Long-Term Care in the USA: Lessons for New Zealand?

Mark Booth
Harkness Fellow in Healthcare Policy
Department of Community Health
Brown Medical School
Brown University

Vince Mor
Professor and Chair
Department of Community Health
Brown Medical School
Brown University

Abstract

The USA and New Zealand have different health and social care systems, but both share a desire for high-quality, affordable, long-term care for elders. In the USA moves have been made to integrate long-term care services within single capitated providers to provide clients with a seamless, easy-to-use service. Other initiatives seek to improve consumer direction by giving clients control of funding their own long-term care services. Such initiatives use different incentives to balance the need for high-quality care with affordability at both an individual and societal level. Initial analysis suggests that satisfaction and outcomes improve under these schemes. The integration models provide an interesting comparison for New Zealand as primary health care changes are further developed.

Introduction

This paper examines long-term care for the elderly in the USA, focusing on two programmes that appear to be producing positive results. Although the paper concentrates on policies in a single country, the USA, long-term care for the elderly is a topic that is challenging all Western societies. Recent policy developments have been seen in a number of countries, such as the United Kingdom, where a Royal Commission Report and an influential policy group report have led to the adoption of a very different approach to the funding of long-term care in Scotland and England (Wanless 2006, Royal Commission on Long Term Care 1999).

Although the health-care and social-care policy contexts for long-term care for the elderly¹ differ considerably between New Zealand and the USA, both countries face similar challenges in relation to the provision of long-term care. Both have seen a decline in the proportion of the age-adjusted population with significant activities of daily living (ADL) impairments, but at the same time anticipate substantial growth in the number of older people, meaning that there will still be a large absolute increase in the number of people requiring assistance. In both countries the workforce is declining, the number of frail elderly requiring significant levels of care is rising, and consumer expectations are increasing. Both countries are facing pressures on funding at both the individual (related to low levels of personal saving) and societal (related to inadequate reserves to cover large future liabilities without increasing tax rates) levels. There is a consensus in both countries that caring for older people within their own homes or community is preferable to doing so in residential care settings. This is coupled with a desire to contain costs while ensuring high standards of quality and safety in a sector dominated by private providers (Bryant et al. 2004, Cornwall and Davey 2004, Ministry of Health 2002, Miller and Mor 2006).

Both the USA and New Zealand have seen the number of people in nursing-home care decline in recent years as more emphasis is put on community-based care. Accurate information on the use of aged-care services in New Zealand is difficult to obtain, but internationally New Zealand spends a relatively small amount of its gross domestic product (GDP) on long-term care services, as illustrated in Figure 1. The graph, using data from 2003, shows that New Zealand spends a similar proportion of its GDP to the USA on long-term care, and that in spite of the decline in residential care both countries are still heavily reliant on residential care.

Figure 1 International Expenditure on Long-Term Care

Source: Huber 2004

The USA has investigated a number of approaches to funding and providing long-term care in recent years. This paper considers two initiatives that are judged to have had positive results, in terms of improving outcomes for the elderly, to see if any lessons can be learnt that are applicable to long-term care in New Zealand. The first initiative (managed long-term care) explores the integration of services to help highly vulnerable elders remain in their own homes; the second (cash and counselling) considers consumer direction of long-term care resources. The paper initially considers the context for long-term care in the USA by reviewing health-care funding and long-term care funding in that country and the links between the two. The two initiatives are then described, and this is followed by a discussion of the implications for New Zealand.

Health-care funding in the USA

The USA has traditionally adopted a policy path that emphasises minimal government intervention, with accompanying low levels of taxation and an emphasis on individualism. Central government tax receipts as a proportion of GDP are, at 33%, among the lowest in the OECD (this compares with an OECD average of 38% and 47% for New Zealand (OECD 2006)). In line with this underlying philosophy, the USA does not have a "universal" government-funded health-care delivery system or insurance scheme for its citizens, but relies on a mixed system with funding from both public and private sources.

