Spiritual

CROSS BORDER EFFECTS OF STATE HEALTH TECHNOLOGY REGULATION

Categories
Published
of 20
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Description
CROSS BORDER EFFECTS OF STATE HEALTH TECHNOLOGY REGULATION jill r. horwitz daniel polsky ABSTRACT Certificate of Need (CON) laws, state laws requiring providers to obtain licenses before adopting health-care
Transcript
CROSS BORDER EFFECTS OF STATE HEALTH TECHNOLOGY REGULATION jill r. horwitz daniel polsky ABSTRACT Certificate of Need (CON) laws, state laws requiring providers to obtain licenses before adopting health-care technology, have been controversial. The effect of CON on technology supply has not been well established. In part this is because analyses have focused on state-level supply effects, which may reflect either the consequence of CON regulation on supply or the cause for its adoption or retention. Instead, we focus on the cross border effects of CON. We compare the number and location of magnetic resonance imaging providers in counties that border states with a different regulatory regime to (1) counties in the interior of states, (2) counties on state borders with the same regulatory regime on both sides, and (3) counties on borders with different regulatory regimes, but with a largeriverontheborder.wefindthereare6.4fewermrispermillionpeopleinregulated counties that border counties in unregulated states than in unregulated counties that border regulated counties. This statistically significant finding that regulatory spillover can be sizable should be accounted for in future research on state-based health technology regulation. In addition, it suggests state experiences may not accurately predict the effects of CONifitwereimplementednationally. KEYWORDS: regulation, state law, certificate of need, MRI, technology, technology licensing JEL CLASSIFICATION: I11, I18, K32, K23, L1, L52 I. Introduction Certificate of Need (CON) laws state licensing regimes that restrict the adoption and modification of health technology by existing providers and new entrants have generated controversy since their adoption in the 1970s. Proponents have urged the adoption of CON in markets like health care where incomplete information and agency problems can lead to overinvestment in high-fixed-cost technologies. Their argument is that direct regulation would work to control costs, constrain investment, and improve quality in markets (Salkever 2000). Opponents cite the anticompetitive nature of such regulation, particularly its susceptibility to capture, and have urged repeal of remaining certificate of need laws (see, e.g., Epstein and Hyman 2013; Cordato 2007). Empirical research has not resolved these debates. Jill R. Horwitz (corresponding author, School of Law, University of California, Los Angeles, and the National Bureau of Economic Research. Daniel Polsky, Perelman School of Medicine and Wharton School, University of Pennsylvania and the Leonard Davis Institute of Health Economics. C 2015 American Society of Health Economists and doi: /ajhe a Massachusetts Institute of Technology American Journal of Health Economics 1(1): AMERICAN JOURNAL OF HEALTH ECONOMICS Here we seek to inform this larger debate by examining the effects of CON regulation on the location of freestanding magnetic resonance imaging (MRI) providers. We hypothesize that if the CON regulation is binding we would see the technology being provided in unregulated states to serve nearby populations in regulated states. Specifically, we test whether MRI providers disproportionately locate in unregulated states in counties that border states regulated by CON licensing laws. Previous research has focused on statewide effects of CON. It is difficult to determine whether CON changes supply decisions or downstream outcomes using this approach because a state s decision to adopt or retain CON regulations can be as much a reaction to existing provider supply as a cause of future provider supply. By focusing on whether counties border a state with a different regulatory regime, we can overcome this limitation by contrasting these counties with three other types of counties: (1) counties in the interior of states that do not face a border at all, (2) counties on state borders with the same regulatory regime on both sides, and (3) counties on borders with different regulatory regimes, but with a large river on the border. In addition, previous research has largely focused on capital-and labor-intensive technologies such as percutaneous coronary interventions (PCI) and open heart surgery. These technologies are typically provided in hospitals, which are both likely to seek a CON for these services and unlikely to respond to regulation of an individual technology by relocating. On the contrary, freestanding MRI providers comprise a relatively fluid market for analysis of the potential effects of technology regulation. These results can inform understanding regarding the effects of regulation on the number of services that are being offered outside of hospitals. Our results show a sizable and statistically