Milq 125

Five Laws for Integrating Medical and Social
Services: Lessons from the United States and

the United Kingdom
WA LT E R N . L E U T Z
Brandeis University, Waltham, Massachusetts
T he integration of acute and long-term care

for persons with disabilities and chronic illnesses has received
support in both the United States and the United Kingdom as a means
to efficiency, user satisfaction, and better outcomes (Davies 1995; Social
Services Inspectorate 1995b; Vladeck 1995). Integration could address
cross-system care problems, including poor coordination of services and
benefits, cost shifting, and frustration for users in accessing services.
The term “integration” has taken on a wide range of meanings, how-
ever, as it can signify anything from the closer coordination of clinical
care for individuals to the formation of managed care organizations
(MCOs) that either own or contract for a wide range of medical and
social support services. Part of the debate, indeed, concerns the degree
to which financial and organizational integration is needed to achieve
clinical integration (Leutz, Greenlick, and Capitman 1994; Wiener and
Skaggs 1995). Nor is there agreement about who should be in charge of
integration, whom it should target, what supportive structures are needed,
or what to call it (Harrington, Lynch, Newcomer, et al. 1993; Leutz,
Greenlick, Ervin, et al. 1995). One purpose of this paper is to draw such
distinctions more clearly.
Integration is defined, for the purposes of this paper, as the search to
connect the health care system (acute, primary medical, and skilled)
The Milbank Quarterly, Vol. 77, No. 1, 1999

© 1999 Milbank Memorial Fund. Published by Blackwell Publishers,
350 Main Street, Malden, MA 02148, USA, and 108 Cowley Road,
Oxford OX4 1JF, UK.
77
78 Walter N. Leutz
with other human service systems (e.g., long-term care, education, and
vocational and housing services) in order to improve outcomes (clinical,
satisfaction, and efficiency). Populations that may benefit from integra-
tion have physical, developmental, or cognitive disabilities—often with
related chronic illnesses or conditions. Integration can occur at the pol-
icy, finance, management, and clinical levels. The means of integration
include joint planning, training, decision making, instrumentation, in-
formation systems, purchasing, screening and referral, care planning,
benefit coverage, service delivery, monitoring, and feedback.
I will review integration efforts in both countries with these goals in
mind:
1. demonstrating which approaches have worked and which have not

2. synthesizing important lessons for policy makers, professionals,
and providers
3. recommending directions for the future
Although there is clearly much left to learn, innovators should build on

the hard base of experience, rather than relearning what works and
repeating past mistakes. To highlight the lessons, they are proposed as
“laws of integration.” First, however, the structures of the US and UK
health care systems are briefly reviewed, particularly the efforts in both
countries to integrate acute and long-term care.
Health Care Systems and Integration in the