The majority (55% in 2004, see Centers for Medicare and Medicaid Services 2006) of funding is from private sources – out-of-pocket expenditures and private health insurance. Premiums for health insurance can be very high, particularly for families. The majority of Americans who are covered by private health insurance are covered by employer-sponsored schemes as part of their employment packages. Clearly the amount of cover available depends on the comprehensiveness of the scheme, which is directly related to its cost. It is estimated that 46.1 million non-elderly US citizens (approximately 18% of the population) lacked any health insurance in 2005 (Kaiser Foundation 2006) and so would face extreme difficulty in accessing non-emergency health-care services. Recent research from the Kaiser Foundation has shown that the number of uninsured is rising as health insurance premiums increase and employers, particularly those employing low-paid workers, cease offering health insurance as a part of remuneration packages (Kaiser Foundation 2006).

The remaining 45% of funding comes through publicly funded sources at the federal and state level, through either the Medicare or Medicaid programmes. Medicare is a federal-level health insurance programme, which provides funding for the elderly (defined as age 65 and over) and those with a permanent disability. This is an entitlement scheme, such that basic coverage is provided as a right (unlike the Medicaid programme, as discussed below) and covers hospital inpatient services – called Part A coverage. Individuals are able to purchase more comprehensive Medicare coverage if they wish, which covers doctors' services and other ambulatory services (known as Part B coverage), as well as a wider range of plans with lower deductibles (Part C coverage). More recently, individuals have also been able to purchase coverage for pharmaceuticals prescribed outside the hospital (Part D). In addition, private insurance is available to top-up Medicare coverage, either by covering co-payments or covering areas not included in Medicare – these are known as "Medigap" policies. Such plans are provided by a large number of different companies each providing a range of different options to consumers. 

Medicaid provides health insurance for low-income pregnant women, children in low-income families up to the age of 19, parents with children under 19 who are on low incomes, low-income people with disabilities, and Medicaid-eligible (financially and medically) Medicare beneficiaries who require long-term care services not covered by Medicare. The Medicaid programme is not an entitlement system, and access depends on eligibility criteria. The federal government establishes general eligibility guidelines, but the actual requirements are developed at the state level and so eligibility for Medicaid will vary from state to state.

It should be noted that any US citizen or visitor who requires emergency health care will be able to obtain it through public hospitals, who treat everyone irrespective of insurance coverage or ability to pay. The treatment provided in such hospitals, particularly in the larger metropolitan areas, is frequently among the best in the world. The key issue for the uninsured is those individuals who suffer from chronic medical conditions or who need non-urgent elective services and are unlikely to receive any treatment. Research consistently points to the fact that the uninsured are less likely to use recommended preventive health services and to be hospitalised for potentially preventable conditions (Kaiser Foundation 2006).

The USA spends a huge amount on health-care services. The 2006 budget for health care was approximately $2 trillion.² Figure 2 shows the relationship between health expenditure per capita and GDP and the degree to which the USA is an outlier in comparison with other OECD countries. There is also increasing concern that, although the amount of expenditure is very high, patient outcomes are not as good as those achieved by other OECD countries (Commonwealth Fund 2006).

Figure 2 Relationship between Health Expenditure and GDP in OECD Countries, 2002

* PPP refers to Purchasing Power Parity and allows for the consistent comparison of different currencies in one base currency.
Source: Ministry of Health 2005

In considering the provision of health services, account must also be taken of the US federal system. Strategic direction for health services is provided at both a federal and a state level. In practical terms this can mean a wide variation in regulations and policies relating to health care. At the federal level there is an absence of a strategic overview or vision for health-care services across the country, and federal–state relationships can often be strained.

Long-Term Care Funding in the US

Long-term care in the US can be funded from one of four sources:

• out-of-pocket payments
• private long-term care insurance
• Medicare
• Medicaid.

Out-of-pocket expenditure accounts for 28% of long-term care costs. The majority of this will be used during a period of "spend down", described below.

Private long-term care insurance has met with very limited success in the US (as in other countries). There are a variety of reasons for this, but the most prominent include (Miller and Mor 2006) high premiums (particularly for older people), the complexity of schemes, a common belief among individuals that the government will pay for their long-term care, and individuals' belief that they will not require long-term care

Medicare funds post-acute care in nursing homes or rehabilitation facilities for a limited time for those individuals who have been discharged from acute-care facilities and require rehabilitation or nursing-care services. Medicare will pay for the first 21 days' stay in such a facility. If an individual goes over 21 days, then they must pay a daily co-pay contribution (in 2006) of $119 for days 21 to 100. Any nursing home stay extending beyond 100 days must be paid by the patient; or, if they have become eligible for Medicaid, it will pay the nursing home. There are an estimated seven million individuals who are "dually eligible"; that is, Medicare beneficiaries who are also eligible for Medicaid services.  

Of the four sources identified above, the Medicaid programme is the single largest contributor to the funding of long-term care. As we have seen, Medicaid was established to provide financial assistance to low-income children and their parents, people with severe disabilities and the elderly. The scheme helps finance care for 55 million people at an annual cost of almost $300 billion. Children make up just over 50% of all enrollees, and expenditure on children is approximately 15%. The elderly and disabled make up 27% of enrollees but account for 66% of expenditure (Centers for Medicaid and Medicare Services 2006). Medicaid is known as a federal/state scheme because expenditure is shared, with federal spending accounting for 57% of Medicaid spending and the states contributing the remaining 43% (Kaiser Foundation 2005). If state expenditure increases, then federal expenditure increases proportionately. The budget for Medicaid has been increasing considerably, bringing federal and state attention to ways to reduce the increase. As Rowland (2005:1439) notes:

Medicaid is the nation's health safety net, but its growing role and increasing cost in the face of state budgetary pressures and the federal deficit have made it a target for reform that could fundamentally reshape the program.

Clearly, in seeking to address increased costs there are two main courses of action: reduce the total amount of funding or make people contribute more. In the case of Medicaid, the latter option is difficult because, by its very nature, the programme looks after those on low incomes. The former option is the one that is being looked at, and in April 2005 Congress passed a proposal to reduce Medicaid funding by $10 billion over five years while giving states more flexibility in using the funding.

If an individual requires long-term care then they would fund needed services "out of pocket" until their assets (excluding their primary residence) and income are below a pre-determined level, which may vary by state (this process is known as spend-down). Following this, the Medicaid programme would fund the ongoing care costs. Currently (2006) 44% of nursing-home care costs are funded through the Medicaid programme, with 28% coming from out-of-pocket payments. The Medicaid component is expected to rise to almost 50% of all nursing-care costs by 2015 (Centers for Medicaid and Medicare Services 2006). It should be noted that approximately two-thirds of nursing-home residents are supported by Medicaid, although most have some residual source of income, which must be used to pay for their care, and the rest is paid by Medicaid.

Managed Long-Term Care

The quest to improve the fit between health and social care for frail elderly people and also to boost service efficiency, effectiveness and quality has focused attention on whole-system approaches to improve the way that sectors, institutions, providers and services work in tandem as part of the long-term care enterprise. (Kodner 2006:384)

Managed care emerged in the 1980s as a movement away from the fee-for-service basis of the Medicare programme. Medicare had seen significant increases in costs for its enrollees, and managed care was seen as a way to encourage cost-effectiveness and quality in the provision of health care. An all-encompassing definition of managed care is difficult, but it envisages a system whereby enrollees would sign up to plans offered by competing providers (known as health management organisations, or HMOs), which would offer defined services (Glied 1999). Typically, such health plans would be associated with capitation payments for health-care providers and, with detailed care plans for providers to follow, an emphasis on prevention and a large role for primary care.

The system of managed care increased in popularity through the 1990s but in more recent years has become less popular. Reasons for the decline in use of managed care include a backlash from health-care professionals who believed it limited their freedom to make the best choices on behalf of their patients, and also from patients who saw the schemes as too prescriptive. The system required high costs of entry for HMOs, which restricted entry into the market and thus reduced competition and associated efficiencies (Nichols et al. 2004). Having said that, Robinson (2001:2622), in an article entitled "The end of managed care", felt that "after a turbulent decade of trial and error, that experiment can be characterized as an economic success and a political failure".

Despite these problems, managed-care programmes continue, and indeed managed care for women and children covered by Medicaid has risen consistently in recent years, with more individuals being given the choice to enter managed-care plans. For long-term care, however, the growth has been very limited, and only 2.3% of current long-term care recipients are in managed-care plans (Saucier et al. 2005). In analysing long-term managed care, however, Saucier et al. noted that while not popular, those states that have initiated long-term managed-care programmes have found them to be effective, and all are planning expansions.

A specific example of a long-term managed-care programme is the Program of All-Inclusive Care for the Elderly (PACE), a community-based programme for frail, chronically ill older people whose functional and cognitive impairments make them eligible for nursing-home care. PACE was developed in San Francisco in the early 1970s as a non-profit programme providing day care to elderly immigrants for whom nursing care was not culturally appropriate. The On Lok Senior Health Services Program in San Francisco was the first such model, formed in 1973. PACE provides a comprehensive and seamless package of services to elderly people that enables most participants to live in their own homes. A key component of PACE is the use of day-care facilities that participants in the programme attend.

Service packages are provided by PACE providers, who are funded by capitation and operate on a not-for-profit basis. Providers assess the care needs of participants, develop care plans to meet those needs, and deliver the required services. The services included in the programme are:

• adult day care that offers nursing; physical, occupational and recreational therapies; meals; nutritional counselling; social work and personal care
• medical care provided by a PACE physician familiar with the history, needs and preferences of each participant
• home health care and personal care
• all necessary prescription drugs
• social services
• medical specialist care such as audiology, dentistry, optometry, podiatry and speech therapy
• respite care
• hospital and nursing-home care, when necessary.³

The capitated financing allows providers to deliver all services that participants need, rather than being limited to those reimbursable under the Medicare and Medicaid fee-for-service systems. PACE providers must ensure that they:

• focus on frail elderly qualifying individuals who require the level of care provided in a nursing facility
• deliver comprehensive, integrated acute and long-term care services
• provide an interdisciplinary team approach to care management and service delivery
• operate under a system of capitated, integrated financing that allows the provider to pool payments received from public and private programmes and individuals
• assume full financial risk.

The final point above reflects a life-time commitment to the enrollee and provides PACE providers with a powerful incentive to aggressively pursue preventive health services, ensure frequent clinical monitoring of clients, and monitor resource allocation across the organisation.

In order to enrol in a PACE programme, participants must be at least 55 years old, live in the PACE service area, and be certified as eligible for nursing-home care by the appropriate state agency. The PACE programme then becomes the sole source of services for Medicare and Medicaid eligible enrollees. There are currently 35 PACE providers operating in 19 different states (National PACE Association 2006).

Evaluations of PACE providers have been positive. Statistically significant improvements have been found in the following areas: fewer hospital visits, fewer nursing home stays, higher levels of satisfaction with care, better health, better quality of life, and lower mortality rates than comparative non-enrolled populations (Chatterji et al. 1998, Kodner 2006, Grabowski 2006). In terms of the costs associated with PACE programmes the evidence is mixed, and it is difficult to determine whether the associated costs are less than they would have been under traditional Medicare or Medicaid schemes (Grabowski 2006).

Cash and Counselling

Cash and counselling is an approach to service provision whereby the locus of control for services is not with the provider of services but with the recipient of services. Under such a model, the recipient of care – together with family caregivers, if appropriate – determines the package of care most appropriate to his or her own needs and purchases that care. Such a model can be seen as having two key advantages.

• From the client's perspective, it should result in more control being retained by the client, and provision of services that are much more focused on an individual's needs. This should result in greater client satisfaction and, possibly, improved outcomes for the client (Foster et al. 2003).
• From an economic perspective, theory would suggest that giving people money to pay for their own care will encourage them to seek out less expensive care options than would have been the case if an agency was providing services on their behalf (Davis 2004). Costs should also be saved by eliminating administration staff, such as case managers.

This model can be contrasted with the traditional model, where professionals provide services to clients based on the professionals' opinion of what the client requires. Under this model, the autonomy of the client is reduced and they have little, if any, input into the type of care they receive. This model of care is also associated with the development of a bureaucracy which often specifies quite rigidly, and to very detailed levels, the services that an individual should receive (Doty 2004).

The cash and counselling scheme started as a Medicaid waiver scheme, initially operating in three states (Arkansas, New Jersey and Florida) from the late 1990s. Under the programme, service users and their families are given individualised monthly budgets, which can be used to purchase a wide range of services. Family members are important, as they may need to act as representatives for elderly relatives who have cognitive impairments such as Alzheimer's disease. The services purchased can include home modifications, technology, and the hire of aides – including family members.

As part of the scheme, the individuals receive advice on how the money could be spent (it should be noted that people are not actually given cash – it is a notional amount that the clients can use). This advice is the "counselling" component of the scheme. The services or items purchased with the cash component must be related to the care of the individual and are not to be used to supplement basic living expenses or to purchase "luxuries". For the three demonstration sites there were fairly stringent assessments to ensure that the clients or their representatives were able to manage the cash effectively. Two of the states required consumers to pass a fiscal skills examination, while the other (Arkansas) individually assessed each consumer for fiscal skills. Unlike the PACE programme, the population here is less dependent and is not at risk of institutionalisation.

Since the demonstration sites were established, a second wave of states has established similar types of programmes. States are attracted to cash and counselling for budgetary reasons – as a possible way of postponing or reducing admission to nursing homes – and also to try to increase satisfaction with services.

In summarising the aims of the original three demonstrations, Dale and Brown (2005 p.xiii) note that:

The premise of the cash and counselling demonstration was that, if consumers were given control over a cash allowance, they would select the types and amounts of care and services to best meet their needs and enhance their lives. When designed the program was expected to cost no more per recipient per month of service than the traditional program.

Initial evaluations have shown that personal care costs have tended to increase because recipients of care used the cash to purchase more care than non-programme members. It was also observed that recipients employed carers who could provide more effective care (often family members). Grabowski (2006) notes that this is a variation of the "woodwork" effect, whereby a more appealing set of home-care benefits increases the propensity to use benefits. This increased amount of personal care has the potential to be offset by reduced nursing-care costs, and in one state, Arkansas, nursing-care costs reduced by 18%. It has also been found that quality of life has increased for clients, and there has been a reported decrease in carer stress (Dale and Brown 2005). In their summary of an evaluation of the Arkansas cash and counselling pilot, Foster et al. (2003:162) noted that:

relative to agency directed services, Cash and Counselling greatly improved satisfaction, and reduced most unmet needs. Moreover, contrary to some concerns, it did not adversely affect participants' health and safety.

Discussion

Both of the initiatives described in this paper aim to ensure that individuals remain in their own homes and that services are cost-effective. In order to do this, they seek to harness both economic and non-economic incentives to improve performance and outcomes. For the economic incentives, the key difference between the two programmes is the level at which the incentives are harnessed. For PACE, the capitation payments incentivise providers to pursue preventive programmes that will keep elderly enrollees out of expensive secondary and tertiary institutions. For cash and counselling, the incentives are with the individual to maximise the most appropriate care for themselves at a minimum cost. Both programmes also have a series of non-economic incentives, reflecting a desire to provide client-focused services in a community/home-based setting. The incentives will include increased professional satisfaction for staff, improved support for carers and increased control over services for clients.

In both cases the programmes would appear to have achieved the goal of keeping clients out of residential care and improving outcomes in a variety of areas, such as reduced morbidity and mortality, improved client-perceived quality of care and reductions in carer stress (Chatterji et al. 1998, Kodner 2006, Dale and Brown 2005, Foster et al. 2003) . It is not clear whether the schemes have reduced costs; nor is it clear whether they will be cost-effective at a macro or system level. This is because the majority of elderly individuals in need of some long-term care will not enter residential care, although they might be more likely to use community-based care if it becomes available, resulting in an expansion of the number of individuals who would use these services but would not enter a nursing home. Any savings associated with reduced costs of home care are likely to be outweighed by the higher numbers of people using home- and community-based services (Grabowski 2006) – the "woodwork" effect. In addition, for services such as PACE, which provides intensive services for very needy clients, the costs of care may actually be higher than nursing-home care.

It should also be noted that the economic incentives may not necessarily produce the intended outcome. In cash and counselling, for the incentives to work effectively the recipient of care must be in a good position to make rational choices. The position of the advisor or counsellor then becomes critical, and the incentives that person has may not always align with those of the client. In addition, there may be incentives for individuals to divert resources to family members who would be paid above market rates, and individuals may spend all of their cash, irrespective of marginal benefit. Similarly, for PACE there may be an incentive to spend up to the limit of the capitation payment irrespective of whether marginal payments are effective, although none of the evaluations have found that this occurs.

For New Zealand, where long-term care is funded through both government and out-of-pocket expenditures and the main policy direction is to – wherever possible – care for people in their own homes, the key issue is whether the introduction of schemes such as managed long-term care and cash and counselling would be beneficial. New Zealand has a history of innovative approaches to funding in health care, and adopted market-based reforms in the 1980s and 1990s. More recent policy directions have, however, been in the direction of providing certainty of care and treatment through the use of universal funding. Recent changes in asset-testing legislation have emphasised this move. The Primary Health Care Strategy has used capitation-based funding for primary care organisations to move them to a more preventive focus.

Such primary care organisations also bring together the diverse range of professionals that sit within primary care to provide a more seamless service to clients. Initiatives such as the Promoting Independence Programme (PIP), operating in the lower North Island, use a key worker system to coordinate rehabilitation programmes for the elderly. Other initiatives such as the Coordination of Services for Elderly (COSE) in Canterbury also seek to operate a seamless service for their elderly clients. The COSE service operates a care coordination model whereby case managers are appointed to individuals to coordinate the services required by the elderly person.

The primary care system in New Zealand is not a managed care system, but it is anticipated that the move to capitation will lead to a greater emphasis on prevention as the incentives in place under a fee-for-service system are removed. In many ways this is similar to the PACE programme, albeit that PACE is concentrated on a single population group. It would be useful to further investigate the links and the potential for service developments between primary care and long-term care in the community. Such a model may be a logical development for New Zealand's primary care organisations. Just as schemes such as the COSE service described above are a natural extension to the work of the current Needs Assessment Service Coordination, the provision of services such as those offered in the PACE programme would be a further extension.

Cash and counselling schemes may be more difficult to adopt in New Zealand given the District Health Board structure, which places the funding of services at a population level. Nevertheless, the increasing culture change in long-term care that is encouraging, among other things, greater client participation in their own care, increased consumer direction of services and services that reflect home-based settings rather than institutional ones, may gradually lead to funding following clients to a lower and lower level. This is again an area in which pilot projects that look at diverting resources to individuals may be useful.

References

Bryant, J., A. Teasdale, M. Tobias, J. Cheung and M. McHugh (2004) Population Ageing and Government Health Expenditures in New Zealand, 1951–2051. New Zealand Treasury Working Paper 04/14, Treasury, Wellington.

Centers for Medicare and Medicaid Services (2006) National Health Expenditure Data, [accessed 30/10/06].

Chatterji, P., N. Burstein, D. Kidder and A. White (1998) Evaluation of the Program for All-inclusive Care for the Elderly (PACE) Demonstration: The Impact of PACE Participant Outcomes, ABT Associates, Cambridge, Massachusetts.

Commonwealth Fund (2006) Why Not the Best? Results from a National Scorecard on Health System Performance,Commonwealth Fund, New York.

