Croatia Health Finance Study

Transcription

1 Report No HR Croatia Health Finance Study Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized April 25, 2004 Human Development Sector Unit Europe and Central Asia Region Document of the World Bank

2 CURRENCY EQUIVALENT (Exchange Rate Effective October 30,2003) Currency Unit = HRK HRK =US$ US$1= HRK GOVERNMENT S FISCAL YEAR January 1 - December 3 1 ACRONYMS AND ABBREVIATIONS CEE ECA ECSHD EU DFID DRG GDP GP HRK Hzzo IBRD IMR MMR MKB MOH MMR OECD PPTP SHA SHI SDR WHO VHI Central and Eastern Europe Europe and Central Asia Europe and Central Asia Human Development Sector European Union Department for International Development Diagnostic Related Groups Gross Domestic Product General Practitioner Croatian Kuna Croatian Institute of Health Insurance International Bank for Reconstruction and Development Infant mortality rate Maternal mortality ratio Croatian diagnostic coding system Ministry of Health Maternal Mortality Rate Organization for Economic Cooperation and Development Croatian case-based hospital payment system System of Health Accounts Supplemental Health Insurance Standardized death rates World Health Organization Voluntary Health Insurance Vice President: Country Director: Sector Director: Sector Manager: Team Leader: Shigeo Katsu Anand Seth Michal Rutkowski Armin Fidler Akiko Maeda.. 11

3 Table of Contents Acknowledgement... Executive Summary..... v11 ix 1. Introduction and Overview of the Croatian Health System... 1 A. Croatian Health System. A Decade in Transition... 1 B. Health System and Overall System Performance... 2 C. Organization of the Health System... 6 II. Health Financing... 7 A. Trends in Financial Flows of Croatian Health System... 7 B. Social Health Insurance... 9 C. Health Insurance Law of D. Private Insurance Market Household Expenditure on Health Care E A. B. C. D. E. F. IV. A. B. C. D. E. F. G. Expenditures on Health services and programs Allocation of Health Resources by Functions Hospital Services Primary Care and Specialists Services Expenditure on Pharmaceuticals Expanding the Scope of Public Health Programs Compensation and Allowances Conclusions and Recommendations Universal Access to Care and Social Protection Diversifying Sources of Contributions Reducing Compensations and Allowances Promoting Quality and Efficiency of Care through Payment Reforms Improving Quality and Efficiency of Care: Supply-side Intervention Promoting Patient Choice through Private Voluntary Health Insurance Improving Planning, Coordination and Consultative Process Annexes Annex 1 Household Budget Survey Expenditures on Health Annex 2. Results of the Introduction of PPTP on Cost and Average Length of Stay

4

5 List of Tables Table 1 : Health Expenditures in CEE Countries Table 2: Health Resources and Utilization Rates. Croatia. EU and CEE countries Table 3: Health Outcomes Data for Croatia. EU. and CEE countries... 4 Table 4: Major Causes of Deaths in Croatia Table 5: Estimated Total Spending on Health. as % GDP Table 6: Per capita spending on Health. total. private and public Table 7: Sources ofrevenue reported by Health Care Providers in Table 8: HZZO revenues and expenditures as % GDP Table 9: Croatia Health Insurance Institute: Revenues and Expenditures (in Table 10: current HRK millions) Annual Real Increase in HZZO Expenditures over Preceding Year Table 11 : Comparison of Health Insurance Contributions in selected Central European Countries Table 12: Changes in co-payment exemptions and central and local Government contribution policies under the new Health Insurance Law Table 13: Definition of Voluntary Health Insurance in EU and Croatia Table 14: Mean household spending on health Table 15: Household Health Expenditures by Income Quintile Groups Table 16: Household Health Expenditures by Own Welfare Status Table 17: Distribution of relative health expenditures by social welfare status Table 18: Comparison of Health Spending by Functions. Croatia and Selected European Countries Table 19: Hospital Rationalization Plan. July Table 20: Comparison of U.S DRGs. Australian DRGs and Croatian PPTPs. Open-Heart Surgery Table 21 : HZZO expenditures on prescription drugs Table 22: Changes in Sick and Maternity Leave Benefits Table 23 Mean Household Spending on Health Care Table 24: Health Expenditures by Location Table 25: Household Health Expenditures by own welfare status Table 26: Household Health Expenditures by Income Quintile Groups Table 27: Household Health Expenditures by Household Types Table 28: Distribution of Household Health Expenditures by Social Welfare Status Table 29: Distribution of relative health expenditures. by income quintile groups Table 30: Distribution of relative health expenditures by social welfare status iv

6 List of Figures Figure 1: Health Expenditure Trends in Central Eastern Europe Figure 2: Life Expectancy and Per Capita Spending on Health in CEE Countries Figure 3: Croatian population over 65 years. percent total population... 5 Figure 4: Measuring the Performance of Health Financing System... 6 Figure 5: Total Per Capita Health Expenditure. in constant and nominal HRK... 8 Figure 6: Beneficiary Composition and Figure 7: Relationship between Voluntary and Statutory Health Insurance Figure 8: Trends in Croatia Hospital Bed Capacity Figure 9: Croatia Hospital Admission Rates and Bed Occupancy Rate Figure 10: Croatia Hospital Average Length of Stay and Cost per Case Figure 11: Average Number of Services and Cost of Polyclinic and Specialist Services Per Beneficiary Figure 12: Average Number of Services and Cost of Specialist Services Figure 13 Number and Average Cost of Drug Prescription covered by HZZO V

7 List of Boxes Box 1 : System of Health Accounts and Definition of Health Expenditure Box 2: Moral Hazard Box 3: Adverse Selection in Insurance Market Box 4: Hospital Prospective Payments (Case-based, Per Stay Payment) vi

8

9 ACKNOWLEDGEMENTS This study was prepared by the team led by Akiko Maeda, Lead Health Specialist (ECSHD), with contributions fiom Prof. Avi Dor, Health Economist (Consultant); Ms. Sanja Madrazevic-Sujster, and Mr. Ivan Drabek at the World Bank Office in Zagreb, Croatia. The analysis of the Croatian Household Budget Survey of 2001 was conducted by Mr. Danijel Nestic, Economist (Consultant). The peer reviewers were Messrs. George Schieber, Sector Manager, Middle East and North Africa Region, and Mr. Daniel Kress, Sr. Health Economist, Middle East and North Africa Region. Translations of the Croatian documents were coordinated by Ms. Marija Sabljak. The team would like to acknowledge all those who met with us in Croatia during the factfinding visit in April 2003 for their generous time and cooperation. vii

10 ... Vlll

11 Background EXECUTIVE SUMMARY In the decade following independence, Croatia s health system has undergone series of health reforms that have helped to transform the fragmented and highly decentralized health system, inherited from former Yugoslavia and battered by five years of the hostilities in the region during , into a health care system that maintains the principles of universality and solidarity. The Health Care Law of 1993 consolidated health financing system under a single public entity, the Croatia Institute for Health Insurance (HZZO), which established the foundation for a revenue-base that has provided universal coverage for the population and has been the main source of health financing in Croatia. Croatia s social health insurance program is based on the principles of solidarity, in which the citizens are expected to contribute according to their ability to pay, and receive basic health services according to need. At the same time, the system recognizes the roles of private insurance and private provision of health care services. The Croatian health system has fared relatively well among the countries in the region: the system has a well-trained health workforce, a well-established system of public health programs and health delivery programs, and good health outcomes in relation to countries at comparable income levels. However, these results have been achieved at a high cost, and the growing deficits in the social health insurance fund has been a major source of concern to the Government. The recent policies to contain costs are leading to perceived and actual problems in access to care, with informal payments The generous benefits and exemptions established during the early growth years have been politically difficult to roll back, while the ageing population and changing epidemiologic profiles will continue to exert an upward pressure on the cost of health care. Croatia s challenge is to channel its already substantial public spending on health care towards cost-effective services and targeting public subsidies to protect the most vulnerable groups in the population. In recognition of the shortcomings of the current system for managing and allocated resources, the Government is currently engaged in a new round of reforms aimed at improving the performance of the health system. Health Reform Initiatives In June 2000, the Ministry of Health (MOH) issued a comprehensive policy statement in a paper entitled The Strategy and Plan for the Reform of the Health Care System and Health Insurance of the Republic of Croatia. The reforms are ambitious and attempt to address simultaneously the following broad objectives: (i) contain the rate of increase in expenditures from the public sources and reduce the payroll contribution rate by limiting benefits and increasing cost-sharing; (ii) improve the efficiency and productivity of services through reorganization and rationalization of the delivery system especially at the tertiary and secondary levels; (iii) enhance the purchasing role of HZZO and its contractual relationship with the health care providers to achieve better alignment of payments with incentives for efficiency and quality; (iv) devolve greater responsibilities to the local authorities (counties and Zagreb municipality) to manage the delivery system at primary and secondary levels, and to improve the continuity of services at these levels; and (v) expand the scope of public health programs focused on prevention and health promotion. The progress on the implementation of these reforms will be reviewed in turn. Containing Costs in Health Services. Due to the Government s overriding concern with the reduction on public debt, greatest attention has been paid to cost containment measures on public expenditure. Croatia currently spends over 9 percent of GDP on health care, of which ix

12 80 percent come from public sources, representing over 15 percent of the total government budget. Croatia s spending level is high in comparison with countries at comparable levels of per capita GDP. The role of HZZO as the single payer offers an effective means for the Government to influence public expenditures, either directly through the annual budget allocation, and indirectly through the contractual relationship between HZZO and the health care providers. Trends in health spending over the last two years ( ) indicate that the Government has succeeded, at least in the short-term, in reversing the growth in public spending on health care. But this achievement is likely to be only temporary, since the factors behind the cost escalation remain in force and are set to intensify in the coming years. The high cost of care in Croatia could be attributed, in part, to its ageing population. Ageing population not only increases the cost of health care, but raises the dependency ratio (ratio of non-active population to active working population). The 2001 Census showed that some 15 percent of the population are above 65 years of age, comparable with many western European countries, and projected to increase in the coming years. Another major source of cost escalation will arise from rapid advances in medical sciences and technology, which continues to expand the scope, as well as the cost, of health care services across the globe. Moreover, the prospects for European accession with an expected increased movement of people, goods and services with the EU member states, will raise expectations and demand for health care services. Under these conditions, Croatia will face a continuing upward pressures on the cost of health care. Health Insurance Law A center-piece of the Government s health financing reform is the new Health Insurance Law which was enacted in The 2002 Law introduced a number of new directions to limit coverage and increase revenues for HZZO, including the introduction of a new co-payment schedule for selected services in the current basic health package, with higher rates for hospital and specialist services, diagnostic tests, and prescription drugs; and offering a new voluntary product, Supplemental Health Insurance (SHI), which would reinstate the full coverage for basic services that required co-payments. The Law also brought under HZZO administration the collection and payment of worker s compensation and occupational health program, which was previously administered by private insurers. Reforms in the Revenue Base of HZZO. On the revenue side, the Govemment s reforms included: (a) reduction in the payroll tax contribution rate for health insurance; (b) establishment of principles for central and local government contributions to social health insurance; and (c) consolidation of the social insurance collection under Treasury. Collectively, these reforms are intended to reduce the tax burden and distortions on the labor market, replace the forgone revenues from payroll tax through other sources including copayments, SHI and government transfers, and improve the financial management of HZZO funds including debt management. Central and Local Government Transfers. The 2002 Law established the principals of central and local government responsibilities for subsidizing the premium and co-payments for special categories of population, such as the unemployed, war veterans and disabled. Central government transfers have been made retroactively to cover the shortfalls in HZZO budget or to cover deficits accumulated by the health care providers, rather than for specific aspects of health insurance according to prospectively agreed set of obligations. In 2002, actual government transfers to HZZO showed a significant decrease over the previous year, with debt financing still being used to cover the shortfall in budget. This outcome suggests that the cost of financing the subsidies has not yet been fully evaluated or included in the budget plan. The parameters for determining subsidy levels will need to be defined for transparency and for planning purposes. X

13 Co-payment Reforms. Co-payment rates have been increased significantly for a range of services to raise revenues. There are a number of issues with this measure. First, copayments are generally introduced to address the moral hazard problem of insurance, that is, to moderate demand for services, mainly to reduce the tendency of patients to make unnecessary and excessive use of services as a result of insurance coverage, rather than to raise revenues. Thus, co-payments should be designed to avoid discouraging use of appropriate and desirable services, such as preventive care or chronic care. As currently designed, the co-payment structure does not adequately address these concerns. For example, chronic patients who face high co-payments due to repeat visits and repeat prescriptions for the treatment of their chronic conditions, the increased co-payments present added financial burden. Furthermore, the exemptions on co-payments are broad-based and undermines the principles of targeted subsidies as well as the effectiveness of co-payments in moderating demand for services. Sumlemental Health Insurance. The Supplemental Health Insurance (SHI) has been introduced partly in order for HZZO to increase its revenues from the increased co-payment rates, and to offer partial relief for beneficiaries who may find co-payments to be barriers to access. But by removing co-payments at the point of care, the SHI re-introduces the moral hazard problem, which co-payments were meant to mitigate in the first place. SHI also presents a potential risk in the form of adverse selection problem: as a voluntary insurance product, it is likely to be purchased by beneficiaries with highest medical costs. The program has turned a net revenue for HZZO in the first year of its implementation, but it still runs the risk of increasing the expenditures in the long-run and undermining the original purpose of SHI, which was to increase the net revenues for the HZZO. Finally, the program adds to the administrative costs to the system for HZZO as well as for the providers who will need to manage another administrative procedure. For these reasons, HZZO is encouraged to explore alternative solutions that could achieve a similar result but with greater administrative simplicity, and which also avoid the problems associated with adverse selection and moral hazard. Consolidation of Collection and Fund Management under Treasury. Another important reform measure introduced in 2002 involves the consolidation of HZZO s health insurance fund under the Treasury account. While HZZO retains its autonomous status, and will continue to manage the basic health insurance and Supplemental Health Insurance hds, the consolidation of the fund was intended to improve fiscal discipline and debt management of the health insurance fund, as well as provide greater liquidity for the fund itself. Since debt repayments and delayed payments to vendors had added significantly to the cost of care, this move should, in principle, improve the financial and administrative efficiency of the health insurance funds. The actual improvements in collection and administrative efficiency from this reform remain to be evaluated. Role of the Private Insurers. The 2002 Law substantially changed the scope of the voluntary private insurance market by making basic (statutory) health services the exclusive domain of the HZZO, and removing the opt-out clause which had previously allowed some citizens to purchase private insurance in lieu of HZZO coverage. Under the Ministry of Health decree, the Government has also temporarily excluded the private insurers from Supplemental Health Insurance until January By removing the opt-out option, the new Law has effectively removed the risk of adverse selection that could have undermined the sustainability of social health insurance, through competition with the private insurance market which could have avoided the coverage of high-risk population. But it has also been a set back for a small but growing private insurance market, which lost a significant share of their health insurance portfolio under the new Law. The potential benefits of private xi

14 insurance in offering greater patient choice and complementing the services covered under statutory insurance have not yet been fully realized under the current system. Impact of Cost Containment on Providers and Users. The relative success at cost containment will need to be measured against its impact on the other objectives of the health system, namely, ensuring access to quality health care services according to need. There is an underlying assumption that part of the cost savings would be achieved through efficiency gains on the provider side. Without such efficiency gains, there is a risk that cost containment measures would result in transferring the financial burden to the providers and users of health care services, thereby undermining access to services and quality of care. If the provider is unable to make the productivity gains to compensate for the decrease in revenues from HZZO, it could reduce volume or quality of services, e.g. by rationing services through longer waiting lists, or by accumulating arrears. Alternatively, it could pass on part of the higher costs to the patients directly, through informal payments. A recent survey shows a growing concerns among the citizens with the Government s latest reform measures, which are perceived as an erosion of the principles of social solidarity and social protection. Informal payments are known to be a widespread practice that affects access to health services at all levels of care, but the magnitude of this problem has not yet been quantified at the national level. The impact of the cost containment measures on the status of the providers and users of health care services warrants careful evaluation and monitoring. In this context, it should be noted that the current system of cash accounting does not provide an accurate picture of the spending trends, while accrual accounting would fully capture the build-up of arrears in the system. Thus, while the system may appear to be containing the increase in expenditures through HZZO, it may in fact be maslung a build up of arrears and under-financing of services on the provider side, which will prove to be costly in the long-run. The Government also faces a potential conflict of interest in its dual role as the financier of the payer (HZZO) and as the owner of the health care providers: by bailing out the deficits accumulated by the providers, the Government weakens the incentives to improve efficiency on the part of the providers, and reduces the effectiveness of payment reforms being introduced by HZZO. Strengthening the Role of HZZO as Purchaser of Health Services. As the national health insurance agency of Croatia, HZZO carries the heavy responsibility of managing the largest health fund on behalf of the citizens of Croatia. To be effective, HZZO will need to transform itself from an administrative entity into a pro-active purchaser of health services, which demands value for money (i.e., quality and efficiency) from the providers. In particular, this would entail a substantial reform in the contracts and payment systems between HZZO and the providers at different levels of care. The current system of capping the number of prescriptions and referrals per beneficiaries and capitated payments offer little incentive to the health care providers to manage their patient care effectively. The recent round of reforms are just beginning to transform the contracts with the providers from a traditional fee-for-service reimbursement system into a prospective payment system with explicit volume and quality performance criteria. In the hospital sector, the case-based payment system (PPTP) was introduced in July 2002 covering initially 33 cases, and will be expanded to other cases. For the primary care sector, a proposal for a new performancebased payment of General Practitioners is currently under review and development. These initiatives are still in the early phase of implementation and will require considerable investments in capacity building and management system development, both on the part of HZZO and the health care providers. Rationalization of Service Deliverv System. Croatia, unlike many of its neighboring former communist countries, has not inherited an excessively over-built health care system. xii

15 Nonetheless, there is significant scope for rationalizing and modernizing the health infrastructure and upgrading the capacity of the health professionals. Inappropriately designed infrastructure, outdated technology and inefficient deployment and management of staff and resources contribute significantly to inefficiencies in the health care delivery system. For the hospital services, the Government has already initiated an ambitious program for rationalization of hospital services and financing system. To be fully effective, these reforms will require parallel revisions in the management of the public health institutions. Introduction of Evidence-Based Medicine to Improve Quality and Efficiency of Health Care. While the Govemment s policies have focused on cost containment, there has been relatively less discussion on the definition of services covered under statutory health insurance, or the acquisition of new medical technologies to be financed through public funds. With the expectation of new cost pressures to be coming from the introduction of new medical technologies and procedures, it will be essential that the Government, HZZO, health care providers and patient representatives work together develop a system for evaluating the safety, appropriateness, clinical efficacy and cost-effectiveness of the new drugs, procedures and technologies, and to base the public budget allocation and regulatory decisions based on the best available evidence. For this purpose, the newly established Quality Assurance network and the Health Technology Assessment unit of the Ministry of Health should be supported and strengthened. Investing in Public Health and Health Promotion Programs. Croatia faces the challenges ageing population, with rising incidence of chronic diseases that require costly and lengthy treatment. A significant portion of chronic disease conditions can be prevented through the promotion of healthy lifestyles, screening and primary and secondary preventive care measures. Investments in appropriate health promotion and preventive care would be expected to have high returns in the long-run, by reducing the cost of treatment and improving the quality of life for the citizens. Government spending on these areas has remained relatively modest, and could be significantly increased. TarPetinP Subsidies and Social Protection Programs. Although the Health Insurance Law of 2002 begins to define more narrowly the categories of beneficiaries eligible for exemptions, it remains to be seen whether the current health financing system, when taken altogether, actually improves the targeting of social subsidies to the most vulnerable groups. The combined effects of the redefined system of exemptions, increased co-payment rates and Supplemental Health Insurance on the net flow of subsidies to beneficiaries are unclear, and deserve careful monitoring. The results of the 2001 Household Survey suggest that prior to the 2002 Health Insurance Law, the pensioners and disabled persons paid a higher share of income on medical care compared with other groups, and could be facing greater financial barriers to health services. Informal payments also appear to be a widespread practice, and present another form of barrier to access that require serious attention. The introduction of SHI provides a partial solution to this problem by reinstating full coverage for those who pay into SHI, but as discussed in paragraph 10, above, it is questionable whether this would be a sustainable or efficient solution. It would also not solve the problem of informal payments, which occur outside the framework of the co-payment structure. Decentralization and the Role of Local Government. As part of the Government s policy for fiscal decentralization, the local govemment is expected to shoulder an increasing share of responsibilities in financing, investment and management of health care services. The decentralization of the health system would include the devolution of planning and coordination of primary and secondary care services to the county authorities, and are intended to improve coordination and continuity of care at the local level. Moreover, the local government is also expected to contribute to the subsidization of certain categories of Xlll

16 HZZO beneficiaries. However, in 2002 local government spent just 3.3 percent of its budget on health services, and its involvement in the health sector remains limited. The local government will need to build its resources and capacity substantially in order to take on these new responsibilities, and the transition process would require careful planning to avoid creating a gap between expectation and capacity. Strategic Planning for Health Care Reform. The need for reform in the health sector is widely accepted among both health professionals and government officials, and this has permitted the successive Governments to introduce a continuous process of reforms in the health sector. Thus, Croatian health reforms have been comprehensive and attempt to address all aspects of the health system. But the proliferation of reform measures has also strained the system s capacity to coordinate, implement and evaluate the reforms effectively, resulting at times in contradictory policies and programs being put in place. The Ministry of Health would play a key role in developing a coherent long-term strategy, for which it will need to build capacity in policy planning, modeling and evaluation, and further to enhance the process of consultation and coordination among the key stakeholders, including the representatives of civil society and professional associations. Substantial resources will be needed to implement the key reform measures, e.g., the development of the performancebased payment systems and rationalization of the delivery system. The cost of implementation will need to be better quantified and included in the medium and long-term development plans. Summary of Recommendations The Government should be commended for tackling a broad range of policy reforms to improve the perfonnance of the health system. As a priority measure, the following actions are suggested for priority attention: a. Improving Budget Planning and Fund Management. While Government appears to have achieved some success in containing public spending on health, the trend may not be fully capturing the effect of cost shifting to the providers and users of health care. That the Government continues retroactively to finance arrears accumulated by the health care providers is a source of concem. The Government will need to ensure that such retroactive financing does not undermine the ongoing efforts to improve efficiency through HZZO payment reforms. Full accrual accounting should be established at all levels to enable the Government to track arrears accurately. Furthermore, the cost of financing subsidies through general revenues should be hlly evaluated and included in the budget plan. b. Targeting subsidies and social protection. The combined effects of increased copayment rates, introduction of SHI, and narrowing of the exemption status on beneficiary access to services should be examined and evaluated systematically, and alternative solutions identified to achieve targeted subsidies and social protection with greater administrative simplicity and transparency. Co-payments should also be re-designed to discourage unnecessary use of health care, but not to be a barrier to accessing appropriate care. In this connection, the impact of informal payments on patient access to services should also be included in the surveys and evaluation procedures. c. Improving quality and efficiency of health services. HZZO s capacity as a purchaser of health care services will need to be substantially strengthened by building on the ongoing reforms in the provider payment system, including the design of the contracts, the introduction of appropriate quality and utilization review process and xiv

17 audit system, and investing in upgrading of the health management information systems in both HZZO and in the health care providers. Parallel reforms in the management and organization of the health care providers will be needed to assure that they are better able to respond to performance-based contracts. Investments in new technologies or decisions to include certain procedures under HZZO statutory insurance should be based on the best available evidence safety, appropriateness and cost-effectiveness (Evidence-based Medicine). d. Decentralization and Local Government Capacitv Building. The local government will require significant capacity building in order to take up its new responsibilities under a decentralized health system. e. Strengthening; Policy, Planning, Monitoring; & Evaluation. Health care reform is a continuous process that requires ongoing monitoring and adjustments based on regular evaluation of the policy effectiveness. Greater resources should be directed to strengthening monitoring and evaluation system, including regular household and facility surveys, citizen s perceptions of the health system. xv

18

19 I. INTRODUCTION AND OVERVIEW OF THE CROATIAN HEALTH SYSTEM In June 2000, the Ministry of Health (MOH) issued a comprehensive policy statement in a paper entitled The Strategy and Plan for the Reform of the Health Care System and Health Insurance of the Republic of Croatia. The Ministry s paper acknowledges that, despite the significant achievements in improving the financing and delivery of health care in the 1990s, the health system continues to face a variety of financial and structural problems. This study is intended to inform the Govemment s ongoing discussions on the implementation of its health reform programs, and will include an analysis of the impact of recent health finance reforms on the performance of the health system and recommendations for priority areas requiring further attention for the next step in the ongoing reform process. A. Croatian Health System - A Decade in Transition In the decade following independence, Croatia s health system has undergone series of health reforms that have helped to transform the fragmented and highly decentralized health system inherited from former Yugoslavia and battered by five years of the hostilities in the region during into a coherent health system that has maintained universal coverage for the population. The Health Care Law of 1993 consolidated health financing system under a single autonomous public agency, the Croatia Institute for Health Insurance (HZZO), which established the foundation for a modem health financing system on a social insurance model based on principles of solidarity and universality. The 1993 Law also introduced the principles patient choice and rights, and recognized the participation of the private sector in both the provision of health care services and health insurance. Although the majority of health care providers remained under public ownership, private providers have grown in number, notably in primary care, dental services, specialized clinics and dispensaries. A small but growing private insurance market has also developed which offered additional insurance coverage for services not covered under the statutory insurance plan and, until 2002, it provided alternative (substitutive) insurance for a limited number of citizens (with annual income above HRK30,OOO) who were permitted to opt out of statutory insurance and purchase private insurance. It is worth noting that the central government continues to play a dual role as the purchaser and provider of health care through its influence on the HZZO funding on the one hand, and its role as the largest owner of the hospital, primary care and public health institutions on the other. The Croatian health system has fared relatively well among the countries in the region: the system has a well-trained health workforce, a well-established system of public health programs and health delivery system, and good health outcomes in relation to countries at comparable income levels. However, these results have been achieved at a high cost and health insurance fund has faced growing deficits in recent years. The generous benefits and exemptions established during the early growth years have been politically difficult to roll back, while the ageing population and changing epidemiologic profiles have contributed to a rapid increase in public spending on health care. The efforts to contain costs in the 1990s were not effective, as HZZO s expenditures continue to outstrip revenues and built up arrears, while attempts to cap costs administratively led to a growing waiting lists and dissatisfaction among the patients and providers. This has prompted the Government to initiate a new round of reforms aimed at containing costs, reducing the tax burden on labor, and increasing revenues through cost sharing. The Govemment s current round of health sector reform measures initiated in 2000 are aimed at achieving a broad set of objectives: (i) contain the rate of increase in expenditures from the 1

20 public sources and reduce the payroll contribution rate by limiting benefits and increase revenues through increased cost-sharing; (ii) improve the efficiency and productivity of services through reorganization and rationalization of the delivery system especially at the tertiary and secondary levels; (iii) enhance the contractual relationship between HZZO and health care providers to achieve better alignment of incentives for efficiency and quality with payments; (iv) devolve greater responsibilities to the local authorities (counties and Zagreb municipality) to manage the delivery system at primary and secondary levels, and to improve the continuity of services at these levels; and (v) expand the scope of public health programs focused on prevention and health promotion. Croatia's challenge is to channel its already substantial public spending towards greater efficiency without jeopardizing the other competing objectives of universality, faimess and equity, quality, and patient choice and satisfaction. B. Health System and Overall System Performance It is estimated that in 2001 Croatia spent about $460 per person on health care, representing about 9.1 percent of its GDP, significantly above the health expenditures of other countries at comparable income levels. Financing from public sources amounted to about 7.1 percent' of GDP, fimded principally by the Croatian Health Insurance Institute (HZZO). Table 1 compares the health spending levels of countries in Central and Eastern Europe in When adjusted for per capita GDP, Croatia's spending on health lies above the CEE trend line, as shown in Figure 1. However, by comparing the per capita health spending level (adjusted for purchasing power parity) with life expectancy, Croatia appears to be doing relatively well among the CEE countries (see Figure 2). These comparisons suggest that while Croatia has a relatively good health outcome for its level of health spending, it is being achieved at a relatively high cost to the economy. Figure 1: Health Expenditure Trends in Central Eastern Europe, Total Health Expenditure in CEE Countries, ,w GDP Per Capita (US%, Ofticial Ewchange Rate), log scale ' This figure excludes cash allowances (sick leave, maternity, etc.) financed by the Croatian Health Insurance Institute. In 2001, cash transfers for sick leave and maternity related compensations accounted for about 17 percent of total HZZO expenditures. 2

21 Table 1: Health Expenditures in CEE Countries, 1998 Source: World Bank Development Indicators, Figure 2: Life Expectancy and Per Capita Spending on Health in CEE Countries, 1998 Life Expectancy Venus Per Capita Spending on Health in CEE Countries, I Cr atia * * I 66 $100 $1,000 Per Capita Health Expenditure, International $ (PPP), log scale Source: Calculated from the World Bank Development Indicators, A comparison of the distribution of health resources and utilization rates in Croatia with other European countries is shown in Table 2. The average numbers of physicians and hospital beds in Croatia are on the low side, and therefore, an oversupply of hospitals and doctors does not appear to be a major issue at the aggregate level. However, there is an imbalance in the distribution of hospital specialties, which is currently being addressed through the Hospital Rationalization Plan. The average length of stay in acute hospital setting is about 8.9 days, above the EU average of 7.7 days and significantly higher than Denmark s average 3

22 of 5.2 days o f stay in acute care hospital. Similarly, Denmark recorded 98 acute days per 100 persons compared with 140 acute days per 100 persons in Croatia. There is considerable scope for improvements in the efficiency of hospital bed use. This comparison is significant in that Denmark s proportion of population over 65 is similar to that of Croatia s. EU Average Denmark France 6.2 (1996) 7.7 (1999) w (1998) (1996) 5.5 (1999) Table 3: Health Outcomes Data for Croatia, EU, and CEE countries. Note. For year 2000, unless otherwise indicated. Sources: WHO EURO, Health For All database, Despite the relatively high level of spending on health care, the Government has been concerned by lack of improvements in some key health outcomes indicators in recent years. 4

23 While Croatia s life expectancy was 73.3 years in 2000, above the average of 72.2 years for Central and Eastern European (CEE) countries, the rate of increase in life expectancy has stagnated over the past few years. This has been attributed largely to a significant increase in death rates due to cardiovascular diseases and cancer, which together accounted for over 75 percent of all deaths in 2000 (see Table 4). Figure 3: Croatian population over 65 years, percent total population s I I, I Table 4: Major Causes of Deaths in Croatia, 2000 Source: Croatian Central Bureau of Statistics. Diagnosis and treatment of cardiovascular diseases and cancer are costly, and an increase in the prevalence of these diseases will increase medical costs. With the over 65 years population already at 15.6 percent of the total population and growing, Croatia will face an increasing upward pressure on medical cost for its ageing population in the coming years. The proximity of European health care system has raised the population expectations of the level and quality of health care. With per capita GDP of approximately US$5,000, Croatia s income level remains significantly below the European average, but its epidemiological and demographic profile approaches those of the EU. Managing the expectation of the population within the affordable limits of national resources will be a major challenge for the health policy makers. Figure 1, below, provides a conceptual fiamework for evaluating the performance of the health financing system: (a) Mobilizing revenues - how well does the system raise revenues 5

24 - does it do so progressively or regressively? Are the revenues collected with administrative efficiency? Are resources redistributed according to the policies of the Government How well does it contribute to the risk-pooling function? (b) Is the system purchasing services to promote efficiency, quality of care and choice? Figure 4: Measuring the Performance of Health Financing System C. Organization of the Health System The Croatian health system is based on the principles of social insurance and comprises mixed public and private provision and financing of services, but is dominated by a single health insurance fund, HZZO. At the central level, the Ministry of Health is responsible for (a) health policy, planning and evaluation, including the drafting of legislation, regulation of standards for health services and training; (b) public health programs including monitoring and surveillance of health status, health promotion, food and drug safety, and environmental sanitation; and (c) operations and investment in teaching hospital and specialized clinics which are owned by the central government. The Ministry of Finance is responsible for the planning and management of government budget, which includes the approval of the central budget transfers to HZZO as well as the Ministry of Health, and the collection of contributions to the HZZO. Through this means, the Ministry of Finance plays a key role in determining the overall level of public spending on health care. The Croatian Institute of Health Insurance HZZO is a public agency responsible for managing the health insurance fund and contracting of services with health care services. As the main purchaser of health services, HZZO also plays a key role in the definition of basic health services covered under statutory insurance and the establishment of performance standards and prices for services covered by HZZO. Local governments own and operate most of the public primary and secondary health care providers, including general hospitals, polyclinics, public health institutes, community health organizations (home care and emergency care units). While these facilities receive operating expenditures through their contracts with HZZO, the local authority is responsible for maintenance of infrastructure, and increasingly for capital investments. Under the Government s decentralization policy, they are expected to play an increasing role in the coordination and management of health services at the county and municipal levels. 6

25 Private providers are mainly in the primary care and specialist services, and many enter into contract with HZZO. Dental care and pharmacies can be included in the primary health centers, either leased or private. Private insurance finances services that are not covered under the statutory insurance scheme and accounted for some 6 percent of the total health spending in II. HEALTH FINANCING A. Trends in Financial Flows of Croatian Health System Croatia spends a relatively high share of GDP on health. As shown in Table 5, below, there was a period of rapid cost escalation in the late 1990s which peaked in year Since 2000, it would appear that the public spending has been contained. The total public health spending in Croatia comprise expenditures by HZZO, which account for most of the direct spending on personal health services, and government spending on public health, research and health administration. Table 5: Estimated Total Spending on Health, as % GDP, Notes: 1. Excludes cash transfers for sick and maternity leave, but includes operating expenses of HZZO. 2. Direct budget of Ministry of Health for policy, regulation, public health, and related activities. 3. Government estimates. As part of the general policy of decentralization, a small but increasing share of public spending on health is being picked up by the local govemments. Local (county) govemments are expected to play a more active part through their expanded contributions to HZZO. In 2002, county governments spent just 3 percent of their revenues on health care compared with nearly 20 percent for education. There is clearly scope for expanding local government contribution to the health services. 7

26 Figure 5: Total Per Capita Health Expenditure, in constant and nominal HRK. Per capita Spending on Health, ,000 I ,500 2,000 1,500 1, Total, nominal HEUS H Total, Constant HRK (1997 price) Total 2,268 2,728 2,885 3,053 3,056 2,899 hvate /a public 1,835 2,276 2,334 2,495 2,450 2,262 See Annex 1 for more discussion on estimation of private spending on health. Reliable data on private spending are not yet available, but government estimates place private spending somewhere around 2 percent of the GDP, or about one fifth of the total health spending. Private spending includes expenditures through private health insurance, direct out-of-pocket spending by households, and voluntary spending by private corporations on health services for their employees. It should be noted that the estimate of 2.0 percent of GDP for the year 2002 could be an underestimation, especially if the cost-containment on the public side resulted in cost-shifting to private spending. If one assumes that private spending remained constant at 2 percent of GDP in 2002, then there was a real decrease in total spending on health care in On the other hand, if private spending increased to compensate for the decline in public spending on health, the total spending on health care might not have decreased. 8

27 /Primary I ~~ Table 7: Sources of revenue reported by Health Care Providers in 2001 Carel Hospital I Health I Pharmacies I ~ * Health Institutes refer to specialist clinics and polyclinics. B. Social Health Insurance As the statutory public entity responsible for managing the social health insurance fund, HZZO administers the single largest funding for health care in Croatia, which accounts for 94 percent of the public spending on health and an estimated 80 percent of the total health expenditure in Croatia. While these measures were initially effective in retuming the system to a financial balance, subsequent rise in expenditures led to deficits, which were periodically covered by transfers from the central budget. Total HZZO expenditures (excluding cash transfers for sick leave) have grown faster than GDP, rising from 6.6 percent of GDP in 1994 to a high of 8.7 percent in In 2001 and 2002, the increase has been contained below the GDP growth rate. Table 8: HZZO revenues and expenditures as YO GDP, compensations and allowances I I I I I I I I I 9

28 Through the 1990s, deficits in HZZO financing necessitated periodic transfer of funds from the central budget to maintain the provision of health services. In 2002, the Government registered short-term liabilities amounting to HRK 3 billion and borrowed HRK 820 million to pay down the old arrears accumulated by the government-owned health care providers. While these arrears have come down from the high level of HRK 4 billion in 1999, the continuing operating deficit of HZZO and the health care providers has been a source of concern. Recognizing the need to improve fiscal discipline in the health sector to reduce the annual deficits which have a negative impact, not only on the health sector but also on the overall fiscal situation, the Government introduced measures to: (a) broaden the sources of revenues; (b) improve fiscal discipline and fund management; and (c) contain cost on the supply side through rationalization of the health delivery structure and provider payment reforms. It should be noted that HZZO expenditures also include substantive cash transfers for sick and maternity leave compensations, which amounted to 1.1 percent of GDP in For reasons described in Box 1, these cash transfers are excluded from the total health expenditure figures. Government estimates that private outof-pocket payments and private voluntary insurance payments account for around 2 percent of GDP, but reliable data on private financing are not yet available. Box 1: System of Health Accounts and Definition of Health Expenditure According to the System of Health Accounts (SHA) guidelines established by the Organization for Economic Development and Cooperation (OECD), the definition of 'kore health care functions" excludes cash transfers, such as sick and maternity leave compensations and related allowances. Therefore, although sick and maternity leave compensations and allowances are administered by HZZO, these categories of expenses are excluded from the total health expenditure in order to maintain international comparability. Other reports on Croatian health expenditures usually include HZZO cash transfers, and this may explain the differences in the reported figures. 10

29 (million HRK, current) TOTAL REVENUES 11,527 12,661 13,081 Contributions 9,778 9,953 10,136 Employer Employee Other 4,434 4,747 4,434 4,229 4,434 4,747 5,118 5, ,859 14,181 10,463 11,381 Source: Croatia Health Insurance Institute, Notes. 1. Stock of short-term liabilities are accounts payable for purchases made in the current fiscal year but paid in the following fiscal year. 2. In this estimation, it was assumed that the stock of shortterm liabilities represented the full stock of accounts payable in the following fiscal year, and that the payments were made in the following year. Receipts from borrowings are included in the 2002 calculation. On the revenue side, accounts receivable were not included in this estimation. A rapid real increase in health expenditures were recorded by HZZO between 1998 and 2000, averaging about 8 percent per annum in real terms, and this rate of increase has outstripped the revenues of HZZO, which have not increased significantly over the same period. From Table 10, it is evident that between 1997 and 2000, spending increase occurred across all categories of health spending, including primary care, hospital services, and pharmaceuticals. A sharp decline in expenditure on drugs in 2002 could be attributed to the higher cost-sharing which was introduced in However, without the data from household expenditure and 11

30 utilization surveys, it is difficult to determine the extent to which cost reductions were attained through productivity gains, or through other means, e.g., rationing of care (long waiting lists for non emergency services) and cost shifting to patients. Table 10: Annual Real Increase in HZZO Expenditures over Preceding Year, Source: Calculated from data provided by HZZO, ComDosition of Beneficiaries. Changes in the composition of the beneficiaries are expected to have an important impact on the financial flows of the health insurance system in the coming years. Whereas the percentage of active workforce contributing to HZZO has remained stable at around one third of all the beneficiaries, the number of beneficiaries falling under the categories of unemployed and pensioner has been increasing over the last few years (see Figure 6). Since these two groups generate the lowest contribution rates, and pensioners are likely to be among the highest users of health services, these changes in the profile of HZZO beneficiaries will likely lead to higher spending and lower revenues. Figure 6: Beneficiary Composition, 1995 and 2002 Distribution of Beneficiaries, 1995 Distribution of Beneficiaries, 2002 Actively Employed and Active Farmers W Pensioners 34% n 0 Unemployed 0 Others including dependents 12

31 ~ Improving Collection Compliance and Debt Management. As part of the overall fiscal reform measures, in 2002 the Govemment introduced a single collection and compliance system under the Treasury for all payroll taxes, including contributions to health insurance. Under the new financing structure, HZZO will retain on its own administrative expenses under an extra-budgetary fund which would support its operating costs. These reforms are aimed at improving compliance on contribution rates and debt management, and are also intended to align HZZO budget within the government s fiscal policy and budget planning process. Moreover, the Govemment s general revenue contribution to HZZO will be linked more clearly to planned and targeted subsidies for selected groups of population, rather than as an ad hoc transfer mechanism to HZZO to cover their debts. The consolidation of the funding under Treasury was also intended to improve the liquidity and debt management of the HZZO funds. Inadequate cash flows have contributed to serious delays in payments to vendors, which probably added to the cost of care as vendors factored into their prices the financial risks associated with expected delays in payment. On the debt management side, Treasury has filled ths year s revenue gap with a HRK 820 million loan which was obtained at a more favorable rate than would have been possible under HZZ0.2 By bringing the HZZO funds under central government oversight, the Govemment intends to bring down HZZO debts and to generate an operating surplus of about HRK 360 million by Contributions and Revenues. The Croatian Health Insurance Institute revenue structure has been heavily dependent on salary contributions of 1.4 million insured employees and employers, whose combined contributions accounted for 80 percent of the HZZO revenues in The payroll contributions have largely subsidized the coverage for the remaining 2.8 million insured population, with central budget transfers covering the deficits. This financing structure has been part of the legacy of the 1993 health refom, when the payroll tax contribution rate had been increased initially to 18 percent in order to cover the severe financial deficits facing the health system in the post-war period. As shown in Table 11, Croatia s payroll contribution rate for health insurance remains among the highest in the region. Communication with the representative from the Treasury Department, April

32 ~ Table 11: Comparison of Health Insurance Contributions in selected Central European Countries Country, year introduced Croatia 1993 Croatia 2000 Croatia 2003 Czech Republic, 1993 Estonia, 1992 Hungary, 1990 Slovakia, 1994 Payroll Tax Rate for Health Salaried (employer: Self-Employed Non-employed employee) 18% (18%: 0%) 16% (7%: 9%) 15 % (15%) 0%) 0.5% (occupational safety) 18% of declared income 18% of a set fixed amount which depends on formal qualification* 18 % of gross benefits plus central budget transfer Central budget transfer Central budget transfer 13.5% (9%:4.5%) 13.5% of declared income Central budget transfer, equal to 13.5% of 80% of statutory minimum wage 13% (13%:0%) 13% of declared Central budget transfer income 14% (11%:3%) plus hypothecated tax of $170 per employee 13.7% (10.0%: 3.7%) 14% of declared income 13.7% of declared income Central budget transfer Central budget transfer, equal to 73% of statutory minimum wage * The base rate is set at HRK 2,318, which is multiplied by coefficients (the total of 9 levels of qualifications) ranging from 1 (non-skilled workers) to 2.8 (doctoral degree holders). Sources: A. Preker, M. Jakab and M. Schneider; and updated information on Croatia provided by HZZO, In 2000, the payroll tax for health insurance was lowered to 16 percent as part of the Government s fiscal policy to reduce tax burden on labor. This was further reduced to 15.0 percent in 2003 and an additional contribution of 0.5 percent for occupational safety and worker s compensation. The decrease in the share funded from salaries is expected to be compensated by (i) an increasing share of central and local budget allocation for the protection of special categories of the population3 covered by HZZO, (ii) increased costsharing; and (iii) new Supplemental Health Insurance which would cover the co-payments. Under the Health Insurance Law of 2002, the responsibilities of the central and local governments has been more explicitly linked with subsidies for the unemployed and The special categories of the population who are exempted from contributions include: children and adolescents below the age of 18, pregnant women, all pensioners, farmers above 65 years of age, unemployed persons, war veterans and socially handicapped persons. These groups make up around 70 percent of the total population. See Table 18 for more details. 14

33 pensioners, while HZZO will be expected to subsidize premiums for children, pregnant women, persons with 80 percent disability and beneficiaries of the social assistance. Diversifylng the revenue base by increasing the share of revenues from general taxation has a number of advantages. First, general taxation draws on a broader revenue base, which could provide a more stable source of revenues, especially during periods of recession and economic instability. Secondly, an increase in general tax contribution would bring HZZO budget under greater public review and accountability, and allow more explicit trade-offs to take place between health care and other public expenditures that reflect the policies of the govemment. As a percentage of total revenue, central budgetary transfers increased from 14 percent in 1998 to 20 percent in 2000, part of which was used for debt settlement. However, in 2002, the central budget transfer decreased significantly to just under 10 percent of the total revenues. As a consequence, HZZO faces a gain increased dependence on contributions from payroll tax to subsidize the health care of the other social cases. This would seem to contradict the stated objective of central government s responsibilities for subsidizing the contribution for the unemployed and pensioners. C. Health Insurance Law of 2002 The Government s new Health Insurance Law, approved in January 2002, aims to improve the financial sustainability of the system by reducing the scope of basic covered services, reorganizing the co-payment system, stimulating the purchase of a voluntary Supplemental Health Insurance (SHI) plan?, and defining the contributions from the central and local govemment budgets. Under the new Law, the Supplemental Health Insurance has been introduced to allow policyholders to purchase policies that cover the new co-payment rates, thereby reestablishing the full level of coverage. At the same time new restrictions have been imposed on the private sector. Briefly, the new law enacts the following: a. Introducing a new co-payment price schedule for selected services in the current benefit package, with higher rates for hospital and specialist services, diagnostic tests, and pharmaceuticals. The categories of beneficiaries exempted from copayments have been reduced to some extent compared with prior years, although major categories of exemptions remain (see Table 12). Increased co-payment rates for high-end procedures are also under consideration. b. Making basic insurance coverage compulsory and exclusively provided by HZZO, thus removing the opt-out clause which had permitted those with income above HRK30,OOO per year to purchase substitutive private insurance in place of HZZO basic plan. c. Offering consumers the option to purchase SHI policies on a voluntary basis that cover co-payments which would restore the full coverage on the basic health services. For the moment, SHI can be offered by HZZO exclusively but this restriction is expected to be lifted in The premium is set at HRK40-HRK80 per month, which can be paid at the individual or employer level. As an added Readers are cautioned that the Croatian use of the tenn, Supplemental Health Insurance, under the Health Insurance Law 2002 should not be confused with the technical definition of the different categories of Voluntary Health Insurance used in the EU context. Under these definitions, the Croatian Supplemental Health Insurance plan would be categorized as complementary VHI. 15

34 incentive to purchase the SHI, tax refund equivalent to the amount of premium for one year is given to any individual or employer that purchases SHI. d. Clarifying the responsibilities of central and local governments in providing subsidies for the social cases. e. Introducing administration of worker s compensation funds for occupational safety under HZZO. The Law represents an important step in rationalizing the health financing system, but it also raises a number of new issues: (a) the effectiveness of the new co-payment system in mitigating excessive utilization (see Box 2: Moral Hazard) is undermined by the broad exemptions as well as the effect of the SHI plan; and (b) as a voluntary plan, SHI is open to adverse selection problems, i.e., the plan is more likely to be purchased by the high-end users (see Box 3: Adverse Selection), which is further exacerbated by the discount policy for the pensioners who are given a 40 percent discount on SHI premiums to encourage their participation. In 2002 about 50 percent of the SHI is being purchased by pensioners, who are most likely to use health care more frequently compared with other groups. In the first year of implementation, SHI revenues have exceeded expenditures, but as the market matures, there is a likelihood that the utilization rates and expenditures for the SHI subscribers would eventually overtake revenues. The experiences of the French health insurance system offer useful lessons for the Croatian SHI.6 Over the last 25 years the French government has used cost-sharing as a means of containing health care expenditure; instead of reducing consumption, however, this strategy has instead encouraged the growth of complementary Voluntary Health Insurance (VHI), with the result that most French people now purchase this type of VHI to reduce the financial burden of out-of-pocket expenditure. France has an exceptionally high level of complementary VHI coverage (85 per cent of the population in 1998) for the reimbursement of co-payments for treatment in the statutory health care system. Although complementary VHI has long been a feature of the French health care system, it has raised considerable equity concerns, since only those who can afford VHI will receive added protection for services that are supposed to be covered under statutory insurance. The French government has addressed these concerns by making complementary VHI available free of charge to individuals with low incomes, but this has undermined the primary purpose of the complementary VHI, which was to raise revenues. The new Law has also had an adverse effect on the private insurance market, which were suddenly closed out of the substitutive insurance market as well as worker s compensation, which used to be administered by private insurers. The changing relationship between the statutory and voluntary insurance from 1993,2002 and 2004 is illustrated in Figure 7. Personal Income Tax law, Official Gazette, allows for premiums for additional health insurance (including premiums for life and voluntary pension insurance) to be tax-deductible expenses from July 1,2001. The rebate is open-ended, and could be renewed each year. See E. Mossialos and S. Thomson, 2002, for more details. 16

35 Box 2: Moral Hazard Moral Hazard is a concept that refers to one of the effect that health insurance has on consumers of medical care. Insurance reduces the net out-of-pocket price of medical services, and thus increases the quantity of services demanded by the beneficiaries. A loss of social welfare occurs because consumers (beneficiaries) do not bear the full costs of medical services they consume, while the price paid by the insurer (to the provider) exceeds the true value to the consumer. For excessive services demands, the cumulative difference between the amount the insurer pays and what the services are worth to the consumer is the Welfare Loss. The area of the shaded triangle below gives the monetary value of the total welfare loss. Price PF PI Welfare Loss Triangle : excess services consumed due to moral hazard are worth less to the consumer than the price paid by the insurer - I I I I I I I I I I units of medical services (Quantity) QE QI Moral Hazard: increase in utilization due to insurance PF = full price without insurance PI = out-of-pocket price in the presence of insurance = (1-copayment rate)?(full price) QF = quantity demanded under full price. Ql = quantity demanded under insurance price. The welfare loss from moral hazard should be weighted against the gain to consumers from having less uncertainty under insurance. Nevertheless policy makers should be aware of the problem of excess utilization and higher medical expenditures that comes with insurance. 17

36 ~ Voluntary ~ 0 Box 3: Adverse Selection in Insurance Market Adverse selection occurs when an individual with poor health acquires low risk medical insurance meant for individuals in good health. This is possible under 'asymmetric information', i.e. a situation in which the subscriber knows more about her own health status than the insurer. With high risk subscribers entering such a plan, the insurer has no choice but to increase premiums, leading lowrisk individuals to exit such a plan, requiring further increase in premiums (Rothschild and Stiglitz, 1976 ). Unable to fully identify the health risk of individuals in markets, insurers may structure their benefits so that high risk individuals are forced to reveal themselves (Van de Ven and Ellis, 2000). Insurers can prevent adverse selection by screening subscribers based on known medical conditions and prior use of medical services (experience rating). This results in another form of risk selection, whereby low risk individuals are preferred, and high-risk individuals are effectively excluded from insurance markets due to prohibitively high premiums. Figure 7: Relationship between Voluntary and Statutory Health Insurance, I Health 1 8 I 2002 Voluntary j 08 r i 2004 Voluntary Basic health services covered exclusively by HZZO Copayments on Basic Health services covered by HZZO Services not covered by HZZO 18

37 Table 12: Changes in co-payment exemptions and central and local Government. Status before 2002 Exempted from co-payments: Children under 18 Unemployed persons Homeland War Veterans and members of their families; Family members of deceased members of the Croatian Army or Police Forces who died as a consequence of wounds received during the war, were imprisoned or are missing, political prisoners and World War I1 veterans; civil invalids Refugees and returnees Pregnant women receiving maternity benefits Voluntary blood donors with more than 50 blood donations and donors of human body parts; Beneficiaries of social care institutions whose costs were either fully or partly covered by social care institutions; Persons with an income from regular employment of less than HRK 1,275 per month; Pensioners with an income of less than HRK 1,700 per month; Disabled persons with permanent physical damage with an income less than HRK 2,550 per month, Persons injured at work or suffering from an occupational disease; Persons suffering from infectious or mental diseases undergoing treatment; Persons participating in organized preventive health care measures; All pensioners and voluntary blood donors with more than 25 blood donation were exempted from co-payment for prescription drugs. Source: Croatia Institute of Health Insurance, 200: alth Ins Exempted from co-payments Children under 18 years Disabled persons, persons with at least 80 percent disability Croatian Homeland War invalids Pregnant women receiving maternity benefits Persons with monthly per capita income less than HRK 1,330 (for 2002) Pensioners living alone with monthly income less than HRK 1,729 Voluntary blood donors with more than 35 (men) and 25 (women) blood donations. 2. General Revenue Subsidies on Contributions General revenue contributions from central and local governments will be transferred to HZZO for selected categories of population in order to compensate for differences between the HZZO reimbursement and full cost of health services. Central Government will compensate HZZO for the following categories of population: Unemployed persons, persons without health insurance over 18; draftees and military reservists during service, farmers over 65 years of age, persons incapable 01 independent living and work, without means of support; War or civil invalids and beneficiaries of survivor s disability pension; Homeland War veterans; secondaq school students and regular university students withoul health insurance coverage. Central or Local Government will compensate healtl. insurance for the following categories of population: Persons participating in public works and civi protection program; members of operation units of fire. fighting brigades; beneficiaries of financia compensation for physical injury according to pensior rermlation. D. Private Insurance Market In 2002, health care providers reported a revenue about HRK 962 million fiom private insurers, representing about 6 percent of the total health expenditure in that year. The Health Insurance Law of 2002 has been a setback in the growth of private insurance market in Croatia. For example, prior to implementation of the 2002 law most of substitutive insurance 19

