Management of Key Purchaser Risks in Devolved Purchase Arrangements in Health Care

Cumming, Jackie Working Paper Management of Key Purchaser Risks in Devolved Purchase Arrangements in Health Care New Zealand Treasury Working Paper, No. 00/17 Provided in Cooperation with: The Treasury, New Zealand Government Suggested Citation: Cumming, Jackie (2000) : Management of Key Purchaser Risks in Devolved Purchase Arrangements in Health Care, New Zealand Treasury Working Paper, No. 00/17, New Zealand Government, The Treasury, Wellington This Version is available at: https://hdl.handle.net/10419/205432 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by/4.0/ Disclaimer: The views expressed are those of the author(s) and do not necessarily reflect the views of the New Zealand Treasury. The Treasury takes no responsibility for any errors or omissions in, or for the correctness of, the information contained in these working papers. TREASURY WORKING PAPER 00/17 Management of Key Purchaser Risks in Devolved Purchase Arrangements in Health Care Jackie Cumming Contract to the Treasury ∗∗ ABSTRACT This paper examines the key risks and risk management strategies associated with devolving the purchase of publicly funded health care to non-governmental organisations, where the aim is to maintain at least the current level of equity of access. The key risk is greater ‘cream-skimming’ or risk selection, particularly with competitive purchase models based on patient choice of purchaser. The paper also examines the risks of poor purchaser performance and cost shifting. Minimising these risks would require considerable government regulation. This paper was commissioned and carried out to inform Treasury’s thinking in helping develop health policy under the National Party Government that was in power from August 1998 to November 1999. It is a background paper and does not provide policy advice, nor does it propose any particular course of action. The Treasury has chosen to make it publicly available in order to encourage public discussion. * Jackie Cumming, Senior Research Fellow, Health Services Research Centre, (Institute of Policy Studies, Victoria University of Wellington, and, Wellington School of Medicine, University of Otago) Management of Key Purchaser Risks in Devolved Purchase Arrangements in Health Care Table of Contents Acknowledgments.....................................................................................................iv About the Health Services Research Centre............................................................iv Disclaimer.................................................................................................................iv Executive Summary ......................................................................................................v Recommendations....................................................................................................vi 1 Introduction .............................................................................................................1 1.1 Methods.............................................................................................................2 2 Devolved purchasing in New Zealand.....................................................................3 2.1 Integrated care..................................................................................................3 2.2 Integrated care and devolved purchasing in New Zealand................................4 2.3 Integrated care and purchaser competition.......................................................5 2.4 Devolved purchasing arrangements in New Zealand........................................5 2.5 Devolved purchasing and policy risks ...............................................................7 3 Risks associated with devolved purchasing............................................................9 3.1 Cream-skimming (risk-selection).......................................................................9 3.2 Poor purchaser performance...........................................................................12 3.3 Cost-shifting ....................................................................................................15 3.4 Risks and devolved purchasing in New Zealand.............................................19 4 Management of risks in devolved purchase arrangements...................................21 4.1 Managing the risks associated with cream-skimming......................................21 4.2 Managing the risks of poor purchaser performance........................................37 4.3 Managing the risks associated with cost-shifting.............................................57 5 Conclusions...........................................................................................................61 5.1 Cream-skimming.............................................................................................61 5.2 Poor purchaser performance...........................................................................65 5.3 Cost-shifting ....................................................................................................66 5.4 An optimal solution to managing these risks? .................................................67 6 References............................................................................................................69 Appendix A Glossary of Terms and Abbreviations......................................................77 Appendix B Integrated care.........................................................................................80 Appendix C Purchaser competition.............................................................................87 Appendix D Purchaser competition in New Zealand...................................................90 Appendix E Standardisation of benefit packages........................................................96 Appendix F Service specification in the Netherlands...................................................97 iii Appendix G Service specification in United States Federal and California HMO legislation..................................................................................................................102 Appendix H Service specification in United States Federal Medigap legislation.......105 Appendix I Health Employer Data Information Set (HEDIS)......................................111 iv Acknowledgments I would like to thank The Treasury for the opportunity to undertake this work. Thanks especially are due to Ken Hand, Annette Bridgman and Kim von Lanthen at the Treasury for the many helpful discussions we held while writing this report. Nicholas Mays, also of the Treasury, provided me with copious comments, which were, as always, extremely insightful. I would also like to thank Professor Ray Robinson, London School of Economics and Political Science, for the valuable comments he made in reviewing this report. About the Health Services Research Centre The Health Services Research Centre aims to promote excellence in academic research, to encourage interaction between researchers and policy makers and to mount policy-focused research as well as evaluative research. The Centre is a joint venture between the Institute of Policy Studies, Victoria University of Wellington and the Wellington School of Medicine, University of Otago. The Health Services Research Centre gratefully acknowledges the role of the Health Research Council of New Zealand in providing core funding for the Centre. Disclaimer The views expressed in this paper are those of the author alone. No other individual or organisation takes responsibility for any errors or omissions in, or for the correctness of, the information contained in this report. v EXECUTIVE SUMMARY This paper was commissioned by The Treasury to provide an overview of the key risks associated with devolved purchasing arrangements and purchaser competition in health care in New Zealand. For the purposes of this paper, devolved purchasing is defined as any purchasing strategy or policy which allocates responsibility for purchasing groups of services (for a particular population) to non-government organisations other than the Health Funding Authority (HFA). Organisations other than the HFA then become responsible for negotiating contracts with other providers and for funding such services from budgets allocated to them. In some cases, the range of services for which purchasing is devolved may be rather limited (e.g., to primary care services) or it may be comprehensive (e.g., cover all health care services). Examples of organisations which might take on this devolved purchasing role include general practitioners (GPs), independent practitioner associations (IPAs), hospital and health services (HHSs), or Iwi organisations, community providers or health insurers. Purchaser competition occurs where people can choose which of a number of devolved purchasing organisations (DPO) has the responsibility to arrange for some or all of their care. The paper does not discuss in depth the overall merits and demerits of moves to devolve purchasing in New Zealand. The paper also does not discuss in detail the potential benefits of purchaser competition, the transaction costs that might arise with such competition, nor the transition issues in establishing purchaser competition. Rather, the author was asked to focus on the following key risks: • cream-skimming – which occurs when health providers or DPOs – either deliberately or by chance – enrol a favourable mix of members, i.e. those with good health and hence a lower chance of making a claim or needing care than the average person; • poor purchaser performance – which occurs when DPOs as agents fail to meet the objectives of principals (e.g., the HFA, the government, taxpayers or communities.); and • cost-shifting – which occurs when costs are charged against a budget that differs from that intended, or from that traditionally charged. Cream-skimming In the current financing environment, successful cream-skimming in New Zealand could lead to: increased overall expenditure; pressure for additional expenditure; inequitable access to services and quality care; limited incentives for technical efficiency and cost-effectiveness; and high levels of profits for some DPOs while others struggle. Cream-skimming has the potential to be a serious problem in New Zealand if we move to payment systems which place providers at greater financial risk and which maintain or further encourage competition between providers or devolved purchasers. The likely extent of cream-skimming cannot be judged at present. One gauge of the significance of the issue, however, is the attention that is being given to regulatory vi reforms in the United States to reduce its prevalence: this suggests that if creamskimming were to become widespread, each of the risks mentioned above is likely to be significant, both in a financial and political sense. Poor purchaser performance The risks associated with poor purchaser performance – relative to good performance – are potentially many. These include: higher overall expenditure for a given level of care or a lower level of care for a given level of expenditure; lower access to services and reduced quality of care; reduced equity of access; increased private financial responsibility for care; limited improvements in health status; and higher overall expenditure as expenditure is shifted to the private sector, which may be less able than the public sector to contain overall costs. The likelihood of such risks is difficult to judge. Any losses from poor performance in delivering value-for-money are difficult to estimate – from all perspectives. The size of the health care market means that systematically poor purchaser performance will lead to significantly lower levels of health or higher expenditures than necessary. This, however, may be balanced by having multiple agencies involved in the delivery of health care: failure on the part of a purchaser to perform may not necessarily be reflected in poor service delivery. Cost-shifting The key risks associated with cost-shifting are also potentially numerous. They include: increased government expenditure on votes other than Vote: Health; increased expenditure for publicly-owned purchaser and provider organisations were costs and patients to be shifted to publicly-owned purchasers and providers; increased private expenditure and resource use were purchasing authorities to make decisions which shift responsibility for financing onto private budgets; increased shunting of patients; higher overall expenditure, if claims in the private sector increase with more care provided privately at higher cost; reduced equity of access if those who cannot obtain publicly-funded care are unable to afford private care; difficulties in planning and budgeting in areas where costs are shifted to; and inappropriate market signals which might arise as a result of cross-subsidies. As far as the author is aware, there is no New Zealand research evidence on the extent of deliberate cost-shifting or its impact. However, cost-shifting is widely believed to take place here. Recommendations Future analysis and research There are a number of areas where further analysis would enable a better judgement to be made about the benefits of further reform in New Zealand health care. Regarding devolved purchasing and purchaser competition, it is recommended that: • the Treasury undertake further analysis of the potential advantages and disadvantages of moves towards integrated care and devolved purchasing in New Zealand. This analysis should compare current purchasing arrangements with devolved purchasing in relation to: purchasing expertise; local responsiveness (of vii providers to DPOs and of DPOs to communities); the removal of the purchaserprovider split which is proposed with devolved purchasing; accountability and governance; conflicts of interest; and levels of transaction costs. Further analysis should involve a) a detailed literature review of United Kingdom and United States experiences with devolved purchasing; and b) formal evaluations of integrated care pilots in New Zealand: these evaluations should draw on key issues noted in this report. • the Treasury complete a more detailed literature review and give more critical thought to the advantages and disadvantages of purchaser competition in New Zealand. This analysis should draw on previous work undertaken for the Health Services Taskforce as well as more recent evidence. Careful attention should be paid to: i the design features of purchaser competition. This should include careful analysis of the implications for equity, efficiency and other goals of: allowing individuals to top-up their government contribution with private financing; ensuring a minimum standard of care; alternative regulatory structures, including the number of standard packages of care; ii the implications for private medical insurance in New Zealand; and iii the transition issues in shifting to purchaser competition. • the Treasury encourage clearer signals to be given about the future direction of New Zealand’s health services, in particular, whether purchaser competition is to be encouraged and what the regulatory arrangements might be with such competition. This will allow providers and emerging DPOs to plan with full knowledge of the future regulatory environment. Cream-skimming It is recommended that if devolved purchasing is thought to be appropriate to develop into competition between providers or competition between DPOs: • the services specified in the HFA’s Service Coverage Document should continue to be incorporated into contracts with providers. This promotes accountability and ensures that provider groupings which compete with each other are required to make available the same set of services1; • current work on specifying services (and eligible patients) continues, adding detail (e.g., condition/treatment pairs; practice guidelines; clinical assessment priority criteria) in relation to services where rationing is most common2. In a competitive purchasing environment, this will promote accountability; promote equity of access; and reduce local service specification and service delivery choices as sources of 1 Some consideration will need to be given to the advantages and disadvantages of standardising user charges (Enthoven 1988, p. 104), to prevent both cream-skimming and cost-shifting. 2 This assumes that the funding of health care in New Zealand remains at present levels and that rationing of particular services (e.g., heart transplants, liver transplants) continues. viii cream-skimming3. It may be that with experience we gain a better understanding of which services need to be specified tightly and which can remain more loosely defined in order to encourage diversity; • an explicit list of services that are not currently purchased by the HFA be developed – this will assist in clarifying boundaries between public and private financing. Such services may need to be specified in some depth (e.g., condition/treatment pairs; practice guidelines; clinical assessment priority criteria). The private insurance market is likely to welcome this development and some further assessment of its impact is essential. This approach makes more explicit the fact that those unable to get or afford private health insurance may miss out on care they might benefit from; • responsibility for technology assessment and national guideline development activities for particular services (e.g., high cost, high volume, experimental services) should be allocated to an existing or new agency. The agency’s roles would be to facilitate guidelines to be used in conjunction with service specification and to implement such guidelines. Again, the role of such guidelines would be to promote accountability and equity of access and to prevent local service specification and service delivery in ways that promote cream-skimming. Possible agencies include: the National Health Committee (incorporating its Guidelines for Guidelines project); the HFA; or a new agency such as the United Kingdom National Health Service National Institute of Clinical Excellence (NHS Executive 1999); and • funding be allocated to DPOs using a risk-rated capitation formula, where the level of funding is either: a) sufficient to cover a broadly defined range of services such that care is available to all those who are deemed likely to benefit, i.e. requiring additional resources in the health sector depending on an assessment of current levels of unmet need4; or b) sufficient to fund a set of services which are defined in detail. The second approach is likely to be required in order to i) remain within current funding levels5; ii) promote equity of access around the country; iii) promote accountability and iv) prevent opportunities for under-servicing and creamskimming6. 3 It will also reduce avenues for under-servicing. 4 The recent experience in relation to booking systems for elective surgery gives an indication of how greater explicitness and the setting of an entitlement may lead to unmet need surfacing (e.g., in New Zealand where a large number of people are waiting to see specialists for assessment) (Kennedy 1998)). See also Light (1998) who argues that the United Kingdom NHS is ‘doubly underfunded’ in that in addition to those already on waiting lists there are likely to be thousands of patients with similar needs who are being kept off waiting lists because their GPs thought the lists were too long (p. 7). 5 This is partly because service specification in and of itself, accompanied by a move to the delivery of care by private agencies engaged in making profits or surpluses, may lead New Zealanders to view the services specified as an entitlement to care and to become more wary of not getting services in a timely way. 6 Contracting with devolved purchasers in terms of volumes of service may not be sufficient to reduce incentives to cream-skim or under-service. 6 from devolved purchasing come in the form of improved processes of care for patients, more services and financial savings13. The devolved purchasing arrangements considered in this paper are: Monopoly HFA – This model represents the status quo. A monopoly purchaser – the HFA – is responsible for the purchasing of a broadly defined range of services for the entire New Zealand population. The HFA contracts with a range of providers to ensure such services are delivered. The HFA uses a mix of contracting arrangements, including contracts with integrated providers, who are paid largely on a capitation basis, and contracts with non-integrated providers who are paid on a feefor-service basis or who are paid according to price-volume contracts with budget caps. Monopoly HFA with specialist purchasing of particular services14 – Where the HFA devolves responsibility for some specific services to particular DPOs, who then become responsible for purchasing those services. Examples might include care for consumers with disabilities. Carved-out services are assumed to be paid for on a capitation basis15. The HFA retains responsibility for purchasing all other services. Regional monopoly integrated purchasers – Where a number of regional monopoly purchasers are responsible for the purchasing of a broadly defined range of services for their regional or sub-regional populations. These purchasers are largely integrated with a few external contracts. It is assumed that such purchasers would be funded on a capitation basis16. Choice of partially integrated DPO – Where individuals make their own choices for coverage, choosing between a wide range of competing DPOs. DPOs arrange for some care only (e.g., primary care or primary care with some secondary care services), and are either publiclyor privately-owned. The HFA contracts for other services (e.g., tertiary services) on behalf of the entire New Zealand population. The DPO market is managed by one or more agencies, through which government funding is channelled. These agencies register DPOs, monitor DPO performance and provide 13 Further work is required to determine whether or not these assumptions would hold with devolved purchasing. 14 Known as ‘carve-outs’ in the United States literature. 15 In some United States examples, carve-out purchasers are often paid an administration fee (based on capitation), and are not subject to financial risk – providers are often paid on a fee-forservice basis. The carve-outs appear to involve very close management of care, e.g. ongoing concurrent review (Goldman, McCulloch et al. 1998; Ma and McGuire 1998). The two papers cited here refer to mental health benefits in employment contracts, ie. for the general population. Whether or not the same arrangements are applicable for those with ongoing serious mental health needs is not clear. 16 An example of this arrangement might be the Marlborough Health Trust model (Hyndman 1998). 