行政院國家科學委員會專題研究計畫 成果報告
醫院權屬、醫療品質與成本和外溢效果
計畫類別： 個別型計畫
計畫編號： NSC93-2415-H-004-013-
執行期間： 93 年 08 月 01 日至 94 年 07 月 31 日
執行單位： 國立政治大學財政系
計畫主持人： 連賢明
報告類型： 精簡報告
處理方式： 本計畫可公開查詢
中 華 民 國 94 年 10 月 24 日
Abstract:
Recently there have been increasing supports of expanding insurance coverage through public programs or universal health insurances (UHI). The adoption of UHI, however, inevitably reduces the need of uncompensated care, one important reason to grant subsidies to nonprofit hospitals through favorable tax treatments. To what extent nonprofit hospitals should be subsidized in the universal insurance regime is thus in question. This paper compares program cost and quality of new stroke cases across hospitals of different ownership status in Taiwan, the latest country adopting UHI. Our results show that stroke patients admitted to nonprofit hospitals receive better quality in terms of survival rates than those admitted to for-profit hospitals. With cost measured in term of medical expenditure, we found no differences by hospital ownership. Furthermore, we find that for-profit hospitals adopt different treatment styles----they have shorter length of stay, and incur higher per-day expenditure than other hospitals. The result indicates that the ownership status may result in various treatment styles which subsequently affect program cost and quality.
摘要 近來擴大政府公共保險和全民保險的議題在美國逐漸受到重視。全民保險的實施不可 避免的會衝擊到慈善醫療的需求，一個政府提供法人醫院租稅減免的重要原因。究竟政府 在全民保險下是否應繼續對法人醫院提供租稅優惠呢？這篇文章比較台灣不同權屬醫院 （法人、私立、財團法人）中風治療的成本和品質。我們的結果顯示在治療費用上，不同 權屬醫院並沒有顯著差別；然而，財團法人醫院相較於私立醫院有較低的死亡率。進一步 的分析發現，法人醫院的住院天數較長，但每天的住院費率較低。這顯示權屬別對治療方 式有相當影響。 關鍵詞： 權屬別, 全民健保, 醫療成果, 醫療費用
報告內容 I. Introduction
Recently there have been increasing supports of reducing the number of the uninsured, either through the expansion of public insurance programs or the adoption of universal health insurance (UHI). In a series of reports, for instance, Institute of Medicine concludes that the lack of insurance results in negative impacts to the uninsured and their families, and urges the government to seek to achieve universal
insurance by 2010 (IOM, 2004). Indeed, U. S. is the only developed county without universal insurance. More than 43 millions are reported uninsured throughout 2002, of which eight out of ten are members of working families (IOM, 2004).
Implementing UHI inevitably affects the existing health care system, however. Currently, substantial subsidies are made to nonprofit hospitals through favorable tax treatments.1 It is argued that the presence of nonprofits not only provides the charity care for the uninsured, but also the quality care for all since nonprofits are less motivated to skimp on patients’ needs or take advantage of uninformed patients due to the non-distributional constraint (Arrow, 1963; Weisbrod, 1988; Hirth, 1999; For a comprehensive survey, see Sloan, 2000). Once UHI is adopted, however, the need for uncompensated care is greatly reduced, and the risk of quality distortion due to price competition is largely mitigated. To what extent the special treatment on nonprofits should be kept under UHI is therefore an important question. Unfortunately, studies of hospital ownership are largely restricted to U. S. experiences. Even for studies drawing from countries adopting UHI, their hospital industries are often comprised predominantly of public or nonprofit hospitals.2 Without being able to compare the hospital performance by ownership under UHI, it is
therefore difficult to properly address this question.
In this article we consider how the ownership status affects the program cost and quality of stroke treatment in Taiwan. We use Taiwan data because Taiwan implemented National Health Insurance (NHI) in 1995, the latest country adopting UHI. In addition, as shown below, each ownership status (nonprofit, public, and for-profit) plays a significant role in Taiwan’s hospital market. We analyze the stroke disease---hemorrhage strokes as well as ischemic strokes---in part because of its importance and cost----stroke is a leading cause for hospital admission, and inpatient expenditures in Taiwan----but also because the severity of the illness is unlikely to be known by an individual prior to its occurrence.3 Therefore, the distribution of disease severity is likely to be independent of an individual’s residence location, which, as
demonstrated below, is essential for instrumental variable estimation in the analysis.
We use longitudinal hospital claims of NHI receipts diagnosed as new stroke cases during 1996 and 2002, matched with comprehensive hospital data over the same period. Since NHI covers almost all stroke related treatment, the hospital’s charged expenditure is very close to the actual treatment cost. Therefore, our cost measure is the health expenditure spent during the treatment episode. Because stroke is the second leading cause of death in Taiwan, we measure the quality of care by the incidence of death within 1, 6, and 12 months after discharge. Conceptually, we could compare medical expenditure and time of death after discharge among new stroke patients admitted to hospitals of different ownership. In practice, however, such a comparison may be misleading since hospitals may be selected based on various
differences related to ownership status. To overcome this selection issue, we employ two alternative methods: propensity score-matching (PSM) method to correct for sample selection bias due to observable
1
Those treatments include the exemption of state and federal corporate income tax, the exemption of state and federal property tax, as well as the access of tax-exempt bonds and the availability of charity deduction.
2 In Canada, 98 percent of hospitals are public facilities. In Germany, public facilities own 51 percent of beds, while
private not-for-profit hospitals own 35 percent. In Netherlands, private for-profit hospitals are even prohibited by law. For more discussions on the proportion of hospitals by different ownership, see Sloan (2000) Sec. 1.
3 Inpatient expenditure for cerebrovascular disease (stroke) is 4.78% in 2003, next to heart disease (7.18%) and
differences, and instrumental variable (IV) method to correct for unobservable differences between the control and treatment groups.
Adjusting for endogeneity, we find that new stroke cases admitted to nonprofits receive lower mortality rates than those admitted to other hospitals. On average, 1, 6, and 12 month survival rates of nonprofits are at least 3% higher than that of for-profit hospitals. By contrast, the quality difference is less obvious between public and for-profit hospitals. In addition, we find no discernible differences in the medical expenditure by hospital ownership; in other words, the lower mortality is not associated with the higher expenditure. Furthermore, we find evidence showing ownership status affects treatment style: patients admitted to for-profit hospitals experience shorter length of stay, and incur higher expenditure per day. Given that nonprofits supply the better care at the cost no significantly higher than for-profit hospitals, our results indicate that the special treatments toward nonprofits could still be supported even under UHI.
This reminder of this article is organized as follows. We continue in Section 2 with a discussion on theories of nonprofit organization and empirical studies of nonprofit hospitals. Section 3 briefly introduces NHI and the hospital market in Taiwan. Section 4 describes the data and the sample we analyze, and Section 5 shows the econometric strategy we adopt. In Section 6, we present results of OLS, IV and PSM estimations. We discuss the robustness checks and mechanisms affecting program cost and outcome in Section 7. We conclude in Section 8.
II. Theories and Empirical Studies on Ownership
A. Theories on the Impact of Ownership
Starting with Arrow (1963), one line of theoretical studies emphasized the importance of the asymmetric information between the consumer and producer in explaining the existence of nonprofits. According to Arrow, the dominance of private nonprofit hospitals is in part due to lower transactions costs between hospitals and patients. Because the hospital care usually involves complex technology and complicated medical knowledge, patients have difficulties in judging the quality of care in advance. Nondistribution constraint embedded in nonprofit organizations---owners of for-profits have legitimate residual claimants while controllers of nonprofits are prohibited from distributing earnings---reduces the incentives to engage opportunistic behaviors, and fill a market niche where consumers are not able to judge effectively the quality of received care.4
Another line of research postulates various sets of objectives such as quality of care (Newhouse, 1970) or charity care (Frank and Salkever, 1991) for the nonprofit firms assuming that the equity capital has been obtained through philanthropic donations, debt, and retained earnings. Newhouse (1970) emphasizes that the quality/quantity-maximizing nonprofit firms will result in productive inefficiency because nonprofit firms set average revenue equal to average cost, rather than marginal revenue equal to marginal cost.
Other studies (e. g. Weisbrod, 1988; Gentry and Penrod, 2000; Lakdawalla and Philipson, 1998) attempt to provide a theoretical justification for the proliferation of nonprofit firms by stressing the decision to trade off between tax advantage allowed under nonprofit status, and unconstrained profit
4 Hansmann (1980) first generalizes this notion and term it the “contract failure theory”. Easley and O’Hara (1983)
formally incorporate the idea into the contract failure theory where they argue that for-profit organizations will produce no output and the manager will simply pocket the whole purchase price in situations where the outputs are unobserved. Consequently, the nondistribution constraint will be valuable and nonprofits may be superior to for-profits. Similar arguments have been made by Hart et al. (1997) and Glaeser and Shleifer (2001) in which they argue that there are so many possible contingencies ex ante that it is impossible to anticipate all of them when forming a contract. Thus, for-profit firms are more likely to exploit non-contractible quality by cutting non-contractible cost to maximize return under incomplete contracts. The nodistribution constraint embedded in nonprofit organizations, on the other hand, softens this incentive and assures higher quality of care.
making under for-profit status. Hansmann (1980, 1987, 1996) further argued that the availability of extensive privileges, including various tax exemptions and charitable deduction, to nonprofits is neither necessary nor sufficient for the initial development of nonprofits. In any case, they all agree that the various subsidies to nonprofits may influence the overall extent of nonprofit activity thereafter. B. Previous Empirical Studies
The empirical literature on the impact of hospital ownership predominately focuses on the health care market in the United States. Sloan et al. (2001) suggest that payments on behalf of Medicare patients admitted to for-profit hospitals during the first 6 months following a hospitalization due to stroke, hip fracture, coronary heart disease or congestive heart failure were higher. Their findings are consistent with Granneman et al. (1986) that for-profit hospitals had costs higher than not-for profit hospitals and
Silverman et al. (1999) that total per capita Medicare spending was higher in markets dominated by for-profit hospitals, but contradicts with earlier results that indicate nonfor-profit hospitals have higher average expenses per day than for-profits but similar average expenses per patient (i.e. Institute of Medicine 1986; Becker and Sloan, 1985).
On the quality front, several studies have suggested the similarities between for-profits and nonprofit hospitals in the health outcomes (Keeler et al. 1992; Sloan et al. 2001; Ettner and Hermann 2001).
McClellan and Staiger (2000) find that for-profits have higher mortality among elderly patients with heart disease, but they explain that the difference could be confounded by the market characteristics.
Nevertheless, their results are consistent with some recent studies on hospital conversion (e.g.: Shen 2002; Picone et al. 2002) which found adverse health outcomes increases after conversion from nonprofit to for-profit status. Overall, the empirical evidence has yielded mixed findings regarding the ownership effect on cost and quality of care. See Sloan (2000) for a detailed review on the impact of ownership status.
III Conclusion
Weisbrod (1988) argues that the presence of nonprofit organization is due to the inability of government to meet the demand of public good (e.g. health care to the indigent) in populations with heterogeneous preferences for public services. As a result, nonprofit organization offers an alternative to the government for providing collective goods and hence should be encouraged through favorable tax treatments. The need for uncompensated care, however, will be considerably reduced after the adoption of UHI. To what extent these subsidies should be granted toward nonprofit hospitals in the universal
insurance regime is thus an important empirical question.
In this study we examine the impact of ownership status on the cost and quality of stroke treatment in Taiwan, the latest country adopting UHI. After adjusting for endogeneity of hospital ownership, we find that stroke patients treated at nonprofit hospitals receive better quality of care in terms of survival rates than admitted to for-profits. On average, the 1, 6, 12 month mortality rates for nonprofits is at least 3% lower than for-profits. Such an increase of quality, however, is not associated with an increase of the treatment cost---no differences in medical expenditure are observed between nonprofits and for-profits. Nonetheless, we find evidence showing the ownership status affecting the treatment style: for profits are more likely to have shorter hospital stays, and incur higher expenditure per day.
A key question is how general our results are. For two reasons we caution the readers planning to extend the results. First, almost every country has some unique features in its health system. Therefore, it is probably difficult to simulate the results unless the differences in health system are properly accounted for. Additionally, as seen earlier, for-profits in Taiwan differ substantially than other hospitals in sizes and accreditations. Such differences may be induced by the medical law prohibiting for-profit companies or
enterprises from owning hospitals. Although our results are robust to different samples and estimation methods, we warn that the results can still be biased due to this particular restriction.
On the other hand, there are reasons to believe the findings may be more general as they look. Unless treated timely, the consequences of all strokes, particularly hemorrhage strokes, are serious. Therefore, the observed cost and quality difference should be at the low end since there should be little differences between providers’ treatment practices based on hospital ownership. Additionally, there have been relatively few developments in stroke treatment. Hence, treatment discrepancies among hospitals of different ownership status should be smaller than other areas that experience rapid innovations (e.g. heart diseases), particularly when public or nonprofits are the first peers to adopt such new technologies.
Another key question is to what extent the cost and quality differences are influenced by the adoption of UHI. Our data show no signs of convergence in these differences over years after implementing UHI.6 As a result, we suspect that UHI causes a long-lasting phenomenon. Nonetheless, it remains possible that UHI produces a one-time effect. A more careful analysis combining the data spanning before and after the implementation of UHI is necessary to determine its true impact.
Our main contribution is to provide the empirical evidence on the cost and quality differences by hospital ownership in the UHI regime. Most studies on hospital ownership are restricted to U. S., whose experiences are too unique to generalize. One important difference, for instance, is that UHI is unavailable in the states. To our knowledge our paper is the first study that analyzes the impact of hospital ownership in a country adopting UHI. Moreover, we believe the findings are also useful in testing the ownership theory. Because UHI greatly reduces the need for uncompensated care, the observed cost and quality differences, if they exist, are likely to attribute to reasons other than the provision of public good. In fact, our findings---nonprofits supply better quality care, and adopt treatment styles similar to public hospitals---are to a large extent consistent with asymmetric information hypothesis. Therefore, there might still be reasons granting subsidies to nonprofit hospitals even under UHI regime.
6 For instance, the 12 month mortality difference between for-profits and non-profits is approximately 2% over the
References
Angrist, J.D., G. Imbens, and D.B. Rubin (1996), “Identification of Causal Effects Using Instrumental Variables”, Journal of American Statistical Association, 91(434): 468-472.
Arrow, K. J. (1963), “Uncertainty and the Welfare Economics of Medical Care”, American Economic
Review, 53(5): 941-973.
Becker, E. R., and F. A. Sloan (1985), “Hospital Onwership and Preference”, Economic Inquiry, 23(1): 21-36.
Becker, S. O. and A. Ichino (2002), “Estimation of Average Treatment Effects Based on Propensity Scores”, Stata Journal, 2(4): 358-377.
Cheng T. M. (2003) “Taiwan's New National Health Insurance Program: Genesis and Experience So Far”
Health Affairs, 22(3): 61-76
Lien, H. M, S. Y. Chou, and J. T. Liu (2004), “Hospital Competition under Universal Health Insurance: the Stroke Treatment in Taiwan”, Mimeo.
Davidson, R and J. M. Mackinnon (1993), Estimation and Inference in Econometrics, Oxford.
Dehejia, R. H. and S. Wahba (2002), “Propensity Score-Matching Methods for Nonexperimental Causal Studies”, Review of Economics and Statistics, 84(1): 151-161.
Easley, D. and M. O’Hara (1983), “The Economic Role of the Nonprofit Firm”, Bell Journal of
Economics, 14(2): 531-538.
Ettner, S. L. and R. C. Hermann, (2001) “The Role of Profit Status under Imperfect Information: Evidence From the Treatment Patterns of Elderly Medicare Beneficiaries Hospitalized for Psychiatric
Diagnoses”, Journal of Health Economics, 20(1): 23-49.
Frank, R. G., and D. S. Salkever (1991), “The Supply of Charity Services by Nonprofit Hopsitals: Motives and Market Structure”, Rand Journal of Economics, 22(3): 430-445.
Gentry, W. M. and J. R. Penrod (2000), “The Tax Benefits of Not-For-Profit Hospitals”, in David Cutler, ed., The Changing Hospital Industry: Comparing Not-for-Profit and For-Profit Institutions, University of Chicago Press, 285-324.
Glaeser, E. L. and A. Shleifer (2001), “Not-for-Profit Entrepreneurs”, Journal of Public Economics, 81(1): 99-115.
Grabowski D. C. and R. Hirth (2003) “Competitive Spillover across Nonprofit and For-profit Nursing Home”, Journal of Health Economics, 22: 1-22
Grannemann, T.W., R.S. Brown and M.V. Pauly (1986), “Estimating Hospital Costs: A Multiple-Output Analysis”, Journal of Health Economics, 5(2): 107-128.
Hansmann, H. B. (1980), “The Role of Nonprofit Enterprise”, Yale Law Journal, 89(5): 835-901. Hansmann, H. B. (1987), “Economic Theories of Nonprofit Organization,” in W.W. Powell, ed., The
Nonprofit Sector, New Haven and London: Yale University Press, 27-42.
Hansmann, H. B. (1996), The Ownership of Enterprise, Cambridge, Massachusetts and London, England: The Belknap Press and Harvard University Press.
Hart, O., A. Shleifer and R.W. Vishny (1997), “The Proper Scope of Government: Theory and An Application to Prisons”, Quarterly Journal of Economics, 112(4): 1127-1162.
Hirth, R. A (1999) “Consumer information and competition between nonprofit and for-profit nursing homes, Journal of Health Economics, 18(2): 219-240
Imbens, G. (2000), “The Role of the Propensity Score in Estimating Dose-Response Functions”,
Biometrika, 87(3): 706-710.
Imbens, G. and J. D. Angrist (1994), “Identification of Local Average Treatment Effects”, Econometrica, 62(2): 467-475.
Institute of Medicine (1986), For-Profit Enterprise in Health Care, Washington: National Academy Press. Institute of Medicine (2004), Insuring American’s Health: Principals and Recommendations, Washington:
National Academy Press.
Keeler, E.B., L.V. Rubinstein, K.L. Kahn, D. Draper, E.R. Harrison, M.J. McGinty, W.H. Rogers and R.H. Brook (1992), “Hospital Characteristics and Quality of Care”, Journal of the American Medical
Association, 268(13): 1709-1714.
Lakdawalla, D. and T. Philipson (1998), “Nonprofit Production and Competition”, NBER Working Paper 6377.
McClellan, M., B.J. McNeil, and J.P. Newhouse (1994), “Does More Intensive Treatment of Acute Myocardial Infarction Reduce Mortality?” Journal of the American Medical Association, 270: 1832-1836.
McClellan, M. and J. P. Newhouse (1997), “The Marginal Cost-Effectiveness of Medical Technology: A Panel Instrumental-Variables Approach”, Journal of Econometrics, 77: 39-64.
McClellan, M. and D. Staiger (2000), “Comparing Hospital Quality at For-Profit and Not-for Profit Hospitals”, in D. M. Cutler, ed., The Changing Hospital Industry – Comparing Not-for-Profit and
For-Profit Institutions, Chicago: University of Chicago Press, 93-112.
Mitka M. (2003), "Tensions Remain over tPA for Stroke", Journal of the American Medical Association, 289: 1363-1364.
Newhouse, J. P. (1970), “Toward a Theory of Nonprofit Institutions: An Economic Model of a Hospital”,
American Economic Review, 60(1): 64-74.
Picone, G.A., S-Y Chou and F.A., Sloan (2002), “Are For-Profit Hospital Conversions Harmful to Patients and to Medicare?” Rand Journal of Economics, 33(3): 507-523.
Rosenbaum, P. R. and D. B. Rubin (1983), “The Central Role of the Propensity Score in Observational Studies for Causal Effects”, Biometrika, 70(1): 41-55.
Silverman, E. M., et. al. (1999) “The Association between For-Profit Hospital Ownership and Increased Medicare Spending”, New England Journal of Medicine, 318: 1100-1107.
Shen, Y. C. (2002), “The Effect of Hospital Ownership Choice on Patient Outcomes after Treatment for Acute Myocardial Infarction”, Journal of Health Economics, 21(5): 901-922.
Shih Y. T., Y. T. Hung, H. Y. Chang, J. P. Liu, H. S. Lin, M. C. Chang, F. C. Chang, and C. A. Hsiung (2004) “The Design, Contents, Operation and the Characteristics of the Respondents of the 2001 National Health Interview Survey in Taiwan.” Taiwan Journal Public Health, 22: 419-430. Sloan, F.A., G.A. Picone, D.H. Taylor Jr., and S-Y Chou (2001), “Hospital Ownership and Cost and
Quality of Care: Is There a Dime’s Worth of Difference?” Journal of Health Economics, 20(1): 1-21. Sloan, Frank (2000), “Not-For-Profit Ownership and Hospital Behavior,” in Culyer and Newhouse (eds)
Handbook of Health Economics, North Holland.
計畫成果自評部份，
