The National Health Insurance (NHI) of Taiwan is a single‐payer, mandatory health insurance system. In 2011, two national policies fostering hospice services for terminal cancer patients took effect. This study explores the impact of these policy changes.
Keywords: Cost, End‐of life care, Hospice, Policy
Abstract
Background.
In 2011, two national policies aiming to foster hospice services for terminal cancer patients took effect in Taiwan. The single‐payer National Health Insurance of Taiwan started to reimburse full hospice services. The national hospital accreditation program, which graded all hospitals, incorporated hospice utilization in its evaluation. We assessed the impact of these national policies.
Methods.
A cohort of 249,394 patients aged ≥18 years who died of cancer between 2008 and 2013 were identified from the National Death Registry. We retrieved utilization data of medical services and compared the health care utilization in the final month of life before and after the implementation of the new policies.
Results.
After the policy changes, hospice utilization increased from 20.8% to 36.2%. In a multivariate analysis adjusting for patient demographics, cancer features, and hospital characteristics, hospice utilization significantly increased after 2011 (adjusted odds ratio [AOR] 2.35, p < .001), accompanied by a decrease in intensive care unit (ICU) admissions, invasive mechanical ventilation (IMV), and cardiopulmonary resuscitation (CPR; AORs 0.87, 0.75, and 0.80, respectively; all p < .001). The patients who received hospice services were significantly less likely to receive ICU admissions, IMV, and CPR (AORs 0.20, 0.12, and 0.10, respectively; all p < .001). Hospice utilization was associated with an adjusted net savings of U.S. $696.90 (25.2%, p < .001) per patient in the final month of life.
Conclusion.
The national policy changes fostering hospice care significantly increased hospice utilization, decreased invasive end‐of‐life care, and reduced the medical costs of terminal cancer patients.
Implications for Practice.
National policies fostering hospice care significantly increased hospice utilization, decreased invasive end‐of‐life care, and reduced the medical costs of terminal cancer patients.
摘要
背景. 台湾当局于2011年出台了两项政策, 旨在面向癌症晚期患者推行临终关怀服务。采用单一支付制度的台湾全民健康保险已开始全额报销临终关怀服务的相关费用。台湾医院认证项目旨在评定所有医院的等级, 其评价项目中也纳入了临终关怀的利用情况。我们评估了上述政策带来的影响。
方法. 我们查阅了台湾死亡登记处的资料, 从中发现249 394例在2008‐2013年期间死于癌症的≥18岁患者。我们还检索了医疗服务利用数据, 并比较了新政策实施前后患者在生命最后一个月内的医疗资源利用情况。
结果. 政策变更后, 临终关怀利用率从20.8%增至36.2%。在一项多变量分析中, 调整患者人口统计学、癌症特征和医院特征后, 结果显示2011年后临终关怀利用率显著增加[调整比值比（AOR）2.35, p <0.001], 与此同时, 进入重症监护室（ICU）、行有创机械通气（IMV）和行心肺复苏（CPR）的患者有所减少（AOR分别为0.87、0.75和0.80；所有 p <0.001）。接受临终关怀服务的患者进入ICU、行IMV和行CPR的可能性显著降低（AOR分别为0.20、0.12和0.10；所有p<0.001）。临终关怀利用使每例患者在生命最后一个月内的调整后费用净减少696.90美元（25.2%, p <0.001）。
结论. 台湾当局关于推行临终关怀的政策变更显著增加了癌症晚期患者的临终关怀利用率, 减少了侵入性临终诊治, 并降低了医疗费用。
Introduction
With the rapid advancement of medical services, the end‐of‐life (EOL) care of cancer patients has become increasingly aggressive in the past 2 decades [1], [2], [3], [4]. This increased aggressiveness might not be beneficial to terminally ill patients and might sometimes be against patient preferences and family perceptions of quality EOL care [5], [6], [7]. In addition, such aggressive medical services near the end of life typically lead to high medical costs [8]. Therefore, hospice care has been advocated for such patients. Hospice service provides social, psychological, and spiritual support to the patient and the family by a specialized team composed of doctors, nurses, social workers, and clinical psychologists to manage EOL symptoms, solve family conflicts, and alleviate psychological issues. Taiwan has promoted hospice care for more than 3 decades [9].
The National Health Insurance (NHI) of Taiwan is a single‐payer, mandatory health insurance system that every person in Taiwan is obliged to enroll in within 2 months after birth. It covers 99.9% of the population [10]. Patients have free access to any health care provider. Copayments are waived for patients with major diseases, including malignancy, so there are no out‐of‐pocket costs from cancer patients once a medical service is reimbursed by the NHI. The NHI began to subsidize hospice home care in 1996 and inpatient care in hospice wards at selected hospitals in 2000 [11]. All hospice admissions were provided by hospitals, so they were considered hospitalization in Taiwan.
In 2011, two national policies fostering hospice services for terminal cancer patients took effect. The NHI began to reimburse the third form of hospice service—the hospital shared‐care—which allowed inpatients to receive hospice care from a shared‐care team without being reallocated to the hospice ward or withdrawing from active anticancer therapy [12]. In addition, the hospice utilization rate was included in the grading criteria of the “Accreditation Program on Cancer Care,” a national hospital accreditation program devised by the Ministry of Health and Welfare of Taiwan [13]. A hospital could not pass the accreditation without hospice shared‐care service. A high hospice utilization rate could increase the chance of the hospital to obtain the highest rating.
We hypothesized that these two national policy changes would significantly increase hospice utilization by terminal cancer patients, decrease the invasiveness of EOL care, and reduce associated medical costs. We thus conducted this population‐based study to explore the impact of these policy changes.
Materials and Methods
Study Source
We retrieved the patients’ death events and causes of death from the National Death Registry database, to which all deaths in Taiwan should be reported. Data regarding patient demographics, utilization of medical services, and medical costs were retrieved from the NHI Research Database (NHIRD). Outpatient clinic and inpatient hospitalization services provided by both the private and public sectors were included in a unified reimbursement system of the NHI. The NHIRD included all medical claims submitted to the NHI and was the study source of many published studies [14], [15], [16], [17], [18].
We retrieved the data regarding the cancer diagnosis from the Taiwan Cancer Registry Database (TCRD) [19]. Data from the three databases were linked by patients’ identification numbers. All major cancer care providers in Taiwan with more than 50 beds (approximately 200 hospitals) are obligated to submit data to the TCRD. By 2012, the TCRD covered 98.4% of new cancer cases [19]. By the end of 2013, 68 hospitals in Taiwan were providing hospice home care, 45 were offering inpatient hospice services, and 96 were offering hospice shared‐care [13].
To comply with the privacy regulations associated with personal electronic data, the personal identities of the patients were encrypted to ensure anonymity. The study data were approved for release by the Data Release Review Board of the Collaboration Center of Health Information Application, Ministry of Health and Welfare, Executive Yuan, Taiwan. The study protocol was approved by the Research Ethics Committee of National Taiwan University Hospital (protocol number 201510125RINB).
Study Population
We searched the databases of the National Death Registry to include patients aged ≥18 years who died of cancer between 2008 and 2013. The diagnosis of cancer was defined by an International Classification of Diseases (ICD)‐9 code of 140‐208 or an ICD‐10 code of C00‐C97 [18, 20, 21].
Study Variables and Outcomes
A patient was classified as a hospice service user if he or she had received hospice service of any type at least once in the final month of life. Nine indicators in the final month of life were collected from the NHIRD to assess the invasiveness of EOL care: intensive care unit (ICU) admissions, invasive mechanical ventilation (IMV), hemodialysis, cardiopulmonary resuscitation (CPR), cancer chemotherapy, more than one visit to an emergency services, more than one hospital admission, more than 14 days of hospitalization, and death in an acute care hospital. Of these nine indicators, ICU admissions and the latter five variables are conventional indicators of aggressiveness adopted by previous studies [1], [2], [4], [22]. We also calculated a composite score incorporating these variables, as previously described [4].
The ICD‐9 Clinical Modification codes derived from the NHIRD were screened for comorbidities included in the Deyo‐Charlson comorbidity index [23], [24]. The basis for considering that the patients possessed a specific comorbidity was when the comorbidity was identified at least twice in the records of the outpatient clinic or at least once during hospitalization within one year before death.
Survival time was the interval between cancer diagnosis and death, as recorded in the TCRD. The marital status, occupation, and income of the patients were identified from the NHI enrollment records. To adjust the hospital characteristics in multivariate analysis, we had to determine the hospital that provided the primary cancer care for the specific patient. A primary care hospital was the one the patient was admitted to or the one that had the most frequent outpatient visits if the patient was not admitted in the final month of life. The characteristics of the primary care hospital, including the region, hospital level, number of beds, and hospice bed availability, were identified. Hospice availability was assessed by hospice bed density, defined as the percentage of hospice beds relative to the annual total cancer deaths in each hospital. The annual crude cancer death rates were derived from the TCRD. Medical costs incurred by the patients were retrieved from the claims database of the NHIRD.
Statistical Analysis
The Cochran‐Armitage trend test was used to examine statistically significant changes in each indicator over the entire study period. In all multivariate regression analyses in this study, a generalized linear mixed model with random intercept was applied to account for patient clustering within hospitals. We divided the study period into two subperiods: before (2008–2010) and after the implementation of the policy changes (2011–2013). To examine whether hospice utilization and the invasiveness of EOL care differed between the two subperiods, binary distribution with logit link function was used and variables including age, gender, marital status, occupation, income, cancer origin, cancer survival time, comorbidities, hospital characteristics, and hospice availability were adjusted. Similarly, a generalized linear mixed model was used to assess the associations between hospice utilization and the invasiveness of EOL care. In this model, however, year of death was also included for adjustment. Normal distribution with identity link function was utilized to examine the adjusted difference in costs between patients with and without hospice care; the adjusted variables were the same as the above, including the year of death. All analyses were performed using SAS statistical software Version 9.3 (SAS Institute Inc., Cary, NC, USA, https://www.sas.com); p < .05 was considered statistically significant. We used the Bonferroni correction to adjust the p value for multi‐comparison analysis.
Results
A total of 249,394 patients were included in our study; 63.2% of them were male, and the mean age was 67.6 ± 14.3 years. The most common origin of cancer was in the lung (20.0%), liver (19.0%), colon and rectum (11.5%), head and neck (5.8%), and stomach (5.5%; supplemental online Table 1).
Hospice utilization in the final month of life increased significantly during the study period (p < .001; Fig. 1A, supplemental online Table 2). The increase from 19.3% in 2008 to 22.2% in 2010 was modest but surged after 2011, rising from 28.7% in 2011 to 41.9% in 2013. On average, hospice utilization increased from 20.8% to 36.2% before and after the policy changes in 2011 (Table 1). According to the multivariate analysis, hospice utilization significantly increased after the new policies were implemented (adjusted odds ratio [AOR] 2.35, 95% confidence interval [CI] 2.30–2.40, p < .001; Table 1). The increase in hospice utilization was consistent across various subgroups of patients, including gender, age, cancer origin, and number of comorbidities (Fig. 2).
Figure 1.
Health care utilization in the final month of life from 2008 to 2013. (A): Hospice and invasive EOL care services. (B): Indicators of aggressive EOL care.
Abbreviations: CPR, cardiopulmonary resuscitation; EOL, end‐of‐life; HD, hemodialysis; ICU, intensive care unit .
Table 1. Health care utilization in the final month of life, 2011–2013 versus 2008–2010.