7 consumers with information on DPO performance. Agencies might include the HFA, employers or regional purchasing co-operatives17. Choice of fully integrated DPO - Agency approach – Where individuals choose between competing DPOs – however, their choices are framed by an agency, through which government funding is channelled. Agencies make the first cut of choices between a range of competing DPOs (e.g., they may seek tenders or choose a narrow range of DPOs), and they monitor, and provide consumers with information on, DPO performance. Agencies might include the HFA, employers or regional purchasing cooperatives. DPOs arrange for comprehensive care, are fully integrated, and are either publiclyor privately-owned. Choice of fully integrated DPO - Individual approach – Where individuals make their own choices for coverage, choosing between a wide range of competing DPOs. Government funding is allocated to DPOs on the basis of consumer choice. DPOs arrange for comprehensive care and are fully integrated, and are either publiclyor privately-owned. The DPO market is managed by one or more agencies, through which government funding is channelled. These agencies register DPOs, monitor DPO performance and provide consumers with information on DPO performance. Agencies might include the HFA, employers or regional purchasing co-operatives. 2.5 Devolved purchasing and policy risks Many of the techniques for devolving purchasing pose little risk on their own. Key uncertainties, however, include: whether any additional transaction costs from devolved purchasing18 are offset by improvements in care or cost-effectiveness; and whether integrated care techniques lead to overall higher costs because of improvements in quality of care, e.g. where clinical guidelines lead to more service and better quality of care. These risks are not discussed here. Devolved purchasing does however raise two important issues: how to provide incentives for purchasers to perform; and how to reduce the potential for cost-shifting where financing/budgets for different types of care remain separate19. 17 Regional purchasing co-operatives are agencies which purchase health care on behalf of employer or other organisations. The concept appears to be a United States one, designed to reduce the costs of purchasing for small and medium sized firms and to enhance purchaser power over providers. 18 For example, the savings on transaction costs were one reason given for the move from four purchasing authorities to one purchasing authority in New Zealand. The one purchasing authority should be cheaper to run than four, while provider groups which operate in more than one region should also find their transaction costs reduced as they need only deal with one agency rather than four. 19 For example, the Marlborough model included private insurance in its financing arrangements. How will it be made clear what is paid for by government funding vs private insurance financing? What will happen with user charges if further services – beyond those currently paid for by the general medical services subsidy – are delivered by general practitioners? 8 Additional risks are also likely with the moves towards placing providers and devolved purchasers at increasing levels of financial risk. The most important of these are in relation to under-servicing and cream-skimming. The former is dealt with in this paper under the issue of purchaser performance; the latter is considered as a stand-alone risk. 9 3 RISKS ASSOCIATED WITH DEVOLVED PURCHASING 3.1 Cream-skimming (risk-selection) 20 Cream-skimming occurs when health providers or DPOs – either deliberately or by chance – enrol a favourable mix of members, i.e. those with a lower risk of making a claim or needing care than the average person21. It is a particular problem where there are capitated payments made to providers and plans, where there is competition between providers and plans to serve particular markets, and where – because premiums for an individual’s care do not necessarily equate with ex ante or ex post financial risk – there is an opportunity for the provider or DPO to make profits/surpluses from selecting patients. In New Zealand, cream-skimming is a risk where a) contracts with providers are on a capitation or budget-holding basis and b) there is competition between providers or DPOs. Although New Zealand has some limited experience with capitation and budget-holding, the extent of cream-skimming in New Zealand is not known. Cream-skimming has been found to offer opportunities for large profits in both the United States and the United Kingdom (Newhouse, Manning et al. 1989; van Vliet and van de Ven 1992; Matsaganis and Glennerster 1994; van de Ven, van Vliet et al. 1994)22. Risk-selection can therefore be a very profitable undertaking. It is, however, very difficult to prove that it is occurring. There is anecdotal and research evidence that cream-skimming occurs frequently in the United States, but little more than anecdotal evidence of its occurrence in the United Kingdom23,24. 20 In this paper, cream-skimming is distinguished from under-servicing, ie. providing fewer services than appropriate. Although under-servicing may be used as a means of cream-skimming, incentives on purchasers and providers to under-service also arise where fee-for-service payments are averaged across a category of care (as in the Diagnostic Related Group payment system). The problem of under-servicing is considered in the section on purchaser incentives. 21 Strictly speaking, compared with the average person within any broad risk pool, e.g., within any particular age-gender pool. 22 In the United States, for example, the Alpha Center notes that insurance statistics consistently show that 35 per cent of people purchasing health insurance will not file a claim in the following five years, while 5 per cent account for more than 50 per cent of health expenditures in the same period (Alpha Center 1997). Similarly, in the United Kingdom Glennerster et al found in their study of one GP fund-holder that the entire fund was expended on 27 per cent of the patients, with the most expensive five patients taking 68 per cent of the expenditure (Glennerster, Matsaganis et al. 1994). 23 The United Kingdom experience in relation to fundholding is in part explained by researchers as a result of: generous budgets; GPs not being at personal financial risk; stop-loss provisions (Le Grand, Mays et al. 1997); the use of historical budgets rather than pure capitation rates to set budgets (Audit Commission 1996); while further explanations could include cultural factors specific to the United Kingdom and not found in the United States. 24 Recent research has found evidence of an increase in the proportion of patients removed from GP fundholders’ registers in Northern Ireland at general practitioners’ requests. Comparisons were made with removals prior to fundholding and with non-fundholders. This may be a means of cream-skimming; equally it may be related to the additional workload and pressures of fundholding (O'Reilly, Steele et al. 1998). 10 The key ways in which competing providers and DPOs can cream-skim include: • specifying benefit packages in particular ways (e.g., including comprehensive maternity care to encourage young couples to enrol; excluding mental health benefits to discourage the enrolment of people likely to suffer from such problems; excluding pre-existing conditions from coverage to discourage those with health problems from enrolling); • refusing coverage altogether to particular individuals or to people employed in certain industries or occupations; or offering coverage at prohibitively expensive premiums; • contracting with providers in particular ways (e.g., not contracting with specialists known to specialise in high-risk conditions; recruiting new specialists with limited patient following); or contracting with providers in particular locations but avoiding others who practice in higher-risk localities (e.g., inner-city areas); • marketing DPOs in particular locations or to particular groups to attract people in good health (e.g., in gyms); and • offering poor quality care, under-servicing, making patients wait for care or developing more stringent protocols for referring patients to specialists for some types of conditions, in order to drive consumers to another DPO. (See: Enthoven and Kronick 1989; Ellwood, Enthoven et al. 1992; Congressional Budget Office 1993; Gauthier, Lamphere et al. 1995; Swartz 1995; Luft 1996; Newhouse 1996; Alpha Center 1997 for further detail.) Risks associated with cream skimming Successful cream-skimming in New Zealand would lead to the following risks25: Crown risks • Increased overall expenditure. This might arise if the approach being used in California to fund care for people shifting from Medicare Cost (traditional fee-for-service insurance) to Medicare Risk (managed care) were used in New Zealand. Medicare Risk premiums are based on the fee-for-service DPO premiums: Risk plans receive 95 per cent of the fee-forservice plans’ premiums. Some Risk plans are obtaining extremely favourable mixes of members when compared with the risk profile of the Cost members. As a result, Risk plans could earn large profits on the premiums they are being paid by 25 It is assumed that government financing of health care remains and that every person is covered by at least one DPO. In an unregulated, privately financed model, cream skimming would mean some people would be unable to obtain any insurance coverage at all, or would only be offered coverage at extremely high premiums. 11 Medicare26. This would seem to imply that overall expenditure on Cost and Risk plan members will increase: as favourable risks sign up with Risk plans, the average risk in the Cost plans rises, leading to increases in Cost plan premiums and therefore further increases in Risk plan premiums. • Pressure for additional expenditure. This may occur if providers or DPOs successfully cream-skim by offering additional services or higher quality care for some people than would otherwise occur27. This might result in pressure to widen coverage and increase quality of care for all New Zealanders. Consumer risks • Inequitable access to a broad range of services and quality care. Cream skimming raises equity concerns, as it reduces the choices of provider or DPO and coverage offered to some people. It may also lead to poor care for some groups where poor quality care or reduced coverage are used as means of cream-skimming. Access to care and quality of care may also be adversely affected for some people if providers or DPOs with an unfavourable mix of risks are forced to reduce services or quality in order to remain within budget. Health sector risks • Diminished incentives for technical efficiency and cost-effectiveness. Cream skimming does this: a) by allowing DPOs to segment the market, thereby reducing the likelihood of individuals switching providers or DPOs because of cost; b) because cream-skimming is likely to be more profitable than promoting efficiency and cost-effectiveness28; and c) because good providers or DPOs may go out of business if they are unable to provide good quality care due to a poor mix of members. • High level of profits for some providers or DPOs while others struggle. Providers or DPOs which successfully engage in cream-skimming may earn abovenormal profits, while others face considerable losses. This is likely to engender 26 In practice, Medicare rules prohibit HMOs from earning more profit on their Risk clients than their commercial clients. If the Health Care Financing Authority (HCFA) estimates capitation payments will lead to excess profits, plans must reduce premiums, out-of-pocket expenses, offer additional benefits or return money to the HCFA. Most decide to reduce premiums or offer additional benefits (General Accounting Office, 1995, p. 8). 27 See footnote 26 above for reasons for why this is happening in California. 28 As noted by the Alpha Center, ‘health plans offer their products knowing that the health or sickness of the people they sell to will have a greater impact on their annual per capita costs than any changes they can make to improve the efficiency of health care delivery’ (Alpha Center 1997, p. 5). See also the points made above (page 18) about the profitable nature of cream-skimming. 12 discontent amongst some providers and DPOs. In addition, efficient, high quality DPOs may be driven out of the market. Likelihood and significance of risks Cream-skimming has the potential to be a serious problem in New Zealand if we move to payment systems which place providers and DPOs at greater financial risk, and maintain or further encourage competition between providers or devolved purchasers. The likely extent of cream-skimming cannot be judged at this point in time. Although the New Zealand system may appear to be characterised by socially-responsible individuals and agencies, the financial incentives inherent in a more commercial and risky environment may change attitudes and behaviour. In the United Kingdom, with the move to fundholding, little evidence of cream-skimming has been found: incentives for cream-skimming may well be muted, however, by the limited financial risks faced by providers. On the other hand, cream-skimming appears common in the United States. One gauge of the significance of the issue, however, is the attention that is being given to regulatory reforms in the United States to reduce its prevalence: this suggests that if cream-skimming were to become widespread, each of the risks discussed above is likely to be significant, both in a financial and political sense. 3.2 Poor purchaser performance The New Zealand health care sector is characterised by a series of agency relationships. For example, the HFA acts as an agent of the government (the principal) in purchasing health care services and in designing contracts to promote efficiency and to minimise Crown risk. Any devolution of purchasing responsibility similarly involves the development of agency relationships. A key task of any principal – the Crown, the HFA, owners or consumers – is to ensure that agents strive to meet the objectives of the principal. Poor performance by a purchaser will therefore mean that the principal’s objectives are not met. Key purchaser objectives include: • to make purchasing decisions which promote technical and allocative efficiency (e.g., maximising health status for the money made available) and equity (e.g., to improve Mâori health status; to promote fair access to care); • to purchase a mix of services appropriate to the population’s needs by assessing population needs and responding to those needs; • to remain within budget and control costs; • to contract in ways which promote technical efficiency and cost-effectiveness and which limit risk; and • to contract in ways which are fair and transparent, and which minimise transaction costs. 13 Risks associated with poor purchaser performance Thus, the risks associated with poor purchaser performance – relative to good performance – might include: Crown risks • Higher overall expenditure for a given level of care or a lower level of care for a given level of expenditure29. • Need to implement and enforce a regulatory function, including oversight of performance. • Pressure to bail out or take-over purchasers which fail to perform. • Pressure to provide additional operational funding to government-owned providers which suffer financial losses in the event of a purchaser failure (e.g., bankruptcy)30. • Legal action against government-owned purchasers which fail to undertake contracting in fair ways. Consumer risks • Higher levels of taxation for a given level of care; or lower access to services and reduced quality of care as a result of poor decision-making or contracting. • Reduced equity of access if particular purchasers perform poorly relative to other purchasers. • Increased private financial responsibility for care: this will also reduce equity of access as care is increasingly only available to those able to pay. • Unresponsive purchasers. • Health status fails to improve or worsens. 29 As the services to be purchased in New Zealand are often unclear, the government is able to shift some of this risk onto consumers (e.g., reduced services, longer waits for health care). 30 If for example a privately-owned purchaser were to run up substantial losses, the government may have to step in to finance government-owned providers which run the risk of large losses as a result. 14 Health sector • Inefficiencies in the delivery of publicly-financed health care, i.e. lower value for money than can be obtained by improved performance. • Excessive profits or losses being made by providers where purchasers fail in their contracting role. • Higher overall expenditure with third-party insurance arrangements which operate in the private insurance sector. Likelihood and significance of risks The likelihood of such risks are difficult to judge relative to the status quo. For example, although New Zealand purchasers have failed to deliver on some issues (Ministry of Health 1999) it is much more difficult to establish if purchasers are failing to deliver optimal value-for-money. Any form of poor purchaser performance which results in the failure of a major purchaser or provider is likely to result in Crown action and potentially expenditure. Business failures in the private sector may impose largely private costs but governments may feel obliged to intervene even where the purchaser or provider is privately-owned. This is because of the importance of health care to the population (and hence the politically sensitive nature of health care), the monopoly positions which purchasers and providers often have in New Zealand, and the fact that most health care expenditure in New Zealand comes from the government. Judging the individual significance of each these risks is also difficult. Any losses from poor performance in delivering value-for-money are difficult to estimate – from all perspectives. The size of the health care market means that systematically poor performance will have a serious impact. This, however, may be balanced out by having multiple agencies involved in the delivery of health care: failure on the part of a purchaser to perform may not necessarily be reflected at a provider level. 15 3.3 Cost-shifting Cost-shifting occurs when costs are charged against a budget which a) differs from that intended or b) differs from that traditionally charged31. The objective of costshifting is to avoid cost-controlling budgetary restrictions and, sometimes, to find ways to increase overall spending. It appears to be easier to achieve where there are openended budgets to shift costs to. Key budget boundaries in the current New Zealand health care system between which costs can be shifted are: • financing boundaries – public financing, private insurance, user charges, out-ofpocket payments and charitable donations; • government budget boundaries – health, social welfare, ACC, and other government votes; • within-HFA boundaries – personal health, disability support, public health, Pharmac; • contractual boundaries – between individual organisations (hospitals, IPAs) and between contracts within organisations (e.g., within hospitals); and • budget boundaries within organisations (e.g., clinical budgets in hospitals; separate pharmaceutical budgets within IPAs). Cost-shifting can be of two main forms. First, it occurs when patient care is inappropriately charged against a particular budget (pure ‘cost-shifting’) or where the location or type of care is changed in order for care to be charged against a particular budget. (This is sometimes called ‘patient-shifting’ (Donaldson and Gerard 1993).) Second, it is defined in some international literature as occurring when the costs of a particular episode of care, or the costs of insurance, are lower for some people and higher for others than is strictly accurate (perhaps known in New Zealand more commonly as ‘cross-subsidisation’). Examples of the first kind of cost-shifting in New Zealand might include: • classifying non-accidents as accidents in order for care to be charged to ACC32; • hospitals supplying prescriptions rather than medication in order to shift costs from hospital budgets to the pharmaceutical tariff; 31 This definition implies that explicit policies which shift costs from, for example, government responsibility to private responsibility are just as much cost-shifting as when costs are inappropriately (e.g., against the spirit of particular legislation or contracts) charged against particular budgets. Light calls the former ‘declassifying’ (Light 1998, p. 13). 32 Reasons for this might be because of lower user charges for primary care or because there is subsidised care available through ACC but not through vote:Health (e.g., physiotherapy and acupuncture are not publicly-financed for non-accident related injuries). 22 B Compensating for risk The main approaches to compensating for a higher or lower than average risk are: • risk-rating capitated premiums. A number of tools have been developed to undertake such risk rating. These include socio-demographic models (e.g., the Adjusted Average Per Capita Cost model used in Medicare in the United States, which adjusts for age, sex, welfare status, institutional status; the population-based funding formula used in New Zealand); prior-use models; self-reported health status models; and disease/diagnosis models (e.g., Ambulatory Care Groups; Diagnostic Cost Groups; Chronic Disease Scores) (Alpha Center 1997; Bowen and Sigoloff 1998). Disease/diagnosis models are argued to provide more accurate estimates of likely cost. Risk assessment and adjustment may take place retrospectively, prospectively or concurrently, at pre-determined intervals. • using a fee-for-service payment mechanism which includes levels of payments adjusted for complexity (as in the DRG approach used in secondary care)37 or moving to blended payment systems which combine elements of capitation and fee-for-service (Newhouse, Beeuwkes Buntin et al. 1997)38. These approaches are not discussed in this paper in any depth; those who are interested are referred to the references supplied here and material from the Treasury workshop held on October 14-15 1998 (Bowen and Sigoloff 1998). 37 A disadvantage of this approach is that it compensates plans directly for what they do, with payment levels based on average existing costs, i.e. Incorporating with new payment levels any existing inefficiencies. Ambulatory Care Group and Diagnostic Cost Group approaches also incorporate current service use information within them, for use in adjusting for higher levels of ‘risk’ (i.e. poorer health status). 38 The more technical literature on incentives in procurement suggests that a mix of payment systems is an efficient means of promoting quality in service delivery (Laffont and Tirole 1993). Hence, such an approach would not only reduce incentives for cream-skimming, but also underservicing (Newhouse, Beeuwkes Buntin et al. 1997). 