significant cross border effect of CON regulation on the location of MRI providers. Among counties located on state borders where one state regulates MRI entry and the other state does not, there is more likely to be an MRIproviderintheunregulatedcountythanintheregulatedcounty.Moreover,weconfirm our results by linking the effect to the relative ease of traveling across state borders; we find the difference between regulated and unregulated border counties to stem almost exclusively on state borders that are not separated by rivers, borders that are presumably easier for potential patients and staff to cross. These results point to several conclusions. First, the effects of state regulation include not only the effects on residents in those states, but also spillover effects that should not be ignored. Previous research on CON has focused on the former and, as Cotet (2012, 203) has argued in another context, when border effects are important, specifications that fail to control for spillovers lead to biased estimates of the impact of the law. In particular, studies that find fewer providers in CON states may overstate the implications of regulation for potential patients who live close to non-con states. On the other hand, this spillover effect may help explain why many studies have been unable to attribute differences in cost or quality to CON regulation. Second, our results suggest that where state regulations have spillover effects on other states, states may not serve as effective sites to evaluate the effectiveness of those regulations. As a result, the effectiveness of state regulations may not inform the effectiveness of the regulation if it were implemented nationally. 102 Cross Border Effects of State Health Technology Regulation // horwitz, polsky The paper proceeds as follows. Section II outlines the purposes and history of CON legislation, summarizes previous research, and provides a conceptual framework for understanding provider location in response to regulation. Section III describes our data and the descriptive statistics. Section IV describes our empirical framework for estimating the cross border effects of CON regulation, reports the results, and describes limitations of the analysis. Section V concludes. II. Health Planning and Certificate of Need Regulation: Background, Previous Research, and Institutional Context A. BACKGROUND Certificate of Need (CON) laws require health-care organizations to obtain permits from a state regulatory agency before building new health-care facilities, offering new medical services, or acquiring certain medical technology. Although the first CON laws were implemented in New York in 1964 and a few other northeastern states shortly thereafter, such programs have their roots in earlier federal health planning efforts, such as the 1946 Hospital Survey and Construction Act (the Hill-Burton Act), meant to increase access to care, improve quality, and control costs (Hamilton 1985). The premise of capital investment restrictions such as CON restrictions was that in markets for health care markets characterized by failures that make price competition an untenable method for providing an efficient level of care regulation controlling the supply of medical providers and health-care services would prevent excess capacity, duplicative and unnecessary service provision caused by provider-induced demand, and spiraling costs (Finkler 1985). Virtually every state adopted CON requirements in response to two federal statutes the Social Security Act of 1972 (which conditioned Medicare and Medicaid payments on reviews for large capital expenditures) (Section 1122, 1972), and the 1974 National Health Planning and Resources Development Act (which offered financial incentives for states to establish review procedures for new clinical services, inpatient technology acquisition, and capital expenditures greater than $150,000). Early state CON requirements applied to inpatient hospital services, but in the 1980s many states expanded their CON regimes to control the growth of ambulatory services such as diagnostic imaging. After 1986, when President Reagan signed the law repealing the National Health Planning Act, many states eliminated some or all of their CON requirements based on evidence that the planning authorities had proved costly to the Federal Government, in the last analysis without benefit, and even detrimental to the rational allocation of economic resources for health care (Yakima Valley 2011, 935). Even more recently, in 2004 the Federal Trade Commission and the US Department of Justice urged the remaining states with CON requirements to reconsider whether these programs best serve their citizens health care needs, in large part because of their anticompetitive risks and failure to contain costs (Federal Trade Commission 2004, 22). Nonetheless, state CON laws remain remarkably stable; states that repealed their CON laws tended to do so in the late 1970s and through the 1980s, with little change since then (NCSL 2011, 2013). 