United States and the United Kingdom
The two countries have very different systems of acute and long-term
care. A half century ago, the United Kingdom created the National
Health Service (NHS), a nationally funded, universal coverage system
for acute, primary medical, and skilled care that is free to users at the
point of delivery. The NHS is administered by the Department of Health,
which works through regional and district health authorities. Hospital
care and community nursing are delivered, respectively, through pub-
licly owned hospital and community “trusts.” General practice (GP)
physicians contract with district health authorities under terms that are
set nationally. Its central authority for operating budgets, capital, and
wages and salaries has allowed the United Kingdom more effective
Integrating Medical and Social Services 79
control over total health care spending than is possible in the United
States. A small, but growing, private insurance and provider sector
specializes in services, like elective surgery, that may be difficult to
obtain through the NHS.
The United States has multiple approaches to financing and deliver-
ing medical care. Most of the working-aged population and their fam-
ilies are insured through employers, who either contract with insurers
and MCOs or purchase care directly from provider networks. Disabled
workers and the elderly are covered by the national Medicare program,
which purchases care from individual providers on a fee-for-service (FFS)
plan and also contracts with a growing variety of MCOs. The very poor
are covered through the federal–state Medicaid program, which in most
states is moving rapidly toward an exclusively managed care contracting
and delivery mode. More than 40 million persons (about one-sixth of
the population) have no health insurance and must fall back on their
own resources, local indigent care funds, or charity care. The majority of
service providers are privately rather than publicly owned, and all types
of providers (hospitals, nursing homes, home health agencies, and even
physician practices) are increasingly being converted from nonprofit–
individual ownership to for-profit–corporate chain ownership.
The two countries’ long-term-care and other human service systems
do not differ as markedly as their health care systems. The states and
localities (counties and cities) that manage and finance such nonmedical
care in the United States find their counterparts in British local author-
ities. Both countries’ state and local systems receive partial national
funding under flexible guidelines. Means testing for long-term care is
also a hallmark of both countries, as is the private provision of many
services (both private provision and means testing have increased during
the last decade in the United Kingdom).
Although their health care systems differ, both countries face some of
the same challenges: increasing numbers of persons with disabilities; the
closure of public hospitals for persons with physical, mental, and de-
velopmental disabilities; increasing costs for acute and long-term care;
shorter acute-care stays that shift care to the community; and fiscal
pressures at national and local levels. Both countries also have had to
struggle with parallel changes in social welfare paradigms, incorporat-
ing the sometimes conflicting ideologies of privatization, devolution,
care management, markets in care, and user–carer rights. (The British
term “carer” is used here, rather than the US term “caregiver,” to refer
80 Walter N. Leutz
to family and friends providing care.) The next two sections review how
the two countries have experimented with new paradigms to integrate
care for populations with special needs.
UK Systems and Initiatives
The UK health and social services systems traditionally have been char-
acterized by organizational separation, like that, for example, between
community and hospital physicians and between NHS medical services
and local authority social services. There also are long-standing efforts,
however, to bridge these divisions through joint consultation commit-
tees and community health councils, as well as through cooperative
operational and purchasing mechanisms, like joint commissioning (Light
1994; Small 1989; Schulz and Harrison 1984; Klein 1979). Such efforts
unite separate bodies (e.g., district health authorities, hospital trusts,
community trusts, and local authorities) to better meet the needs of
population subgroups that use multiple services.
The last decade saw major health and social services reforms in the
United Kingdom. Beginning in the late 1980s and early 1990s, the
Conservative government introduced internal markets in the NHS, pri-
vatization of long-term-care service delivery, and stronger means testing
for long-term-care services offered by local authorities. Professional lead-
ership was deemphasized in favor of managerialism in both sectors.
Although it is unlikely that full-blown MCOs will be established in the
United Kingdom, the ideas of demand management, medical manage-
ment, and care delivery are popular (Smith 1997). Several major reforms
were initiated with a bearing on integration:
• The Community Care Act (1990): strengthened care management

in long-term care and directed health authorities and local author-
ities to cooperate in planning and purchasing services for certain
population subgroups, like persons with learning disabilities (Greig
1997; Social Services Inspectorate 1995b; Waddington 1995).
• GP fundholding (1990): allowed private GP practices to finance
and contract for more than their own services. Although it does not
allow GPs to purchase long-term care from local authorities, they
can purchase home health and nursing-home care (National Health
Service Management Executive 1993; Weiner and Ferriss 1990).
• Total Purchasing Pilots (TPPs) (1995): expanded the range of health

services commissioned by GPs through 53 pilot projects. The ex-
panded services are still paid for by district health authority bud-
gets, but the program allows GPs to participate in planning for
long-term-care populations (Mays 1997b; Myles, Popay, Wyke,
et al. 1997).
In December, 1997, the Blair government released a blueprint for

NHS reform that will abolish fundholding and the internal market and
create 500 primary groups; each will have about 50 GPs and will serve
100,000 patients (Department of Health 1997). The groups will be
responsible for more planning (e.g., for regional health improvement
plans) and, by virtue of these increased responsibilities, could them-
selves eventually become free-standing primary care trusts. The reforms
may reflect the fact that the market-based reforms envisioned in the
creation of GP fundholding have not found fertile soil in the tradition-
bound NHS (Klein 1995; Light 1997a). I will discuss below the ways
in which planning and purchasing through new, larger groups of GPs
will present its own set of opportunities and challenges.
US Systems and Initiatives