38 was sold by one company, Addenda Insurance. Addenda claims to have lost a significant part of its portfolio as a consequence of the 2002 Law. Currently two insurers dominate the market for voluntary insurance: Addenda with about 20,000 lives and Croatia Osiguranje (Croatia Insurance)8 with under 10,000 lives. Premiums at Addenda are experience rated, i.e., are based on the medical condition of the individual or the combined level of risk of employees in a private firm contracted. Theoretically, individuals with severe pre-existing conditions will be able to obtain an Addenda policy, but at a prohibitively high rates. Addenda already has a network of participating private doctors and facilities. Croatia Insurance plan offers a number of policies, with actuarially fair premiums to be determined by the maximum level of private services allowed within each type package that will be offered; premiums will be differentiated only on the basis of age cohort. Thus, both firms are capable of offering insurance policies based on sound actuarial principles. In addition there are over 25 smaller general insurance companies that could become more marginal participants in these markets. The regulatory framework for medical insurance market remains largely undeveloped in the Croatian health system. This could lead to problems from market failure. In particular, private insurers would have incentives to select only low risk subscribers, which could result in a two-tiered system with high-risk individuals having limited access to private markets. A number of regulatory measures can be introduced to mitigate the adverse selection problem, but these may have limited effectiveness and would involve additional transaction costs. Examples include introduction of risk adjustments, restrictions on medical underwriting and exclusions for pre-existing conditions, and requiring community rated premiums for VHI. It should be noted that the introduction of these regulations would be incompatible with the EU Insurance law, which prohibits Governments from applying material regulation in the insurance sector that could restrict competition among insurer^.^ The EU Law has focused on financial regulation of the private insurance sector in order to assure solvency of insurers and their ability to meet obligations, while the insurance product or benefits plan has been left largely unregulated. Thus, insurers are permitted to introduce pre-existing condition and other exclusions, are not required to meet minimum standard clauses (as in the US.), and are not permitted to adjust risks through community rating. Similarly there is no regulation of prices and premiums, since these are considered material regulation which is not permitted under the EU Law. The regulatory framework for VHI in the European Union will become an increasingly important aspect of public policy towards VHI, largely as a result of a series of European Commission directives aimed at creating a single market for life and non-life insurance in the European Union. With respect to the development of the private market for VHI in Croatia, accession to the EU would entail both advantages and disadvantages: the deregulated environment of the EU fits well with the current market in Croatia, where private insurers are already participating in the market in an unregulated environment. Accession to the EU will also serve to enhance competition in the long run, given the potential for new entrants from Interview with Dr. Dario Lovric, President and CEO, Addenda, April 3,2003. Addenda is a fully private company, while Croatia Insurance is a government-owned insurance company, with about 52 percent market share of the general insurance market in the country. For purposes of this discussion we refer to both fums as private. Under the EU Law, consumer protection for VHI is limited to financial safeguards against the negative consequences of insolvency. The only exceptions allowed will be for substitutive VHI, for which the government may invoke the general good clause in order to adopt or maintain regulations designed to protect public interests, as long as they do not unduly restrict the right of establishment or the freedom to provide services. 20

39 EU into the local market for VHI. While consumers could benefit from greater competition which might lead to lower prices and greater choices of services, concems related to the market failure problems, such as adverse selection, remain. However, in the absence of substitutive insurance market, the risk is significantly reduced as the entire population would be covered for basic services under statutory insurance. A review of EU experience shows that the role of Voluntary Health Insurance (VHI) has been relatively limited in countries that have a well-established statutory health insurance system. In the EU countries, VHI have been used mainly for improved amenities care (e.g., private hospital rooms), faster access (e.g., use of private clinics of hospitals to avoid the waiting lists in public providers), and greater choice of providers. In some countries, such as France, VHI has also been used extensively to provide complementary coverage for co-payments and other cost-sharing requirements of the statutory health insurance programs. Market participation rates throughout the European Union (EU) remain relatively low, with expenditures on VHI account for less than 10 percent on total health expenditures in most states. In the context of European Union and countries that have a statutory (compulsory) health insurance system in place, the role of Voluntary Health Insurance are classified as substitutive, complementary or supplementary. o The use of these terms in the Croatian context is shown in Table 13, below. Table 13: Definition of Voluntary Health Insurance in EU and Croatia European Union Context Substitutive insurance: Coverage for services that would otherwise be covered by the state or national health care system Comulementarv insurance Coverage for services only partially covered by the state (e.g., co-payments imposed by the statutory health insurance). Suuulementarv Insurance Coverage for services not covered by statutory health insurance, e.g., to provide faster access to selected services, offering greater consumer choice, and for non-medical amenities Croatian Context Substitutive insurance Not permitted to be covered by VHI. Suuulemental Insurance Insurance coverage for co-payments required by Basic Health Insurance This is covered by private insurance for all services not covered under the Basic Health Insurance. E. Household Expenditure on Health Care Government estimates that the private spending contributes about 2.0 percent of GDP, or one fifth of the total spending in health care in Croatia. Results of the Household Budget Survey conducted in 2001 suggest that direct household spending on health care accounted for about 1.2 percent of GDP. Reported revenues from health care providers indicate that reimbursements from private insurers accounted for about 0.7 percent of GDP. Thus, out-ofpocket payments and private insurance together account for approximately 2 percent of GDP. lo E. Mossialos and S. Thomson,

40 It should be noted that the HBS results may have underestimated the actual household spending, due to a relatively long recall period (four months) used for outpatient services and medical products." Table 14: Mean household spending on health, 2001 Household expenditure per capita GDP per capita Household expenditure on health, as % GDP 22,092 HRK 36,712 HRK 1.22% The analysis of the HBS data do not reveal any regressive spending pattems by income quintiles with respect to mean per capita health spending Table 15. However, when groups are divided by their social welfare status, it is evident that the pensioners and disabled persons incur the highest out-of-pocket expenditures. This is not surprising given that they face chronic conditions that require frequent and repeated use of health services and products. Table 15: Household Health Expenditures by Income Quintile Groups, 2001 Note: Quintile groups are created according to total income per capita. Source: 2001 Household Budget Survey. 11 Typically, recall periods of two to four weeks are used for these expenditure items. 22

41 Table 16: Household Health Expenditures by Own Welfare Status, 2001 Notes: All amounts are attributed to the persons according to her household expenditures per capita. Welfare status is based on the individual s most frequent activity status in the last 12 months, except for disabled persons where status is defined by self-reported disability to work or receiving invalidity pension or receiving other invalidity benefits. Source: 2001 Household Budget Survey. A more revealing pattem emerges when the variance on expenditures is analyzed for different categories of households. For example, retired persons spent on average 779 Kunas, or 3.54 percent of the total household expenditures on health. When the distribution of expenditures are analyzed for this category of household, it reveals that some 7 percent of the retired persons spent more than 10 percent of their household budget on health care. This would suggest that a significant number of pensioners face a major financial burden, and that the existing health insurance system does not adequately provide protection for these groups of beneficiaries. More details are provided in Annex 1, Table 17: Distribution of relative health expenditures by social welfare status 2001 Note: Social welfare status is based on the most frequent activity status in the last 12 months, except for disabled persons where status is defined by self-reported disability or receiving invalidity pension or receiving other disability benefits. Data source: 2001 Household Budget Survey. 23

42 Informal payments outside of formal co-payments appears to be a fairly widespread phenomenon. Preliminary results from a 2002 study conducted by the Institute for Public Health on informal payments in Zagreb suggest that some 44 percent of respondents who used health services indicated that they made some form of informal payments. An earlier study, conducted in 1994 (Mastilica, M. and Bohikov, J. 1999) in Split and Zagreb also confirm that co-payments accounted for only a small share of households expenditures on health care. Thus exemptions on co-payments may provide only a limited relief fi-om the financial burden of medical care for the low income households. The other forms of direct payments include, for example, payments for private consultations, non-prescription drugs, and informal payments for physicians. As indicated elsewhere, the implementation of more stringent cost-containment measures introduced by HZZO in 2002 could have led to further cost-shifting by the providers to private payers. Implications of each new policy measures on equity will need to be closely monitored, evaluated and used to inform hture decisions on the design of the health insurance system. III. EXPENDITURES ON HEALTH SERVICES AND PROGRAMS A. Allocation of Health Resources by Functions The 1993 reforms introduced a capitation payment for the primary care sector, and a point system for the hospital sector. Subsequently, the Ministry of Health added the global budget caps for hospitals and reference pricing for drugs in Although these measures appear to have had some moderating effect on the rising cost of care, their effectiveness in promoting productivity and assuring quality of care have been limited.12 The following section will review the impact of the recent reform measures on the cost and volume of services in the hospital services, primary care and specialist services, and pharmaceutical. Table 18: Comparison of Health Spending by Functions, Croatia and Selected European Countries I I I I I I I Sources: OECD Ecosante health database, 2001; Croatia figures were calculated by the author based on available data on government, private insurance and household expenditure data. l2 For example, the introduction of the global cap and point system for hospital services may be leading to rationing of services through waiting list for certain high-cost services, such as cardiac surgery and Percutaneous Transluminal Coronary Angioplasty (PTCA). Delays in treatment could have an adverse effect on the patients health outcome. 24

43 B. Hospital Services The Government has recognized that the current Croatian hospital system shows significant imbalances and inefficiencies exist in the distribution and types of acute inpatient care beds and other forms of patient care settings. To address these shortcomings, the Ministry of Health has initiated a Hospital Rationalization and Financing Plan (see Table 19) which will intensify the reforms in this sector over the next 3 years, and for which the central government may allocate up to HRK 800 million to support its implementation. The goal of the Rationalization Plan is to have hospitals operating with no losses by For each county hospital, an agreement will be reached between the Ministry of Health and the county on a hospital rationalization plan. Under this agreement, the central government will finance investments and services up to a fixed level, above which the counties will be responsible for securing additional financing on their own. The ongoing comparison of costs and resources between similar departments across the country, as well as the comparison of costs for 43 diagnostic groups, are intended to provide an empirical basis for the Ministry to determine, in negotiations with the each county, the staffing, service and financing level for each county hospitals. Hospital Pavment System. The current point-based hospital payment system is essentially a fee-for-service reimbursement system which reimburses hospitals on the basis of inputs rather than outcomes, and does not allow hospital management to be rewarded for productivity gains.i3 In addition, each hospital is limited by a global cap which includes financial penalties if a hospital exceeds its ceiling. The current payment system has created motivation for a hospital to keep the beds full and extend the length of stay, since high occupancy results in steady funding based on the per diem reimbursement. Low occupancy rate also increases the risk that the ceiling on the global cap would be lowered by HZZO. The current contract arrangement makes it difficult to adjust staffing levels in response to shortened lengths of stay and other efficiency gains, since staffing costs remain fixed. Thus, cost overruns are likely to result in the imposition of arbitrary internal controls, e.g., by restricting the use of medications or procedures, rather than a response to improve productivity, such as reorganization of staffing and other systemic reforms. While global hospital funding provides for financial control in the broader system, parallel reforms in hospital management and realignment of the incentive structure are essential in order to protect access and assure appropriate quality of care. l3 The hospital payment system consists of a three separate components: (i) hotel services are paid by a flat payment per bed-day; (ii) physician services are paid by procedure using the WHO point system; (iii) pharmaceuticals and other materials are paid separately, depending on the cost of each item. 25

44 Specialization Acute Care Hospital Beds Internal medicine Infectious Disease Psychiatry Neurology Dermatology Physical medicine and rehabilitation Radio-therapy and oncology General surgery Child surgery Maxillofacial surgery Orthopedics Neurosurgery Urology Ophthalmology Ear, Nose & Throat Pediatrics Gynecology TOTAL Table 19: Hospital Rationalization Plan, July 2002 Number of Beds per Contract 4, , , ,318 2,124 16,819 Proposal , , , ,083 1,955 15,675 Difference ,144 Chronic Disease Hospital Beds Physical medicine and rehabilitation Psychiatry Chronic child diseases Chronic pulmonary diseases Diseases of elderly people TOTAL 2,140 3, ,494 1,841 2, , Source: Ministry of Health, Health Reform,eject, July

45 Figure 8: Trends in Croatia Hospital Bed Capacity, Croatian Hospital Bed Capacity, ,, , , , , I Year T Total Hospital Beds t Beds per 1,000 Population Figure 9: Croatia Hospital Admission Rates and Bed Occupancy Rate, I U Hospital Admission Rates % pop +Bed Occupancy Rate 1 27

46 The Govemment has been attempting to control costs in the hospital sector through a combination of bed capacity reduction and hospital payment reforms. As shown in Figure 8, the total bed capacity has been gradually reduced over the past decade. The average length of stay has remained relatively unchanged in the 1990~~ whereas admission rates have been increasing steadily. Subsequently, the government introduced a global ceiling to hospital budget in 1999 to strengthen the cost-containment of the hospital sector. On one hand, this measure appears to have reduced the average cost per inpatient case, but it has contributed to longer waiting lists for some interventions. The rationing may be taking place that could have a negative impact on quality of care. The hospitals currently report on patient information on admission and discharge diagnoses coded using the MKB-10, a Croatian variant of the International Classification of Diseases - 10th Edition (ICD-IO). No secondary diagnoses, such as complications or co-morbidities, are coded, nor are any procedure codes using MKB provided. However, other patient encounter information is recorded using the Blue Book, which offers a detailed listing of most types of patient care events.14 The point system is used to report on patient visits to general practitioners, specialist (polyclinic) visits and hospitalizations. The current point-based hospital payment system is essentially a fee-for-service reimbursement system which reimburses hospitals on the basis of inputs rather than outcomes, and does not allow hospital management to be rewarded for productivity gains. In addition, each hospital is limited by a global cap which includes financial penalties if a hospital exceeds its ceiling. The current payment system has created motivation for a hospital to keep the beds full and extend the length of stay, since high occupancy results in steady funding based on the per diem reimbursement. Low occupancy rate also increases the risk that the ceiling on the global cap would be lowered by HZZO. The current contract arrangement makes it difficult to adjust staffing levels in response to shortened lengths of stay and other efficiency gains, since staffing costs remain fixed. Thus, cost overruns are likely to result in the imposition of arbitrary intemal controls, e.g., by restricting the use of medications or procedures, rather than a response to improve productivity, such as reorganization of staffing and other systemic reforms. l4 This point system is a modification of the WHO point system. Diagnostic investigations and clinical procedures including specific surgical procedures are identified by a five-digit code provided along with the name of the test or procedure. In addition, a listing of medications used in treatment is provided. 28

47 Figure 10: Croatia Hospital Average Length of Stay and Cost per Case, Comparison of Average Cost of Care and Average Length of Stay of Croatian Hospital Cases, , , , , , =Average expense per hospital case +Average length of stay In July 2002, the Government introduced a case-based payment system based on the average cost weights determined for some 33 cases selected on the basis of interventions that are either high cost, high volume, or have a long a waiting list. Under this system, referred to as PPTP, HZZO negotiates volume contracts prospectively with all hospitals for these selected interventions, using case-based reference price in order to encourage a more efficient use of resources. This strategy is intended to reduce the waiting list while improving control over the total costs. As shown in Annex 2, the initial results indicate that the introduction of PPTP has had an impact in reducing the average length of stay in most of the interventions. In principle the movement toward a prospective case-based payment system should improve the ability of the system to reward outcomes rather than inputs, there are limitations in the current approach that will have to be addressed in future reforms. A fundamental flaw in the current PPTP system is the flexibility given to the hospital to choose whether to bill HZZO under PPTP or under the point-based system on a case-by-case basis. This leads to a new form of gaming, whereby the hospital is implicitly guaranteed the highest rate, thereby defeating the cost-controlling motivation behind the reforms. The use of broad based case groupings, as opposed to more detailed Diagnostic-Related Groups (DRGs), as shown in Table 20, creates incentives for a more traditional form of strategic behavior (creamskimming) whereby the hospital attempts to avoid high-risk high cost patients by dumping them on other providers. The Government intends eventually to move fully towards a comprehensive prospective caseadjusted payment system based on DRGs. This will represent an important step in rationalizing incentives in the system, and will alleviate some of the above concerns. At the same time DRGs create an incentive for another form of gaming known as code creep (see 29

48 Box 4). However, the U.S. experience shows that the problem of DRG creep can be kept low with minimal monitoring effort (Carter et al., 1990). Table 20: Comparison of U.S DRGs, Australian DRGs and Croatian PPTPs, Open- Heart Surgery U.S. Medicare Australian DRGs Croatian PPTP DRGS (Version 4.1) 106 Coronary Bypass with PTCA 107 Coronary Bypass with cardiac catheterization 108 Other cardio thoracic procedures 109 Coronary Bypass without cardiac Catheterization F05A F05B F06A F06B Note: 1. CC = complicating condition. Coronary Bypass with Invasive Cardiac Investig. Procedure with catastrophic cc I Coronary Bypass with Invasive Cardiac Investig. Procedure without Catastrophic CC Coronary Bypass without Invasive Cardiac Investig. Procedure with Catastrophic or Severe CC Coronary Bypass without Invasive Cardiac Investig. Procedure without Catastrophic or Severe CC \ J Coronary Bypass Surgery To address these issues the following recommendations are made: a. MOH and HZZO should retract the option to opt out of case-based methods wherever it is implemented. In the case of legitimate, high cost cases (i.e. due to medical severity and complications) an objective criteria for additional compensation to the hospital should be established. For example, the outlier payment rule under the Medicare PPS system can serve as a model. b. To the extent possible, more advanced case-based systems should be implemented in the first place, to avoid unnecessary switching costs. While it may be necessary to implement simpler coding systems initially, this should be viewed only as an intermediate step until information systems are brought up to par. Moreover, with appropriate training of nursing and medical staff, more advanced coding should be possible. Interviews of health professionals in Croatia indicated that in order to implement coding under the current PPTP classification system, more detailed knowledge by appropriate staff is required in the first place. 30

49 c. A related constraint is the limited scope available to the hospital management to respond to the new performance-based payment systems. Hospital management will need to have greater flexibility and autonomy in decision-making in order to achieve the desirable productivity gains. For example, the introduction of performance-based payments will need to be negotiated with the trade unions, whose support will be essential in the success of any hospital reorganization initiatives. BOX 4: Hospital Prospective Pavments (Case-based, Per Stav Pavmentl: Prospective Payments have become the most common method of payment for hospital services by both public and private insurers in high and middle-income countries. In this system, payment rate is set in advance (usually at the start of the administrative year) for each category o f inpatient care, and the payment covers all services provided between admission and discharge. For example, the rate for a hospitalization for appendectomy might be fured at $2,000, regardless of the number of hospital days, lab tests, or drugs provided to a particular patient. This system creates the incentives to control costs by avoiding unnecessary services (efficiency), since the difference between the payment rate and the actual cost of care is additional resources for the hospital. From the perspective of the payer, payment rates must be set low enough to achieve cost-savings, but not too low, so as to avoid gaming or strategic behavior by the hospital, which would look for ways to avoid high-risk and high cost cases (a process known as cream-skimming) or to under-provide services to such patients. The latter problem can be mitigated by setting the payment categories in as detailed a way as possible that closely mirror the clinical practices. A most common classification system used for acute hospital care is Diagnostically Related Groups (DRG). Since their introduction as part of the Medicare Prospective Payment System in the U.S. in 1983, many versions have come into being, such as the Australian Refined DRGs (AR-DRG). Currently, there are 527 Medicare DRGs, and 662 AR-DRG. Although differences in classification between the two systems exist (see Table 20), the larger number of AR-DRG is due to the inclusion of additional procedures, which in the U.S. would normally be provided outside of the hospital, rather than to a more detailed enumeration of a given procedure. For instance, in the U.S., colonoscopy (a screening technique for colon cancer) is almost always performed on an outpatient basis, and therefore is not assigned a DRG code. Under AR-DRG there are four different colonoscopy codes under acute hospital setting. Therefore AR-DRGs would be more suitable for adoption in countries that have similar health service organization. In many middle income countries broader groupings with fewer categories have been considered as a starting point for introduction of DRG. For instance in neighboring Slovenia, as in intermediate step in overhauling hospital financing, only 10 Groups were initially intr~duced. ~ Fewer groups tended to create more opportunities for gaming by hospitals and do not provide adequate information on hospital performance. form of gaming known as DRG-creep, or upcoding, designed to artificially inflate the payment- to the hospital. Slovenia is now rapidly moving towards implementation of modified AR-DRG, which would mitigate the tendency. Hospital payment reforms should be monitored not only for their impact on the hospital sector, but on their impact on other levels of care. Box 4, below, describes some of the unexpected consequences of the introducing the DRG system in the USA hospital sector. The Ministry of Health is currently implementing a restructuring of county hospital in order to introduce day-surgeries and other infrastructure improvements on the efficiency of Introduction of AR-DRG in Slovenia being implemented concomitantly with hospital global and sub-global budgeting. See Don Hindle,

50 inpatient care. The hospital restructuring is accompanied by a reconfiguration of the county health system to expand the altemative care services that will provide community-based postacute care and rehabilitation services. These altemative care providers are intended to provide post-acute care in a community-setting for the patients discharged from the hospital, and assure appropriate follow up care. A detailed review of this county-level activities should provide valuable information for the national reform process. C. Primary Care and Specialists Services The 1993 Health Law established general practitioners as the first contact and gatekeeper for the entire population. Although the system is gradually moving towards a family doctor system, primary health care is still organized around several medical specialties. About a third of primary care doctors specialize in general medicine, a smaller proportion in family medicine, and the rest are organized around pediatrics, gynecology and emergency medicine. Annually, each citizen signs up with a specific General Practitioner (family doctor), who becomes responsible for all his or her primary care and who controls access to the rest of the care system. The GP receives an annual capitation fee for this service, and the fees are adjusted primarily by the age of the individual patient, with locational allowances paid to GP serving in remote communities. Each GP is expected to carry a minimum of about 1,700 patients per year on a roster. This figure is low compared with GP rosters of patients in European countries, but the lower figure was established deliberately to encourage physicians to work in the under-served areas. Regional disparities remain in the distribution of primary care physicians, despite attempts to provide some incentives to serve in underserved areas. The Government has encouraged the privatization of primary health care (except for emergency services and public health services). By 1999 there were some 2,570 private medical doctors, representing about 25 percent of the total medical doctors, up from just 600 private doctors in As shown in Figure 12, below, the utilization rate of primary care services has been expanding steadily since The shift to capitation payments and privatization of primary care physicians was intended to give them incentives for more efficient and effective care. In their role as the gatekeeper, primary care physicians should play an influential role in determining the costs of health care by prescribing drugs and referring patients for specialist or hospital care. However, the current capitation payment system does not provide incentives that support the gate-keeping functions of the GPs. HZZO imposed limits on the average number of prescriptions and referrals per patient for the physician s practice. These limits are simple administrative counts of services and do not reflect the cost of these services or the service needs based on clinical or epidemiological evidence. Since prescription and referral limits are flexibly imposed, obtaining exemption to exceed the limit is permitted if the physician can provide evidence of patient needs. 32

51 Figure 11: Average Number of Services and Cost of Polyclinic and Specialist Services Per Beneficiary, Utilization Rate of Primary Care Services, i r 15 Y p * a 8.- c ~ 1 * 5 5 e Average visits per beneficiary t Average cost per beneficiary 1 The current system does not provide a basis for utilization reviews to monitor and evaluate the referral pattems of the GPs. In recent years, utilization rate of specialist services have continued to rise (see Figure 11). The current reform measures are intended to achieve the following: a. The incentives for GPs will introduce utilization and quality reviews, and performance-based payment system that encourage quality of care while discouraging unnecessary prescriptions and referrals. b. Re-introduce group practice and family medicine teams as the principle mode of delivery, and the contracts with HZZO will be developed to reinforce these trends. c. Support the development of clinical information systems and management tools that will enable utilization and quality reviews with minimal administrative burden. 33

52 Figure 12: Average Number of Services and Cost of Specialist Services, Polyclinic and Specialist Services - Average Cost and and Number of Services, i 35 I Year 0 Average number of services per beneficiary I 4 Average cost per service i One of the Government strategies for reform is to delegate greater responsibility for health services planning, coordination and financing to the county level. Achieving clinically appropriate and efficient use of resources would require improved coordination of care at the patient level, better integration with community-based services, and greater focus on prevention, early detection and better management of diseases. The Government's health reform program at Koprivnica-Knzevic County aim to promote better integration of health services and achieve efficiency and quality improvements by: reducing the dependence on acute inpatient, expand the availability of altemative care settings, including home and community-based care setting, and enhance the incentives and capacities of primary care providers to manage patient care, including better case management through appropriate referrals, prescriptions, diagnostics and laboratory services. These could include referrals, diagnostic and lab services, prescriptions, case (disease) management and, possibly, discretionary hospital care. A key feature of the county health system is the Community Health Organization (CHO), which is expected to play a critical role in the coordination of home and community based care (Sales, 2002). The CHO will build upon the elements of the existing community health centers that have traditionally supported the coordination of services in the past. The new CHO will develop those capacities to foster a more patient-focused seamless service from home to primary care to hospital and back to home. The CHO is intended to work closely with hospital, primary care, public health and social welfare to deliver (a) home care nursing and paramedical of hospital discharge planning, and (c) 34

53 coordination and management of the county emergency services, including the establishment of a new dedicated call center to provide the county with a single emergency help-line telephone number. The changes in the emergency services are intended to transform the present pre-hospital service into a single, integrated, emergency ambulance service for the county, including an effective triage of all emergency calls to ensure emergency ambulances are not used for non-emergency patient transfers, which has been one of the major source of high cost and inefficiency in the system. The concept of the CHO offers an opportunity for improving the coordination o f services across primary, community-based, and secondary hospital services. The potential benefits are significant: the new delivery structure could improve service quality and continuity of care while potentially reducing cost of healthcare and increasing effectiveness. The expected reduction in hospital length of stay cannot be achieved without an effective primary care network and CHO to coordinate post-acute care. Decentralization offers an opportunity to strengthen the planning and coordination at the county-level and improve continuity of care at the local level. A major challenge is to align the new coordinating functions with financial incentives in order to assure continuity of care among primary care, hospital sectors, and the CHO. The existing health financing system reimburses different levels of care separately and do not include incentives to promote continuity of care or encourage better coordination among different levels of care. D. Expenditure on Pharmaceuticals Spending on pharmaceuticals represent another major source of cost-escalation in any health care system, and health insurance agencies around the world have typically resorted to tight monitoring and regulation of coverage (e.g. positive lists), pricing policies (including copayment rates), and close monitoring of rational prescribing practice by the providers. In Croatia, once a drug is registered, its manufacturer may apply to have the drug placed on the Positive List for coverage under HZZO. The drug is reviewed by the Positive List Committee, comprising representatives fkom HZZO, Institute of Public Health, and clinical pharmacologists and clinicians, who evaluate the application in terms of clinical efficacy and affordability, and determine the percentage of the drug price for co-payment. It has been noted that the Positive List Committee lacks staff with background in health economics or pharmaco-economics to assess the economic evaluation presented in the application (Health Insurance Commission, Report on Pharmacoeconomic Concepts, March 2003). A proposal is currently under consideration to strengthen the economic evaluation process through capacity building of the economic evaluation skills within the HZZO department or through other entities such as the University of Zagreb or an independent National Drug Agency (NDA), which would be responsible for undertaking an independent economic evaluation of the drugs and providing technical advice to the Positive List Committee, including the estimation of the total cost of listing the drugs based on epidemiological profile and prescribing patterns. The drug preference price was introduced in 1999 as an additional means of containing the rise in expenditure on drugs.i6 The prices of drugs are determined biannually by comparing wholesale prices from three reference countries (Slovenia, France, Czech Republic and two additional countries where necessary). An initial assessment by Health Insurance Commission of Australia compared the Croatian wholesale prices of a sample of 45 drugs l6 Reference drug pricing establishes fixed reimbursement rates for products assigned to the same group, i.e., are interchangeable because they have similar therapeutic action or produce similar clinical outcomes. 35

54 against the Australian price for comparable product. The results show that 36 percent of drugs were below and 64 percent were above the Australian prices. This comparison was made since the Australian reimbursement prices are some percent of the world average (Health Insurance Commission, 2003). While this represents an assessment on a limited sample, the exercise indicates that there might be opportunities for further cost savings by changing the reference pricing process, including possible changes in the list of reference countries. However, the relatively small size of the Croatian market would present a disadvantage in negotiating prices with the manufacturers. One possible solution, for future consideration, would be to form a regional purchasing group with the neighboring countries to improve the purchasing power. Figure 13 Number and Average Cost of Drug Prescription covered by HZZO, O =Number of drugs prescriptions per beneficiary 1 1 t Average expense on drug prescriptions (constant 1997 price) International experiences have shown that in order to achieve an effective control over the total expenditures on drugs, pricing controls would have to be combined with explicit demand-side incentives and volume controls (see, for example, Mrazek, 2002). To control the volume of prescriptions, HZZO has imposed a ceiling of no more than 5 prescriptions per beneficiaries and limited the number of drugs per prescription, but exceptions are allowed for special cases. As indicated in Figure 13, above, the ceiling on prescriptions per beneficiary has not been effective as the average number has increased from 5 prescriptions in 1996 to 7 prescriptions in HZZO will need to introduce measures that will align demand-side incentives that encourage appropriate and safe prescribing and use of pharmaceuticals based on clinical needs, rather than on administratively established limits. At present HZZO does not review the prescribing practices or include them as part of the performance indicators for 36

55 payments, although these indicators are currently under consideration in the revised primary care payment system. For this purpose, HZZO will need to: (a) establish rational prescribing targets and strengthen utilization and quality reviews by HZZO; and (b) disseminate prescribing protocols, promote continuing medical education for physicians and pharmacists, and public education on the safe use of pharmaceuticals. Prescription drugs (HRK, millions) Prescription drugs (HRK, millions in constant 1997 price) As % of total HZZO expenditure on health ,345 1,831 1,529 2,238 2,096 1,664 1,345 1,689 1,356 1,864 1,728 1, % 17.9% 13.4% 16.9% 16.5% 14.1% As shown in Table 21, HZZO s expenditures on drugs have decreased significantly between 2001 and This may reflect the effect of increased co-payment rates under the 2002 Health Insurance Law, rather than effective reduction in the volume of prescriptions. This is evident from the fact that the average prescription per beneficiary has not declined over the same period. The drug expenditure data shown represent only those expenditures covered by HZZO, and do not capture private out-of-pocket spending. The extent to which the cost containment has been achieved by shifting costs directly to the patients needs to be monitored from both access and quality perspectives, for example, it would be important to know whether this might be leading an inappropriate use of drugs (e.g. not completing the course of treatment) from the perspective of patient safety and appropriate care. E. Expanding the Scope of Public Health Programs Croatia has a long-standing history and tradition in public health programs under the leadership of the Andrija Stampar School of Public Health. Under the guidance of the national Public Health Institute, county public health institutes have been responsible for undertaking traditional public health functions, such as health surveillance, hygiene and infectious disease control activities. In recent years, the public health system has been expanding programs that address the new health risks arising from high incidence of noncommunicable diseases and injuries. Since almost all the causes underlying premature deaths due to cardiovascular diseases, cancer and injuries can be influenced by changes in behavior and habit, an effective prevention program targeted to these changes could significantly reduce the burden of disease, and hence, the cost on health care system. The alarming increase in smoking prevalence especially among the youths has recently prompted the Croatian Government to launch a tobacco control program, with assistance from the IBRD Loan. A preliminary evaluation of the program indicates a significant decrease in smoking prevalence has been achieved since the start of the anti-smoking campaign. Other types of health promotion interventions that support appropriate and rational use of health services and drugs are expected to receive significant support to complement other measures designed to influence demand for health care. The local public health institutes should be supported more broadly in playing a wider role at the local level in ensuring that the public health functions are well integrated into the evolving health systems. The public health institutes at national and local levels, in close 37

56 collaboration with HZZO, should ensure that the incentives for providers are consistent with the optimal care and public health goals. Public health institutes could also provide support to the county government in monitoring and evaluating the performance of the local health system. These activities would be consistent with the expanding role of local governments in public health and health care services, but they will require significant capacity building of the local public health institutes in the context of the new responsibilities and functions of the local government. In this context, the relatively low level of public spending on public health programs by both central and local governments is a concern, and should be addressed in the future budget allocation process. F. Compensation and Allowances Croatia continues to provide for one of the most generous sick leave compensation by international standards. Since the state takes on almost the entire risk of added labor costs due to illness or maternity, there is little incentive on the part of the employers and employees to be judicious in the use of sick benefits. As a result, there are indications that the current system is subject to abuse, often as a result of collusion between employer and employee, who may use the sick benefit for other purposes, e.g., in lieu of unemployment benefits. Under the new Health Insurance Law, some modest reduction in the level of compensation has been introduced (see Table 22) but the benefits remain essentially unchanged. A more comprehensive analysis should be undertaken to explore alternative strategies for providing protection for workers that might transfer a greater share of responsibilities to employers and employees, while targeting the use of state budget to categories of population who are most vulnerable (e.g., the disabled). A cost benefit analysis could show the tradeoffs between maintaining this compensation system against other benefits that could provide support to workers and families without imposing a rigid cost structure of the current compensation scheme. For example, the development of day-care facilities and introduction of flexible working hours are examples of measures used in the industrialized countries that offer flexibility to both the employer and employee. Table 22: Changes in Sick and Maternity Leave Benefits, 2002 Original Status Replacement rate 80 percent up to 6 months, using the average net wage in the previous six months as a reference, 90 percent up to 1 year and 100 percent for special cases. New Health Insurance Law 2002 Replacement rate 70 percent using the average net wage in the previous six months as a reference, and 100 percent for special cases (sick-leave due to injuries sustained at work or occupational diseases; sick-leave due to injuries and complications in connection with pregnancy and birth; exercising the right to mandatory maternity leave; care of an ill child younger than three years of age; sick-leave due to transplant of life organs on behalf of the second party; in the case of infectious disease). No wage reimbursement for the first three days of sick leave. From 4 to 42 days, it is financed by employer and from 43 days by HZZO. Maximum sick leave duration is 12 months. 38

57 Since policy formulation and analysis of sick and maternity benefits are more appropriately considered part of employment policies rather than of health care, consideration should be given to moving its administration outside of the purview of the health insurance fund. This will permit, on the one hand, to integrate the sick benefits into labor and social welfare programs, while releasing HZZO to concentrate its resources on developing the expertise and capacity in its core functions, namely, the financing of health services to ensure access to cost-effective and high quality medical care for the covered population. IV. CONCLUSIONS AND RECOMMENDATIONS The broad range of strategies being employed coupled with improved health and financial indicators suggest that the Govemment is seriously attempting to improve its policy environment in order to provide universal access to quality health care for its the citizens. Croatia s challenge is to channel its already substantial public spending on health care to strengthen the capacity to manage its resources and strengths with greater effectiveness. Recent reforms appear to have succeeded in containing the increase and even bringing about a reversal in the level of public spending on health. However, demographic and epidemiological transition points to a continuing in the demand for health care. Globalization and integration with the EU market would also subject Croatia to increasing pressures to adopt high-end medical technologies. The pressure on public resources to spend more on health will intensify in the coming years, and will have to be met with prudent allocation of resources and continuous efforts to improve productivity wherever possible. Irrespective of improvement already made in the financing and organization of health services, under these constraints providing universal access to care and maintaining quality of services will be a challenge. In this section, we summarize the main impact of the recent health financing reform in terms of the key objectives of the health system: universality and access, progressivity and fairness in contribution, efficiency, quality, and patient choice and satisfaction. Further, we present recommendations for undertaking the next steps that would be needed to build upon these ongoing reforms. A. Universal Access to Care and Social Protection The health financing system has been effective in providing universal access to health care. Though small on average, increase in co-payment rates could have negative repercussions on access to care for certain groups in the population. It is evident from the household expenditure data (HBS 2001) that a significant percentage of the households with pensioners or disabled persons face high out-of-pocket costs. This would be consistent with the notion that the elderly and disabled persons have chronic conditions that make the particularly vulnerable to large co-payments for expensive treatment and continuous care. With the disproportionately high degree of cost sharing placed on such groups, the current statutory insurance plan may not always provide adequate protection and coverage for those most in need. Co-payments should be maintained in place to mitigate excessive utilization of services. But under the current pricing structure, co-payments on basic health services are left open-ended, and this has presented a particularly heavy financial burden on certain groups of beneficiaries, notably those with chronic diseases. Thus, by offering a relief from the financial burden of co-payments, SHI has therefore been particularly welcomed by advocates 39

58 of patients with chronic disease^.'^ However, SHI introduces other distortions to the system,l* and can only be viewed as a stop-gap measure for addressing the needs of the chronic patients. Alternative solutions should be considered, such as ceilings on co-payments for different beneficiaries, which is a commonly used policy in many social health insurance programs. Another issue that requires attention is the extensive system of exemptions on co-payments and premiums on the HZZO plan. Despite efforts to reduce the number and categories of citizens exempted from premiums and co-payments, the recent round of reforms have not yet made significant progress in this area. These broad-based exemptions effectively reduce the net contributions from those who are able to pay, and therefore undermines the principles of progressivity in contribution according to ability to pay. Exemptions should be targeted only to those groups who are unable to access to services otherwise. Adding multiple exemptions and discounts not only reduces the contribution rates, but exacerbates the lack of transparency in the process. Greater transparency would therefore help to persuade the public to accept the removal of exemptions if this is seen to be done in a fair and open manner. In this connection expanding the consultative process with the beneficiaries and employers would be essential for mobilizing public support. Finally, it would be essential to obtain a better understanding of the scope of informal payments which appear to be widespread and to affect access to health care for a significant number of citizens. The recent cost-containment measures have shifted greater financial risks to the provider and the patients, and unless these are offset by productivity gains, it is possible that the households would be bearing a major share of that risk. Careful monitoring of the impact of these reforms on actual household expenditures would help to identify problems early and suggest solutions to mitigate these effects. B. Diversifying Sources of Contributions Continued reliance on payroll tax places a heavy burden on the productive labor force and on the economy. The Government has made some progress in this regard by gradually reducing the contribution rate from 18 percent in 1998 to 15.5 percent in 2003, with M her reductions envisioned in the coming years. Under the Health Insurance Law of 2002, these reductions were expected to be compensated through other revenues, including transfers from central and local governments which would cover part of the subsidies for pensioners, unemployed, disabled and other social cases. Despite the reduction in payroll contribution rates, the total revenues from payroll has actually increased over this period, suggesting that a combination of improved collection and reduced exemptions may be having an effect. What is not clear is the extent to which transfers from general revenues would be used to substitute payroll tax for subsidizing the social welfare cases. To date, actual transfers from general revenues have not covered the full cost of subsidies to these groups. In fact, in 2002 the central budget transfer to HZZO dropped to 9 percent of the total revenues, down from 16 percent o f the total HZZO revenues in At the same time, the Government borrowed HRK 820 million (6 percent of total revenues) to pay down the old arrears built up by the government-owned health care providers toward their suppliers. The end result is that the health financing through social insurance remains primarily financed through payroll tax and deficit financing, which contradicts the intended aim of the reforms. This outcome suggests that the cost of financing l7 Based on author's interview with the representative of the Croatia Patient Societies for Diabetes, Zagreb, April l8 By removing co-payments to all subscribers, SHI reintroduces the moral hazard problem, with added transaction costs. 40

59 the subsidies has not yet been fully evaluated or included in the budget plan. The parameters for determining subsidy levels will need to be defined for transparency and for planning purposes. C. Reducing Compensations and Allowances A significant portion of HZZO funds are used to finance compensations and allowances, including sick and maternity leave, and in 2002 included worker's compensation was added under HZZO. In 2002, expenditures on this category of spending accounted for some 14 percent of total HZZO expenditure. Since these compensations introduce distortions and rigidity in the labor market, the continuation of these benefits under the social insurance program should be re-evaluated in the context of the national labor policy. A reduction in these allowances is likely to have a far greater impact in reducing public spending than any marginal productivity gains within the health sector. Furthermore, based on the experiences of other industrialized countries with ageing population, Croatia will likely face new demands on social insurance resources in order to finance long-term care for the elderly. In anticipation of these future demands on social insurance, Croatia should find means of reducing the generous allowances for sick and maternity leave in order to make room for future programs in elderly care. D. Promoting Quality and Efficiency of Care through Payment Reforms International experiences show that cost, volume and quality of services will need to be actively managed simultaneously, and monitored closely in order to achieve improvements in productivity and quality. Administratively imposed limits, such as capping the referrals and prescriptions per beneficiaries at the primary care level are not effective means for improving productivity or quality and HZZO is encouraged to move away from the reliance these methods of cost containment. The Government has recognized the need for reforms in the payment contracts between HZZO and the providers, and have initiated a number of important reforms aimed at aligning provider incentives with desirable performance standards. a. HZZO is currently preparing the introduction of performance-based payment for primary care providers which would add a fee-for-service component to the capitation payment system. The proposed payment system would include utilization and quality reviews which encourage appropriate pharmaceutical prescribing practices and referrals to specialists, hospitals and other levels of care. These efforts provide an excellent starting point for payment reform. One of the challenges is to find an appropriate payment system that would support the integration of care at the county-level. The current efforts are still focused on payments for individual providers. It is recommended that alternative forms of payment, e.g., through a network of primary and secondary care providers, that encourage continuity and patient-centered care. b. As Croatia moves towards implementing the hospital prospective payment system (PPTP) more widely, the government needs to ensure that incentives are aligned throughout the system in a consistent manner. The current option allows free movement between PPTP and fee-for-service payments, which could lead to gaming of the system and undermine the cost-containment objectives of the MOH and HZZO. It is recommended that the 'opting option' in PPTP be significantly curtailed. Partial "blending" (weighting) of PPTP and fee-for-service can be used to compensate hospitals for high-risk, high cost cases. There is also a need to develop a 41