103 AMERICAN JOURNAL OF HEALTH ECONOMICS B. PREVIOUS RESEARCH ON CON Previous research evaluating the effects of direct regulation such as CON has generated mixed results (Salkever 2000). However, research comparing the experiences of regulated to unregulated states has often concluded that CON has been ineffective at limiting supply, controlling costs, or improving quality (Conover and Sloan 1998). The earliest research on the effects of CON on costs, investments, and service diffusion in hospitals generated disparate results. Russell (1979) found some evidence that CON limited investments in specialty services. However, in sum, early research tended to find that capital expenditure controls such as CON did not constrain costs, but may have limited the supply of beds within individual hospitals (Salkever 2000). For example, Salkever and Bice (1979) studied the effects of regulation in those states that were early adopters of CON programs and found no conclusive evidence that CON had effects on costs or quality. Another early study found mixed results on the effects of CON, but largely concluded that regulation not only failed to control costs but may well have increased costs and labor inputs (Sloan and Steinwald 1980). Later studies were similarly mixed. Some found that CON increased costs (Lanning, Morrisey, and Ohsfeldt 1991; Antel, Ohsfeldt, and Becker 1995), while another found that CON programs reduced per capita acute care spending by 5 percent, but did not affect total per capita spending, suggesting that providers responded to cost-containing regulation by increasing costs elsewhere (Conover and Sloan 1998). Only a few studies focused on CON regulation for ambulatory services such as diagnostic imaging. These analyses were conducted soon after regulation of outpatient and freestanding medical services was implemented and focused exclusively on technology diffusion rates. For example, using data from the American College of Radiology, Hillman and Schwartz (1985) conducted telephone interviews with MRI installers and employees in the marketing departments of MRI manufacturers to track all the early adoption of MRI conducted through CON applications; they found that MRI diffused more slowly than computed tomography (CT) in freestanding facilities, and concluded that Medicare s prospective payment system and CON regulations explained the difference. Steinberg, Sisk, and Locke (1985) made similar claims, as well as identifying clinical, technical, and other economic issues as the reason that MRIs diffused comparatively slowly with only 108 machines in place by the end of The most recent CON research has focused on a few hospital-based services, particularly invasive cardiac treatments, and has also generated mixed results regarding its effectiveness at controlling costs, limiting quantity, and improving quality. Cutler, Huckman, and Kolstad (2010) concluded that the removal of CON in Pennsylvania was welfare neutral; in that case, increased entry into cardiac surgery led to a redistribution of patients to higher-quality surgeons, an effect that approximately offset the losses due to increased fixed costs. Studies of the effects of CON on the volume of cardiac services typically found that the average number of procedures performed per hospital in states without cardiac CON was significantly lower than in regulated states (Ho 2004; Vaughan-Sarrazin et al. 2002). Studies further link the relatively high volume at hospitals in regulated states with better health outcomes. For example, Vaughan-Sarrazin and colleagues (2002) found 104 Cross Border Effects of State Health Technology Regulation // horwitz, polsky lower mortality in regulated states, and Popescu, Vaughan-Sarrazin, and Rosenthal (2006) found that patients in regulated states were less likely to need revascularization services than those in unregulated states. Moreover, Ho et al. (2007) demonstrated that the number of hospitals offering intensive cardiac services (coronary artery bypass grafting and PCI) was lower in states with CON than in others, and that CON was associated with fewer PCIs per capita. In more recent work, Ho, Ku-Goto, and Jollis (2009) used a difference-indifference approach that substantiated some of Ho s earlier findings. They analyzed states that discontinued CON between 1989 and 2002 and found no change in utilization rates after the elimination of regulation. Some researchers have considered the effects of CON on access, for example finding that the loosening of CON rules was associated with increased access to cardiac care for African-Americans as well as with reductions in health disparities (DeLia et al. 2009). Our research, in the context of this previous work, is significant in at least four important respects. First, to our knowledge there is no previous research on the effects of CON on practice location. Second, strategic practice location may result in spillovers that have not previously been accounted for in this body of work evaluating state-specific cost and quality outcomes resulting from CON regulation. Third, because CON regulations occur at the state level, previous