Revascularization Utilization: Title and subTitle BreakToo Much of a Good Thing?

The emergence of specialty hospitals that focus on lucrative procedural aspects of medicine has generated a heated debate among policy makers^{1 - 2} that largely involves 4 issues: patient selection (ie, “cherry-picking” of healthier and wealthier patients by specialty hospitals); quality of care in specialty and general hospitals; impact of specialty hospitals on the financial health of general hospitals; and influence of specialty hospitals on utilization and health care costs.

The debate intensified in 2003 with the passage of a temporary congressional moratorium (allowed to expire in 2005) on new specialty hospital construction. However, empirical data addressing the 4 issues remain limited. Consistent evidence suggests that specialty hospitals admit patients with lower acuity and fewer comorbidities than general hospitals.^{3 - 4} Studies assessing quality have found that risk-adjusted rates of adverse outcomes in specialty hospitals are similar or somewhat lower (perhaps 10%-20%) compared with general hospitals.^{5 - 6} While anecdotal reports about the negative influence of specialty hospitals on the financial health of general hospitals abound, limited data suggest that, thus far, the impact has been small.⁷

In this issue of JAMA, Nallamothu and colleagues⁸ provide intriguing new data suggesting that increases in the use of coronary revascularization were 2.5 to 3 times higher in health care markets that experienced entry of a new physician-owned specialty hospital compared with markets without specialty hospitals, including those markets in which new revascularization programs were established at general hospitals. The differences reflected a much lower decline in use of coronary artery bypass graft surgery and a larger increase in use of percutaneous coronary intervention (PCI). The increase in PCI was particularly striking among patients without acute myocardial infarction, a group for whom PCI may provide less benefit, but that accounted for more than two thirds of all PCI procedures.

One potential explanation for these findings is that utilization in markets with specialty cardiac hospitals reflects that astute investors chose to open specialty hospitals in markets that were already experiencing rapid growth in demand for revascularization. However, additional analyses conducted by the authors provided little support for this possibility. Alternatively, the growth in utilization in markets with new specialty hospitals may be directly attributable to procedures performed in the new specialty hospitals. Indeed, at the end of the observation period, in the study by Nallamothu et al, specialty hospitals had approximately twice the volume as new cardiac programs in general hospitals and accounted for more than a third of all revascularization procedures performed in their markets.

The current study would seem to support the hypothesis that specialty hospitals directly drive utilization of coronary revascularization. However, immediately drawing such a conclusion requires careful consideration and caution, particularly given the lack of direct information in the current study about the clinical appropriateness of procedures. Although new specialty hospitals may directly increase utilization by performing procedures for patients who might receive only marginal benefit from having interventions, it is also possible that specialty hospitals indirectly increase utilization by drawing patients away from existing general hospitals through competition. In turn, existing revascularization programs may respond to losses in patient volume with their own efforts to generate new business, further fueling utilization.

Even though it is not possible to dissect the fundamental drivers of utilization from the study by Nallamothu et al, or the influence of increased utilization on patient outcomes, the results need to be considered in light of 3 important issues currently confronting the US health care system. First, and perhaps most important, increasing evidence suggests a general lack of association between more aggressive management practices and greater health care expenditures and better patient outcomes at a population level.⁹ At the individual-patient level, the recently published Occluded Artery Trial (OAT) found that PCI performed more than 3 days following an acute myocardial infarction provided no benefits relative to medical therapy.¹⁰ Yet payment policies promoted by Medicare and other third-party payers have created large financial incentives favoring procedural interventions over careful medical management and disease prevention.¹¹ These policies may change with the recently enacted update to the Medicare physician fee structure that will increase reimbursement for evaluation and management visits, while reducing reimbursement for many procedures, including PCI.^{12 - 13} Changes in the relative reimbursement for cognitive and procedural services and further refining of prospective payment to more equitably reimburse hospitals that care for more complex patients would likely mitigate the financial incentives driving specialty hospital growth. Nonetheless, potential drivers of physician interest in specialty hospitals, other than financial motivation, must be recognized, including a desire by physicians to gain greater control over hospital operations and quality,¹⁴ which, in turn, are related to the most fundamental issues of medical professionalism and physician job satisfaction.

Second, the study by Nallamothu et al must be considered in the context of the ongoing debate over the definition of what is (and what is not) a specialty hospital and the importance of physician ownership in this definition. Specialty hospitals have traditionally been identified as freestanding hospitals with unique Medicare billing numbers that focus on a discrete disease or set of procedures; the presence of unique Medicare billing numbers has been critical because it has allowed for easy identification of specialty hospitals using administrative data sources. Yet general hospitals are increasingly developing both “hospitals-within-hospitals” and freestanding single-specialty hospitals of their own that may not have unique Medicare billing numbers¹⁵ and may blur the distinction between specialty and general hospitals.

While Nallamothu et al focused exclusively on physician-owned specialty hospitals, which have been the primary source of public concern, not all specialty hospitals have physician owners.⁶ Even among physician-owned specialty hospitals, contractual relationships and governance are highly variable.³ Some hospitals are characterized by a small number of physician owners, each of whom owns a significant proportion of the hospital. Others are characterized by multiple physician owners who, even in aggregate, own a very small proportion of the overall hospital.

Third, the results of the study by Nallamothu et al should be considered in light of the data on physician- and hospital- (or supply-) induced demand. Physician ownership of ancillary services has been associated with overuse of these services, raising questions about whether physicians are placing their own financial interests ahead of their patients' best interests.¹⁶ This concern has provided the impetus for the federal ban on physician ownership of facilities such as pharmacies and home health agencies.¹⁷ The current findings raise important questions about the appropriateness of the “whole hospital” exemption loophole that permits physician ownership of specialty hospitals.

The emergence of specialty hospitals is in an early state of evolution but may represent the beginning of a fundamental reorganization in the ways in which hospitals are structured and care is delivered. Specialization already permeates most sectors of the US economy and is associated with both increased efficiency and product quality.¹⁸ Although there is no fundamental reason hospital care should differ, the current findings suggest that physician ownership of specialty hospitals may be problematic if such ownership increases the use of services for patients with marginal indications. As specialty hospitals evolve, vigilance will be needed to determine if benefits are being delivered as promised and if untoward effects on the delivery system are emerging. In the meantime, all hospitals will need to look carefully at specialty hospitals to see what, if any, lessons can be gleaned from their successes and failures.