In contrast to the United Kingdom’s public reorganization of largely
public systems, US policy makers must achieve integration largely
through private managed care models. Despite great variation in the
details, three generic models are being developed: those that integrate
acute care; those that integrate long-term care; and those that try to
integrate both.
Acute-care initiatives include:
• Medicare HMOs: These organizations receive a capitation for ben-

eficiaries who enroll voluntarily. About 10 percent of Medicare
beneficiaries have joined to date (Welch 1996). The 1997 Bal-
anced Budget Act contains provisions to expand sponsorship op-
tions and enrollment.
• Medicaid HMOs: Until recently, most states exempted the aged,
blind, and disabled from mandates that other beneficiaries face to
join HMOs. This is changing in many states, including some (Min-
nesota and Tennessee) that are stimulating providers by requiring
82 Walter N. Leutz
both state workers and Medicaid beneficiaries to join contracted

plans (U.S. General Accounting Office 1996; Iglehart 1992).
Initiatives designed to consolidate long-term care include:
• Gatekeeping systems: A single agency—sometimes public and

sometimes private—in each region manages entry to all important
sources of long-term-care funding. At least two states (Oregon and
Massachusetts) permit management agencies to assume some fi-
nancial risk (Kane, Kane, Haye, et al. 1996; Leutz, Sadowsky, and
Pendleton 1992).
• Managed long-term care: Arizona Medicaid contracts with one
organization in each county (usually the county itself ) to manage
long-term care on a capitated basis (McCall and Korb 1994). Flor-
ida has a two-site demonstration, in which HMOs receive a capi-
tation covering all long-term care for beneficiaries who enroll
voluntarily.
The third category, initiatives that integrate the financing and de-
livery of both acute and long-term care on a capitation basis, covers both
full-population and “carve-out” models. The former seek to enroll both
frail and nonfrail persons. The latter enroll only persons with disabili-
ties. Examples of full-population models include the following:
• Social HMOs: These demonstration MCOs were launched in 1985

and enroll a cross-section of Medicare beneficiaries over age 65. All
enrollees receive Medicare benefits, while the frail subgroup is
eligible to receive a finite long-term-care supplement that is con-
trolled by a care management unit (Leutz, Greenlick, and Capit-
man 1994; Leutz, Greenberg, Abrahams, et al. 1985).
• Minnesota Senior Health Options (MSHO): a Medicaid demon-
stration project that allows dually eligible beneficiaries to choose
among competing MCOs for all covered acute- and long-term care
(Minnesota Department of Human Services 1997).
Examples of carve-out models include the following demonstration

MCOs:
• On Lok/Program for All-inclusive Care for the Elderly (PACE)

enrolls community residents over age 65 who meet nursing-home
entry requirements and who are eligible for both Medicare and
Medicaid (dual eligibles). All the enrollees’ acute and long-term
care is managed by a multidisciplinary day center team (Ansak and
Zawadski 1984; Branch, Coulam, and Zimmerman 1995; Kane,
Illston, and Miller 1992).
• The Wisconsin Partnership enrolls either elders or working-aged
adults with physical disabilities. All the enrollees’ physicians’ ser-
vices, hospital care, and long-term care are coordinated by the
MCO (Hamilton 1995).
This brief summary of the US and UK systems and integration ini-