Cornwall, J. and J. Davey (2004) Impact of Population Ageing in New Zealand on the Demand for Health and Disability Support Services, and Workforce Implications, Ministry of Health, Wellington.

Dale, S. and R. Brown (2005) The Effect of Cash and Counselling on Medicaid and Medicare Costs: Findings for Adults in ThreeStates, Final Report, Mathematica Policy Research, Princeton.

Dale, S. and R. Brown (2006) "Reducing nursing home use through consumer-directed personal care services" Medical Care, 44(8):760–767.

Davis, K. (2004) "Consumer driven health care: Will it improve health system performance?" Health Services Research 39(4, part 2, p.1219–1234).

Doty, P. (2004) Consumer Directed Care: Effects on Family Caregivers, Family Caregiver Alliance Policy Brief, Washington, DC.

Foster, I., R. Brown, B. Phillips, J. Schore and B. Carlson. (2003) "Improving the quality of Medicaid personal assistance through consumer direction" Health Affairs, [accessed 24/10/06].

Glied, S. (1999) Managed Care, National Bureau of Economic Research Working Paper 7205, Cambridge, Massachusetts.

Grabowski, D. (2006) "The cost-effectiveness of non-institutional long-term care services: Review and synthesis of the most recent evidence" Medical Care Research and Review 63(13–28).

Huber, M. (2004) "Estimating expenditure on long-term care" presentation to the 6th Meeting of Health Accounts Experts and Correspondents for Health Expenditure Data. 30 September – 1 October, Paris, France.

Kaiser Foundation (2005) Long-Term Care: Understanding Medicaid's Role for the Elderly and Disabled, Kaiser Foundation Commission on Medicaid and the Uninsured, Menlo Park, California.

Kaiser Foundation (2006) Health Coverage for Low Income Americans: An Evidence Based Approach to Public Policy, Kaiser Foundation Commission on Medicaid and the Uninsured, Menlo Park, California.

Kodner, D. (2006) "Whole system approaches to health and social care partnerships for the frail elderly: An exploration of North American models and lessons" Health and Social Care in the Community, 14(5):384–390.

Miller, E. and V. Mor (2006) Out of the Shadows: Envisioning a Brighter Future for Long-Term Care in America, Brown University, Providence.

Miller, E. and W. Weissert (2003) "Strategies for integrating Medicare and Medicaid: Design features and incentives" Medical Care Research and Review, 60(2):123–157.

Ministry of Health (2002) Health of Older People Strategy, Ministry of Health, Wellington.

Ministry of Health (2005) Health Expenditure Trends in New Zealand 1992–2003,Ministry of Health, Wellington.
National Pace Association (2006) Who, What and Where is PACE, [accessed 27/11/06].

Nichols, L., P. Ginsburg, A. Berenson, J. Christianson and R. Hurley (2004) "Are market forces strong enough to deliver efficient health care systems? Confidence is waning" Health Affairs, 23(2): 8–21.

OECD (2006) Annex Table 33: General Government Net Financial Liabilities, [accessed 27/11/06].

Robinson, J. (2001) "The end of managed care" Journal of the American Medical Association, 285(20): 2622–2628).

Rosenbaum S. (2002) "Medicaid" New England Journal of Medicine 346(8):635–640.

Rowland, D. (2005) "Medicaid: Implications for the health safety net" New England Journal of Medicine,353(14):1439–1441.

Royal Commission on Long Term Care (1999) With Respect to Old Age: Long Term Care – Rights and Responsibilities, The Stationery Office, London.

Saucier, P., B. Burwell and K. Gerst (2005) The Past, Present and Future of Managed Long Term Care, University of Southern Maine, Maine.

Wanless, D. (2006) Securing Good Care for Older People: Taking a Long-term View, King's Fund, London.

¹ This paper looks at long-term care for the elderly and does not consider long-term care for the younger disabled or other groups. For ease of use, the term "long-term care" here refers to long-term care for the elderly.

² All figures in this paper are in US dollars.

³ PACE Medical Directors Handbook 2006.