research has been limited in their ability to control for the fact that states that retained CON regulations are different from those that did not in ways that are relevant to current health technology markets. Our research focuses on border counties. This allows us to use interior counties to remove any state-specific fixed effects. We also can control for border-specific effects because not all borders face regime changes. Fourth, existing research focuses on hospital-based services, neglecting ambulatory services such as free standing MRI that are relatively easy to relocate compared to hospitalbased services. The lower costs of the MRI technology allows providers to be more nimble. AnewMRIunitcostsbetween$1and$3million(thepriceofacommon1.5teslawhole body scanner is $1.25 million), while costs for cardiac catheterization laboratories are in the range of $3 million and up (Cosmus and Parizh 2011). 1 Moreover, hospital-based technologies are provided to relatively sick patients, those who are either inpatients or likely have more limited ability to travel compared to patients seeking diagnostic imaging at a freestanding center. On the contrary, even though increasedtravelforpatientsmaybecostly andatsomedistancepatientsmayforgocare entirely patients can likely more easily travel to freestanding MRI offices than to hospitals. This is especially true given that some MRI machines can be located anywhere one can site a trailer, particularly with the development of compact scanners. As a result of these differences, MRI providers can be relatively nimble in their responses to regulation andcompetition.weassumethattherelativeeaseofentryonthesupplysideandpatient travel on the demand side makes MRI markets good markets in which to consider 1 Costs range depending on the type of equipment, including whether it is new or used. The most recent New York applications for new laboratories that we could find were roughly $3.5 million. Mount Sinai Hospital requested the purchase of two new labs for a total of $7,139,016 (New York State Department of Health 2010) and Orange Regional Medical Center requested one new lab for $3,462,325 (New York State Department of Health 2008). 105 AMERICAN JOURNAL OF HEALTH ECONOMICS responsiveness to state technology regulation. Given the growth of services provided on an outpatient basis, primarily due to an increase of service provision outside of hospitals (MedPac 2012), studying freestanding providers like MRI will yield increasingly generalizable results. C. INSTITUTIONAL CONTEXT Ourstudyrestsonseveralassumptionsabouttheresponsivenessofentrepreneursand firms to direct technology regulation. First, we hypothesize that a provider considering serving a population in a regulated area would prefer to locate the business in a nearby unregulated location if one is available as long as travel costs from the regulated area are minimal. We make this assumption because opening a new imaging center in a regulated state can add additional costs to starting a business such as financial costs to assemble a CON application, time costs of waiting for approval, and uncertainty regarding whether a potential competitor with a CON license would challenge the application. But providers weigh these costs against the benefits of locating where there is unmet demand for care, demand that may be greater in regulated areas near unregulated areas where licensing costs can be easily avoided. Second, the financial benefit from locating on the other side of a state border depends crucially on whether patients are covered by their insurance plans when they see providers across state lines. Although insurers do not typically sell insurance across state lines, they do typically include out-of-state providers in their preferred networks when the market spans multiple states. 2 Insurers also do generally reimburse for out-of-state care, albeit with less favorable cost-sharing, when providers are outside of their preferred network. Finally,onemightreasonablyassumethatthethreatofentryacrossastateborder would push states to eliminate CON if neighboring states have done so. In this case, regulating may mean allowing a neighboring state to capture tax and labor benefits that travel across borders with new business. And, as can be seen in Figure 1, the states that have retained CON are clustered in a few regions of the United States, leaving a limited number of states with MRI CON that face these border issues. There are, however, several reasons why states may maintain CON. For example, health-care cost control may be a more important or salient issue for politicians. Or, even if politicians might wish to eliminate direct regulation, the organizations that benefit from CON such as large hospitals and major teaching centers also have considerable influence over politicians. 2 Although this may not be true for every insurer, there is sufficient evidence that insurance coverage is not a barrier for out-of-state care. First, beneficiaries in traditional Medicare can obtain care
Search
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks