International Journal for Quality in Health Care, 2015, 27(3), 214–221 doi: 10.1093/intqhc/mzv020 Advance Access Publication Date: 15 April 2015 Article
Article
What affects local community hospitals’ survival in turbulent times? HUNG-CHE CHIANG1 and SHIOW-ING WANG2 1
Division of Environmental Health and Occupational Medicine, National Health Research Institutes, Zhunan Town, Miaoli County, Taiwan, R.O.C, and 2National Environmental Health Research Center, National Health Research Institutes, Zhunan Town, Miaoli County, Taiwan, R.O.C Address reprint requests to: Shiow-Ing Wang, 35, Keyan Road, Zhunan Town, Miaoli County, Taiwan, R.O.C. Tel: +886-37-246166; Fax:+886-37-584-730; E-mail: [email protected] Accepted 13 March 2015
Abstract Downloaded from by guest on November 17, 2015
Objectives: Hospital closures became a prevalent phenomenon in Taiwan after the implementation of a national health insurance program. A wide range of causes contributes to the viability of hospitals, but little is known about the situation under universal coverage health systems. The purpose of present study is to recognize the factors that may contribute to hospital survival under the universal coverage health system. Study design: This is a retrospective case–control study. Local community hospitals that contracted with the Bureau of National Health Insurance in 1998 and remained open during the period 1998– 2011 are the designated cases. Controls are local community hospitals that closed during the same period. Methods: Using longitudinal representative health claim data, 209 local community hospitals that closed during 1998–2011 were compared with 165 that remained open. Variables related to institutional characteristics, degree of competition, characteristics of patients and financial performance were analyzed by logistic regression models. Results: Hospitals’ survival was positively related to specialty hospital, the number of respiratory care beds, the physician to population ratio, the number of clinics in the same region, a highly competitive market and the occupancy rate of elderly patients in the hospital. Teaching hospitals, investor-owned hospitals, the provision of obstetrics services or home care, and the number of medical centers or other local community hospitals may jeopardize the chance of survival. Conclusions: Factors-enhanced local hospitals to survive under the universal coverage health system have been identified. Hospital managers could manipulate these findings and adapt strategies for subsistence. Key words: hospital survival strategies, hospital closure, local community hospitals, universal coverage health system
Introduction In 1995, Taiwan launched its National Health Insurance (NHI) program. Its major accomplishments include universal coverage, comprehensive and uniform benefits, freedom of choice, affordability and a high quality of care [1–5]. Details of the program have been presented elsewhere [6, 7]. However, since the introduction of NHI, there has
been a marked decline in the number of hospitals. In 1995, there were 787 hospitals in Taiwan. By 2011, the number had declined to 507 (Fig. 1), a drop of 35.58%. The decline is particularly steep in the case of local community hospitals. In 1995, local community hospitals accounted for 49.24% of insured beds and 34.55% of outpatient medical benefit claims. In 2011, 29.88% insured beds and 10.56%
© The Author 2015. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved
214
Hospital closure in Taiwan • Universal Healthcare Coverage System
215
Figure 1 Number of hospitals in Taiwan by year, 1986–2011. Source of data: Ministry of Health and Welfare, R.O.C. http://www.mohw.gov.tw/EN/Ministry/Statistic. aspx?f_list_no=474&fod_list_no=3486 (25 August 2013, date last accessed).
Methods Data sources We use data from 1998 through 2011 from the National Health Insurance Research Database (NHIRD) to survey the factors contributing to local community hospital survival. The NHIRD, a valuable population-based database, contains substantial information on the utilization of enrollees (99% coverage rate) and a longitudinal time frame for cohort design. In addition, the establishment, expansion and closure of hospitals are restricted by ministry regulations and by contractual obligations with the Bureau of National Health Insurance (BNHI). Thus, hospital-related information is accurate and timely.
The Research Ethics Committee of the National Health Research Institutes approved this study (No. EC1011003-E) and waived the requirement of informed consent, because the dataset is devoid of identifiable personal information.
Study design This is a retrospective case–control study. We singled out local community hospitals as our study subjects, because they are a key to the goal of creating equal opportunities for health for all inhabitants. In addition, hospitals of this type are especially vulnerable to competitive pressures due to their small size, aging facilities and limited services. Local community hospitals that contracted with the BNHI in 1998 and remained open during the period 1998–2011 are the designated cases. Controls are local community hospitals that also contracted with the BNHI in 1998 but which closed during the same period. To reduce interference, hospitals that experienced a merge, changed ownership or were renamed are not included for analysis. Hospitals located on Taiwan’s remote islands were also excluded due to their different competitive environments and reimbursement policies. Fig. 2 shows the selection process. We gathered related information on the control group 1 year prior to closure; for those hospitals that remained open, data on each variable were gathered in 2010. Based on literature findings and the unique situation in Taiwan, four dimensions—institutional characteristics, degree of competition, characteristics of patients and financial performance—were examined comprehensively in the present study. The variables included in each dimension were defined as follows. Institutional characteristics Type of hospital, ownership, status of teaching hospital, bed size and services provided by the hospital were all included in evaluating a hospital’s institutional characteristics. Degree of competition The number of hospitals in a given region and the ratio of physicians to the area’s population may serve as proxies for degree of competition.
Downloaded from by guest on November 17, 2015
of outpatient medical benefit claims were handled by local community hospitals [8]. Studies have identified numerous factors correlated with hospital viability and hospital closure; these factors can be grouped into four dimensions: financial performance [9–13], institutional characteristics [12–21], degree of competition [10–12, 16–18, 22, 23] and characteristics of patients [9, 13, 24]. Because of Taiwan’s unique medical environment, however, it would be premature to conclude that those factors would also apply here. For example, Chinese herbal medicine, complementary medicine and alternative medicines are commonly used in Taiwan [25]. Additionally, patients in Taiwan prefer going to large hospitals, even if they suffer from only minor illnesses [7, 26]. And some patients deem that they are not cured without prescription medicines [27]. Local hospitals serve as anchors for their communities. Hospital closures may compromise accessibility, put patient health and even lives at risk, lessen equality in the health system and lead to misallocations of health resources—all of which are against the primary goal of the universal coverage system. Thus, the issue of hospital viability has important implications for access to care and for social justice. Although a wide range of factors contributing to the viability of hospitals have been examined, little is known about the contribution of these factors under a universal coverage health system. Thus, the present study attempts to identify the factors that contribute to hospital survival under the universal coverage health system in Taiwan.
216
Chiang and Wang
Downloaded from by guest on November 17, 2015
Figure 2 Flowchart for study subject selection.
We also employ the Herfindahl–Hirschman Index (HHI), which is widely used to represent competition. We divide Taiwan’s medical market into three levels [28]: strong competition (HHI < 1000), moderate competition (1000 ≤ HHI < 1800) and low competition (HHI ≥ 1800). Data on the physician to population ratios and the number of other facilities were compiled from a governmental annual report [29].
Financial performance The variables standing for financial performance include outpatient claims, inpatient claims, outpatient benefit claims and inpatient benefit claims.
Statistical analysis Characteristics of patients Patients’ age, sex, place of residence and status of chronic disease were calculated from the NHIRD.
We first examined the differences in influencing factors between the case and control groups by χ 2 tests and t-tests. Additionally, after testing for collinearity, logistic regression was performed to explore
Hospital closure in Taiwan • Universal Healthcare Coverage System whether institutional characteristics, degrees of competition, characteristics of patients and financial performance were associated with survival of hospitals. We used SPSS Statistics to conduct analyses (Version 20.0, IBM Corp., Armonk, NY, USA). A P-value of 20 000 visit hospitals over 100 times yearly. In addition, elderly patients account for one-third of the total NHI expenditure [30, 31]. Thus, a higher occupancy rate of elderly patients may indicate higher financial stability. For several decades, the medical industry has moved toward specialization rather than generalization. Our findings support this observation for Taiwan. From the perspective of hospital management, the main benefit of specialization is efficiency. Focus on a specialty can allow the clustering of facilities and efficient allocation of resources in ways leading to the acquisition of advanced technology and the
provision of high-quality services, which in turn boost the chances of survival in turbulent times. In addition, patients believe that specialists can accumulate valuable experience through their frequent treatment of specific diseases in specialty hospitals. Without a referral system, patients can consult any specialist they desire. However, this may impede physicians from taking a holistic approach when treating patients. Furthermore, if few doctors show an interest in some specialties, shortages may result in certain fields. For example, specialists in internal medicine, surgery, obstetrics and gynecology, and pediatrics are in increasingly short supply in Taiwan [32]. The association between the number of respiratory care beds and the viability of hospitals may reveal the impact of health policy and regulations. Medical expenditures on patients with prolonged ventilation consumed around 3–5% of the total NHI expenditure. Thus, in 2000, Taiwan implemented an integrated care program for ventilatordependent patients. This program encourages local community hospitals to integrate with other medical centers or regional hospitals to establish wards for the care of relatively stable ventilator-dependent patients [7]. Integrated cooperation will allow local community hospitals to have steady revenue, enhancing their chance of survival. The aspect of competition gave some apparently paradoxical results. The likelihood of success was higher in areas with higher physician to population ratios, highly competitive markets or more clinics in the same region. On the other hand, the existence of medical centers or other local community hospitals in the same medical region may threaten the viability of hospital. The phenomenon may indicate that the number of physicians had not yet reached the threshold of competition. In 2010, Taiwan had 24.1 physicians per 10 000 population, which was higher than the ratios in Singapore (19) and Canada (21); similar to that of the United States (24); and lower than the ratios in Britain (27), Germany (37) and Switzerland (41) [33]. However, the issue of how many physicians Taiwan should have under its universal coverage health care system is beyond the range of the present study. In addition, the result may also imply that clinics cannot substitute for local community hospitals. Thus, a higher physician to population ratio and having more clinics in the same region may not necessary increase the competition a local community hospital faces. We also found that medical centers are the main rivals of local community hospitals for resources and patients, while regional hospitals seem have but little impact on them. Our findings demonstrate that a higher occupancy rate of male patients, the provision of obstetrics or home care, and teaching hospitals were negatively associated with local community hospitals’ survival. The results are in line with our expectations. The total fertility rate of Taiwan dropped to 0.9 in 2010, the lowest rate in the world [34]. Because of this low fertility, providing obstetrics services may be costly (because of labor costs to meet the need for 24-h standby services), provide a low return on investment (low utilization) and may raise the risk of medical malpractice suits (from eager parents). Several studies suggest that men have lower medical care service utilization and lower associated expenditures than women [35, 36]. Thus, a higher occupancy rate of male patients may not help hospitals stabilize revenue. Taiwanese may be resistant to teaching hospitals for seeking medical advice due to worries about being a guinea pig or about having to express private medical matters with students or residents (even though those students and residents will be doctors in the future). Financial factors and ownership of hospitals have relatively little influence on a hospital’s viability in Taiwan, compared with hospitals in other countries [10, 13, 37]. We argue that this might be the consequence of the universal coverage health system. NHI pays providers according to a uniform fee schedule, which means that the revenue does not vary across ownership but is affected by quantity. However,
Downloaded from by guest on November 17, 2015
Hospitals that closed and those that remained open had significant differences (Table 1). A higher proportion of hospitals that remained open were general or specialty hospitals, were public or had not-forprofit ownership, were providers of comprehensive services (except home care) and possessed more beds than controls (P < 0.05). Remaining open hospitals were located in medical regions with a higher physician to population ratio, had more additional facilities and were less often situated in a highly competitive market (P < 0.05) than controls. A comparison of their patients revealed that remaining open hospitals had proportionately more patients who were elderly, male, lived in the same medical region as the hospital and suffered from chronic diseases (P < 0.001). Unsurprisingly, hospitals that remained open also had better financial performance (P < 0.001).
217
218
Chiang and Wang
Table 1 Characteristics of the hospitals (n = 374) Remaining open hospitals N = 165 n (%)/mean ± SD
χ 2/t-test
21 (10.0) 31 (14.8) 157 (75.1)
31 (18.8) 27 (16.4) 107 (64.8)
6.583*
11 (5.3) 198 (94.7)
15 (9.1) 150 (90.9)
2.089
13 (6.2) 7 (3.3) 189 (90.4)
25 (15.2) 17 (10.3) 123 (74.5)
16.976***
153 (75.7) 39 (19.3) 139 (68.8) 75 (37.1) 59 (29.2) 135 (64.6)
165 (100) 90 (54.5) 154 (93.3) 88 (53.3) 31 (18.8) 117 (70.9)
46.192*** 49.473*** 33.923*** 9.660** 5.327* 1.673
75.18 ± 104.1 9.72 ± 47.06 4.30 ± 13.40
184.19 ± 202.4 55.33 ± 226.7 23.11 ± 33.35
−6.629*** −2.775** −7.295***
8.16 ± 5.25
10.50 ± 5.49
−4.191***
1.63 ± 2.62 4.38 ± 6.19 40.93 ± 41.95 1539.5 ± 2243.8
2.41 ± 3.69 6.46 ± 7.79 54.05 ± 54.43 2615.1 ± 3656.2
−2.402* −2.874** −2.633** −3.500**
171 (81.8) 24 (11.5) 14 (6.7)
116 (70.3) 36 (21.8) 13 (7.9)
7.910*
15 325 (14.8) 76 929 (74.1) 11 613 (11.2)
12 775 (10.5) 85 516 (70.1) 23 733 (19.4)
3403.763***
44 717 (43.1)
55 756 (45.7)
158.408***
27 350 (26.3)
46 925 (38.5)
3736.900***
44 021 (42.4) 71 820 (69.1)
51 364 (42.1) 86 580 (71.0)
1.917 87.459***
52 686 ± 56 132 4380 ± 6417 1097 ± 1566 2922 ± 5162
82 999 ± 83 085 9560 ± 11 369 1902 ± 2474 7724 ± 10 129
−4.199*** −5.561*** −3.755*** −5.820***
a General hospitals are those with >100 beds and which are engaged in medicine, surgery, pediatrics, obstetrics and gynecology, anesthesiology, radiology and other medical services. Specialty hospitals include specialist hospitals, psychiatric hospitals and special hospitals. b Market share calculation of Herfindahl–Hirschman Index is based on outpatient benefit claims. c The National Health Insurance Administration has listed 98 ailments as chronic illnesses. http://www.nhi.gov.tw/webdata/webdata.aspx? menu=18&menu_id=683&webdata_id=444. *P < 0.05. **P < 0.01. ***P < 0.001.
Downloaded from by guest on November 17, 2015
Institutional characteristics Type of hospitala General hospital Specialty hospital Others Teaching hospital Yes No Ownership Public hospitals Not-for-profit Investor owned Diversified services Inpatient services Hemodialysis services Preventive health services Obstetrics services Home care/rehabilitation Emergency services Number of beds Acute beds Chronic beds Respiratory care beds Degree of competition Physician to population ratio ( per 10 000) Other facilities Medical centers Regional hospitals Local community hospitals Clinics Herfindahl–Hirschman Indexb Strong competition Moderate competition Low competition Characteristics of patients Age (y) 0–14 15–64 65+ Sex Male Disease patterns With chronic illnessc Residence In the same township In the same sub-medical region Financial performance Outpatient claims Outpatient benefit claims (10 000 RVUs) Inpatient claims Inpatient benefit claims (10 000 RVUs)
Closed hospitals N = 209 n (%)/mean ± SD
Hospital closure in Taiwan • Universal Healthcare Coverage System
219
Table 2 Logistic regression of hospital survival (n = 374) Variables
Exp (B)
95% CI of Exp (B)
8.999 2.087
14.976 2.678
2.555–87.772 0.704–10.194
6.688
0.075
0.011–0.535
2.503 4.256
0.208 0.179
0.030–1.454 0.035–0.918
0.000 1.157 0.669 5.937 21.756 1.082
5.6E9 1.617 1.826 0.307 0.079 1.592
0.000 0.674–3.881 0.431–7.732 0.119–0.794 0.027–0.229 0.663–3.825
2.369 4.390
1.003 1.019
0.999–1.008 1.001–1.036
28.789
1.476
1.280–1.701
21.727 2.622 9.519 22.334
0.134 0.771 0.960 1.003
0.058–0.312 0.563–1.056 0.936–0.985 1.002–1.004
3.776 4.249
4.981 5.801
0.986–25.156 1.090–30.865
9.664 4.831 0.132
1.071 0.968 1.003
1.026–1.119 0.940–0.996 0.988–1.018
0.696 2.093
1.012 0.977
0.984–1.040 0.948–1.008
2.173 1.331 0.031 0.000
1.000 1.000 1.000 0.000
1.000–1.000 1.000–1.001 1.000–1.000
Cox and Snell R 2 = 0.513; Nagelkerke R 2 = 0.686. Hosmer–Lemeshow goodness-of-fit test: χ 2 = 12.673, df = 8, P = 0.124.
uniform payment does not take into consideration that declines in local community hospitals’ admissions and patient volumes, especially when accompanied by cost increases, put those hospitals in an inferior position.
The impact of universal coverage on hospital closure Creating equal opportunities for health for all inhabitants is an overarching goal for Taiwan’s universal coverage health system. Local community hospitals are a key element to achieving that goal. However, some care-seeking habits of patients and certain elements of the design of the NHI program may be harmful to the survival of local hospitals.
First, there are no gatekeeper constraints. Because enrollees have the freedom to choose their provider without the constraints of a gatekeeper or referral system, this may encourage doctor-shopping [38, 39]. Second, co-payments are so low that patients have no incentive to choose local hospitals over medical centers. Third, the design of the payment system favors medical centers. NHI reimbursement is based on accreditation; thus, there is unequal pay for the same job. Medical centers can also reduce the cost of medicine per unit through bulk purchases, whereas local hospitals usually lack such bargaining power. In addition, hospitals may increase quantity to gain revenue, but the floating point value subsequently decreases. Local community hospitals are more likely to suffer from the impact of the ‘prisoner’s dilemma’ scenario.
Downloaded from by guest on November 17, 2015
Institutional characteristics Type of hospital General hospital (reference) Specialty hospital Others Teaching hospital No (reference) Yes Ownership Public hospitals (reference) Not-for-profit Investor owned Diversified services (reference = does not provide) Inpatient services Hemodialysis services Preventive service Obstetrics services Home care/rehabilitation Emergency services Bed size Acute beds Respiratory care beds Degree of competition Physician to population ratio Other facilities in the same region Medical centers Regional hospitals Local community hospitals Clinics Herfindahl–Hirschman Index Low competition (reference) Moderate competition Strong competition Characteristics of patients Elderly prevalence (%) Male prevalence (%) With chronic illness prevalence (%) Residence In the same township In the same sub-medical region Financial performance Outpatient claims Inpatient claims Inpatient benefit claims Constant
Wald statistic
220
Limitations and strengths
Conclusions Factors-enhanced local hospitals to survive under the universal coverage health system have been identified. The determinants of local community hospital survival under a universal coverage health system were quite different than those elsewhere. Based on the results, we recommend that the managers of local community hospitals adopt several strategies to enhance the chance of survival, such as cooperating with rival medical centers to deliver integrated health care, strengthening their own niches and creating a senior friendly environment. The results found in the study can provide early-warning signs of closure that will allow policy planners and hospital managers to coordinate needed resources and execute programs to prevent adverse effects on the community. To meet community needs when medical resources are scarce, it is necessary to clarify the crucial factors in the sustainability of local community hospitals.
Funding This study was supported by grants from National Health Research Institutes, National Health Institutes, Taiwan (PH-103-SP-06: Building hospital vertical integration models to solve the problem of unbalanced medical system) and used data from the National Health Insurance Research Database, which is overseen by the National Health Research Institutes. The interpretations and conclusions contained herein do not necessarily represent those of the National Health Research Institutes in Taiwan.
References 1. Reinhardt UE. Humbled in Taiwan. BMJ 2008;336:72. 2. Lu JF, Hsiao WC. Does universal health insurance make health care unaffordable? Lessons from Taiwan. Health Affairs 2003;22:77–88. 3. Davis K, Huang AT. Learning from Taiwan: experience with universal health insurance. Ann Intern Med 2008;148:313–4.
4. Williams I. Health care in Taiwan: why can’t the United States learn some lessons? Dissent 2008;55:13–7. 5. Krugman P. Pride, prejudice, insurance. New York Times. 2005. http:// query.nytimes.com/gst/fullpage.html?res=9500E2D9143EF934A35752C1 A9639C8B63 (11 September 2013, date last accessed). 6. Cheng TM. Taiwan’s new national health insurance program: genesis and experience so far. Health Aff 2003;22:61–76. 7. National Health Insurance Administration, Ministry of Health and Welfare. National Health Insurance in Taiwan 2013–2014 annual report. 2014. http://www.nhi.gov.tw/Resource/webdata/26703_1_2013-2014. pdf (18 June 2014, date last accessed). 8. Department of Health, Executive Yuan, R.O.C. Hospital and Clinic statistics. 2011. http://www.mohw.gov.tw/EN/Ministry/Statistic.aspx?f_list_no= 474&fod_list_no=3486 (25 August 2013, date last accessed). 9. McLafferty S. Neighborhood characteristics and hospital closures. A comparison of the public private and voluntary hospital systems. Soc Sci Med 1982;16:1667–74. 10. McNeil D, Williams R. Wide range of causes found for hospital closures. Hospitals 1978;52:76–81. 11. Mayer JD, Kohlenberg ER, Sieferman GE et al. Patterns of rural hospital closure in the United States. Soc Sci Med 1987;24:327–34. 12. Topping S. Hospital closures: what is the story? Health Serv Manage Res 1991;4:148–58. 13. Sager A. Why urban voluntary hospitals close. Health Serv Res 1983;18: 451–75. 14. Rosenstein AH. Hospital closure or survival: formula for success. Health Care Manage Rev 1986;11:29–35. 15. Hsieh HL, Chiang TL. The determinants of hospital closure and opening in Taiwan (in Chinese). Chin J Public Health 1994;13:453–8. 16. Liu JH, Chiang TL. Factors affecting opening and closure of small hospitals in Taiwan (in Chinese). Taiwan J Public Health 2001;20:27–33. 17. Succi MJ, Lee SY, Alexander JA. Effects of market position and competition on rural hospital closures. Health Serv Res 1997;31:679–99. 18. Longo DR, Chase GA. Structural determinants of hospital closure. Med Care 1984;22:388–402. 19. Bonney RS. Hospital survival strategies for the 1980s. Am J Hosp Pharm 1983;40:1483–8. 20. Berry RE Jr. Returns to scale in the production of hospital services. Health Serv Res 1967;2:123–39. 21. Fareed N. Size matters: a meta-analysis on the impact of hospital size on patient mortality. Int J Evid Based Healthc 2012;10:103–11. 22. Mullner RM, Rydman RJ, Whiteis DG et al. Rural community hospitals and factors correlated with their risk of closing. Public Health Rep 1989;104:315–25. 23. Lillie-Blanton M, Felt S, Redmon P et al. Rural and urban hospital closures, 1985–1988: operating and environmental characteristics that affect risk. Inquiry 1992;29:332–44. 24. Hernandez SR, Kaluzny AD. Hospital closure: a review of current and proposed research. Health Serv Res 1983;18:419–36. 25. Shih CC, Su YC, Liao CC et al. Patterns of medical pluralism among adults: results from the 2001 National Health Interview Survey in Taiwan. BMC Health Serv Res 2010;10:191. 26. Hong WH, Cheng SH, Chang RE et al. Changes in the proportion of crossregion admissions in Taiwan, 1985–1995. Chin J Public Health 1998;17: 388–94. 27. Lee SP. Medicine survey (in Chinese). Pharmacist Weekly. 2012. http:// ys34485257.imgshelf.com/1797/1797-1-1.htm (18 June 2014, date last accessed). 28. Lu JR, Hsieh CR. An analysis of the market structure and development of Taiwan’s hospital industry (in Chinese). Taiwan Econ Rev 2003;31: 107–53. 29. Ministry of Health and Welfare. The statistical annual report of medical care institutions status and hospitals utilization. http://www.mohw.gov. tw/EN/Ministry/DM2.aspx?f_list_no=474&fod_list_no=3444 (12 May 2014, date last accessed). 30. Health Promotion Administration, Ministry of Health and Welfare. 2007 Taiwan longitudinal study on aging (TLSA). https://olap.hpa.gov.tw/
Downloaded from by guest on November 17, 2015
Merging or consolidating with other organizations would be a possible survival strategy for local community hospitals. Eliminating these hospitals may jeopardize the chance to depict the whole picture of hospital closure. Moreover, mergers, consolidation and changes in leadership may all alter a hospital’s organizational culture and the management style of hospital executives, which may be important factors in hospital survival. Unfortunately, the study could not obtain the relevant information to examine this. In addition, the financial performance used here may not fully explain the real financial status of hospitals or their cost of operation. Moreover, information on the quality of services and of physician–patient relationships was not available. Data of the case group gathered in the year 2010 may create bias in computing the degree of competition or hospital size. Furthermore, rural and urban counties might have diverse demographic composition and, accordingly, different needs; the difference of viability between urban and rural hospitals will be explored in future work. Despite these limitations, our longitudinal population-based dataset strengthens the credibility of the results. Exhaustive variables were incorporated to reflect the real medical environment that local community hospitals face. We also elaborate on Taiwan’s experiences in the shifts in health care services after the implementation of NHI, and we highlight the impact of the universal coverage health system on hospital closure. The present study provides valuable information to illuminate the prospects for equal, affordable and accessible health care, especially for countries that have or hope to introduce universal coverage health programs.
Chiang and Wang
Hospital closure in Taiwan • Universal Healthcare Coverage System en_US/Search/02_ListSummary.aspx?menu=100000000008&mode=5 2007 (20 May 2014, date last accessed). 31. Bureau of National Health Insurance. The National Health Insurance Statistics. 2010. http://www.nhi.gov.tw/English/webdata/webdata.aspx? menu=11&menu_id=296&WD_ID=296&webdata_id=4010 (20 May 2014, date last accessed). 32. Chiang TL. The difficulty to enroll internal medicine, surgery, obstetrics and gynecology, and pediatrics physician (in Chinese). United Daily News. 2012. http://2020hfa.blog.ntu.edu.tw/2012/11/27/%E5%85%A7%E5% A4%96%E5%A9%A6%E5%85%92%E9%86%AB%E7%99%82%EF %BC%8C%E7%82%BA%E4%BD%95%E5%9B%9B%E5%A4%A7% E7%9A%86%E7%A9%BA%EF%BC%9F/ (23 September 2014, date last accessed). 33. The World Bank. Data of physicians. http://data.worldbank.org/indicator/ SH.MED.PHYS.ZS?order=wbapi_data_value_2010+wbapi_data_value& sort=asc (8 June 2014, date last accessed).
221
34. Department of Household Registration Affairs, Ministry of the Interior. Fertility rates of childbearing age women. 2010. http://sowf.moi.gov.tw/ stat/year/elist.htm (23 September 2014, date last accessed). 35. Bertakis KD, Azari R, Helms LJ et al. Gender differences in the utilization of health care services. J Fam Pract 2000;49:147–52. 36. Lee CH, Chou YJ, Chen LS et al. Utilization of ambulatory chinese medical services under the National Health Insurance in Taiwan. Taiwan J Public Health 2004;23:100–7. 37. Mullner RM, McNeil DL, Andes S. Hospital closure: who will be at risk in the upcoming decade? Healthc Financ Manage 1986;40:42–8. 38. Chiou SJ, Wang SI, Liu CH et al. Outpatient-shopping behavior and survival rates in newly diagnosed cancer patients. Am J Manage Care 2012;18:488–96. 39. Wang MJ, Lin SP. Study on doctor shopping behavior: insight from patients with upper respiratory tract infection in Taiwan. Health Policy 2010;94:61–7.
Downloaded from by guest on November 17, 2015
Article
What affects local community hospitals’ survival in turbulent times? HUNG-CHE CHIANG1 and SHIOW-ING WANG2 1
Division of Environmental Health and Occupational Medicine, National Health Research Institutes, Zhunan Town, Miaoli County, Taiwan, R.O.C, and 2National Environmental Health Research Center, National Health Research Institutes, Zhunan Town, Miaoli County, Taiwan, R.O.C Address reprint requests to: Shiow-Ing Wang, 35, Keyan Road, Zhunan Town, Miaoli County, Taiwan, R.O.C. Tel: +886-37-246166; Fax:+886-37-584-730; E-mail: [email protected] Accepted 13 March 2015
Abstract Downloaded from by guest on November 17, 2015
Objectives: Hospital closures became a prevalent phenomenon in Taiwan after the implementation of a national health insurance program. A wide range of causes contributes to the viability of hospitals, but little is known about the situation under universal coverage health systems. The purpose of present study is to recognize the factors that may contribute to hospital survival under the universal coverage health system. Study design: This is a retrospective case–control study. Local community hospitals that contracted with the Bureau of National Health Insurance in 1998 and remained open during the period 1998– 2011 are the designated cases. Controls are local community hospitals that closed during the same period. Methods: Using longitudinal representative health claim data, 209 local community hospitals that closed during 1998–2011 were compared with 165 that remained open. Variables related to institutional characteristics, degree of competition, characteristics of patients and financial performance were analyzed by logistic regression models. Results: Hospitals’ survival was positively related to specialty hospital, the number of respiratory care beds, the physician to population ratio, the number of clinics in the same region, a highly competitive market and the occupancy rate of elderly patients in the hospital. Teaching hospitals, investor-owned hospitals, the provision of obstetrics services or home care, and the number of medical centers or other local community hospitals may jeopardize the chance of survival. Conclusions: Factors-enhanced local hospitals to survive under the universal coverage health system have been identified. Hospital managers could manipulate these findings and adapt strategies for subsistence. Key words: hospital survival strategies, hospital closure, local community hospitals, universal coverage health system
Introduction In 1995, Taiwan launched its National Health Insurance (NHI) program. Its major accomplishments include universal coverage, comprehensive and uniform benefits, freedom of choice, affordability and a high quality of care [1–5]. Details of the program have been presented elsewhere [6, 7]. However, since the introduction of NHI, there has
been a marked decline in the number of hospitals. In 1995, there were 787 hospitals in Taiwan. By 2011, the number had declined to 507 (Fig. 1), a drop of 35.58%. The decline is particularly steep in the case of local community hospitals. In 1995, local community hospitals accounted for 49.24% of insured beds and 34.55% of outpatient medical benefit claims. In 2011, 29.88% insured beds and 10.56%
© The Author 2015. Published by Oxford University Press in association with the International Society for Quality in Health Care; all rights reserved
214
Hospital closure in Taiwan • Universal Healthcare Coverage System
215
Figure 1 Number of hospitals in Taiwan by year, 1986–2011. Source of data: Ministry of Health and Welfare, R.O.C. http://www.mohw.gov.tw/EN/Ministry/Statistic. aspx?f_list_no=474&fod_list_no=3486 (25 August 2013, date last accessed).
Methods Data sources We use data from 1998 through 2011 from the National Health Insurance Research Database (NHIRD) to survey the factors contributing to local community hospital survival. The NHIRD, a valuable population-based database, contains substantial information on the utilization of enrollees (99% coverage rate) and a longitudinal time frame for cohort design. In addition, the establishment, expansion and closure of hospitals are restricted by ministry regulations and by contractual obligations with the Bureau of National Health Insurance (BNHI). Thus, hospital-related information is accurate and timely.
The Research Ethics Committee of the National Health Research Institutes approved this study (No. EC1011003-E) and waived the requirement of informed consent, because the dataset is devoid of identifiable personal information.
Study design This is a retrospective case–control study. We singled out local community hospitals as our study subjects, because they are a key to the goal of creating equal opportunities for health for all inhabitants. In addition, hospitals of this type are especially vulnerable to competitive pressures due to their small size, aging facilities and limited services. Local community hospitals that contracted with the BNHI in 1998 and remained open during the period 1998–2011 are the designated cases. Controls are local community hospitals that also contracted with the BNHI in 1998 but which closed during the same period. To reduce interference, hospitals that experienced a merge, changed ownership or were renamed are not included for analysis. Hospitals located on Taiwan’s remote islands were also excluded due to their different competitive environments and reimbursement policies. Fig. 2 shows the selection process. We gathered related information on the control group 1 year prior to closure; for those hospitals that remained open, data on each variable were gathered in 2010. Based on literature findings and the unique situation in Taiwan, four dimensions—institutional characteristics, degree of competition, characteristics of patients and financial performance—were examined comprehensively in the present study. The variables included in each dimension were defined as follows. Institutional characteristics Type of hospital, ownership, status of teaching hospital, bed size and services provided by the hospital were all included in evaluating a hospital’s institutional characteristics. Degree of competition The number of hospitals in a given region and the ratio of physicians to the area’s population may serve as proxies for degree of competition.
Downloaded from by guest on November 17, 2015
of outpatient medical benefit claims were handled by local community hospitals [8]. Studies have identified numerous factors correlated with hospital viability and hospital closure; these factors can be grouped into four dimensions: financial performance [9–13], institutional characteristics [12–21], degree of competition [10–12, 16–18, 22, 23] and characteristics of patients [9, 13, 24]. Because of Taiwan’s unique medical environment, however, it would be premature to conclude that those factors would also apply here. For example, Chinese herbal medicine, complementary medicine and alternative medicines are commonly used in Taiwan [25]. Additionally, patients in Taiwan prefer going to large hospitals, even if they suffer from only minor illnesses [7, 26]. And some patients deem that they are not cured without prescription medicines [27]. Local hospitals serve as anchors for their communities. Hospital closures may compromise accessibility, put patient health and even lives at risk, lessen equality in the health system and lead to misallocations of health resources—all of which are against the primary goal of the universal coverage system. Thus, the issue of hospital viability has important implications for access to care and for social justice. Although a wide range of factors contributing to the viability of hospitals have been examined, little is known about the contribution of these factors under a universal coverage health system. Thus, the present study attempts to identify the factors that contribute to hospital survival under the universal coverage health system in Taiwan.
216
Chiang and Wang
Downloaded from by guest on November 17, 2015
Figure 2 Flowchart for study subject selection.
We also employ the Herfindahl–Hirschman Index (HHI), which is widely used to represent competition. We divide Taiwan’s medical market into three levels [28]: strong competition (HHI < 1000), moderate competition (1000 ≤ HHI < 1800) and low competition (HHI ≥ 1800). Data on the physician to population ratios and the number of other facilities were compiled from a governmental annual report [29].
Financial performance The variables standing for financial performance include outpatient claims, inpatient claims, outpatient benefit claims and inpatient benefit claims.
Statistical analysis Characteristics of patients Patients’ age, sex, place of residence and status of chronic disease were calculated from the NHIRD.
We first examined the differences in influencing factors between the case and control groups by χ 2 tests and t-tests. Additionally, after testing for collinearity, logistic regression was performed to explore
Hospital closure in Taiwan • Universal Healthcare Coverage System whether institutional characteristics, degrees of competition, characteristics of patients and financial performance were associated with survival of hospitals. We used SPSS Statistics to conduct analyses (Version 20.0, IBM Corp., Armonk, NY, USA). A P-value of 20 000 visit hospitals over 100 times yearly. In addition, elderly patients account for one-third of the total NHI expenditure [30, 31]. Thus, a higher occupancy rate of elderly patients may indicate higher financial stability. For several decades, the medical industry has moved toward specialization rather than generalization. Our findings support this observation for Taiwan. From the perspective of hospital management, the main benefit of specialization is efficiency. Focus on a specialty can allow the clustering of facilities and efficient allocation of resources in ways leading to the acquisition of advanced technology and the
provision of high-quality services, which in turn boost the chances of survival in turbulent times. In addition, patients believe that specialists can accumulate valuable experience through their frequent treatment of specific diseases in specialty hospitals. Without a referral system, patients can consult any specialist they desire. However, this may impede physicians from taking a holistic approach when treating patients. Furthermore, if few doctors show an interest in some specialties, shortages may result in certain fields. For example, specialists in internal medicine, surgery, obstetrics and gynecology, and pediatrics are in increasingly short supply in Taiwan [32]. The association between the number of respiratory care beds and the viability of hospitals may reveal the impact of health policy and regulations. Medical expenditures on patients with prolonged ventilation consumed around 3–5% of the total NHI expenditure. Thus, in 2000, Taiwan implemented an integrated care program for ventilatordependent patients. This program encourages local community hospitals to integrate with other medical centers or regional hospitals to establish wards for the care of relatively stable ventilator-dependent patients [7]. Integrated cooperation will allow local community hospitals to have steady revenue, enhancing their chance of survival. The aspect of competition gave some apparently paradoxical results. The likelihood of success was higher in areas with higher physician to population ratios, highly competitive markets or more clinics in the same region. On the other hand, the existence of medical centers or other local community hospitals in the same medical region may threaten the viability of hospital. The phenomenon may indicate that the number of physicians had not yet reached the threshold of competition. In 2010, Taiwan had 24.1 physicians per 10 000 population, which was higher than the ratios in Singapore (19) and Canada (21); similar to that of the United States (24); and lower than the ratios in Britain (27), Germany (37) and Switzerland (41) [33]. However, the issue of how many physicians Taiwan should have under its universal coverage health care system is beyond the range of the present study. In addition, the result may also imply that clinics cannot substitute for local community hospitals. Thus, a higher physician to population ratio and having more clinics in the same region may not necessary increase the competition a local community hospital faces. We also found that medical centers are the main rivals of local community hospitals for resources and patients, while regional hospitals seem have but little impact on them. Our findings demonstrate that a higher occupancy rate of male patients, the provision of obstetrics or home care, and teaching hospitals were negatively associated with local community hospitals’ survival. The results are in line with our expectations. The total fertility rate of Taiwan dropped to 0.9 in 2010, the lowest rate in the world [34]. Because of this low fertility, providing obstetrics services may be costly (because of labor costs to meet the need for 24-h standby services), provide a low return on investment (low utilization) and may raise the risk of medical malpractice suits (from eager parents). Several studies suggest that men have lower medical care service utilization and lower associated expenditures than women [35, 36]. Thus, a higher occupancy rate of male patients may not help hospitals stabilize revenue. Taiwanese may be resistant to teaching hospitals for seeking medical advice due to worries about being a guinea pig or about having to express private medical matters with students or residents (even though those students and residents will be doctors in the future). Financial factors and ownership of hospitals have relatively little influence on a hospital’s viability in Taiwan, compared with hospitals in other countries [10, 13, 37]. We argue that this might be the consequence of the universal coverage health system. NHI pays providers according to a uniform fee schedule, which means that the revenue does not vary across ownership but is affected by quantity. However,
Downloaded from by guest on November 17, 2015
Hospitals that closed and those that remained open had significant differences (Table 1). A higher proportion of hospitals that remained open were general or specialty hospitals, were public or had not-forprofit ownership, were providers of comprehensive services (except home care) and possessed more beds than controls (P < 0.05). Remaining open hospitals were located in medical regions with a higher physician to population ratio, had more additional facilities and were less often situated in a highly competitive market (P < 0.05) than controls. A comparison of their patients revealed that remaining open hospitals had proportionately more patients who were elderly, male, lived in the same medical region as the hospital and suffered from chronic diseases (P < 0.001). Unsurprisingly, hospitals that remained open also had better financial performance (P < 0.001).
217
218
Chiang and Wang
Table 1 Characteristics of the hospitals (n = 374) Remaining open hospitals N = 165 n (%)/mean ± SD
χ 2/t-test
21 (10.0) 31 (14.8) 157 (75.1)
31 (18.8) 27 (16.4) 107 (64.8)
6.583*
11 (5.3) 198 (94.7)
15 (9.1) 150 (90.9)
2.089
13 (6.2) 7 (3.3) 189 (90.4)
25 (15.2) 17 (10.3) 123 (74.5)
16.976***
153 (75.7) 39 (19.3) 139 (68.8) 75 (37.1) 59 (29.2) 135 (64.6)
165 (100) 90 (54.5) 154 (93.3) 88 (53.3) 31 (18.8) 117 (70.9)
46.192*** 49.473*** 33.923*** 9.660** 5.327* 1.673
75.18 ± 104.1 9.72 ± 47.06 4.30 ± 13.40
184.19 ± 202.4 55.33 ± 226.7 23.11 ± 33.35
−6.629*** −2.775** −7.295***
8.16 ± 5.25
10.50 ± 5.49
−4.191***
1.63 ± 2.62 4.38 ± 6.19 40.93 ± 41.95 1539.5 ± 2243.8
2.41 ± 3.69 6.46 ± 7.79 54.05 ± 54.43 2615.1 ± 3656.2
−2.402* −2.874** −2.633** −3.500**
171 (81.8) 24 (11.5) 14 (6.7)
116 (70.3) 36 (21.8) 13 (7.9)
7.910*
15 325 (14.8) 76 929 (74.1) 11 613 (11.2)
12 775 (10.5) 85 516 (70.1) 23 733 (19.4)
3403.763***
44 717 (43.1)
55 756 (45.7)
158.408***
27 350 (26.3)
46 925 (38.5)
3736.900***
44 021 (42.4) 71 820 (69.1)
51 364 (42.1) 86 580 (71.0)
1.917 87.459***
52 686 ± 56 132 4380 ± 6417 1097 ± 1566 2922 ± 5162
82 999 ± 83 085 9560 ± 11 369 1902 ± 2474 7724 ± 10 129
−4.199*** −5.561*** −3.755*** −5.820***
a General hospitals are those with >100 beds and which are engaged in medicine, surgery, pediatrics, obstetrics and gynecology, anesthesiology, radiology and other medical services. Specialty hospitals include specialist hospitals, psychiatric hospitals and special hospitals. b Market share calculation of Herfindahl–Hirschman Index is based on outpatient benefit claims. c The National Health Insurance Administration has listed 98 ailments as chronic illnesses. http://www.nhi.gov.tw/webdata/webdata.aspx? menu=18&menu_id=683&webdata_id=444. *P < 0.05. **P < 0.01. ***P < 0.001.
Downloaded from by guest on November 17, 2015
Institutional characteristics Type of hospitala General hospital Specialty hospital Others Teaching hospital Yes No Ownership Public hospitals Not-for-profit Investor owned Diversified services Inpatient services Hemodialysis services Preventive health services Obstetrics services Home care/rehabilitation Emergency services Number of beds Acute beds Chronic beds Respiratory care beds Degree of competition Physician to population ratio ( per 10 000) Other facilities Medical centers Regional hospitals Local community hospitals Clinics Herfindahl–Hirschman Indexb Strong competition Moderate competition Low competition Characteristics of patients Age (y) 0–14 15–64 65+ Sex Male Disease patterns With chronic illnessc Residence In the same township In the same sub-medical region Financial performance Outpatient claims Outpatient benefit claims (10 000 RVUs) Inpatient claims Inpatient benefit claims (10 000 RVUs)
Closed hospitals N = 209 n (%)/mean ± SD
Hospital closure in Taiwan • Universal Healthcare Coverage System
219
Table 2 Logistic regression of hospital survival (n = 374) Variables
Exp (B)
95% CI of Exp (B)
8.999 2.087
14.976 2.678
2.555–87.772 0.704–10.194
6.688
0.075
0.011–0.535
2.503 4.256
0.208 0.179
0.030–1.454 0.035–0.918
0.000 1.157 0.669 5.937 21.756 1.082
5.6E9 1.617 1.826 0.307 0.079 1.592
0.000 0.674–3.881 0.431–7.732 0.119–0.794 0.027–0.229 0.663–3.825
2.369 4.390
1.003 1.019
0.999–1.008 1.001–1.036
28.789
1.476
1.280–1.701
21.727 2.622 9.519 22.334
0.134 0.771 0.960 1.003
0.058–0.312 0.563–1.056 0.936–0.985 1.002–1.004
3.776 4.249
4.981 5.801
0.986–25.156 1.090–30.865
9.664 4.831 0.132
1.071 0.968 1.003
1.026–1.119 0.940–0.996 0.988–1.018
0.696 2.093
1.012 0.977
0.984–1.040 0.948–1.008
2.173 1.331 0.031 0.000
1.000 1.000 1.000 0.000
1.000–1.000 1.000–1.001 1.000–1.000
Cox and Snell R 2 = 0.513; Nagelkerke R 2 = 0.686. Hosmer–Lemeshow goodness-of-fit test: χ 2 = 12.673, df = 8, P = 0.124.
uniform payment does not take into consideration that declines in local community hospitals’ admissions and patient volumes, especially when accompanied by cost increases, put those hospitals in an inferior position.
The impact of universal coverage on hospital closure Creating equal opportunities for health for all inhabitants is an overarching goal for Taiwan’s universal coverage health system. Local community hospitals are a key element to achieving that goal. However, some care-seeking habits of patients and certain elements of the design of the NHI program may be harmful to the survival of local hospitals.
First, there are no gatekeeper constraints. Because enrollees have the freedom to choose their provider without the constraints of a gatekeeper or referral system, this may encourage doctor-shopping [38, 39]. Second, co-payments are so low that patients have no incentive to choose local hospitals over medical centers. Third, the design of the payment system favors medical centers. NHI reimbursement is based on accreditation; thus, there is unequal pay for the same job. Medical centers can also reduce the cost of medicine per unit through bulk purchases, whereas local hospitals usually lack such bargaining power. In addition, hospitals may increase quantity to gain revenue, but the floating point value subsequently decreases. Local community hospitals are more likely to suffer from the impact of the ‘prisoner’s dilemma’ scenario.
Downloaded from by guest on November 17, 2015
Institutional characteristics Type of hospital General hospital (reference) Specialty hospital Others Teaching hospital No (reference) Yes Ownership Public hospitals (reference) Not-for-profit Investor owned Diversified services (reference = does not provide) Inpatient services Hemodialysis services Preventive service Obstetrics services Home care/rehabilitation Emergency services Bed size Acute beds Respiratory care beds Degree of competition Physician to population ratio Other facilities in the same region Medical centers Regional hospitals Local community hospitals Clinics Herfindahl–Hirschman Index Low competition (reference) Moderate competition Strong competition Characteristics of patients Elderly prevalence (%) Male prevalence (%) With chronic illness prevalence (%) Residence In the same township In the same sub-medical region Financial performance Outpatient claims Inpatient claims Inpatient benefit claims Constant
Wald statistic
220
Limitations and strengths
Conclusions Factors-enhanced local hospitals to survive under the universal coverage health system have been identified. The determinants of local community hospital survival under a universal coverage health system were quite different than those elsewhere. Based on the results, we recommend that the managers of local community hospitals adopt several strategies to enhance the chance of survival, such as cooperating with rival medical centers to deliver integrated health care, strengthening their own niches and creating a senior friendly environment. The results found in the study can provide early-warning signs of closure that will allow policy planners and hospital managers to coordinate needed resources and execute programs to prevent adverse effects on the community. To meet community needs when medical resources are scarce, it is necessary to clarify the crucial factors in the sustainability of local community hospitals.
Funding This study was supported by grants from National Health Research Institutes, National Health Institutes, Taiwan (PH-103-SP-06: Building hospital vertical integration models to solve the problem of unbalanced medical system) and used data from the National Health Insurance Research Database, which is overseen by the National Health Research Institutes. The interpretations and conclusions contained herein do not necessarily represent those of the National Health Research Institutes in Taiwan.
References 1. Reinhardt UE. Humbled in Taiwan. BMJ 2008;336:72. 2. Lu JF, Hsiao WC. Does universal health insurance make health care unaffordable? Lessons from Taiwan. Health Affairs 2003;22:77–88. 3. Davis K, Huang AT. Learning from Taiwan: experience with universal health insurance. Ann Intern Med 2008;148:313–4.
4. Williams I. Health care in Taiwan: why can’t the United States learn some lessons? Dissent 2008;55:13–7. 5. Krugman P. Pride, prejudice, insurance. New York Times. 2005. http:// query.nytimes.com/gst/fullpage.html?res=9500E2D9143EF934A35752C1 A9639C8B63 (11 September 2013, date last accessed). 6. Cheng TM. Taiwan’s new national health insurance program: genesis and experience so far. Health Aff 2003;22:61–76. 7. National Health Insurance Administration, Ministry of Health and Welfare. National Health Insurance in Taiwan 2013–2014 annual report. 2014. http://www.nhi.gov.tw/Resource/webdata/26703_1_2013-2014. pdf (18 June 2014, date last accessed). 8. Department of Health, Executive Yuan, R.O.C. Hospital and Clinic statistics. 2011. http://www.mohw.gov.tw/EN/Ministry/Statistic.aspx?f_list_no= 474&fod_list_no=3486 (25 August 2013, date last accessed). 9. McLafferty S. Neighborhood characteristics and hospital closures. A comparison of the public private and voluntary hospital systems. Soc Sci Med 1982;16:1667–74. 10. McNeil D, Williams R. Wide range of causes found for hospital closures. Hospitals 1978;52:76–81. 11. Mayer JD, Kohlenberg ER, Sieferman GE et al. Patterns of rural hospital closure in the United States. Soc Sci Med 1987;24:327–34. 12. Topping S. Hospital closures: what is the story? Health Serv Manage Res 1991;4:148–58. 13. Sager A. Why urban voluntary hospitals close. Health Serv Res 1983;18: 451–75. 14. Rosenstein AH. Hospital closure or survival: formula for success. Health Care Manage Rev 1986;11:29–35. 15. Hsieh HL, Chiang TL. The determinants of hospital closure and opening in Taiwan (in Chinese). Chin J Public Health 1994;13:453–8. 16. Liu JH, Chiang TL. Factors affecting opening and closure of small hospitals in Taiwan (in Chinese). Taiwan J Public Health 2001;20:27–33. 17. Succi MJ, Lee SY, Alexander JA. Effects of market position and competition on rural hospital closures. Health Serv Res 1997;31:679–99. 18. Longo DR, Chase GA. Structural determinants of hospital closure. Med Care 1984;22:388–402. 19. Bonney RS. Hospital survival strategies for the 1980s. Am J Hosp Pharm 1983;40:1483–8. 20. Berry RE Jr. Returns to scale in the production of hospital services. Health Serv Res 1967;2:123–39. 21. Fareed N. Size matters: a meta-analysis on the impact of hospital size on patient mortality. Int J Evid Based Healthc 2012;10:103–11. 22. Mullner RM, Rydman RJ, Whiteis DG et al. Rural community hospitals and factors correlated with their risk of closing. Public Health Rep 1989;104:315–25. 23. Lillie-Blanton M, Felt S, Redmon P et al. Rural and urban hospital closures, 1985–1988: operating and environmental characteristics that affect risk. Inquiry 1992;29:332–44. 24. Hernandez SR, Kaluzny AD. Hospital closure: a review of current and proposed research. Health Serv Res 1983;18:419–36. 25. Shih CC, Su YC, Liao CC et al. Patterns of medical pluralism among adults: results from the 2001 National Health Interview Survey in Taiwan. BMC Health Serv Res 2010;10:191. 26. Hong WH, Cheng SH, Chang RE et al. Changes in the proportion of crossregion admissions in Taiwan, 1985–1995. Chin J Public Health 1998;17: 388–94. 27. Lee SP. Medicine survey (in Chinese). Pharmacist Weekly. 2012. http:// ys34485257.imgshelf.com/1797/1797-1-1.htm (18 June 2014, date last accessed). 28. Lu JR, Hsieh CR. An analysis of the market structure and development of Taiwan’s hospital industry (in Chinese). Taiwan Econ Rev 2003;31: 107–53. 29. Ministry of Health and Welfare. The statistical annual report of medical care institutions status and hospitals utilization. http://www.mohw.gov. tw/EN/Ministry/DM2.aspx?f_list_no=474&fod_list_no=3444 (12 May 2014, date last accessed). 30. Health Promotion Administration, Ministry of Health and Welfare. 2007 Taiwan longitudinal study on aging (TLSA). https://olap.hpa.gov.tw/
Downloaded from by guest on November 17, 2015
Merging or consolidating with other organizations would be a possible survival strategy for local community hospitals. Eliminating these hospitals may jeopardize the chance to depict the whole picture of hospital closure. Moreover, mergers, consolidation and changes in leadership may all alter a hospital’s organizational culture and the management style of hospital executives, which may be important factors in hospital survival. Unfortunately, the study could not obtain the relevant information to examine this. In addition, the financial performance used here may not fully explain the real financial status of hospitals or their cost of operation. Moreover, information on the quality of services and of physician–patient relationships was not available. Data of the case group gathered in the year 2010 may create bias in computing the degree of competition or hospital size. Furthermore, rural and urban counties might have diverse demographic composition and, accordingly, different needs; the difference of viability between urban and rural hospitals will be explored in future work. Despite these limitations, our longitudinal population-based dataset strengthens the credibility of the results. Exhaustive variables were incorporated to reflect the real medical environment that local community hospitals face. We also elaborate on Taiwan’s experiences in the shifts in health care services after the implementation of NHI, and we highlight the impact of the universal coverage health system on hospital closure. The present study provides valuable information to illuminate the prospects for equal, affordable and accessible health care, especially for countries that have or hope to introduce universal coverage health programs.
Chiang and Wang
Hospital closure in Taiwan • Universal Healthcare Coverage System en_US/Search/02_ListSummary.aspx?menu=100000000008&mode=5 2007 (20 May 2014, date last accessed). 31. Bureau of National Health Insurance. The National Health Insurance Statistics. 2010. http://www.nhi.gov.tw/English/webdata/webdata.aspx? menu=11&menu_id=296&WD_ID=296&webdata_id=4010 (20 May 2014, date last accessed). 32. Chiang TL. The difficulty to enroll internal medicine, surgery, obstetrics and gynecology, and pediatrics physician (in Chinese). United Daily News. 2012. http://2020hfa.blog.ntu.edu.tw/2012/11/27/%E5%85%A7%E5% A4%96%E5%A9%A6%E5%85%92%E9%86%AB%E7%99%82%EF %BC%8C%E7%82%BA%E4%BD%95%E5%9B%9B%E5%A4%A7% E7%9A%86%E7%A9%BA%EF%BC%9F/ (23 September 2014, date last accessed). 33. The World Bank. Data of physicians. http://data.worldbank.org/indicator/ SH.MED.PHYS.ZS?order=wbapi_data_value_2010+wbapi_data_value& sort=asc (8 June 2014, date last accessed).
221
34. Department of Household Registration Affairs, Ministry of the Interior. Fertility rates of childbearing age women. 2010. http://sowf.moi.gov.tw/ stat/year/elist.htm (23 September 2014, date last accessed). 35. Bertakis KD, Azari R, Helms LJ et al. Gender differences in the utilization of health care services. J Fam Pract 2000;49:147–52. 36. Lee CH, Chou YJ, Chen LS et al. Utilization of ambulatory chinese medical services under the National Health Insurance in Taiwan. Taiwan J Public Health 2004;23:100–7. 37. Mullner RM, McNeil DL, Andes S. Hospital closure: who will be at risk in the upcoming decade? Healthc Financ Manage 1986;40:42–8. 38. Chiou SJ, Wang SI, Liu CH et al. Outpatient-shopping behavior and survival rates in newly diagnosed cancer patients. Am J Manage Care 2012;18:488–96. 39. Wang MJ, Lin SP. Study on doctor shopping behavior: insight from patients with upper respiratory tract infection in Taiwan. Health Policy 2010;94:61–7.
Downloaded from by guest on November 17, 2015