60 methodologies for identifling 'outlier' cases, which would qualify for blended payments. The 33 payment categories under PPTP represent an excellent starting point for development a comprehensive case-based payment system. However, the limited number of categories also imply a wide range of medical severity of patients within each grouping, resulting in winners and losers among hospitals due to reasons unrelated to efficiency. Implementing well-established grouping systems will help in mitigating this problem. The experience of other middle-income countries, such as Slovenia in adopting Australian DRGs countries could be studied for possible adaptation and implementation in Croatia. Review of the co-uavment system. When used appropriately, co-payments can be used as a policy lever to reduce excessive utilization (moral hazard) in certain expenditure categories. While HZZO and the Croatian legislature have emphasized the role of co-payments as a means of revenue-sharing, their potential role in limiting utilization may not be fully recognized. Future research and policy evaluation should aim to target services where inefficiencies may exist, particularly in routine outpatient services, such as office visits. At the same time, care must be taken to avoid imposing high co-payments for preventive services, as under-provision of such services may result in cost spillovers. Primary examples include preventive prescription drugs (e.g., hypertension treatments, anti-diabetic drugs). Finally, some alternative approaches to SHI should be explored to reduce the burden of copayments on patients suffering from chronic diseases or disabilities. Building the Cauacitv of HZZO. In order to support the rational allocation of budget, HZZO will need to enhance its own capacity to undertake economic evaluations and actuarial analyses of its financial flows. Such analyses will be necessary to establish premiums and co-payments rates to ensure that a desirable level of revenues is obtained, and that any new health insurance product (e.g., SHI) is financially viable. These estimations would also help to establish more reliable estimates of the subsidies required fkom the central and local budgets, for example, by taking into account expected changes in the demographic composition of the beneficiaries; changes in the utilization and costs of health services due to new investments in technology or behavioral responses to new payment incentives. UugradinP of Health Management Information Systems. Modem health care service involves a complex production process, with considerable variability inherent in the clinical procedures, practices and outcomes. This complicates the evaluation of price, quality and productivity of health care services, and necessitates the development of a comprehensive health management information system (HMIS) that supports a multidimensional evaluation process. A HMIS will be needed to generate accurate, timely and consistent (i.e. comparable) information on the use, cost and quality of care across multiple institutions. For example, management of the performance-based payment systems for hospitals and primary care providers will require a database that can link cost and utilization data with standard benchmarks for productivity and quality of care. The next generation of provider payment system for the health care providers in Croatia will require significant upgrading of the information systems on both the payer (HZZO) and provider perspectives. HZZO has already established a strong information system network that provides effective support for the existing payment systems, but HZZO will need to upgrade its information system to support the more sophisticated utilization and quality reviews of the health care providers required under the new performance-based contracts. The new provider payment systems will also necessitate a parallel upgrading the information systems in the health care providers contracted under HZZO. In recognition of the importance of information systems in health are, the Ministry of Health has embarked on the development of a National Strategy for Health Information System, and investments in HMIS 42

61 will likely feature prominently in supporting the next generation of health sector reforms in Croatia. Defining standards of care based on evidence. Quality of health care services is difficult to define in terms of productivity and effective outcome, such as reduction in morbidity and mortality, and depend on both the application of evidence-based medicine and consultative process among the professionals and other key stakeholders to arrive at a consensus on quality standards and measures (e.g., clinical guidelines, pharmaceutical prescribing guidelines, standards for adopting drugs on registry of on HZZO positive list, etc.). Under the leadership of MOH, a number of to initiatives are being taken to develop standards of care and products by different specialties, and program settings. However, many of these activities need better coordination and an appropriate consultative and review process with all the key stakeholders, including the Ministry of Health and HZZO. Audits. To be effective, provider payment systems require both a strong management information system to provide appropriate and timely information on utilization and performance, and an effective audit mechanism to ensure compliance and promote good practice. The audit system currently in place in HZZO emphasizes punitive measures rather than supportive measures that promote good practices and continuous quality improvements. Due to lack of standard protocols and guidelines, the decisions on appropriate care tend to be based on the personal experiences of the doctor. There is a need to coordinate the various activities to introduce standards of care among the key stakeholders, including HZZO, the professional associations, and the providers, and to design and audit system which emphasizes quality improvements rather than on punitive measures. Increasing Allocations to Public Health Programs. Investments in public health programs are expected to have significant long-term benefits by reducing the incidence of costly diseases, improving health outcomes and quality of life of the population, and limiting the high cost of treatment and rehabilitation of patients. Since a significant share of health care costs in Croatia are attributable to non-communicable disease and chronic conditions, resources should be targeted to support public health programs that address lifestyle and behavioral changes that reduce risks to health, such as good diet, physical exercise, and smoking cessation. Croatia already has a strong capacity to maintain public health, and these strengths should be exploited to the full in support of effective health promotion and prevention programs. E. Improving Quality and Efficiency of Care: Supply-side Intervention In its role as the owner of the major network of health care providers, the Govemment has embarked on two major initiatives designed to address inefficiencies in the existing health care delivery structure. First is the rationalization of hospitals, which aims to reduce the overall number of beds, and improve the distribution and types of hospital services and staff to meet the changing epidemiological profile of the population. This reform would also include consolidation of specialties (centers of excellence) to improve economy o f scale and scope. This is an important investment that will have long-term implications on the cost of care, since the hospitals could become the operating efficiency and quality of care. The second initiative is the decentralization of the primary and secondary health services to the local government, with the intention of improving coordination and continuity o f care at the local level. The central government is collaborating with the local government to invest in reorganization and modemization of care. The new decentralized delivery system will require strengthening payment and performance monitoring systems that would send appropriate incentives for the local network of providers. The role of branch offices o f the 43

62 HZZO as the purchasing authority could be enhanced for better managing the payment system in the decentralized setting. F. Promoting Patient Choice through Private Voluntary Health Insurance Despite recent setbacks in the private insurance market due to the 2002 Health Insurance Act, there are opportunities for the private insurers to expand and improve the quality and choice of care for the beneficiaries. Currently there is no system in place for monitoring the prices and products of Voluntary Health Insurance schemes. Government should give priority attention to strengthening its capacity to monitor VHI market. There will be greater need for monitoring of the VHI products as coverage expands and other insurers enter the market. Specifically, following actions are recommended: a. The capacity of the existing Office of Insurance Regulation should be expanded to monitor and evaluate new VHI products, and disseminate information to the consumers. This function is essential for ensuring that the insurers do not resort to risk selection to enhance their profitability and that potential and existing subscribers have access to adequate levels of information to allow them to make appropriate choices. b. The impact of the Supplemental Health Insurance and other VHI products on utilization of health services should be monitored closely, e.g., to assess whether VHI coverage provide an incentive for providers to treat VHI patients differently from those not covered, or whether VHI patients use their health service differently from those who are not covered. This would ensure that any impact of VHI on equity and appropriate use of care would be detected early and allow appropriate actions to be taken. c. With regard to the regulation of VHI, it is recommended that this option be evaluated against the expected timing of EU accession. Since EU Insurance Law restricts VHI regulation to matters related to financial solvency and does not permit other forms of regulation such as community rating, it may not be advisable to introduce these regulations at this time if they have to be repealed again in the near future. Nevertheless, Government should monitor and evaluate VHI products and make this information widely available to consumers, as indicated in (a) and (b) above. The combination of statutory basic coverage and the availability of additional services through VHI could balance the competing objectives of ensuring access to care with consumer sovereignty and choice. Private insurance should be promoted in that context. G. Improving Planning, Coordination and Consultative Process The need for reform in the health sector is widely accepted among both health professionals and govemment officials, and this has permitted successive Governments to introduce a continuous process of reforms in the health sector. Thus, Croatian health reforms have been comprehensive and attempt to address all aspects of the health system. But the proliferation of reform measures has also strained the system s capacity to coordinate, implement and evaluate the reforms effectively, resulting at times in contradictory policies and mixed incentives in various programs. The Ministry of Health should strengthen its role in developing a coherent long-term strategy, build capacity for policy planning, modeling and evaluation. In addition it will have to enhance the process of consultation and coordination among the key stakeholders, including the civil society and professional associations. 44

63 ANNEX 1 : HOUSEHOLD EXPENDITURES ON HEALTH - RESULTS OF THE HOUSEHOLD BUDGET SURVEY 2001 The following section is based on the analysis by Danijel Nestic using the results of the 2001 Household Budget Survey. The data source is the 2001 Household Budget Survey, and the main aggregates calculated are close to officially released figures (CBS First Release No from 11 July, Description of HBS Data The sample frame used in the survey dates from pre-war period and will require adjustments using the latest census data. The Government has been updating the sample frame using the census data carried out in April but these were not available at the time this report was prepared. Expenditures on health were collected for the following categories: a. b. C. Medical products, appliances and equipment (pharmaceutical products, thermometers, bandages, contraceptive devices, orthopedic devices, corrective eyeglasses and contact lenses, hearing aids...) Out-patient services (consultation of physicians in general or specialist practice, dental services, services of medical analysis laboratories and X-ray centers, services of medical auxiliaries, ambulance services, etc.) Hospital services (accommodation, food and drink, ambulance transport, provision of medicine and medical products, medical services, administrative expenses) Health expenditures include the total amounts that households actually pay directly and does not include reimbursement made by the social welfare organizations or by private insurers. Household expenditures on final consumption and total income are defined in accordance with standard statistical practice for HBS purpose, but both without imputed housing rent. The recall period is four months, which is likely to be too long for reliable recall of outpatient services and medical supplies. In household surveys on health care spending, recall period for these items are usually indicated between two weeks to four weeks maximum. Results of HBS 2001 Data In the first step, income and expenditures were calculated at the level of household. In the second step, household income/expenditure is divided by the number of household members. The resulting estimated amount is attributed equally to every household member. In summary, all amounts are expressed in kuna values (HRK) per capita per year. Unit of observation in all calculation is the individual. 45

64 Table 23 Mean Household Spending on Health Care, Mean HRK per capita per year % of total household expenditures Household health expenditures per capita o/w medical products o/w out-patient services o/w hospital services % (0.09) 1.24% (0.05) 0.71% (0.06) 0.07% (0.02) Household expenditure per capita GDP per capita Household expenditure on health, as % oj 22,092 HRK 36,712 HRK 1.22% GDP Note: Standard errors are reported in parentheses (clustered on household identifiers). GDP per capita is estimated by dividing total GDP in 2001 (162.9 billions of kuna) by population (4.437 millions). Data sources: 2001 Household Budget Survey (expenditure figure) and CBS Monthly Statistical Report No Table 24: Health Expenditures by Location Note: Urbadrural division is based on interviewer s assessment (given within HBS). Source: 2001 Household Budget Survey. 46

65 Table 25: Household Health Expenditures by own welfare status, Unemployed Retired Other inactive Disabled Age<= 15 Tota , % 419, , % 277,308 21, % 818,712 20, % 581,598 19, % 312,818 20, % 687,886 22, % 4,238,764 Table 26: Household Health Expenditures by Income Quintile Groups Note: Quintile groups are created according to total income per capita. Source: 2001 Household Budget Survey. 47

66 Table 27: Household Health Expenditures by Household Types, 2001 ita, % One person hh, between 30 and 64 years One person hh, 65 years plus % 4.29% adults, no dependent children, at least one adult adults, no dependent children, both adults under 65 Other hh without dependent children Single parent hh, one or more dependent children 2 adults, one dependent children % 2.27% 2.14% 1.11% 1.70% adults, two dependent children % adults, three or more dependen children Other hh with dependent children Tota , % 1.60% 2.02% ,238,764 48

67 Table 28: Distribution of Household Health Expenditures by Social Welfare Status, 2001 I I Total Note: Social welfare status is based on the most fiequent activity status in the last 12 months, except for disabled persons where status is defined by self-reported disability or receiving invalidity pension or receiving other disability benefits. Data source: 2001 Household Budget Survey. 35% c % 5% 0% Per capita spending on health per year, Kunas I 49

68 Table 29: Distribution of relative health expenditures, by income quintile groups Within-group relative frequencies by quintiles I 0% 0-1 % 1-3% 3-5% 5-10% % +15% Total Jote: Quintile groups are defined according to total expenditures per capita. Data source: 2001 Household Budget Survey. 40.0% 35.0% 30.0% CI C 25.0% Q.- C E" 20.0%.- 0 CI 3 n L 15.0% i3 10.0% 5.0% 0.0% 0% 0-1% 1-3% 3-5% 5-10% 10-15% 1520% 20-30% +40% % household spending on health (not drawn to scale) I + Quintile Quintile 2 + Quintile 3 -e- Quintile 4 + Quintile

69 Table 30: Distribution of relative health expenditures by social welfare status. Lelative expenditures1 Within-group relative frequencies Retired 0 0-2% 2-5% 5-1O% + 10% Total 23.5% 32.8% 51.O% 46.6% 18.6% 14.2% 5.1% 3.9% 1.8% 2.4% 100.0% 100.0% 16.6% 3 2.4% 26.5% 17.8% 6.8% 100.0% 35.0% 30.0% 25.0% C c 20.0% Q P c.- E.- 0 g E % 10.0% 5.0% I 0% 0-1 % 1-3% 3-5% 5-10% +IO% % household spending on health (not drawn to scale) -e- Employed - -m- - Self-employed ---a--- Unemployed -@- Retired --31t Disabled 51

70 ~ - R n 3 - D c) ", t D u oc ~ 3 3 T 7 3 Q\ d