23 C Limiting risk The following approaches can be used to limit risk, and hence to reduce the incentives to cream-skim: • Establishing separate risk pools (including carve-outs). For example, individuals with particular conditions (e.g., disabilities) or requiring particular care (e.g., kidney dialysis, organ transplants) may be covered by a separate financing arrangement. DPOs or providers may get paid on a fee-for-service basis to provide such care, or such care may be paid for by an alternative agency (e.g., the HFA or the Ministry of Health in New Zealand). van Barneveld, van Vliet and van de Ven have proposed a similar approach for the Netherlands, where plans would predetermine a small fraction of members whose costs would be pooled and reimbursed separately (van Barneveld, van Vliet et al. 1996). These approaches have the advantage of limiting the financial risk that plans or providers may take on if they register particular individuals. Hence, incentives to cream-skim may be reduced. Without clear boundaries, however, cost-shifting becomes a risk, while alternative systems must also be established to fund and administer separate schemes. • Limiting financial risk39. In this case, a financial limit may be placed on the risk an individual provider or DPO is responsible for in regards to particular patients. For example, the United Kingdom general practice fundholding scheme limited risk to £5,000 initially: the District Health Authority became responsible for care costing beyond this amount. This approach may, however, fail to reduce incentives to cream-skim for patients likely to cost a lot, but less than the financial limit. Costshifting may be an issue if the limit is set such that it provides few incentives for considering the costs and cost-effectiveness of care. • Establishing separate clinics. This final approach allows those unable to obtain good care to choose a separate clinic to provide care. For example, in the United States those without insurance coverage can obtain emergency care from statefinanced and state-owned hospitals: however the care tends to be too little, too late. It is possible to imagine that specialist clinics might be established for those with particular conditions to ensure quality care is provided (e.g., for mental health or alcohol and addiction services). The care may be paid for separately (as above) or costs sought from an individual DPO or provider. Separate payments will reduce the incentives to cream-skim. However, cost-shifting could become a significant problem; a separate system is needed to administer the arrangements; those accessing care in this way may feel stigmatised by the need to access such care; and access to care may be viewed as inequitable if quality of care differs from care available to others. D Defining service entitlements Service specification – in the form of a standard package (or standard package s ) of benefits – is an important tool for reducing opportunities for cream-skimming. Service specification prevents plans or providers from specifying or offering services in ways which encourage particular groups to enrol or disenrol, or which discourage particular 39 Sometimes known as an outlier or reinsurance scheme. 24 groups from enrolling. Hence, wherever there is competition between plans or providers where benefit design can influence risk selection, service specification is an important tool for reducing at least one avenue for cream-skimming40. Service specification at a broad level is common in many health care systems41, and a standard package of benefits is largely viewed as an essential element of proposals which promote competition between purchasers (see for example Enthoven and Kronick 1989; Pauly, Danzon et al. 1992; The White House Domestic Policy Council 1993; Enthoven 1994; White 1995). Service specification is not common in countries like New Zealand and the United Kingdom42. The reasons for this may be that coverage is intended to be fairly comprehensive; that expenditure determines service availability rather than service specification determining expenditure; and that government ownership of key providers (e.g. hospitals) promotes accountability in place of service specification. Although New Zealand has developed a set of service obligations in the past few years (Health Funding Authority 1998; Shipley undated), these obligations do not guarantee an entitlement to care: purchasers have limited budgets and must make choices between services set out in service obligations (Cumming 1994). Approaches to specifying services Services can be defined in a number of ways. The main approaches and their advantages and disadvantages are set out in Table 4.143. No specification A first approach is to have no specification at all. For example, New Zealanders may have a broad understanding of the services not covered by the publicly financed health care system (e.g. most adult dental care, optometrist services, most cosmetic services), but little clarity about the services which are actually provided, the circumstances under which specific services are or are not available (e.g. kidney 40 Other reasons for service specification are set out in Appendix E. 41 European countries often specify broadly the services to be made available under social insurance schemes (see Appendix F for the Netherlands example). Canadian provinces each specify service coverage (Health Canada 1997). In some countries, service specification takes the form of fee schedules. For example, Australia’s Medicare system has a Benefits Schedule, as does Germany’s main social insurance system. Many countries also have specified pharmaceuticals lists which act as a form of service specification. See also Glaser (1991); Lenaghan (1997) for examples of service specification. 42 New Zealand has some aspects of service specification, e.g. the pharmaceutical tariff; fee-forservice payment schedules in primary care. 43 The material here focuses on the specification of services . It is also possible to specify access requirements (e.g., travel times, user charge maximums, waiting times) and standards for quality of care (including for example access to medically qualified doctors rather than simply nursing staff and access to specialist care rather than generalist care alone). Such further specification would improve accountability and comparability of plans, but would increase the costs of specification and remove some aspects of access and quality of care as aspects of care on which plans could compete. Politically, however, governments may find it extremely difficult not to include such aspects of care in specified coverage. 25 dialysis, heart transplants), and the time they might wait for care. Most decisions about who gets what care when are, with this approach, taken by medical professionals. This approach does not require the administrative costs which might be incurred through service specification. It also allows plans or providers to decide on relative priorities at service and patient levels and allows flexibility in the choice of care provided (e.g., choices between medications; choices between hospital care or care at home). 26 Table 4.1 Approaches to defining services44 Possible approaches Advantages Disadvantages No specification. • Low administrative costs. • Providers or DPOs determine priorities and services. • Consumer flexibility in choice of care. • Difficult to determine capitation amounts which are related to coverage expectations. • Lack of accountability in terms of what is covered and whether individual consumers will receive care. • Likely differences in access to and the type of care available in different regions, localities, DPOs and providers. • Consumers find it difficult to compare DPOs. • Offers the ability for providers and DPOs to under-service those consumers whom they wish to discourage from joining or encourage to leave. Publicly specified in terms of providers. • Low administrative costs. • Providers or DPOs determine priorities and services. • Consumer flexibility in choice of care – though limited to particular providers. • Limits service delivery to particular providers: this may reduce costeffectiveness. • Difficult to determine capitation amounts which are related to coverage expectations. • Lack of accountability in terms of what is covered and whether individual consumers will receive care. • Differences in access to and the type of care available in different regions, localities, DPOs and providers. • Consumers find it difficult to compare DPOs. • Offers the ability for providers and DPOs to under-service those consumers whom they wish to discourage from joining or encourage to leave. Publicly specified quite generally.45 • Establishes broad range of services covered. • Allows for better estimation of capitation amounts (taking into consideration coverage). • Offers opportunities to consider breadth of service coverage (e.g., what is in and what is out). • Lack of accountability for providing care to individual patients in particular circumstances and in relation to modalities of treatment. • Consumers may still find it difficult to compare DPOs if they perceive there to be differences in care actually offered. • Offers the ability for DPOs to underservice those consumers whom plans wish to discoura g e from j oinin g or 44 There is the potential for one or more plans to be devised. Care would need to be taken that different plan structures do not encourage differences in health status. The more plans, the greater the administrative costs – both in designing plans and in ensuring consumers receive only the care they are entitled to under the plan they have chosen. In addition, in each case, providers or plans may or may not be allowed to offer services in addition to those specified and to offer better quality of care than that set out in regulation. This would provide incentives to improve coverage, but may be to the detriment of equity goals and may be used to cream-skim. 45 Represents the present situation in New Zealand. 27 encourage to leave. • Politicisation of decision-making process and political risks in specifying coverage. • Some administration costs in defining services, monitoring compliance by plans, and establishing review procedures. • Limits choice by individuals to decide on coverage which best suits their needs. 28 Table 4.1 (cont.) Approaches to defining services Possible approaches Advantages Disadvantages Publicly specified in detail (e.g., with practice guidelines). • Promotes clarity of entitlement for consumers. • Promotes accountability at a detailed level. • Allows for detailed estimation of capitation amounts. • Facilitates detailed comparisons between competing plans. • Offers opportunities to consider effectiveness and efficiency in designing the core – i.e. allows a much greater focus on the health benefits associated with particular services, reducing excessive use of technologies, and taking into account wider societal costs. • Analysis of effectiveness and efficiency undertaken only once – at a national level. • Promotes equity of access for equal need. • Administratively complex. • Politicisation of decision-making process and political risks in specifying coverage. • Reduces choice and flexibility at patient level. • May not be fully feasible given information gaps. • Limits choice by individuals to decide on coverage which best suits their needs. • May view such detailed specification as an entitlement. Privately specified. • Allows consumers to choose packages which best meet their needs (health care and affordability). • Allows specification in ways which may promote cream skimming – likely to be more profitable than improving efficiency. • High administrative costs as each DPO specifies own package. • Consumers may find it difficult to compare DPOs – may reduce incentives for efficiency if -in the absence of good comparative information – consumers use price or coverage as a guide for choosing between plans. • Competition between DPOs may lead to excessive use of new, unproven technologies if consumers choose plans based on availability of new technologies. Principles specified (e.g., medically necessary services). • Flexibility for individual providers and consumers/patients. • Leaves final determination up to a regulatory or legislative body • May involve costly litigation. 29 However, in a budget-constrained system which cannot always be comprehensive and with long waiting times, a lack of specification makes it difficult to determine budgets and capitation amounts for plans or providers. This is because it is unclear exactly which services should be made available and when they ought to be provided. The lack of clarity can lead to arguments about whether the funding is adequate or whether the DPO or provider is failing to deliver adequate care. Hence, this approach can also make it difficult to hold plans or providers to account. Having no specification of services leaves decisions about service priorities and the care which individuals are offered in the hands of professionals or managers. This may provide flexibility, but it may also mean that community values about priorities are not adequately considered. In a competitive system with multiple plans or providers paid on a capitation basis, a lack of specification makes it very difficult for consumers to compare plans and it allows plans and providers to offer services in ways which promote creamskimming46. Publicly specified in terms of providers In many health care systems, it is common to define services publicly, usually in law or regulation. In addition, it is also common to define the types of providers from whom care is available rather than the specific services which are available. For example, specification may be in terms of ‘physician services’ or ‘hospital services’ (see for example, Glaser 1991; Health Canada 1997; Lenaghan 1997). This approach has the advantage of low administrative costs for governments: costs are limited to specifying – in law – groups of providers from whom covered care can be obtained, and to covering any administrative and legal costs which arise in monitoring, auditing or challenging providers and plans who appear to be not complying. Individual patients may also incur costs in seeking compliance. In addition to the disadvantages under the no specification approach (see the section on ‘no specification’, above), this method of specification limits service delivery to particular providers. This may reduce cost-effectiveness if alternative providers who can provide care equally or more effectively but at lower cost are excluded from coverage47. Publicly specified in general terms A third approach has public specification (in the form of legislation or regulations), but with services specified in general terms. For example, coverage may include ‘maternity services’, ‘dental services’, ‘diagnostic services’, ‘medical and surgical services including referral services’ and ‘preventive services’. This approach is close 46 In practice, in an unregulated private market, competitive plans define their own packages of benefits. 47 For example, if nurses can provide more cost-effective care than doctors for some conditions; if dental therapists can provide check-ups and clean teeth more cost-effectively than dentists. 30 to that used in New Zealand in the Service Coverage document (Minister of Health 1998). This approach: establishes the broad range of services covered; allows for better estimation of capitation amounts; and offers governments the opportunity to consider the breadth of service coverage48. If there is compulsory coverage, this approach also ensures that everyone is covered for a broad set of services: there can be no free-riders. One disadvantage with this approach (as with the above two approaches) is that it leaves decisions about the modalities of care (e.g., specialist vs generalist care) in the hands of plans and providers: this may generate public concern at inadequate care which might be offered by some plans or providers. It also leaves decisions about the actual services made available to specific patients in the hands of plans and providers (e.g., is post-natal home care available within the maternity service?; are heart transplants available?; which patients can obtain kidney dialysis services?). Not only might this lead to under-servicing in any health care system, but in a competitive, capitation model: consumers may still find it difficult to compare plans; and it provides a means of cream-skimming (through selection of specific services to encourage or discourage particular groups from joining and through the ability to under-service). In addition, the approach requires some administration in defining broadly the services to be covered. This approach is likely to result in politicisation of decision-making49. Consumers’ choices of the package which best suits their financial and health needs will be limited by this approach. 48 For example, the Netherlands government has tried in recent years to remove physiotherapy services and in vitro fertilisation services from coverage of the Health Insurance Act (Mulder 1995). Dental care was also proposed to be removed but the government had to reinstate it (personal communication, Nicholas Mays). 49 In the Netherlands example given above, the government has found it very difficult to remove some services due to strong provider and public resistance (Mulder 1995). 31 Publicly specified in detail With this method, public specification in legislation or regulations continues, but the services are specified in some detail. This might include: the modalities of care which are available (e.g. medical and surgical services delivered in a hospital by specialists); the specific services which are available (e.g., pre-natal care, labour care and postnatal care within a maternity service; immunisation and cervical cancer screening within a preventive care service); the number of visits or length of care (e.g., six prenatal sessions; post-natal care for two weeks after the birth of a baby); condition/treatment pairs (e.g., Xenical for defined levels of obesity but not for those who are just ‘over weight’); clinical guidelines (e.g., a series of steps to try before a patient is offered expensive hypertension medication) or priority criteria (e.g., for elective surgery. See for example (Hadorn and Holmes 1997; Hadorn and Holmes 1997)). Such detailed specification promotes good clarity of entitlement for consumers and accountability of plans and providers. With detailed utilisation information, it will allow for improved estimation of capitation payment amounts. Within a competitive environment, this approach facilitates consumer comparison of plans because the detailed service specification will reassure consumers that all plans must offer the same care (i.e., a cheaper DPO is not one which is offering narrower coverage). As the service specification process is undertaken publicly only once, administration costs are reduced compared with a private approach (see below). In addition, the process of specifying services offers opportunities to consider at a detailed level the effectiveness of services and the mix of services which promotes allocative efficiency50. Equity is promoted to some extent because every individual is entitled to the same set of services51. This approach is administratively complex. It requires a process for considering services and for specifying them at a detailed level; for updating the list; and requires monitoring, audit and complaints mechanisms if the specification is to have any force. This means the approach may be very costly, particularly if practice guidelines are developed as part of the service specification52. Designing a specific set of services will involve politicisation of a decision-making process; and it may reduce individual 50 For example, a practice guidelines approach allows consideration of the circumstances under which particular services are offered to individual patients; and the overall process enables society to consider the trade-offs between particular services such as having a new preventive service or more heart transplants. 51 Promoting equity of access would also require specifying the details for accessing care, e.g. travel time to specific services, waiting times. 52 Consider for example that the National Health Committee has developed a number of practice guidelines in New Zealand, each of which involved a number of clinicians and others. Off-theshelf guidelines from overseas might cut the costs in New Zealand; however, such guidelines might not be used if clinicians do not have ownership of them. New Zealand has the advantage of being relatively small so that almost all clinicians might be involved in designing guidelines in their specialty areas; on the other hand, the costs of guidelines cannot be spread over a large population. 38 operation). A significant amount of work is underway to develop health care outcome indicators to monitor the performance of providers (Bowling 1991; Benson 1992; Wilkin, Hallam et al. 1992; Jenkinson 1994; Boyce, McNeil et al. 1997) and these same indicators could be used to monitor the performance of purchasers62. However, there are also problems in using such measures to monitor either provider or purchaser performance. First, it is difficult to determine the contribution that care makes relative to no intervention at all. Second, an accurate assessment of health care outcomes relies on adequate adjustment for differences in health status. Some adjustments – age and gender – are relatively straightforward. Usually proxy measures are used to adjust for different levels of need, e.g. socio-economic status indicators, although the development and use of such tools are in their infancy. Stronger adjustments using case complexity, severity and co-morbidities are, however, increasingly argued for: these require reasonably detailed patient-based information systems. Without such adjustments, the relative importance of case mix and health care is difficult to judge. Unfortunately, administrative data sets often do not contain sufficient information for adequate health status adjustment63. Third, as health care outcome measures are usually disease-specific – and as New Zealand has a very wide range of services covered by the health vote (including disability support services) – it can be very difficult to build up an overall picture of a provider’s or DPO’s performance. There can also be prolonged delays in the data being released and problems with a small number of cases – with resultant statistical limitations, especially once particular diagnoses are considered individually (Iezzoni and Greenberg 1994). In the United States there is also considerable concern about the costs associated with initiatives to adjust for differences in health status (Iezzoni and Greenberg 1994). Health care outputs, processes and inputs Given the problems noted above, monitoring the performance of purchasers might be more appropriate at the output and process level. As with health care outcomes, adjustments are still needed, however, at the registered population level to compare the level of outputs delivered to those registered with different providers or DPOs. Such adjustments are made to identify different likely levels of need which might explain differences in, for example, utilisation of particular services. The monitoring of levels of outputs is easier the more comprehensive is the care within an organisation’s sphere of responsibility. With fragmented providers, for example, comparing levels of outputs (and health care outcomes) requires an understanding of the context in which care takes place: a hospital which admits patients who are dying will have different rates of output (and different rates of 62 Using such indicators at the purchaser level gets around some of the problems of using indicators for monitoring the performance of non-integrated providers, where the performance of one provider may be affected by the circumstances in which the care takes place (Brown, McCartney et al. 1995). 63 This raises the question of who collects the data. For example, clinicians would have to hand out questionnaires to consumers, ensure they are filled in accurately and are returned; the data then needs to be analysed and interpreted carefully. 39 mortality) than one where patients are cared for at home or in a hospice. With a comprehensive purchasing agent, however, such problems are minimised. Another alternative is to monitor inputs, although this, at least in terms of capital and labour inputs, has been out of fashion for some time (Trebilcock 1995). In New Zealand, inputs are still used to monitor performance in relation to community mental health services (Performance Management Unit 1998), while the Health and Disability Commissioner used input measures as part of her inquiry into care at Canterbury Health’s Emergency Department (Health and Disability Commissioner 1998). Recent work on practice guidelines – many of which specify patient characteristics or symptoms which are identified as leading to specific outcomes – suggests that holding providers accountable for the steps they undertake in deciding what care to offer to service users, or to whom they deliver services, is important in promoting effectiveness, and therefore allocative efficiency. The increasing interest in practice guidelines suggests a role for a new approach to accountability and monitoring of provider performance in the form of audit to ensure compliance with guidelines (Sheldon and Borowitz 1993). It is possible to imagine holding devolved purchasers accountable at this level of detail, in order to assess performance in promoting effectiveness. Furthermore, with measures of outputs, processes and inputs, some agency needs to take responsibility for ensuring those outputs, processes and inputs purchased have been shown to, or are believed to, contribute to desirable health outcomes. A key issue in holding purchasers accountable for outcomes, outputs, processes or inputs relates to the sheer number of potential outcomes, outputs, processes or inputs which are produced in the health sector. This problem is worsened when it is recognised that different groups – e.g., health professionals, service users, carers, families or society – may view outcomes in different ways. Specifying a large number of measures in order to ensure all services are being provided or to track through changes in service delivery over time may prove expensive and time-consuming. On the other hand, specifying a limited number of measures may encourage purchasers to focus their attention only on those measures, to the detriment of other measures and services, and conceal changes in the distribution of resources which may be viewed by society as undesirable. HEDIS 64 At a purchaser level, the most well known performance monitoring approach is that of the United States National Committee for Quality Assurance Health Plan Employer Data and Information Set (HEDIS). HEDIS ‘is a set of standardised performance measures designed to ensure that purchasers and consumers have the information they need to reliably compare the performance of managed health care plans’ (National Committee for Quality Assurance 1997). HEDIS has one key advantage 64 Light also notes the existence of the so-called Primary Care Assessment Survey (PCAS), for use in monitoring primary care quality and service, noting it is an inexpensive and patient-based approach to monitoring care processes. 40 over approaches based on provider-focused administrative data sets: it is based on population measures rather than encounters between physicians and service users; in other words, the measures go beyond those who selectively choose to visit a physician (Hanchak 1996). (See Appendix I for the 1999 HEDIS measures.) HEDIS is a relatively recent innovation. Hanchak (1996) notes a number of serious limitations, including problems with the data sets and collection methods and continued differences of opinion over the adjustment factors for risk and for different demographic factors such as socio-economic status (Hanchak 1996). (See also Borfitz (1995).) Kenkel (1996) notes that after some ‘five years and millions of dollars were spent…a pilot report card project in 1994 …revealed the inability of most health care organizations to collect and record patient information accurately’. Another issue in relation to measurement tools such as HEDIS – which apply also to comparisons of mortality and morbidity – are that the individual scores ‘seesaw up and down depending on the polling techniques used. Even the season in which a survey is conducted may skew results’ (Kenkel 1996). Such analysis suggests it is very difficult to be fully certain that differences in health or health care outcomes are attributable to the care offered by purchasers or providers. This is particularly the case in terms of outcomes indicated by many performance measurement systems, which cannot be relied upon on their own to provide good information on the relationships between outputs and outcomes. A broader evaluative approach is often likely to be necessary to tease out underlying factors (Iezzoni and Greenberg 1994; Brown, McCartney et al. 1995): investigation rather than castigation being the more appropriate goal. Conclusions Much of the literature in this area is moving towards a view that the on-going analysis of data using administrative data bases is expensive. Such measurement tools require numerous quality control measures in terms of definitions and data coding, while also needing much greater information on patient and provider or DPO characteristics than is currently collected in New Zealand. It is not yet clear that the benefits from such measurement tools are worth the expense involved. Furthermore, the need for further investigation limits the use of such tools as measures of accountability in the short-term. It appears better to identify and focus on particular priority areas, covering a range of different attributes of DPO performance, and to identify if there are meaningful outcomes measures which can be used in contracts with purchasers and providers65. Desired outcomes should be within the control of purchasers, they must be meaningful and not able to be gamed, and they must reflect sector priorities. 65 The first step is to identify outcome measures which are appropriate and meaningful (e.g., percent smokers in a population); a second step is to gather information from purchasers on current outcomes levels; a third step is to set a target which is meaningful and achievable and which, if it requires additional resources, is a priority. 41 In the absence of outcome measures, continued monitoring of outputs is required. Priority outputs should be monitored, and developed into a reporting framework such as HEDIS. Once again, however, adjustments must be made for differences in health status or health need in different communities. Measures must not be able to be gamed and they must reflect sector priorities. Monitoring of performance against practice guidelines (i.e. audit) may also be an appropriate step to take in New Zealand. Provided that appropriate care is generally agreed and based on whatever evidence exists, this approach may allow for the faster identification of problems. Adopting this approach will however be seen as reducing doctor autonomy and intruding into the doctor-patient relationship. Finally, we can continue to monitor performance against particular tasks which purchasers are expected to undertake, such as minimising their own costs, consulting widely with providers and the public, undertaking contract negotiations in good faith, developing and implementing guidelines and so on. B Rewards and sanctions There are a number of ways in which employees and firms can be motivated to perform. In addition to the profit incentive (see below), the following methods can be used to motivate people and organisations: • rewarding achievement; • recognising good performance; • providing new challenges; • providing an interesting job; • giving responsibility to employees; • providing advancement (promotion); and • rewarding using salaries and benefits (Osborne and Plastrik 1997, p.147); Tools for performance management include: • performance awards – e.g., non-financial recognition; • ‘psychic’ pay – e.g., paid time off, new equipment, study leave, research funding; • bonuses – to individuals or teams; • gain sharing – gives a guaranteed portion of financial savings as long as specified targets are met; • shared savings – gain sharing for organisations, or teams; 42 • performance pay – links pay to performance; • performance contracts and agreements – gives the opportunity to get rid of nonperforming managers or organisations; • efficiency dividends – reducing agencies’ budgets by a small amount each year, while maintaining output levels; and • performance budgeting – outputs and outcomes specified along with a budget (Osborne and Plastrik 1997, p.146). All of these approaches might be used in non-competitive markets, e.g., in the case of the HFA, DPOs or hospitals. In all situations, however, key performance expectations must be established, they must be realistic and information is required to measure performance against those expectations. Hence the issues raised in the section on benchmarking performance need to be considered at the same time as rewards and sanctions66. Much of the attention in recent years in health care has focused on financial rewards and sanctions. Budget-holding in New Zealand and the United Kingdom, for example, involves the setting of a budget and offers providers the ability to use savings for new services. There is a reasonable amount of evidence which suggests that health care agencies do respond to financial incentives – although the same literature also notes that ‘doing a good job’ and ‘recognition by one’s peers’ are also powerful incentives (see the material reviewed in Cumming 1996). Research into individual provider behaviour does suggest responsiveness to changing financial incentives. For example, Mooney reports on changes in moving from a capitation to a mixed capitation/fee-for-service payment system in Copenhagen: doctors provided many more services in their own surgeries and reduced referrals and hospital admissions, in response to receiving fee-for-service payments for specific conditions (Mooney 1994). The key issue here is: are the changes efficient? Given our limited knowledge of the relative efficiency of providing services in hospitals compared to doctors’ surgeries, this is unclear. This is a key problem in health policy: we need to be clear about what we are trying to achieve and to be sure that the incentives support those goals. Research on GP fund-holding in the United Kingdom suggests: that fundholders reduced the rate of growth in prescribing rates in the early stages of fundholding; mixed evidence in relation to changes in rates of referrals to hospitals and emergency care; and a rise in the number of clinics provided in GP practices (Goodwin 1998). Other studies have found more muted responsiveness to financial incentives. Whynes 66 Once performance information systems begin to be used for financial rewards, or to award contracts to particular providers or plans, they must be robust enough to withstand legal challenge. 43 and Baines cite evidence of equivocal results, and argue that GPs, for example, have limited ability to directly respond to income incentives (Whynes and Baines 1998). In relation to hospitals, United States evidence suggests that payment methods – such as the prospective payment system (PPS, using diagnostic related groups or DRGs) – can lead to reduced costs when compared with traditional payment methods (such as reimbursement of actual costs). However the evidence on quality of care is less comprehensive, and suggests that in some cases, quality of care has worsened. Furthermore, there is also evidence of undesirable effects – for example, cost-shifting to budgets not included in caps; and DRG-creep, where the case complexity of the service mix provided increases as coders choose the most expensive code against which to charge care67. Financial incentives have clearly made a difference in the United States insurance market. HMOs are widely recognised as reducing costs, although the separate impact of financial incentives from practice style are hard to identify. Furthermore, such gains may well have been easier to make in the bloated United States system. The worst fears of the impact of HMOs on quality of care have perhaps not been realised: there are HMOs which provide better, the same or worse care than indemnity plans (Robinson and Steiner 1998). There is, however, a concern that poor care may be more likely for those who are chronically ill, low income enrollees in worse health, impaired or frail social HMO enrollees and Medicare home health patients, many with chronic conditions and diseases (Managed Health Care Improvement Task Force 1998, p. 104). Conclusions Any method of rewarding or sanctioning providers or DPOs provides both positive and negative incentives. The key steps are to determine the priority goals and objectives for providers or DPOs, to identify meaningful measures or indicators, to negotiate rewards and sanctions in the event that goals and objectives are or are not met, and to monitor carefully any potentially undesirable effects (e.g., on classification, health care outcomes, or in relation to cost-shifting). Some baseline monitoring of service outputs or processes may be required to ensure that providers or DPOs do not skimp elsewhere. However, although these approaches seem to be easily implemented they are in fact complex and require careful thought and implementation. C Competition for the market: franchising Franchise bidding takes financial incentives one step further: by rewarding purchasers with the opportunity to take an entire market. Thus, the ability to earn extra revenue and the threat of loss of business may provide powerful incentives to minimise costs and to provide quality care. 67 As the DRG system is based on current resource use and average cost, the extent to which it promotes efficiency is questionable – an approach based on identifying better practice may be a more appropriate starting point (Donaldson and Gerard 1991; Donaldson and Magnussen 1992). 44 Williamson, in his work on the economic institutions of capitalism, noted that uncertainty and the complexity of health care services may mean that franchising in health care is difficult (Williamson 1985). In spite of an exhaustive literature search, I was unable to find much evidence about franchising in the health literature. One example of franchising in health care is however the Arizona experience. In 1982, Arizona was the first state in the United States to get approval to develop and implement a mandatory statewide Medicaid managed care system. Prior to this, health care for low-income people was provided and funded by county governments, with care provided by county hospitals and clinics or through contracted providers. The managed care programme, Arizona Health Care Cost Containment System (AHCCCS), began with acute medical care services, based around prepaid capitated financing of private health plans68. The programme initially had a number of problems: Jacobson (1992) suggested that making utilisation information publicly available would allow an increasing number of participants into the franchise bidding process and would perhaps assist in reducing the potential for insolvencies. According to the General Accounting Office (GAO) (1995), the programme has gradually improved its administrative performance, bidding had become very competitive and the approach has grown to encompass a wide range of services (including long term care for people with disabilities)69,70. It is also estimated to have made significant savings. The programme does, however, involve higher administration costs compared to other states, although these costs in the Arizona case ‘more than pay for themselves in net program savings’ (General Accounting Office 1995, p. 2). Although the Arizona experience appears to suggest franchising via a bidding approach can be successful, it is not clear how relevant this experience is in the New Zealand environment. The Arizona population was 3.75 million in 1990, about the same population as New Zealand has now. However, details of the previous health payment system – with which the savings from the AHCCCS approach are compared – are not clear. Further, the Medicaid market is only a part of the entire market in which health plans operate: hence loss of the Medicaid market may not 68 Although all Medicaid beneficiaries may specify which available plan they will join, only about 56 per cent actively pick a plan. The rest are assigned by AHCCCS (General Accounting Office 1995). 69 In the bidding process, more weight is given to access and quality factors than to capitation rates (i.e. the price bid); plans are required to meet certain standards for primary care coverage and financial and operational performance is routinely monitored. The bidding process considers: the extent of health plan’s provider network, including number, type, geographic location of physicians; member and provider services, quality management and administrative tasks; and financial ability to meet contract terms. A minimum allowable capitation rate is also established, prohibiting low bids that might force health plans to curtail services and adversely affect the quality of care provided. At the time of the 1995 GAO report, a new quality management system was to be implemented. 70 Contracts stipulate the type and location of providers in each county, with minimum providerenrollee ratios. The State also requires that plans make transportation available where necessary. The State limits the number of contracts awarded in each county: to increase the chances of missing out and therefore increasing the incentives to submit low bids. 45 necessarily imply that a firm loses all its market share. This may mean that entry is relatively easy and exit from the market not catastrophic: in the New Zealand situation if the whole population were covered by this arrangement, entry and exit may not be as easy and hence competition much more limited. It is interesting that the approach allows consumers to choose between plans, but that only a slight majority of the population chooses to make a choice. This may indicate the strength of an approach relying on a bidding process run by a regulatory body in addition to consumer choice. D Purchaser competition Market-based reforms in health care begin with an assumption that competition amongst providers and/or amongst purchasers will provide incentives for good performance. In the market for providers, where consumers buy services directly and purchasers buy services on behalf of consumers, competition is argued to provide price information so that comparisons can be made between alternative providers. Those providers offering quality care at a good price will be offered additional work, and will thrive. This process is argued also to offer incentives for reducing costs and encouraging cost-effectiveness and good quality care. Similarly, in the market for health plans, consumers are argued to make choices based on price and perceived quality of care offered by a health DPO. Those plans which offer good quality care at a reasonable price will expand their market share and profitability and all plans will have incentives to reduce costs and encourage costeffective and good quality care. The most developed proposals for competition between purchasers are those designed in the United States. Many of these proposals envisage a significant role for some agencies in managing the market in order for it to operate in ways which promote allocative and technical efficiency and equity. The key question for consideration here is: to what extent will competition between purchasers reduce the risks of poor purchaser performance, without placing significantly greater risks on the sector? The following sections consider evidence on the effects of competition in health care to assist in answering this question. Unfortunately, in answering this question for New Zealand, the key problem is that much of the experience of the effects of competition between providers and purchasers comes from the United States, which is moving towards competitive models from a very different environment to that which exists in New Zealand. Thus, any benefits of competition in the United States are usually compared to the previous United States system: this does not supply us with good information about how purchaser competition might compare with the status quo in New Zealand. This is particularly the case when we are sure that further moves to devolve purchasing will lead to higher management costs in New Zealand. Thus there needs to be significant quality or efficiency gains to offset these costs. 46 Competition at the provider level In many countries, the use of competition as an incentive in socially-organised health care is relatively new, particularly beyond primary care services. The United States has the most extensive experience of competition, but much of this evidence comes from provider markets71. Studies in the 1980s of the effects of competition between hospitals focused on traditional competition in the United States hospital markets. This involved hospitals competing for physicians who would undertake to admit their patients to a particular hospital, guaranteeing hospitals a number of patient admissions. Such competition was based largely on amenities and technology. Price competition played little role in such a market as a result of the third-party payment system. The effects of competition between hospitals in such circumstances would depend on the relative bargaining position of hospitals and physicians. The more hospitals, the higher expenditures may be in order to attract physicians; the more physicians, the lower expenditure as hospitals need not compete as much to attract physicians and their patients (Zwanziger and Melnick 1996). Many cross-sectional studies did indeed find evidence of higher expenditures in areas with more hospitals (for a review see Cumming n.d.). In addition, there was evidence of increased duplication of equipment and specialties in competitive markets as compared to non-competitive markets (Bruce and Jonsson 1996). In the late 1980s, in the United States, the introduction of selective contracting and prospective payment72 altered incentives in the hospital industry. Once again, the effects in theory depend on the relative bargaining position of three parties: purchasing authorities, physicians and hospitals. In this case, the greater the number of hospitals the more credible are threats to shift contracts between hospitals; the greater the competition between insurers, the less dependent are hospitals on any one particular insurer and the better the hospital’s bargaining position; the greater the competition between physicians and the less ‘bonding’ between hospitals and physicians, the greater the ability for hospitals to reduce expenditure without fear of losing physician loyalty. In addition, insurers must consider the effects on consumers of changing contracts; who in turn will take into consideration ‘switching costs’, i.e. the costs of changing insurers in order to remain with a particular physician and the costs of changing physicians in order to remain with a particular insurer (Zwanziger and Melnick 1996). Thus, provider competition in the United States – in the presence of selective contracting and a massive oversupply of beds – is reaping benefits in the form of cost reductions. There is also evidence of competition improving performance at the hospital level where GP fundholders in the United Kingdom have threatened to move 71 Providers in the United States have traditionally not operated within a cash-constrained environment. 72 Selective contracting is where a purchaser is able to select particular providers to contract with to provide care. The opposite approach – universal contracting – occurs where a purchaser must contract with ‘all willing providers’. In the United States, prior to about 1983, universal contracting was the norm. Prospective reimbursement occurs when providers are paid in advance for care to be provided. The DRG system is an example of a prospective reimbursement system. 47 contracts elsewhere. This pressure may have been successful as a result of the attractiveness of marginal income available through cost per case contracts (Le Grand, Mays et al. 1998). Provider competition in New Zealand It is unclear if the experiences of the United States and the United Kingdom can be replicated in New Zealand, particularly when considering hospital markets. Greater competition in New Zealand amongst hospitals in the presence of selective contracting may provide incentives for improved performance in the longer term, but increased competition in the form of more hospitals may lead to higher average prices: if occupancy rates fall; where expensive equipment is duplicated in particular centres; and if specialists’ earnings are evened out over the private and public sectors. Evidence on economies of scale in New Zealand suggests that significant gains could be reaped by down-sizing larger hospitals and merging smaller hospitals (Devlin and O’Dea, 1998), suggesting that in rural areas, competition between hospitals is unlikely to be efficient. The implications for competition of down-sizing larger hospitals are unclear – given uncertainties over the effects of duplication and lower occupancy rates. In any event, the results may not apply to particular services, especially where the relationship between outputs and outcomes may be strong. Competition in New Zealand is thus more likely to be efficient amongst primary and community care providers; growth in the availability of secondary care beds in the private sector may, however, signal that some competition at the margins is possible in elective care also (see Appendix D), particularly if there is significant spare capacity in the private sector. The effects of competition in provision of care are relevant when considering the potential effects of introducing competition at the purchaser level. With a limited number of hospitals providing secondary and tertiary services, combined with a limited supply of specialists, some hospitals will have monopoly power in negotiations with purchasers; while specialists will have monopoly power in negotiations with hospitals: competing purchasers may therefore find it more difficult than a monopsony purchaser to restrain costs. Competition between health plans – what is the evidence? A number of studies have examined differences in utilisation, expenditure and quality of care between traditional indemnity insurance plans and HMOs (for reviews see Miller and Luft 1994; Luft 1996; Robinson and Steiner 1998). These studies clearly show, in comparison with traditional plans: HMOs have lower utilisation rates for intensive services, lower costs and no worse health outcomes (except in mental health); fewer HMO enrollees are satisfied with the quality of care and patientphysician interactions; and HMO enrollees have greater satisfaction with costs. Of more interest in this context is the effect of managed care competition on traditional indemnity insurance activities and costs73. Miller and Luft in 1994 noted there had 73 Some commentators note that HMO activities have spread to traditional indemnity insurance. Robinson and Steiner for example cite reference to a new type of organisation ‘managed indemnity insurance’ (Robinson and Steiner 1998, p. 18). 54 establishment of professional purchasers – as in the United Kingdom and New Zealand – might not be a more effective means of challenging the performance of providers (Reinhardt, 1997, p. 63)91. The relative advantages and disadvantages of each approach deserves further consideration – and piloting – in New Zealand. As the section on cream-skimming discussed, it is likely that the incentives on purchasers or health plans to perform will very much depend on the extent to which cream-skimming is profitable and how well health plans can segment the market. Success at this may leave some consumers with very few choices in relation to coverage, and may result in consumers choosing packages of care based on coverage rather than price. A key issue in designing managed competition relates to the role of financial incentives for consumers in choosing between health plans. An early version of managed competition envisaged competition only in relation to quality of care. In other versions of managed competition, however, consumers would pay some of the cost of coverage: for example, only the cheapest DPO might be fully subsidised, with consumers paying additional amounts for the same coverage offered by the more expensive plans. This is argued to provide a financial incentive to choose carefully; but might also be seen as compromising ‘social solidarity’ objectives depending on the extent to which choice becomes affected by ability to pay (Reinhardt 1997). An issue which is not covered to this point is, however, that of the responsiveness of organisations to changing tastes and technologies. In spite of the considerable importance that might be attached to this concept of efficiency, the literature does not make much of it in the context of competition. Of considerable interest is the extent to which competing health plans and providers respond more quickly to market signals than in a non-competitive environment; and whether or not such responsiveness is compatible with any health maximisation or cost-containment goals. A final point is also the extent to which consumer responsiveness is improved by competition. Clearly, some insurers in United States markets are viewed as responsive to some needs: however, the same responsiveness may be viewed less positively if used to cream-skim and to sell ineffective and allocatively inefficient services. The Managed Care Improvement Task Force in California noted that consumers who have a choice of plan are more likely to be satisfied with their plan than consumers with no choice (Managed Health Care Improvement Task Force 1998, p. 205). This needs to be balanced against the limits on choice due to a) restricted use of providers and b) the effects of increasing levels of ‘management’ of individual care. In addition, the more vulnerable populations are unlikely to be beneficiaries of increased choice (Managed Health Care Improvement Task Force 1998, p. 128). 91 It is likely, however, that arguments will be made about the extent to which purchasers can adequately represent consumer views in purchasing health care, and about how to promote good purchaser performance. 55 E. Investment in purchaser expertise and knowledge As a final point, Light raises the issue of how to promote effective commissioning. Using evidence from the United States, he argues that: • commissioning organisations need to be large and strong – requiring marketplace clout to take on inefficiencies and a large population base across which (i) to configure clinical services in cost-effective ways, (ii) to bear risk, (iii) to support a highly skilled team to undertake contracting, (iv) to spread administrative and transaction costs, (v) to avoid inequalities and service fragmentation and (vi) to advance prevention and health gain aggressively; • commissioning teams need to be smart, well trained and technically supported; • re-engineering clinical care for cost-effectiveness takes time and money; and requires more attention to be paid to the diffusion of successful and cost-effective new ways of doing things; and • commissioning through primary care has serious drawbacks, because they do not have: sufficient clout; technical skills and infrastructure; time and training; and the ability to address inequities and wasteful practices in primary care itself (Light 1998). These issues need further serious consideration in the New Zealand context if we are serious about ensuring good purchaser performance. In particular, more attention needs to be paid to developing the purchasing skills of both professionals and managers. There is no guarantee that good purchasing will develop of its own accord. Conclusions Managing the risks of poor purchaser performance is complicated by the information problems which plague the health and disability sector. The move in 1998 to a single purchaser makes it difficult to benchmark purchaser performance in New Zealand (even though one strong purchaser may be better than having a number of weaker purchasers), while the lack of a clear counterfactual – what would have happened if there were a ‘better’ purchaser – makes assessment of a purchaser’s performance difficult. The current purchaser performance framework monitors particular tasks which the HFA is expected to complete, and monitors changes in the levels of outputs over time. Each of the first four approaches set out here for managing purchaser performance – monitoring and benchmarking performance; using sanctions and rewards; franchising; and competition between purchasers – requires identification of priority goals and the collection and interpretation of indicators. A significant investment is required to establish indicators, and to adjust them adequately for different levels of health status. A focus on health care outputs and processes is more likely to be feasible in both the short and long terms. An approach based on practice and other guidelines offers advantages in using evidence on effectiveness and best practice, but requires auditing at an individual practitioner or institutional level. 56 Having identified key performance goals and indicators which are tightly defined and audited, performance rewards and sanctions can also be designed to ensure that purchasers focus on these goals. Care needs to be taken that purchasers do not neglect activities in services which are not monitored, however. The literature suggests that financial incentives can have an important effect on behaviour, but the existence of a myriad of health sector goals can make it hard to ensure that all goals are met with the same set of incentives. Increasing the number of purchasers may reduce the risks of poor purchaser performance if a single purchaser is systematically performing poorly, but smaller purchasers may not have the same level of expertise or power over providers in order to challenge provision. Competition for the market provides an alternative approach to managing purchasing risks. At present, this would seem an unlikely scenario, given the limited expertise in purchasing in New Zealand. Even in the longer term, the small size of the New Zealand market may limit opportunities for franchising to primary care purchaser/providers. The costs involved in managing a franchise process may not be worth the gains which might be made. Full-scale purchaser competition in New Zealand – between plans offering comprehensive services – appears an unlikely scenario, given our small and widely dispersed population. Competition between plans may well improve incentives for better performance, but competition requires much better information to be collected and to be made available at a local level. It is not clear if the risks associated with such competition would be worth the potential gains involved (see Appendix D), even if cream-skimming and cost-shifting risks can be minimised. Further analytical work, drawing on the international literature, comparing the potential benefits of a national purchaser with competition between purchasers is needed urgently to inform this debate. 57 4.3 Managing the risks associated with cost-shifting Key approaches to managing the risks associated with cost-shifting include: A. Integrating budget responsibility. B. Clarifying boundaries between budgets. C. Developing practice and referral guidelines. D. Private ownership of purchasers and providers. A Integrating budget responsibility In New Zealand, the move in 1993 to integrate primary, secondary and disability support funding for health care was in part due to concerns about cost shifting92. Theoretically, this integration of budget responsibility might encourage the purchaser to ensure that inefficient cost-shifting (within the budget) does not occur. Whether this happens will depend on a) the incentives on the purchaser to provide the best value for money and b) the purchaser’s specification and monitoring of contracts (and use of the techniques set out below). It could be argued that introducing competition at the purchaser level will provide greater incentives to ensure that inefficient cost-shifting does not occur within the purchaser’s budget, and such incentives will be further enhanced if financial incentives at the provider level to support this (e.g., via ownership pressure, bonuses). So, for example, a fully integrated purchaser-provider, with staff on salaries with bonus incentives or as part-owners, competing against other similar organisations, might have incentives to minimise cost-shifting within the organisation. Staff-model HMOs in the United States and budget-holding arrangements in New Zealand and the United Kingdom, for example, are based on this approach93. What is less clear is how individuals within organisations react if, for example, separate budgets and budget responsibility within an organisation are established. Even within an organisation which holds a budget across a wide range of services, as soon as a key incentive is to remain within a budget, cost-shifting may become both an internal and external management problem. However, even if integration did reduce cost-shifting within a DPO or organisation, there is potential for cost-shifting outside the DPO or organisation. Furthermore, the integration of budgets is unlikely to occur at the higher levels of the system: i.e. between personal health, disability support and public health; between government votes; and between public and private financing. 92 Originally, ACC health care was also to be integrated into the RHAs’ budgets, but this never occurred. 93 There is some United Kingdom evidence on the potential of fund-holding to alter the pattern of service use across previous budget boundaries (personal communication, Nicholas Mays). 58 B Clarifying boundaries between budgets One way of reducing the potential for cost-shifting is to clarify boundaries between budgets. This requires clear definitions of what is publicly financed vs privately financed; between personal health, public health and disability support; and between contractual responsibilities94. In a number of other countries, the boundary between what is publicly financed and what is privately financed is clearer than in New Zealand95. But countries like New Zealand, the United Kingdom and Spain do not clearly specify the services which are to be covered by their national health insurance systems. They are not entitlementbased systems but they do attempt to be comprehensive. The general level of specification of contracts between purchasers and providers and between the government and its citizens, combined with a lack of guarantee for access and with budget caps, means that cost-shifting between budgets can occur easily. Hence, an important way in which cost-shifting may be discouraged is to better define the boundaries associated with publicly financed health care. Although this can partly be achieved by ensuring people or services are only covered by one set of arrangements (e.g., those over a certain income can go private if they wish), it also requires better boundaries between core coverage and supplementary insurance. This is discussed in more depth in the section above on service specification (also see the discussion below on practice guidelines). Similar issues may arise if alternative risk pools are established. For example, GP responsibility for care for an individual patient may be capped at a particular amount. Clearly, this is a fairly clear boundary, which can be audited96. Risk pools established in other ways may not have such clear boundaries. These may be based around particular diseases (e.g., renal care, AIDS)97; past use of services; or there may be completely separately funded providers who offer care in particular areas (e.g., to those living in low income areas). All are problematic: the first and second offer 94 Reducing cost-shifting between government votes would probably require complete budget integration, or allocation of responsibility for flow-on costs to each sector. For example, if health fails to keep a person with a psychiatric problem out of trouble and the individual ends up in prison, the costs could be charged against the health budget. 95 For example, in Germany and the Netherlands, those insured are covered fully by either the private or the public system. They cannot straddle both insurance systems, except for supplementary coverage for amenities. Only those below a certain income level must belong to the social insurance system; those above can choose the private system, but they then cannot return to the social insurance system. In Canada, an individual could choose to stay outside the public system, but he or she would then be responsible for paying for all their own care. Private insurance for services covered under the national health insurance system is prohibited. In the United States Medicare system, the definition of coverage is such that the boundary between Medicare and Medigap insurance is clear: the latter covers things such as: co-payments for particular services; foreign care; care in the home; preventive care services such as cholesterol and diabetes screening and prescription drugs. 96 This approach may require a definition of appropriate care however to ensure that the care offered is not frivolous nor outside a concept of core. 97 As in the United States: renal care is funded through Medicare for all United States citizens. 59 incentives to diagnose more cases and to offer more services; the third requires a boundary to prevent plans or providers from neglecting certain areas in the knowledge that care will be made available by someone else. Cost-shifting in relation to user charges is also a potentially large problem in New Zealand. Pharmac may simply be shifting costs onto individuals with its reference price scheme, although if patients’ needs can still be adequately met with the remaining fully subsidised medicine or if drug companies cut their prices and become more efficient, Pharmac policies are indeed improving overall efficiency. With a move to capitation payments for GP care for those with community services cards, general practitioners who feel they are earning insufficient revenue to cover costs have the ability to shift costs to those continuing to pay fee-for-service charges: these charges are limited only by competition, the effects of which are uncertain in New Zealand. Furthermore, attempts to shift care out of hospitals and into GP offices may also increase cost-shifting to those paying user charges. Cost-shifting to those paying user charges might be reduced by: • regulating all user charges by putting a zero or maximum amount that can be charged per visit or episode of care. For example, in the United States, HMO legislation limits user fees for care and it is illegal for physicians and patients to contract privately for the provision of services already covered by Medicare. Physicians and hospitals must accept Medicare’s scheduled fees for these services. This approach would have the additional advantage of maintaining affordability of care. GPs in New Zealand have shown great reluctance for these charges to be eliminated or regulated, however, and the approach of eliminating user fees would in all likelihood result in higher expenditure for the government, at least in the short term98; and • putting in place incentives not to charge user fees. For example, in Australia, general practitioners are encouraged not to charge patients additional fees. Doctors can accept the 85% Medicare reimbursement of the schedule fee as full payment. Alternatively, the doctor can charge the patient a fee: however, the amount above the government’s fee cannot be privately insured against (Donaldson and Gerard 1993). C Developing practice and referral guidelines In many circumstances, it may be possible to develop practice or referral guidelines which set out the responsibilities of different providers in caring for patients with particular conditions. This approach may not only reduce inappropriate cost shifting, but may also enhance quality of care and reduce under-servicing. In developing the lead maternity carer approach in New Zealand, a set of protocols was attached to provider contracts setting out the circumstances under which patients were to be transferred to specialists. Thus an attempt was being made to clarify the boundaries between providers. To my knowledge, no research has been undertaken to determine the effects of these guidelines, but the approach appears to offer some promise. 98 In the longer term, if good care is provided in the primary care sector, hospital admission rates may fall as a result of such a policy. 60 Shared care plans developed between hospital services and GPs in caring for people with ongoing mental illness provide another example: if the plans are developed based on agreed parameters, and can be monitored, the opportunities for cost-shifting could be reduced. The approach relies on good communication between providers as well as goodwill in following the guidelines through in practice, and in negotiating around any difficulties which occur. Financial incentives might accompany such guidelines if necessary, and may be accompanied by an audit process to ensure they are being followed. D Private ownership of purchasers and providers In New Zealand, many hospitals are currently owned and operated by the government. Within the current environment, public hospitals have residual risk passed on to them: they argue they must treat most of those coming through their doors (especially urgent cases), but they do not get paid for service levels above those contracted for. The government bears this financial risk, in the form of deficits. Potentially, this financial risk could increase if competing providers and DPOs are able to cost-shift to public hospitals. A similar situation may arise at purchaser level: where the government-owned HFA is in competition with privately-owned DPOs, and where particular care were left to the responsibility of the HFA or as a result of cream-skimming. Thus, if the approaches laid out above are not successful at preventing cost-shifting, the government may be tempted to argue for privatisation in order to shift residual risk to the private sector (at a price, of course). Yet governments play a significant role in health care in all developed countries. Even in the United States, nearly 50 per cent of health care is funded by government agencies. Furthermore, governments play key roles in regulating health care: for example, by requiring hospitals to provide care to all urgent cases or regulating the growth of expenditure or premiums. These key government roles suggest that even if all purchasers and providers are privately owned, governments may find it difficult not to become involved where purchasers or providers are in trouble. Conclusions Integration of budgets may reduce incentives to cost-shift within organisational budgets, and this may especially be an effective approach to managing cost-shifting where financial incentives also support cost containment within the overall organisational budget. Yet there appears to be no available evidence to support this. In any case, cost-shifting outside of the organisational budgets will remain a problem. The key approaches to managing cost-shifting appear to be to design contract boundaries in ways which discourage cost-shifting. Regulatory or contract approaches may well be the simplest approaches, whereby contract boundaries or referral guidelines could be used to monitor cost-shifting. 61 5 CONCLUSIONS The discussion in the previous chapters has focused on the key risks associated with general moves to devolve purchasing in New Zealand. In this section, a summary table (Table 5.1) is provided of the key risks that might arise in each of the devolved purchasing arrangements which may develop in New Zealand. Conclusions are then drawn about each of the risks discussed above, both in general and in the context of the information summarised in Table 5.1. 5.1 Cream-skimming Moves to capitate primary care providers or place them at increased financial risk, along with moves to promote competition between primary care providers and DPOs, will provide incentives to cream-skim. The likely extent of such cream-skimming is unknown, but if it does occur systematically it has the potential to increase overall expenditure, encourage pressure for additional expenditure, lead to inequities in access and quality of care, provide diminished incentives for technical efficiency and cost-effectiveness, and lead to differences in profitability for DPOs. Cream-skimming is not an issue in regional monopoly DPO models, and may not be a problem with carve-out purchasers, but it is likely to become an issue within the monopoly HFA model (as primary care providers are moved onto capitated or other risk contracts), and particularly with models which offer consumer choice between DPOs. Methods to control cream-skimming include encouraging larger risk pools; compensating for risk; limiting risk; defining service entitlements and regulating the purchaser or provider market. This paper has focused particularly on defining entitlements and regulating the insurer or provider market. 62 Table 5.1 Summary of risks and management options by devolved purchasing arrangements Cream-skimming Management Options Poor purchaser performance Management Options Cost-shifting Management Options Monopoly HFA (Status Quo) •Not applicable •However, increased use of capitation to pay providers may see providers engaging in cream-skimming, – see risks below under Choice of DPO – Individual model •Not applicable •See below, Choice of DPO – Individual model Crown risk •Higher overall expenditure and pressure for increased governmentfunded health resources •Pressure to bail out purchasers or providers •Potential legal action Consumer risk •Poor access to care •Poor quality care •Higher taxation •Higher private expenditure •Poor access to care for those unable to afford private care •Lower health status than would otherwise be the case •Under-servicing of some consumers where costs greater than fee paid Health sector •Poor value for money •Excessive profits or losses where contracting is poor •Higher overall expenditure •Likely to generate ongoing policy disputes about monolithic purchaser and no consumer choice All risks •Monitoring of purchaser by Crown •Use of performance sanctions and rewards •Franchising •Competition between purchasers •Smart purchasing All raise performance measurement and information issues Competition between purchasers raises additional risks (see below, Choice of DPO models) Crown risk •Greater expenditure in other votes (e.g., ACC, justice, DSW) •Greater expenditure for crown-owned providers where risk is shifted from other providers Consumer risk •Greater private expenditure if access to publicly-funded services is reduced (must go private) •Lower access to care if access to publicly-funded services is reduced (and cannot afford to go private) •Greater shunting of consumers between providers •Greater private expenditure in the form of user charges Health sector risk •Higher overall expenditure as claims in the private sector increase with more care provided privately at possibly higher cost •Reduced equity of access as those who cannot obtain publicly-funded care are unable to afford private care •Reduced or improved costeffectiveness where the type of care provided is driven by budget constraints rather than need or costeffectiveness •Difficulties in planning and budgeting in areas where costs are commonly shifted to Inappropriate market signals which might arise as a result of crosssubsidies (although such crosssubsidies may also enhance equity of access). Crown risk •HFA directed to consider impact of purchasing on other votes and to report on such risks •HFA directed to specify contracts tightly and to monitor service delivery in order to prevent cost-shifting Consumer risk •Crown specify maximum user charges (politically difficult in relation to GPs) •Crown specify and monitor access to services (service specification issues - see text), including protocols for referral Health sector risk •Encourage integration of budgets across substitutable sets of services 63 Table 5.1 (cont) Monopoly HFA with specialist purchasing of particular services (carve-outs) •Carve-out purchasers’ consumer profile is likely to be riskier on average than the HFA profile •Carve-out purchasers may try to engage in cream-skimming of the most serious cases with a flat capitation rate – this would seem unlikely however if it is clear which patients are to be cared for by which carveouts Hence risks may include: Crown risk •Higher expenditure or reduced other services where most serious cases become responsibility of HFA Consumer risk •Must fall back on HFA – may lead to differences in care between those in HFA and those in carveout care •HFA clearly specify those covered by the carved-out contracts •Monitoring of risk profile of carve-out purchasers •Limit financial risk associated with higher risk people (see text) •Risk rating •Regulations - contracts, with sanctions (see text) •As above, but •Overall risk may be reduced if some purchasers perform well while others do not •Ability to make comparisons between carved-out purchasers in different parts of the country, subject to performance monitoring issues (see text) •Devolved purchaser performance becomes responsibility of HFA •Issues as above for Monopoly HFA •As above: in particular, carve-outs may try to shift costs elsewhere As above, Monopoly HFA Regional Monopoly Purchasers •Not applicable •Not applicable •As Monopoly HFA •Purchaser performance can be compared with other regional monopoly purchasers •Issues as above, Monopoly HFA •Integration may reduce cost-shifting; otherwise, as above, Monopoly HFA Consumer risk Purchasers may refuse to care for the sickest consumers from another region - this is unlikely •Clear specification of purchaser responsibility to care for those from another region Choice of partially integrated DPO •As below, Choice of DPO - Individual model •As below, Choice of DPO - Individual model •As Monopoly HFA, perhaps limited by consumer choice which may force improved performance •Purchaser performance can be compared with other regional monopoly purchasers •Issues as above, Monopoly HFA •Cost-shifting from ICOs to HFA may be a problem - see above, Monopoly HFA •As above, Monopoly HFA 70 Clayton, C. and C. Scott (1998). Integrated Care Organisations: Delegated or Contestable Purchasing? (Draft). Wellington, Clayton and Associates Ltd, Scott and Associates Ltd. Congressional Budget Office (1993). Managed Competition and Its Potential to Reduce Health Spending. Washington, D.C., Congress of the United States, Congressional Budget Office. Consumers' Institute (1999), The best insurers. Consumer: 27-28. Crampton, P. (1999), Third sector primary health care. In National Health Committee, Discussion Papers on Primary Health Care. National Health Committee, Wellington. Crengle, S. (1999), Maori Primary Care Services. In National Health Committee, Discussion Papers on Primary Health Care. National Health Committee, Wellington. Culyer, A. J. (1991), The promise of a reformed NHS: an economists angle, British Medical Journal , 302(25 May), pp. 1253-6. Cumming, J. (n.d.). Competition in Health Services. Wellington, Health Services Taskforce (unpublished paper). Cumming, J. (1994), Core services and priority setting: the New Zealand experience, Health Policy , 29(1,2), pp. 41-60. Cumming, J. (1996). Contract Incentives for Crown Health Enterprises: A Literature Search. Wellington, Health Services Research Centre. Cumming, J. and C. Scott (1998), The role of outputs and outcomes in purchaser accountability: Reflecting on New Zealand experiences, Health Policy , 46, pp. 53-68. Cutler, D. M. and L. Sheiner (n.d.) Managed care and the growth of medical expenditures. Frontiers in Health Policy Research 1. A. M. Garber. Cambridge, Massachusetts, National Bureau of Economic Research and MIT Press: 77-115. Devlin, N. and D. O'Dea (1998). Hospitals. The structure and dynamics of New Zealand industries. M. Pickford and A. Bollard, Dunmore Press. Donaldson, C. and K. Gerard (1991). Minding our Ps and Qs? Financial incentives for efficient hospital behaviour. Health Policy 17: 51-76. Donaldson, C. and K. Gerard (1993). Economics of Health Care Financing: The Visible Hand. Basingstoke, Macmillan. Donaldson, C. and J. Magnussen (1992). DRGs: the road to hospital efficiency. Health Policy 21(47-64). Ellwood, P., N. Anderson, et al. (1971). Health maintenance strategy. Medical Care 291(May): 250-256. 71 Ellwood, P. M. and A. C. Enthoven (1995). Jackson Hole Revisited. Health Affairs 14(2): 24-39. Ellwood, P. M., A. C. Enthoven, et al. (1992). The Jackson Hole initiatives for a twenty-first century American health care system. Health Economics 1: 149-168. Enthoven, A. (1988). Managed Competition of Alternative Delivery Systems. Journal of Health Politics, Policy and Law 13(2): 305-321. Enthoven, A. C. (1988). Theory and Practice of Managed Competition in Health Care Finance. Amsterdam, Elsevier Science Publishers B.V. Enthoven, A. C. (1993). The History and Principles of Managed Competition. Health Affairs 12(Supplement): 24-48. Enthoven, A. C. (1994). On the ideal market structure for third-party purchasing of health care. Social Science and Medicine 39(10): 1413-1424. Enthoven, A. C. and R. Kronick (1989). A consumer-choice health plan for the 1990s. New England Journal of Medicine(5 and 12 January): 29-37 and 94-101. Enthoven, A. C. and S. J. Singer (1995). Market-Based Reform: What to Regulate and by Whom. Health Affairs 14(1): 105-119. Evans, R. (1984). Strained Mercy: The Economics of Canadian Health Care. Toronto, Butterworth. Foster, G. (1994). Gains from tendering elective surgery. Health Manager 1(3): 9-10. Gauthier, A. K., J. A. Lamphere, et al. (1995). Risk selection in the health care market: a workshop overview. Inquiry 32(Spring): 14-22. General Accounting Office (1995). Arizona Medicaid: Competition Among Managed Care Plans Lowers Program Costs. Washington, D.C., United States General Accounting Office. General Accounting Office (1995). Medicare Managed Care: Growing Enrollment Adds Urgency to Fixing HMO Payment Problem. Washington, D.C., United States General Accounting Office. General Accounting Office (1998). Medicare: Many HMOs experience high rates of beneficiary disenrollment. Washington, D.C., United States General Accounting Office. Glaser, W. A. (1991). Health Insurance in Practice. San Francisco, Jossey-Bass. Glennerster, H., M. Matsaganis, et al. (1994). GP fundholding: wild card or winning hand? Evaluating the NHS Reforms. R. Robinson and J. Le Grand. London, King's Fund Institute: 74-107. 72 Goldman, W., J. McCulloch, et al. (1998). Costs and use of mental health services before and after managed care. Health Affairs 17(2): 40-52. Goodwin, N. (1998). GP fundholding. Learning from the NHS Market: A Review of the Evidence. J. Le Grand, N. Mays and J.-A. Mulligan. London, King's Fund: 43-68. Hadorn, D. and A. Holmes (1997), The New Zealand priority criteria project. Part 2: coronary artery bypass graft surgery, British Medical Journal , 314, pp. 135-138. Hadorn, D. and A. Holmes (1997), The New Zealand priority criteria project. Part I: Overview, British Medical Journal , 314, pp. 131-134. Health Funding Authority (1998) What can I expect? Health and disability support services in New Zealand 1998/99 , Health Funding Authority, Wellington. Haas-Wlison, D. and M. Gaynor (1998). Physician networks and their implications for competition in health care markets. Health Economics 7: 179-182. Hall, J., A. Sheill, et al. (1993). Health outcomes: a health economics perspective. Sydney, Centre for Health Economics Research and Evaluation. Hanchak, N. A. (1996). Managed care, accountability and the physician. Medical Clinics of North America 80(2): 245-261. Health and Disability Commissioner (1998). Canterbury Health Ltd. Auckland, Health and Disability Commissioner. Health Canada (1997). Canada Health Act Annual Report 1996-1997, Minister of Public Works and Government Services. Health Funding Authority (1998). The Next Five Years in General Practice. Auckland, Health Funding Authority, Northern Office. Health Funding Authority (1998). Service Integration: Guidelines for the Development of Integration Demonstration Projects. Wellington, Health Funding Authority. Hyndman, J. (1998). The Marlborough Healthcare model. Health Manager 5(3): 17-18. Iezzoni, L. and L. Greenberg (1994). Risk adjustment and current health policy debates. Risk adjustment for measuring health care outcomes. L. Iezzoni. Michigan, Health Administration Press: 347-408. Jacobson, C. K. (1992). Health policy goals and firm performance: a transaction cost analysis of the Arizona Medicaid experiment. Health Insurance and Public Policy: Risk, Allocation and Equity. M. K. Mills and R. H. Blank. Connecticut, Greenwood Press. Jenkinson, C. (1994). Measuring health and medical outcomes: an overview. Measuring health and medical outcomes. C. Jenkinson. London, UCL Press: 1-6. 73 Kenkel, P. (1996). The new HEDIS: boon or burden? Health Systems Review 29(6): 17-19. Kennedy, E. (1998). 96,000 still waiting to see specialists says report. The Dominion. Wellington: 3. Kronick, R., D. C. Goodman, et al. (1993). The marketplace in health care reform: the demographic limitations of managed competition. New England Journal of Medicine 328(2): 148-152. Laffont, J.-J. and J. Tirole (1993). A Theory of Incentives in Procurement and Regulation. Cambridge (MA). Le Grand, J., N. Mays, et al. (1998). The reforms: success or failure or neither? Learning from the NHS Internal Market: A Review of the Evidence. J. Le Grand, N. Mays and J.-A. Mulligan. London, King's Fund: 116-143. Le Grand, J., N. Mays, et al. (1997). Models of Purchasing and Commissioning: Review of the Research Evidence. London, Department of Social Policy and Administration, London School of Economics and Kings Fund. Lenaghan, J., Ed. (1997). Hard Choices in Health Care: Rights and rationing in Europe. London, BMJ Publishing Group. Light, D. W. (1998). Effective Commissioning: Lessons from Purchasing in American Managed Care. London, Office of Health Economics. Lovatt, D. (1996). Lessons from contestability. Health Manager 3(1): 12-15. Luft, H. S. (1996). Modifying Managed Competition to Address Cost and Quality. Health Affairs 15(1): 23-38. Ma, C.-t. A. and T. G. McGuire (1998). Costs and incentives in a behavioral health carve-out. Health Affairs 17(2): 53-69. Managed Health Care Improvement Task Force (1998a), Improving Managed Health Care in California: Volume One: Executive Summary, Managed Health Care Improvement Task Force. Managed Health Care Improvement Task Force (1998b). Improving Managed Health Care in California: Volume Two: Main Report, Managed Health Care Improvement Task Force. Managed Health Care Improvement Task Force (1998c). Improving Managed Health Care in California: Volume Three: Appendices, Managed Health Care Improvement Task Force,. Matsaganis, M. and H. Glennerster (1994). The threat of 'cream skimming' in the post-reform NHS. Journal of Health Economics 13(31-60). 74 Mays, N. and J. Dixon (1996). Purchaser Plurality in UK Health Care. London, King’s Fund. McKenzie Webster (1997). Investigation into the Ownership and Purchase Issues for the Crown in Relation to Integrated Care Developments in New Zealand. Wellington. Miller, R. H. and H. S. Luft (1994). Managed care plan performance since 1980: a literature analysis. Journal of the American Medical Association 271(19): 1512-1519. Minister of Health (1998). Service Coverage Document: Appendix D to the Health Funding Authority's Funding Agreement with the Minister of Health. Wellington, Minister of Health. Ministry of Health (1995). Tertiary Health Services in New Zealand. Wellington, Ministry of Health. Ministry of Health (1997). Purchasing for Your Health: 1995/96. Wellington, Ministry of Health. Ministry of Health (1998). Licensed Privately Operated Hospitals in New Zealand. Wellington, Ministry of Health. Ministry of Health (undated). Review of 1994/95 RHA Contracting. Wellington, Performance Monitoring and Review, Ministry of Health. Ministry of Health (1999), HFA Performance Report: Quarter Three 1998/99. Ministry of Health, Wellington. Ministry of Health, Welfare and Sport. (1995). Health Insurance in the Netherlands. Rijswijk, Ministry of Health, Welfare and Sport. Mooney, G. (1994). Key Issues in Health Economics. Hemel Hempstead, Harvester Wheatsheaf. Mulder, J. (1995). Priority setting in health care: the Dutch experience. Seminar on Priority Setting in Health Care, Birmingham. Mulligan, J.-A. (1998). Health authority purchasing. Learning from the NHS Internal Market: A Review of the Evidence. J. Le Grand, N. Mays and J.-A. Mulligan. London, King's Fund: 20-42. National Committee for Quality Assurance (1997). HEDIS 3.0 Executive Summary. National Interim Provider Board (1992). Providing Better Health Care for New Zealanders: Report to the Government and the New Zealand Public. Wellington, National Interim Provider Board. New Zealand Medical Council (n.d.). The NZ Medical Workforce in 1987. Newhouse, J. P. (1996). Reimbursing health plans and health providers: Efficiency in production versus selection. Journal of Economic Literature 34(September): 12361263. 75 Newhouse, J. P., M. Beeuwkes Buntin, et al. (1997). Risk adjustment and Medicare: taking a closer look. Health Affairs 16(5): 26-43. Newhouse, J. P., J. Manning, W G, et al. (1989). Adjusting capitation rates using objective health measures and prior utilization. Health Care Financing Review 10(3): 41-54. NHS Executive (1999) Faster Access to Modern Treatment: How NICE Appraisal Will Work , NHS Executive, Leeds. O'Reilly, D., K. Steele, et al. (1998). Effect of fundholding on removing patients from general practitioners' lists: retrospective study. British Medical Journal 317: 786-786. Organisation for Economic Co-operation and Development (1996). New Zealand. Paris, Organisation for Economic Co-operation and Development. Osborne, D. and P. Plastrik (1997). Banishing Bureaucracy: The Five Strategies for Reinventing Government. Massachusetts, Addison-Wesley. Pauly, M. V., P. Danzon, et al. (1992). Responsible National Health Insurance. Washington, D.C., American Enterprise Institute for Public Policy Research. Pauly, M. V., P. M. Danzon, et al. (1991). A Plan for 'Responsible National Health Insurance'. Health Affairs 10(1): 5-25. Performance Management Unit (1998). Purchasing For Your Health 1996/97. Wellington, Ministry of Health. Physician Payment Review Commission (1996). Annual report to Congress. Washington, D.C., Physician Payment Review Commission. Plank, D. (n.d). Supplementary Insurance. Wellington, The Treasury. Plank, D. (n.d.). The Health Care Plan Market. Wellington, The Treasury. P rime Health Ltd (1998), Revised proposal to the Health Funding Authority: Western Bay of Plenty Integrated Care Organisation. Released by the HFA under the Official Information Act. Downloaded from <http://www.hfa.govt.nz> Reinhardt, U. E. (1997). Accountable health care: is it compatible with social solidarity? London, Office of Health Economics. Robinson, R. and A. Steiner (1998). Managed Health Care. Buckingham, Open University Press. Sheldon, T. and M. Borowitz (1993). Changing the measure of quality in the NHS: from purchasing activity to purchasing protocols. Quality in health care 2: 149-150. 76 Shipley, J. n.d. Policy Guidelines for Regional Health Authorities 1996/97 , Minister of Health, Wellington. Southern Cross Healthcare (1994). Here's how to select the healthcare plan that best suits you. Swartz, K. (1995). Reducing Risk Selection Requires More Than Risk Adjustments. Inquiry 32(1): 6-10. The White House Domestic Policy Council (1993). The President's Health Security Plan. New York, Times Books. Trebilcock, M. (1995). Can government be reinvented? The state under contract. J. Boston. Wellington, Bridget Williams Books: 1-35. Upton, S. (1991). Your Health and the Public Health: A Statement of Government Health Policy. Wellington, Minister of Health. van Barneveld, E. M., R. C. J. A. van Vliet, et al. (1996). Mandatory high-risk pooling: an approach to reducing incentives for cream-skimming. Inquiry 33: 133-143. van de Ven, W. P. M. M., R. C. J. A. van Vliet, et al. (1994). Risk-adjusted capitation: Recent experiences in the Netherlands. Health Affairs 13(1): 120-148. van Vliet, R. C. J. A. and W. P. M. M. van de Ven (1992). Towards a capitation formula for competing health insurers. An empirical analysis. Social Science and Medicine 34(9): 1035-1048. White, J. (1995). Competing Solutions: American Health Care Proposals and International Experience. Washington, D.C., The Brookings Institution. Whynes, D. K. and D. L. Baines (1998). Income-based incentives in UK general practice. Health Policy 43(1): 15-31. Wilkin, D., L. Hallam, et al. (1992). Measures of Need and Outcome for Primary Health Care. Oxford, Oxford Medical Publications. Williams, B., J. Coyle, et al. (1998). The meaning of patient satisfaction: an explanation of high reported levels. Social Science and Medicine 47(9): 1351-1359. Williamson, O. E. (1985). The Economic Institutions of Capitalism. New York, The Free Press. Wilson, G. (1995). Health purchasing - a regional health authority perspective. Public Sector 18(1): 11-14. Zwanziger, J. and G. A. Melnick (1996). Effects of competition in the hospital industry: evidence from California. Competitive approaches to health care reform. R. J. Arnould, R. F. Rich and W. D. White. Washington, D.C., The Urban Institute Press: 111-138. 77 APPENDIX A GLOSSARY OF TERMS AND ABBREVIATIONS Adverse selection Where individuals choose a devolved purchasing organisations or provider on the basis of the services provided, and some plans or providers end up with a registered population with higher health needs than on average. With a standard, unadjusted capitation formulae, such plans or providers will find it difficult to provide good care within the budget made available. Capitation Where individual providers or organisations are paid an amount per person registered with them, for providing care over a period of time (a month, a year). A flat capitation rate is one where the amount paid per person is the same for each and every person registered (e.g., $200 per person per year). Cost-shifting Where costs are charged against a budget which a) differs from that intended or b) differs from that traditionally charged. Cream-skimming (Risk-selection) Where plans or providers actively engage in practices which lead to a registered population which has lower health needs on average, in order to reduce expenditures. They may do this by encouraging healthy people to join, and discouraging those with poorer health from joining. Devolved purchasing Any purchasing strategy or policy which allocates responsibility for purchasing groups of services (e.g., pharmaceuticals, secondary care services, community services) to non-government organisations other than the Health Funding Authority (HFA). Organisations other than the HFA then become responsible for negotiating contracts with other providers and fund such services from budgets allocated to them. 78 Efficiency101 Technical efficiency is obtained when effective services are provided at least resource cost, i.e. using the minimum amount of resources necessary. Cost-effectiveness is obtained when resources are concentrated on effective services, provided at least cost, that offer the best payoff in terms of health. Enrolment Where individuals enrol with a particular DPO or provider, whose role it is to organise the care of those who are enrolled. Equity Providing fairer access to care and improving the health of the most disadvantaged. Health Funding Authority (HFA) The principal funder of health and disability support services in New Zealand. The HFA is a crown agency, and receives its funding from the Government. Health Maintenance Organisations (HMOs) United States health organisations, which receive capitation funding from employers and individuals to insure and arrange the delivery of health care to individual members. Health Care Outcome A change in the health of an individual or population attributable to a particular health service, e.g. reduction in blood pressure; restoration of mobility following a hip operation. Health Outcome A change in the broader health of an individual or population, e.g. increase in average life expectancy, reduction in vaccine-preventable disease, reduction in cancer rates. Independent Practitioner Associations (IPAs) Umbrella organisations representing groups of GPs who negotiate budgets and contracts with the HFA on behalf of those GPs. Integrated care See managed care. 101 See Culyer (1991) for further detail. 79 Managed care A term used to describe a range of practices and organisations, usually ‘any system of health service payment or delivery arrangement where the health DPO attempts to control or coordinate use of health services by its enrolled members in order to contain health expenditures, improve quality or both’ (Physician Payment Review Commission 1996 cited in Managed Health Care Improvement Task Force 1998). Monopsony Where a single purchasing authority is responsible for purchasing. Risk-rating Where flat capitation rates are adjusted for differences in the health status of populations or individuals. For example, payments for older people may be 1.5 times payments for all other people, as a result of having a higher ‘risk’ of needing health care. Risk-selection See cream-skimming. 86 Organisational Form Key characteristics Social Health Maintenance Organisation (SHMO) • Demonstration projects • Involves ‘non-physician case co-ordinators’ • Combines social and health care, both acute and long term, into a single casemanaged delivery system • Targeted at elderly Medicare beneficiaries (65 years or over) • Aims for an integrated approach to older people’s health and social care to improve appropriateness of care and to reduce long-term care expenditure • Co-ordinators assess needs and arrange for provision of services • Salaried co-ordinators are subject to utilisation review but bear no financial risk • Constrained by range of services included in SHMO package, but is comprehensive as includes: hospital, physician and home health services; chronic care benefits such as personal care, homemaker services; nursing home coverage; expanded benefits such as prescription drugs, eye glasses and dental care • SHMO is able to limit the number of highly dependent new enrollees Programme of AllInclusive Care for the Elderly (PACE) • PACE enrollees are all at risk of nursing-home placement • Most are low income, 55 years or over • Members are enrolled on a Medicaid capitation basis, with Medicare covered services billed on a fee-for-service basis • Each site may negotiate Medicaid/Medicare capitation payments with providers • Sites may develop benefit packages not allowed under existing programmes • Full range of primary, acute and long-term services • Multi-disciplinary case management and day case centres feature predominantly Verticallyintegrated providers • A new form of organisation • Same group owns and manages services in primary, secondary, post-acute and sometimes long-term care sectors • Such groups contract with insurers (usually fee-for-service) to offer services along full continuum of care • Referrals generally made within the system • Overall clinical practice is managed centrally • Designed to promote delivery of care in the most cost-effective manner possible • Use range of integrated care techniques, especially utilisation review Source: summarised from Robinson and Steiner (1998). 87 APPENDIX C PURCHASER COMPETITION Purchaser competition – where DPOs compete with each other for members – is only rarely promoted without a significant regulatory framework around it to promote equity and efficiency. Most commonly, purchaser competition is debated in the form of ‘managed competition’. The managed competition literature is largely focused on reform proposals for the United States, although other countries have adopted aspects of the model. In the United States, the idea of encouraging competing pre-paid, capitated integrated networks of purchaser/providers was first put forward by Paul Ellwood in 1971 (Ellwood, Anderson et al. 1971). Among the first formal proposals were those suggested by Herman and Ann Somers in 1972 and 1973; their proposals included a top-down global budget and was entitled ‘regulated competition’. The more well-known managed competition proposals were developed by Paul Ellwood and Alain Enthoven and a group known as the ‘Jackson Hole Group’. Members of the group have been meeting since the mid-1970s and have developed a comprehensive set of policy proposals for managed competition reform (Ellwood, Enthoven et al. 1992; Ellwood and Enthoven 1995). Much of their work informed the Clinton plan (known formally as the President’s Health Security Plan), other United States Federal reform plans which were hotly debated in the early 1990s, and some United States health care reform at the State level. Managed competition ‘Management’ (or regulation) of competition occurs in various ways in these proposals, but usually includes: • the establishment of purchasing co-operatives. Such co-operatives are designed to: allow the pooling of risk by taking on the responsibility for managing the market for health plans for individuals and people employed by small firms; achieve economies of scale; cut the administration burden for small firms by offering an informed purchaser who can act on behalf of small firms; offer choice of plans to consumers; and manage competition amongst plans. Their roles include: selecting health plans for consumers to choose between; providing information to consumers about, and monitoring the performance of, health plans; interpreting benefits contracts; and resolving complaints (Enthoven 1993); • the encouragement of vertically integrated purchaser/providers (‘health plans’, i.e. DPOs), with providers employed by or contracted to health plans. In some versions, each provider (especially primary care physicians) is only employed by or contracted to one health plan, in order that provider practice be influenced by only one plan103; 103 Enthoven (1988) describes the following inefficiencies if an IPA model were used with each physician belonging to ten plans: ‘each doctor would have to deal with the utilisation controls and fee 88 • consumer choice of health plan, i.e. competition between health plans for consumers, which might also involve some form of consumer contribution towards premiums in order to encourage cost-conscious choice. In most proposals, consumers would have the choice of switching plans only once every year; • mechanisms to ensure that health plans take on all consumers who apply, to ensure that everyone is covered and that health plans don’t exclude consumers for e.g., pre-existing conditions (‘open enrolment’); • standard packages of benefits. Some versions would prohibit balance billing and supplementary insurance covering co-payments and additional services; • financing arrangements in which individuals pay premiums which are community rated, i.e. which are the same within age-gender bands, or where variations in premiums are limited; • pre-paid capitation payments to health plans in order to encourage efficiency in the use of resources; • in some managed competition models, an expenditure cap would also be placed on total health expenditure. There is however some controversy about how the cap would fit with managed competition proposals, which are designed in part to help determine an appropriate level of health expenditure. However, some commentators believe that without an expenditure cap, limits on expenditure growth will not be guaranteed by managed competition alone; and • strong relationships between plans and providers of health services. In some cases, purchasers and providers may be integrated into the same organisation; in other cases, contractual arrangements link purchasers and providers. These relationships are often assumed to include arrangements which share risk with providers (see Appendix B, Table B1.2 for examples of the organisational forms of integrated care organisations). (See for example Enthoven 1988; Enthoven 1988; Enthoven and Kronick 1989; Pauly, Danzon et al. 1991; Pauly, Danzon et al. 1992; Congressional Budget Office 1993; Enthoven 1993; Enthoven 1994; Ellwood and Enthoven 1995; Enthoven and Singer 1995; White 1995). These arrangements imply the development of health plans and the provision of information to consumers to help them choose between health plans. Usually, they are also structured to ensure that all people are covered: hence the requirements for open enrolment. schedules of ten health plans, none of which would command his [sic] loyalty. If one health plan persuaded a doctor to adopt a more efficient health practice, the benefits would be likely to be spread immediately over all ten plans, reducing the incentive of any plan to make the effort to pursue innovation at the provider level’. 89 Premium payments under managed competition In addition, proposals usually note that in a competitive market with profitmaximising insurers, consumers will be charged premiums based on individual and family risk. Many proposals break the link between financing and payment to health plans at this point: either by collecting premiums via taxes or social insurance schemes and allocating them to plans via a separate process; or by requiring community rating of premiums (i.e. not based on risk). These arrangements are designed to promote equity by improving affordability, but they have their own risks: chiefly the encouragement of cream-skimming. If capitation payments to health plans are not adjusted for the risk associated with individual consumers, then plans have an incentive to exclude higher risk consumers from joining up (an ex ante approach); and incentives to actively encourage consumers found to be high risk to disenrol (an ex post approach). The policy response to these problems varies in proposals for managed competition, in large part relating to the relative weight that analysts place on market versus regulatory responses. Benefits of managed competition The perceived (theoretical) benefits from managed competition are usually stated in contrast to the existing United States health care arrangements. Benefits for the United States are universal coverage, and enhanced incentives for promoting efficiency arising from the integrated nature of the health plans; the prospective payment system which would replace fee-for-service medicine; competition between plans; and the expenditure cap. In the United States context, the reforms may lead to less choice (or provider), more limited access to providers, fewer services and slower access to new technologies, greater involvement of plans in provider practice and lower provider incomes (Congressional Budget Office 1993). There are, however, some uncertainties about the extent to which competition between competing health plans will promote innovation, improved efficiency and improved services and choice for consumers. This will depend on the way in which the plans choose to compete and to contract with providers and the extent to which competition can develop for different population groups, including high risk groups and those in geographically isolated communities (Congressional Budget Office 1993; Plank n.d). Furthermore, a key driver of improvements in care will be the choices that consumers make in relation to plans. A lack of experience in choosing between plans and a lack of good information on plan performance may limit any potential benefits of competition. 90 APPENDIX D PURCHASER COMPETITION IN NEW ZEALAND Characteristics of the existing New Zealand health care system The implications of devolved purchasing and purchaser competition must be considered in relation to the context in which such changes takes place. New Zealand’s health care sector has some unique features which must be carefully considered in analysing the risks and benefits of such moves. These features include: • largely government financing of health and disability care, with an increasing proportion of expenditure financed privately. For the purposes of this paper it is assumed that health care in New Zealand will continue to be financed via a mildly progressive tax system and that government spending will remain at around 75 per cent of total health care spending. Thus, the issue of ensuring that all New Zealanders have access to health care becomes one of ensuring that all New Zealanders are registered with an DPO, that there is an DPO on which they can fall back or which is charged for unregistered consumers who need care104; • historically, a health care system which is government financed and organised with the aim of maximising the welfare of society. Consumer choice is largely limited to choice of primary care provider; whether to purchase private insurance; and whether to pay privately for care not covered by the publicly funded system (e.g., osteopathy, acupuncture, chiropractic). Where moves are made to devolve purchasing responsibility which requires registration with only one primary care provider, this will involve New Zealanders in a new way of thinking about health care. Similarly, in thinking about purchasing competition, New Zealanders are not used to an insurance-based model which offers choices in coverage or purchasing agent, while providers are only just becoming used to a contracting model; • a small, geographically dispersed population. This may offer limited opportunities for competition between DPOs. There are likely to be trade-offs between the need for large plans to manage risk where there is comprehensive coverage (i.e., including secondary and tertiary care, long-term care), and a desire for competition, a range of DPOs to suit culturally diverse populations, 104 It is possible for the system to move to a social insurance model, with premium payments rather than taxes. These can be set at a set percentage of income or can be adjusted for risk (as is done with ACC). It is also possible to remove the government from the financing role, with DPOs collecting premiums themselves. These options would change the nature of financing of health care. Key disadvantages would include issues of affordability the more premiums are determined based on risk, and rising transaction costs from each DPO having to collect its own premiums. See (Upton 1991) for a discussion on financing issues. 91 and networks in which individual practitioner decisions have more than just small consequences (Haas-Wlison and Gaynor 1998)105; • a large number of providers, particularly in primary and community health and disability care. These arrangements are often argued to result in duplication, gaps and a lack of continuity of care. The implications of this for New Zealand are a) there are likely to be high transaction costs in encouraging DPOs to contract for a comprehensive set of services (for example, without further integration in management or ownership, DPOs (just like the RHAs and HFA) would have to contract with numerous providers; considerable amalgamation of largely independent providers is required to move to comprehensive DPOs if complex and costly contracting arrangements are to be avoided; and b) continued separation of providers offers opportunities for cost-shifting; • a lack of horizontal integration of providers; an ability for some providers to charge patients unlimited fees in the form of user charges; and an availability of supplementary insurance coverage. Each ‘boundary’ offers excellent opportunities for cost-shifting; • a small number of secondary care providers, largely government-owned; there are very few tertiary providers (e.g., for organ transplants); and in some rural areas there may also only be few providers (e.g., GPs). This also raises issues about the likely extent of competition, and suggests there may be high transaction costs where competing DPOs contract for comprehensive care (e.g., the more specialised hospitals will have to have contracts with a number of DPOs); • significant user charges for some groups in the population for primary care. This raises issues relating to affordability of primary care, and may act to prevent improvements in cost-effectiveness because patients will resist attempts for more care to be delivered in primary care settings. This boundary also provides opportunities for cost-shifting; • good cost control over government-financed secondary and tertiary care; demand-driven primary care which leads to uncontrolled primary care expenditure106; • beyond licensing requirements, a limited government regulatory framework for health care. The lack of regulation in New Zealand health care may make it more difficult to introduce new regulations where they are seen as appropriate; 105 Luft notes that since market reforms have been introduced in California, health maintenance organisations have been consolidating, such that in 1996 four large plans accounted for 66% of total enrolment (Luft 1996). Haas Wilson and Gaynor note that it is predicted that within the next 5-10 years there will be only three to seven health care networks in California (Haas-Wlison and Gaynor 1998). 106 The HFA is currently consulting on plans to shift GPs from fee-for-service to capitated contracts. Pharmac is also presently using market mechanisms to reduce pharmaceutical expenditure. Both improve the government’s ability to contain its health care expenditure. 92 • poor information systems and a lack of integration between systems. Although secondary care providers are used to providing detailed information on outputs, primary care providers appear to be somewhat reluctant to share such information; • the inclusion of disability support services (DSS) in the New Zealand health care system. This may have implications for risk pooling and the spreading of risk; and • government ownership of key hospitals, with private ownership in primary care and disability support. This distinction may offer opportunities for cost-shifting, while a public distaste for further private involvement in health care is likely to make further moves to integrate care politically unpopular. Implications for purchaser competition The studies reported in the main body of the paper suggest that the effects of competition between health care providers is heavily dependent on the market structure – the relative position of insurers, physicians and hospitals -, and approaches to contracting. Given the right circumstances however – including a large number of competing hospitals, a good supply of physicians and selective contracting – competition can lead to reduced prices. There is however limited evidence on the effects on quality of care, while the United States evidence is very likely to reflect a massive oversupply of beds, built up in earlier times when costs were simply reimbursed. In relation to purchaser incentives, it appears that HMO competition has altered the approaches used by traditional insurers; however, the implications of competition between HMOs compared to no competition remain unclear, as do the implications of the combination of competing HMOs and a number of competing hospitals compared, for example, to a situation of bilateral monopoly. A key question for consideration in New Zealand is: how much competition between providers and purchasers might develop in New Zealand? Competition in New Zealand health care There are a number of characteristics that markets must fulfil to be called ‘competitive’: freedom of entry and exit, perfect information about prices and quality; and a large number of firms and consumers. Whether or not a market is ‘competitive’ contrasts to whether or not a market is ‘contestable’, that is whether or not there are significant barriers to entry that prevent new providers entering the market. Provider competition New Zealand is a small country with a geographically dispersed population. Thus, there is likely to be little competition for the most technologically advanced hospital 93 services; in fact, the Tertiary Services Review has suggested that positive and strong relationships between outputs and outcomes warrant fewer hospitals providing such care (Ministry of Health 1995). Thus, for liver transplants, for example, only one provider (Auckland Health Care) exists in New Zealand. The National Interim Provider Board considered in 1992 that competition in 24hour acute hospital care was limited, and may well stay that way given the need for ‘continuous provision of staffing, spare beds and equipment capacity to cope, at zero notice, with unpredictable needs’ (National Interim Provider Board 1992, p. 56). This situation remains so today. It would appear that the potential for competition for medical and surgical services is greater now than at the beginning of the 1990s. Medium-sized private hospitals offering a range medical and surgical services appear to be available in: Whangarei; Auckland (Epsom, Glenfield); Hamilton; Tauranga; Rotorua; Matamata; New Plymouth; Napier; Palmerston North; Wanganui; Wellington (Newtown; Lower Hutt, Crofton Downs); Masterton; Blenheim; Christchurch; Dunedin; Invercargill107. The level of service offered at each of these hospitals requires further research, however. In addition, the Calan Group is currently building a state-of-the-art hospital in Ellerslie, in Auckland. In their 1997 paper, Ashton and Press examined the degree of market concentration in selected secondary care services: tonsillectomies or adenoidectomies; prostatectomies; hip replacements; knee replacements; cataract removal; angioplasties; and coronary artery bypass grafts. They considered the geographic area from which each provider draws its patients and the extent of market concentration within those areas using the Hirschman-Herfindahl index. They found the seven markets to be reasonably concentrated, although none consisted of a monopoly provider, but noted that these results confirmed ‘the expectation that patient flows generally reflect residential patterns’ (Ashton and Press 1997, p.54). Ashton and Press concluded that potential efficiency gains from competition might be expected to be less under this concentrated market structure than under a more competitive structure. They noted however the ability of the then four RHAs to wield monopsony power, which might constrain providers to some extent. In a market with competition at the purchaser level as well however this ability would be weakened and we might expect more concentrated markets to have higher prices than less concentrated markets. 107 Private hospitals offering medical and surgical care, listed in Ministry of Health (1998). In addition, there are specialist clinics in Auckland for eye surgery, endoscopy; and special day surgery clinics at the North Shore, Whangarei, Tauranga, Rotorua, Hastings and Christchurch. Southern Cross owns hospitals in Auckland (3), Hamilton, Tauranga, Rotorua, Napier, New Plymouth, Wanganui, Palmerston North, Wellington, Christchurch and Invercargill (Southern Cross Healthcare 1994). The Ministry of Health report does not provide detail on the services provided beyond ‘surgical, medical’. It is not clear for example if some of these providers offer largely cosmetic surgery. 94 As noted by Ashton and Press, increased participation of the private sector in health care markets in New Zealand could reduce concentration. In making purchasing decisions, the RHAs were able to engage in competitive tendering for some secondary care services. However, their experiences with such competition were mixed. For example, prices sometimes rose as a result of competitive tendering, because CHE deficits continue to be subsidised by the government and CHEs sometimes sought to fully cost their services and add a further margin of around 5-10 per cent when tendering for additional contracts (Lovatt 1996). Private sector prices were often been uncompetitive (Wilson 1995; Ministry of Health undated). There were also concerns over potential discontinuities in service delivery (Foster 1994; Lovatt 1996; Ministry of Health undated), difficulties in specifying the services RHAs wish to buy with sufficient precision (Organisation for Economic Co-operation and Development 1996), and difficulties in defining the boundaries between contracts (Lovatt 1996). Secondary care markets might be argued to be likely to become more competitive only in the main centres. The entrance of the Calan group into the Auckland market certainly suggests that barriers to entry may not be as great as might have been expected. Yet the question remains as to the overall effect of such competition on prices in the sector. Duplication of expensive equipment and falling average numbers of patients treated with such equipment can lead to rising prices per patient treated, and in the extreme cases, to falling quality as the expertise and experience is reduced in each centre as numbers treated fall. An additional factor is the extent to which more cost-effective ways of treating patients can develop as a result of competition between, for example, hospitals, GPs and other providers. International trends here include the shift from inpatient to day-patient and out-patient care; the movement of elderly people into nursing homes; the movement of people with ongoing mental health needs into communities in many countries; using generalists rather than specialists to provide care for people with serious mental illnesses; and the increasing ability of professionals to undertake surgical operations in smaller clinics. There is limited evidence that each of these trends is cost-effective, with much work to be done to evaluate such changes fully108. Reinhardt for example argues that much of the perceived improvement in cost-effectiveness in the United States comes from not fully considering the differences between marginal and average costs and that ‘many transfers to alternative sties often would not make sense at all if the were evaluated on the basis of truly incremental costs’ (Reinhardt 1997, p. 38, emphasis in original). This suggests that hospitals will continue to bear the overhead costs, and will either try to recover them via higher prices, including across other services, or will incur ever increasing deficits until they are able to reduce the overhead costs themselves. This analysis is, however, only applicable in the short-run. In the longer term, as hospitals are able to close wards completely and quit buildings, it may well be more efficient to have care delivered in the community. This may take a number of years to achieve. 108 Some of these changes, for example the move in the United States to outpatient care rather than inpatient care, are the result of changes in payment methods and cost shifting to uncapped budgets. 95 Perhaps a key issue in New Zealand is the supply of specialists. Work is currently underway to reduce barriers to the supply of specialists in New Zealand; this work will be crucial to expanding choice and competition. However, it is not clear to me that we will ever be in a position of over-supply of specialists. In the presence of selective contracting – either from primary care providers or purchasers – competition between specialists may well improve quality of care. Whether or not it will reduce costs in the publicly funded health care sector may depend on changes in the salary levels of specialists. Currently specialists earn much higher rates of income in the privately-funded sector than they do in the publicly-funded sector: any moves to a situation where they must work in one or other sectors will probably lead to higher costs in the publicly-funded sector109. New Zealand’s primary and community care markets110 may be much more competitive than our secondary care markets111. For example, in most urban centres there is a choice of GP for diagnostic and primary treatment care. There are a number of pharmacists and laboratories in most larger centres. Women have a choice between midwives and GPs in many centres, although there is speculation that the choices are reducing in some parts of the country. There is increasing interest in up-skilling GPs to provide care for patients previously cared for in hospitals: the move to a certificate for GPs for mental health is an example. However, many rural towns have little or no choice when it comes to even GP care. GPs in such towns are increasingly being asked to up-skill in order to stabilise patients before patients are transported to the larger centres; but a number of rural towns arguably have tremendous difficulty in recruiting and retaining GPs. Moves to allowing nurses to prescribe some medicines, for example, may improve access for some people to care, but with limits to be placed on prescribing, service gaps in some towns are likely to remain. Telemedicine may improve access to some specialist care in the future, but the effects are still uncertain. 109 Unless the supply of specialists increases dramatically. 110 Care which is delivered in a community setting as opposed to a hospital. 111 In some areas, however, there are workforce training issues to be considered before we can be confident that good quality services are delivered in the community. Community mental health and Mâori providers are good examples. 102 APPENDIX G SERVICE SPECIFICATION IN UNITED STATES FEDERAL AND CALIFORNIA HMO LEGISLATION Federal HMO legislation Health Maintenance Organisations in the United States are governed by both federal and state legislation. Federal legislation requires HMOs to provide ‘without limitation as to time or cost other than those prescribed by or under this subchapter, basic and supplemental services to its members’ (Section 330e (b)). Basic health services means: • physician services (including consultant and referral services by a physician); • inpatient and outpatient hospital services; • medically necessary emergency hospital services; • short-term (not to exceed twenty visits), outpatient evaluative and crisis intervention mental health services; • medical treatment and referral services (including referral services to appropriate ancillary services) for the abuse of or addiction to alcohol and drugs; • diagnostic laboratory and diagnostic and therapeutic radiologic services; • home health services; and • preventive health services (including (i) immunizations, (ii) well-child care from birth, (iii) periodic health evaluations for adults, (iv) voluntary family planning services, (v) infertility services, and (vi) children’s eye and ear examinations conducted to determine the need for vision and hearing correction). ‘Such term does not include a health service which the Secretary, upon application of a health maintenance organization, determines is unusual and infrequently provided and not necessary for the protection of individual health. The Secretary shall publish in the Federal Register each determination made by him under the preceding sentence.’ (Section 300e-1.Definitions (1).) Supplemental health services means: ‘any health service which is not included as a basic health service’ (Section 300e-1. Definitions (2).) 103 California HMO legislation California’s Health and Safety Code, Chapter 2.2 deals with health care service plans. (Known as the Knox-Keene Health Care Service Plan Act of 1975.) A health care service plan is ‘any person who undertakes to arrange for the provision of health care services to subscribers or enrollees, or to pay for or reimburse any part of the cost for such services, in return for a prepaid or periodic charge paid by or on behalf of such subscribers or enrollees’. (Section 1345. Definitions (f).) Basic health services means: • physician services, including consultation and referral; • hospital inpatient services and ambulatory care services; • diagnostic laboratory and diagnostic and therapeutic radiologic services; • home health services; • preventive health services; and • emergency health care services, including ambulance services and out-of-area coverage. (Section 1345. Definitions (b).) Each plan must provide to subscribers and enrollees all of the basic health care services (although some plans may be exempt). In addition, a series of additional services are also specified. For example, Section 1367.2. relating to Alcoholism, chemical dependency or nicotine use states: ‘On or after January 1, 1990, every health care service plan that covers hospital, medical, or surgical expenses on a group basis shall offer coverage for the treatment of alcoholism under such terms and conditions as may be agreed upon between the group subscriber and the health care service plan’. Similar statements apply for: • comprehensive preventive care of children; • insurance coverage for those who are blind or partially blind; • diabetic daycare self-management education programs; • mastectomy coverage – including prosthetic devices or reconstructive surgery; • mammography (if the policy covers mastectomy and reconstructive surgery); 104 • prenatal diagnosis of genetic disorders of the fetus (if the policy covers maternity services); • insurance coverage for those who have physical or mental illnesses; • diethylstilbestrol; and • direct reimbursement of medical transport providers. Note that Section 1373. (h) (1) allows plans to exclude professional mental health services, but if they are covered coverage should include care offered by a psychiatric health facility and that reasonable efforts should be made to offer members the services of licensed psychologists, but failure to do so shall not constitute a misdemeanor (Section 1373. (h) (5)). Section 1374.4. Maternity benefits also notes that no plan providing maternity benefits can contain exclusions, reduction or other limitations to coverage, deductible or coinsurance for involuntary complications of pregnancy, unless such provisions apply generally to all benefits paid under the plan. Sections 1373.14, 1374.7, 1374.10 also have provisions for coverage including long-term care facility services or home-based care, to not exclude those with significant destruction of brain tissue (including Alzheimer’s disease) and those with genetic disability traits. (Note: other similar requirements apply to plans other than HMOs.) 105 APPENDIX H SERVICE SPECIFICATION IN UNITED STATES FEDERAL MEDIGAP LEGISLATION Medicare Supplement Contracts (California Health and Safety Code, S 1358.18, pp. 126-129) (All contracts relate to Medicare-eligible coverage Parts A and B, and are in addition to the core package not in lieu of the core package.) Plan A (‘core’ package) Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible Plan B Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible Plan C Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus 106 Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) Medicare Part B deductible Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) Plan D Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus Specified at-home recovery benefit (at-home short-term assistance with activities of daily living for those recovering from an illness, injury or surgery) Plan E Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus 107 Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus Specified preventive medical care (an annual clinical preventive medical history and physical examination and patient education; fecal occult blood test or digital rectal examination or both; mammogram; dipstick urinalysis for hematuria, baceriuria and proteinuria; pure tone, air only, hearing screening test, administered or ordered by a physician; serum cholesterol screening every five years; thyroid function test; diabetes screening; influenza vaccine administered at any appropriate time during the year and tetanus and diphtheria booster every to years; any other tests or preventive measures determined appropriate by the attending physician; actual charges up to 100% of Medicare-approved amount for each service up to a maximum of $120 annually) Plan F Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) Medicare Part B deductible Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus 100% of Medicare Part B Excess charges 108 Plan G Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) 80% of Medicare Part B Excess charges Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) Specified at-home recovery benefit (at-home short-term assistance with activities of daily living for those recovering from an illness, injury or surgery) Plan H Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) 80% of Medicare Part B Excess charges Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus 109 Basic outpatient prescription drug benefit (50% of outpatient prescription drug charges after a $25 calendar year deductible to a maximum of $1250 per enrollee per year) Plan I Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) 100% of Medicare Part B Excess charges Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus Basic outpatient prescription drug benefit (50% of outpatient prescription drug charges after a $25 calendar year deductible to a maximum of $1250 per enrollee per year) Specified at-home recovery benefit (at-home short-term assistance with activities of daily living for those recovering from an illness, injury or surgery) Plan J Hospitalization from 61st to 90th day (Medicare Part A) Medicare lifetime inpatient reserve day expenses (Medicare Part A) Upon exhaustion of the Medicare Part A inpatient coverage including lifetime reserve days, expenses up to a lifetime maximum of an additional 365 days Reasonable cost of the first three pints of blood (Medicare Parts A and B) Coinsurance (Medicare Part B), subject to Part B deductible plus Medicare part A inpatient deductible plus 110 Skilled nursing facility care (actual billed charges up to coinsurance amount from the 21st through to 100th day) 100% of Medicare Part B Excess charges Specified medically necessary emergency care in a foreign country (80% of Medicare-eligible expenses, within first 60 days of each trip outside the United States ($250), subject to a deductible and lifetime maximum benefit of $50,000) plus Extended prescription drug benefit (50% of outpatient prescription drug charges after a $250 deductible up to a maximum of $3000 per enrollee per year Specified preventive medical care (an annual clinical preventive medical history and physical examination and patient education; fecal occult blood test or digital rectal examination or both; mammogram; dipstick urinalysis for hematuria, baceriuria and proteinuria; pure tone, air only, hearing screening test, administered or ordered by a physician; serum cholesterol screening every five years; thyroid function test; diabetes screening; influenza vaccine administered at any appropriate time during the year and tetanus and diphtheria booster every to years; any other tests or preventive measures determined appropriate by the attending physician; actual charges up to 100% of Medicare-approved amount for each service up to a maximum of $120 annually) Specified at-home recovery benefit (at-home short-term assistance with activities of daily living for those recovering from an illness, injury or surgery). 111 APPENDIX I HEALTH EMPLOYER DATA INFORMATION SET (HEDIS) HEDIS® 1999 Reporting Set Measures by Domain Domain Description of changes EFFECTIVENESS OF CARE Childhood Immunization Status* Minor modifications Adolescent Immunization Status* Minor modifications Advising Smokers to Quit * Minor modifications Flu Shots for Older Adults* Minor modifications Breast Cancer Screening* No changes Cervical Cancer Screening* Minor modifications Prenatal Care in the First Trimester* Minor modifications Low Birth-Weight Babies No changes (not required) Check-Ups After Delivery* Minor modifications Beta Blocker Treatment After a Heart Attack* Minor modifications Cholesterol Management After Acute Cardiovascular Events New measure Eye Exams for People with Diabetes* Minor modifications Comprehensive Diabetes Care New measure (voluntary) Follow-Up After Hospitalization for Mental Illness* Minor modifications Antidepressant Medication Management New measure The Health of Seniors Specifications in HEDIS ’99, Vol. 6 ACCESS/AVAILABILITY OF CARE Adults' Access to Preventive/Ambulatory Health Services Minor modifications Children's Access to Primary Care Practitioners Minor modifications Availability of Primary Care Providers Measure retired Availability of Behavioral Health Care Providers Measure retired Availability of Obstetrical and Prenatal Care Providers Measure retired Initiation of Prenatal Care Minor modifications Low Birth-Weight Deliveries at Facilities for High-Risk Deliveries & Neonates No changes (not required) Annual Dental Visit No changes Availability of Dentists Measure retired Availability of Language Interpretation Services No changes