All odds ratios were adjusted for age, gender, marital status, occupation, income, cancer origin, cancer survival time, comorbidities, hospital characteristics, and hospice availability.
By the Bonferroni correction for the multi‐comparison analysis, p < .0026 was considered statistically significant.
Statistically significant.
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; ICU, intensive care unit.
Figure 2.
Hospice utilization in the final month of life from 2008 to 2013, by patients’ gender (A), age (B), cancer origin (C), and number of comorbidities (D).
During the study period, invasive EOL care, including ICU admissions, IMV, and CPR, significantly decreased (all p < .001; Fig. 1A, supplemental online Table 2). After the implementation of the policy changes, ICU admissions decreased from 17.7% to 15.7% (AOR 0.87, 95% CI 0.85–0.89, p < .001), IMV decreased from 16.7% to 13.0% (AOR 0.75, 95% CI 0.73–0.77, p < .001), and CPR decreased from 8.3% to 6.6% (AOR 0.80, 95% CI 0.77–0.82, p < .001; Table 1). Hemodialysis remained low (≤1.1%) during the study period (supplemental online Table 2).
Other conventional indicators of aggressive EOL care, such as cancer chemotherapy and more than one hospital admission, did not significantly decrease during the study period (Fig. 1B, Table 1). Instead, more patients had more than one visit to the emergency services or long hospital stays after 2011 (Table 1). Overall, 88.9% of the patients had at least one indicator of aggressive treatment, and the composite score remained stable at 2.1–2.2 throughout the study period (supplemental online Table 2).
We then analyzed whether hospice utilization was associated with invasive or aggressive EOL care. According to the multivariate analysis, the patients who received hospice services were significantly less likely to have ICU admissions (AOR 0.20, 95% CI 0.19–0.21, p < .001), IMV (AOR 0.11, 95% CI 0.11–0.12, p < .001), CPR (AOR 0.09, 95% CI 0.08–0.10, p < .001), and hemodialysis (AOR 0.17, 95% CI 0.15–0.21, p < .001) during the final month of life (Table 2). By contrast, the patients who received hospice care were more likely to have hospital admissions (AOR 1.64, 95% CI 1.61–1.68, p < .001), be admitted for more than 14 days (AOR 1.54, 95% CI 1.50–1.57, p < .001), and die in an acute care hospital (AOR 1.27, 95% CI 1.23–1.30, p < .001) during the final month of life.
Table 2. Health care utilization in the final month of life for the patients who received hospice care and for those who did not.
All odds ratios were adjusted for year of death, age, gender, marital status, occupation, income, cancer origin, cancer survival time, comorbidities, hospital characteristics, and hospice availability.
By the Bonferroni correction for the multi‐comparison analysis, p < .0026 was considered statistically significant.
Statistically significant.
Abbreviations: AOR, adjusted odds ratio; CI, confidence interval; ICU, intensive care unit.
The proportion of medical costs spent in the EOL care for cancer decedents relative to the annual total NHI expenditure decreased from 0.66% in 2008 to 0.61% in 2013 (p = .02; Fig. 3). The mean medical costs in the final month of life of the individual patients with and without hospice care were U.S. $2282.70 and U.S. $2762.30, respectively. Hospice utilization was associated with an adjusted net saving of U.S. $696.90 (25.2%, p < .001) per patient in the final month of life.
Figure 3.
Annual total NHI expenditure, EOL care costs, and the proportion of EOL care costs relative to the total NHI expenditure for cancer decedents from 2008 to 2013.
Abbreviations: EOL, end‐of‐life; NHI, National Health Insurance; USD, U.S. dollars.
Discussion
In this population‐based study, we determined that hospice utilization in the final month of life significantly increased after the implementation of the national policies fostering hospice care in 2011. After adjustment for other confounding factors, the cancer patients were 2.35‐fold more likely to receive hospice care after the policy changes. Hospice utilization was significantly associated with less utilization of invasive EOL care such as ICU admissions, IMV, and CPR. In addition, the medical costs in the final month of life were significantly lower for the patients who received hospice.
The two new policies implemented in 2011 were significantly associated with increased hospice utilization. Although palliative care has been promoted in Taiwan for more than 3 decades [9], hospice utilization for cancer decedents increased slowly; less than 20% of patients with cancer received hospice care before 2008 [25]. Reimbursing hospice shared‐care service allowed the patients to receive inpatient palliative care without having to be transferred to hospice wards [12]. The specialized multidisciplinary hospice care team could visit patients admitted outside the hospice wards and help with symptom palliation, psychological support, coordination of family care, etc. The problem with the limited availability of hospice beds (652 total hospice inpatient beds in 2010) was thus overcome. The incorporation of hospice utilization rates in the hospital accreditation criteria provided an additional incentive [13].
ICU admissions among cancer patients in the final month of life have increased in the U.S. and Canada [1], [3], [26]; however, our study revealed a significant decrease in Taiwan. According to the multivariate analysis, hospice care was associated with 80% fewer ICU admissions. For the patients who did not receive hospice care, the ICU admission rates remained stationary (supplemental online Fig. 1). In the early 2000s, when only one sixth of cancer decedents received hospice care in Taiwan, ICU admissions among cancer patients were still rising [27]. Therefore, the decrease in ICU admissions most likely resulted from increased hospice utilization, which was compatible with previous reports in the U.S. that hospice care could significantly reduce ICU admissions and invasive EOL procedures [28].
Despite the decrease, the mean ICU admission rate of 15.7% in Taiwan between 2011 and 2013 was still higher than that in Canada (9.8%), the Netherlands (7.0%), and other developed European countries, albeit lower than that in the U.S. (27.2% for elderly patients) [29]. These data may be attributed to the higher ICU bed density in Taiwan (30.4 per 100,000 population) than in Canada and in most European countries [30], [31]. Because we showed that hospice utilization was associated with an approximately 80% decrease in chance of receiving ICU admission, we should try to further increase the hospice utilization in terminal cancer patients to increase its impact. However, there was no obvious decrease in ICU admission and IMV use from 2012 to 2013. Whether the effect of policy changes on invasive EOL care had plateaued should be studied further.
Earle et al. introduced indicators of aggressive EOL care that can be retrieved from administrative data [2], [22], and a set of six indicators was used in the present study [4]. In our cohort, 88.9% of the patients had at least one indicator event, which is markedly higher than 30% in the U.S. and 22.4% in Canada [2], [3]. Whereas hospice care decreased invasive EOL care in Taiwan, it failed to reduce the conventional indicators of aggressiveness, except for ICU admissions, and it was paradoxically associated with increased hospitalization frequency, longer hospital stay, and more deaths in acute care hospitals. The high health care accessibility, universal insurance coverage, and extremely low out‐of‐pocket pay in Taiwan may have contributed to these findings. For example, terminal cancer patients usually pay almost nothing for hospital admission, which resulted in high rates of dying in hospitals and long hospital stays. In addition, inpatient hospice care in Taiwan was all provided by hospitals and counted as hospitalizations. These findings also suggest reconsideration of the current inpatient‐centered hospice care in Taiwan.
Our study demonstrated that hospice care was associated with a significant reduction in the medical costs incurred by patients with terminal cancer. This finding agrees with that of previous studies [8], [21], [32], [33], [34]. Studies from Taiwan have revealed that hospice home care and shared‐care were associated with 19.8% and 13.3% reductions in medical costs, respectively [33], [34], with most costs saved in the final 3 months of life [34]. With the inclusion of all three forms of hospice services, our analysis showed an even more substantial cost reduction of 25.2% in the final month of life. Considering the paradoxically increased hospitalization frequency, longer hospital stay, and more deaths in acute care hospitals associated with hospice care, the overall reduction in costs was remarkable. Despite the increasing incidence of cancer death in Taiwan [35], we demonstrated that the proportion of the NHI expenditure on cancer EOL care decreased with increasing hospice utilization.
The strength of this study lies in its population‐based design, the large number of patients, and the inclusion of all cancer types. Previous studies evaluating the impact of hospice care have been restricted to a selective population, such as patients of a certain age [2], [28], [32], insurance enrollment [2], [28], [32], cancer types [2], residential regions [2], [8], or types of hospice services received [33], [34]. Furthermore, the changes in the national policies in 2011 provided a natural experiment for assessing the countrywide impact of hospice care.
Nonetheless, our study has limitations. We could not determine the causality of our findings. The increase in hospice utilization, although significantly associated with the time of policy changes, could still be the result of other unmeasurable factors. Whether the decision on EOL care was based on patient preferences, physician recommendations, or resource availability was unclear. The temporal sequence of hospice care relative to the indicator events was not assessed; patients might receive hospice care after the indicated events have occurred. However, this limitation would result in underestimating the impact of hospice care rather than undermining the associations observed.
Conclusion
This large population‐based study demonstrated that the national policies fostering hospice care increased hospice utilization, decreased invasive EOL care, and reduced medical costs.
See http://www.TheOncologist.com for supplemental material available online.
Acknowledgments
This study was supported by the Ministry of Science and Technology, Taiwan (MOST‐103‐2314‐B‐002‐181‐MY2, MOST‐103‐2314‐B‐002‐090, MOST‐103‐2314‐B‐002‐092), the National Taiwan University Hospital, Taipei, Taiwan (NTUH‐103‐002314, NTUH.105‐S2954), and the National Center of Excellence for Clinical Trial & Research, National Taiwan University Hospital, Taipei, Taiwan (NCTRC201208, NCTRC201603).
Contributed equally.
Footnotes
For Further Reading: Chin‐Chia Wu, Ta‐Wen Hsu, Chun‐Ming Chang et al. Palliative Chemotherapy Affects Aggressiveness of End‐of‐Life Care. The Oncologist 2016;21:771–777.
Implications for Practice: Palliative chemotherapy is used for patients with incurable cancer toward the end of life (EOL). Aggressiveness of EOL care and hospice care are related to the quality of life of these patients. This study of data from the Taiwanese National Health Insurance Research Database found that palliative chemotherapy led to more aggressive EOL care and less hospice care. There is a need to provide patients with terminal cancer access to care information that best meets their needs, especially those patients who receive palliative chemotherapy.
Author Contributions
Conception/Design: Yu‐Yun Shao, Emily Han‐Chung Hsiue
Provision of study material or patients: Mei‐Shu Lai
Collection and/or assembly of data: Ho‐Min Chen
Data analysis and interpretation: Yu‐Yun Shao, Emily Han‐Chung Hsiue, Ho‐Min Chen
Manuscript writing: Yu‐Yun Shao, Emily Han‐Chung Hsiue, Chih‐Hung Hsu, Ann‐Lii Cheng
Final approval of manuscript: Yu‐Yun Shao, Emily Han‐Chung Hsiue, Chih‐Hung Hsu, Chien‐An Yao, Ho‐Min Chen, Mei‐Shu Lai, Ann‐Lii Cheng
Disclosures
Ann‐Lii Cheng: Bayer Inc., Eisai, Ono Pharma, Bristol‐Myers Squibb, Novartis, Merck & Co., Inc., Merck Serono (C/A). The other authors indicated no financial relationships.
(C/A) Consulting/advisory relationship; (RF) Research funding; (E) Employment; (ET) Expert testimony; (H) Honoraria received; (OI) Ownership interests; (IP) Intellectual property rights/inventor/patent holder; (SAB) Scientific advisory board
Supplementary Information
References
- 1. Earle CC, Landrum MB, Souza JM et al. Aggressiveness of cancer care near the end of life: Is it a quality‐of‐care issue? J Clin Oncol 2008;26:3860–3866. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Earle CC, Neville BA, Landrum MB et al. Trends in the aggressiveness of cancer care near the end of life. J Clin Oncol 2004;22:315–321. [DOI] [PubMed] [Google Scholar]
- 3. Ho TH, Barbera L, Saskin R et al. Trends in the aggressiveness of end‐of‐life cancer care in the universal health care system of Ontario, Canada. J Clin Oncol 2011;29:1587–1591. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Tang ST, Wu SC, Hung YN et al. Determinants of aggressive end‐of‐life care for Taiwanese cancer decedents, 2001 to 2006. J Clin Oncol 2009;27:4613–4618. [DOI] [PubMed] [Google Scholar]
- 5. Tang ST, Liu TW, Wen FH et al. A decade of changes in preferences for life‐sustaining treatments among terminally ill patients with cancer. J Natl Compr Canc Netw 2015;13:1510–1518. [DOI] [PubMed] [Google Scholar]
- 6. Wright AA, Keating NL, Ayanian JZ et al. Family perspectives on aggressive cancer care near the end of life. JAMA 2016;315:284–292. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Wright AA, Zhang B, Ray A et al. Associations between end‐of‐life discussions, patient mental health, medical care near death, and caregiver bereavement adjustment. JAMA 2008;300:1665–1673. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Cheung MC, Earle CC, Rangrej J et al. Impact of aggressive management and palliative care on cancer costs in the final month of life. Cancer 2015;121:3307–3315. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Lai YL, Su WH. Palliative medicine and the hospice movement in Taiwan. Support Care Cancer 1997;5:348–350. [DOI] [PubMed] [Google Scholar]
- 10.National Health Insurance Administration Ministry of Health and Welfare . National health insurance in Taiwan 2013–2014 annual report. 2016. Available at http://www.nhi.gov.tw/Resource/webdata/28139_1_National%20Health%20Insurance%20in%20Taiwan%202013-2014%20(bilingual).pdf. Accessed March 29, 2016.
- 11. Tang ST, Chen ML, Huang EW et al. Hospice utilization in Taiwan by cancer patients who died between 2000 and 2004. J Pain Symptom Manage 2007;33:446–453. [DOI] [PubMed] [Google Scholar]
- 12.National Health Insurance Administration Ministry of Health and Welfare . 201–2016 handbook of Taiwan's national health insurance. Available at http://www.nhi.gov.tw/English/webdata/webdata.aspx?menu=&menu_id=&webdata_id=4489&WD_ID. Accessed March 29, 2016.
- 13.Health Promotion Administration Ministry of Health and Welfare Taiwan. 2014 health promotion administration annual report. 2016. Available at http://www.hpa.gov.tw/English/ClassShow.aspx?No=201412160001. Accessed February 4, 2016.