tiatives reveals both common goals and significant contrasts in systems.
In the remaining sections, I will analyze how these and other initiatives
have worked, with an eye toward finding parallels and lessons for future
policy and practice. It appears that the Labour party will lead the United
Kingdom away from the Conservatives’ plans to cede considerable con-
trol of health services delivery to powerful private organizations (Wilkin,
Butler, and Coulter 1997); the new primary groups will face manage-
ment, information, and contracting challenges analogous to those con-
fronting large HMOs. As the United States advances toward more
structured integration within and between managed care systems (Wiener
and Skaggs 1995), it should consider the experiences of UK medical and
long-term-care personnel who have been tackling the same issues in
public managed systems. In the interest of provoking action and dis-
cussion, I have formulated the lessons and parallels as five “laws of
integration.” Following the presentation and discussion of these laws, I
will conclude with lessons and recommendations.
Laws of Integration
1. You can integrate all of the services for some of the people, some of
the services for all of the people, but you can’t integrate all of the
services for all of the people.
This paraphrase of Abraham Lincoln’s insight about fooling the people
addresses choices about narrow versus broad efforts to integrate. To date,
the first part has received the most attention, and it is epitomized by the
ideal of the interdisciplinary team that is empowered to cut across sys-
tems (e.g., being authorized to offer acute and long-term care and to
provide housing services in On Lok/PACE). This strategy has raised one
84 Walter N. Leutz
question that has received much attention (who needs integration?) but
has diverted attention from another (what should be done for everybody
else?). A comprehensive approach to integration will shape whole sys-
tems to enable them to respond to the varied needs of all persons with
disabilities.
Who needs what level of integration cannot be answered with em-
pirical data, but some dimensions of need and the operational domains
of service systems that may shape the answer can be suggested. First,
need dimensions include the stability and severity of patients’ condi-
tions, duration (short, medium, long, end-of-life), urgency of the inter-
vention (is it needed today, next week, next month?), scope (number and
complexity of service and benefit systems), and the users’ or carers’
capacity for self-direction. It seems logical that, at the extreme, sub-
groups characterized by limited capacity for self-direction and by long-
term, severe, unstable conditions that require urgent interventions from
both acute and related systems may benefit from more integration.
Groups that score on the other end of these dimensions might be can-
didates for less. Second, there are operational domains in which inte-
gration can be structured in service systems: systems to identify persons
with disabilities; clinical practices that are responsive to the needs of
these persons; management of transitions across settings; information
gathering and exchange; case management; management of funds from
multiple payment sources; and coordination of benefits.
Table 1 illustrates the first law by describing three levels of integra-
tion: linkage, coordination, and full integration. The table is designed
to answer the question of how the acute care and the “other side” (e.g.,
long-term care, education) work together. The first seven entries com-
pare the three levels with regard to the seven different operational do-
mains for integration. The final entry lists the six dimensions of need
and gives profiles that can be addressed by each level of integration.
“Linkage” allows individuals with mild to moderate or new disabil-
ities to be cared for appropriately in systems that serve the whole pop-
ulation without having to rely on outside systems for special relationships.
Linkage begins with population screening to identify emergent needs.
This can be done through population surveys (like those performed in
the Social HMOs) (Leutz, Abrahams, Greenlick, et al. 1988), analysis of
clinical and administrative data, and in the course of clinical practice.
Clinicians are taught to understand the basic special needs of different
categories of persons with disabilities, for example, by knowing to check
for sensory impairments in patients with Down syndrome (Valk, Akker,
Maskaant, et al. 1997). Personnel from nonmedical systems also estab-

lish links, for example, by noticing when clients are not taking medi-
cations and contacting medical personnel with the information. When
more serious conditions are identified, professionals know where it is
appropriate in other systems to send people and how to ensure that they
get there. They are ready to provide more information about care when
asked or to make inquiries for the benefit of patients who need to know
more about topics that affect them. In systems with linkage, providers
on both sides understand who is responsible for payments for each type
of service and who is eligible for what benefits, and they follow the rules
of coverage and payment rather than trying to shift costs and responsi-
bility. Because systematic linkage has seldom been fully implemented in
the current system, its pursuit promises to improve system performance
(Kerr, Fraser, and Felce 1996).
At the second level, “coordination,” explicit structures and individual
managers are installed to coordinate benefits and care across acute and
other systems. Coordination is a more structured form of integration
than linkage, but it still operates largely through the separate structures
of current systems. Coordination focuses on persons receiving benefits
simultaneously or sequentially from two systems of care on either a
short- or a long-term basis. The main tasks are coordinating benefits
(where does health coverage end and long-term care begin?); coordinat-
ing use of services; sharing clinical information in a planned manner;
managing transitions between settings; and assigning primary respon-
sibility for coordinating care. It goes beyond linkage by developing and
implementing policies in these areas. The Social HMO sites have gen-
erally operated at the linkage and coordination levels of integration. For
example, when frail members are hospitalized, the care management
unit is informed of admissions so that home care is terminated. Dis-
charge planners are trained to identify current or potential clients for
case management and to know the eligibility criteria for long-term-care
services. After discharge, communications are established between skilled
home health and long-term home care to share responsibilities and
transition to long-term care (Leutz, Abrahams, Greenlick, et al. 1988).
In summary, coordination identifies points of friction, confusion, or
discontinuity between systems and establishes structures and processes
to address problems (Social Services Inspectorate 1995b).
The third level, “full integration” creates new programs or units
where resources from multiple systems are pooled. Examples include the
interdisciplinary team at the PACE day center and joint commissioning
86
TA B L E 1
Levels of Integration: How Do You Work with the “Other Side”?
Operations Linkage Coordination Full integration