- 14. Tsan YT, Lee CH, Ho WC et al. Statins and the risk of hepatocellular carcinoma in patients with hepatitis C virus infection. J Clin Oncol 2013;31:1514–1521. [DOI] [PubMed] [Google Scholar]
- 15. Lu LC, Shao YY, Kuo RN et al. Hospital volume of percutaneous radiofrequency ablation is closely associated with treatment outcomes for patients with hepatocellular carcinoma. Cancer 2013;119:1210–1216. [DOI] [PubMed] [Google Scholar]
- 16. Tsan YT, Lee CH, Wang JD et al. Statins and the risk of hepatocellular carcinoma in patients with hepatitis B virus infection. J Clin Oncol 2012;30:623–630. [DOI] [PubMed] [Google Scholar]
- 17. Shau WY, Shao YY, Yeh YC et al. Diabetes mellitus is associated with increased mortality in patients receiving curative therapy for hepatocellular carcinoma. The Oncologist 2012;17:856–862. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Lin CP, Hsu CH, Fu WM et al. Key opioid prescription concerns in cancer patients: A nationwide study. Acta Anaesthesiol Taiwan 2016;54:51–56. [DOI] [PubMed] [Google Scholar]
- 19. Chiang CJ, Lo WC, Yang YW et al. Incidence and survival of adult cancer patients in Taiwan, 2002–2012. J Formos Med Assoc 2016;115:1076–1088. [DOI] [PubMed] [Google Scholar]
- 20. Sharp L, Deady S, Gallagher P et al. The magnitude and characteristics of the population of cancer survivors: Using population‐based estimates of cancer prevalence to inform service planning for survivorship care. BMC Cancer 2014;14:767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Winter H, Cheng KK, Cummins C et al. Cancer incidence in the south Asian population of England (1990–92). Br J Cancer 1999;79:645–654. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Earle CC, Park ER, Lai B et al. Identifying potential indicators of the quality of end‐of‐life cancer care from administrative data. J Clin Oncol 2003;21:1133–1138. [DOI] [PubMed] [Google Scholar]
- 23. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD‐9‐CM administrative databases. J Clin Epidemiol 1992;45:613–619. [DOI] [PubMed] [Google Scholar]
- 24. Quan H, Sundararajan V, Halfon P et al. Coding algorithms for defining comorbidities in ICD‐9‐CM and ICD‐10 administrative data. Med Care 2005;43:1130–1139. [DOI] [PubMed] [Google Scholar]
- 25. Tang ST, Huang EW, Liu TW et al. A population‐based study on the determinants of hospice utilization in the last year of life for Taiwanese cancer decedents, 2001–2006. Psychooncology 2010;19:1213–1220. [DOI] [PubMed] [Google Scholar]
- 26. Teno JM, Gozalo PL, Bynum JP et al. Change in end‐of‐life care for Medicare beneficiaries: Site of death, place of care, and health care transitions in 2000, 2005, and 2009. JAMA 2013;309:470–477. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27. Tang ST, Wu SC, Hung YN et al. Trends in quality of end‐of‐life care for Taiwanese cancer patients who died in 2000–2006. Ann Oncol 2009;20:343–348. [DOI] [PubMed] [Google Scholar]
- 28. Obermeyer Z, Makar M, Abujaber S et al. Association between the Medicare hospice benefit and health care utilization and costs for patients with poor‐prognosis cancer. JAMA 2014;312:1888–1896. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Bekelman JE, Halpern SD, Blankart CR et al. Comparison of site of death, health care utilization, and hospital expenditures for patients dying with cancer in 7 developed countries. JAMA 2016;315:272–283. [DOI] [PubMed] [Google Scholar]
- 30. Adhikari NK, Fowler RA, Bhagwanjee S et al. Critical care and the global burden of critical illness in adults. Lancet 2010;376:1339–1346. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Cheng KC, Lu CL, Chung YC et al. ICU service in Taiwan. J Intensive Care 2014;2:8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Campbell DE, Lynn J, Louis TA et al. Medicare program expenditures associated with hospice use. Ann Intern Med 2004;140:269–277. [DOI] [PubMed] [Google Scholar]
- 33. Lin WY, Chiu TY, Ho CT et al. Hospice shared‐care saved medical expenditure and reduced the likelihood of intensive medical utilization among advanced cancer patients in Taiwan–A nationwide survey. Support Care Cancer 2014;22:1907–1914. [DOI] [PubMed] [Google Scholar]
- 34. Chen LF, Chang CM, Huang CY. Home‐based hospice care reduces end‐of‐life expenditure in Taiwan: A population‐based study. Medicine (Baltimore) 2015;94:e1613. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Ministry of Health and Welfare Taiwan. Ministry of health and welfare ‐ statistics ‐ 2013 statistics of causes of death. 2014. Available at http://www.mohw.gov.tw/EN/Ministry/Statistic.aspx?f_list_no=474&fod_list_no=5044. Accessed February 4, 2016.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.