Screening Screen or survey populations Screen flow at key points (e.g., Not important except to
to identify emergent needs hospital discharge) to find receive good referrals
those who need special (changing needs identified
attention (e.g., CM, MD and met through team
consult) members)
Clinical practice Understand and respond to Know about and use key Multidisciplinary teams man-
special needs of PWDs in workers (e.g., discharge age all care
primary care, LTC, educa- planners) to link
tion, etc.
Walter N. Leutz
Transitions/service delivery Refer and follow up Smooth the transitions be- Control or directly provide
tween settings, coverage, care in all key settings
and responsibility
Information Provide when asked; ask when Define and provide items/ Use a common record as part
needed reports routinely in both of daily joint practice and
directions management
Integrating Medical and Social Services
Case management None Case managers and linkage Teams or “super” case manag-
staff (e.g., an MD rep on ers manage all care
the CM team)
Finance Understand who pays for each Decide who pays for what in Pool funds to purchase from
service specific cases and by guide- both sides and new services
lines
Benefits Understand and follow eligi- Manage benefits to maximize Merge benefits; change and
bility and coverage rules efficiency and coverage redefine eligibility
Need dimensions
Severity Mild/moderate Moderate/severe Moderate/severe
Stability Stable Stable Unstable
Duration Short to long term Short to long term Long term or terminal
Urgency Routine/nonurgent Mostly routine Frequent urgency
Scope of services Narrow–moderate Moderate–broad Broad
Self-direction Self-directed or strong infor- Varied levels of self-direction May accommodate weak self-
mal and informal direction and informal
Abbreviations: CM, care management; LTC, long-term care; PWD, person with disabilities.
87
88 Walter N. Leutz
teams set up to facilitate deinstitutionalization of dually diagnosed pa-

tients (Waddington 1995) and other challenging populations (Cam-
bridge 1997). Rather than better coordinating services and benefits of
existing systems, the fully integrated program gains control of resources
to define new benefits and services that it controls directly. Rather than
sharing information across systems, its providers work from a common
record.
The final entry of table 1 outlines the types of needs that might be
served by linkage, coordination, and full integration. It posits that link-
age is the most appropriate way to identify new needs and to begin
integrating services for many persons with disabilities who have both
mild or moderate impairments and stable medical and functional con-
ditions that are unlikely to destabilize and thus to require urgent at-
tention. Linkage candidates may receive services in both medical and
long-term-care or other settings, but they require little in the way of
ongoing communication or coordination, and these patients themselves
or their families may be willing and able to manage. Coordination may
be appropriate for a broad range of persons with moderate to severe
conditions who receive short-term or long-term special services and
whose care is usually, but not always, routine. A critical difference from
linkage is that, when urgent or complex circumstances arise, there are
systems and personal relationships in place to identify changes and to
ensure coordination, particularly when the user’s or family’s capacity for
self-direction is limited. Full integration probably is appropriate only
for a small subset of each group of persons with disabilities. Full inte-
gration may need specialized types of interventions, expedited access to
them, close collaboration between knowledgeable professionals, and the
like. Defining the persons who belong in such programs, assembling the
necessary services, and allocating appropriate resources (enough to do
the job but not so much as to short-change other groups) are among the
most pressing issues in integration policy and practice.
To develop integration priorities more systematically, the needs of
various subgroups can be identified by analyzing utilization patterns,
health conditions, and user concerns about the quality of health care,
long-term care, and other important service systems. Depending on the
population, the additional benefit systems might include long-term
care, behavioral health, education, housing, or vocational rehabilitation.
For example, persons with learning (developmental) disabilities often
have more pressing concerns, like the need for housing (Prosser 1997)
and supports for vocational, educational (Social Services Inspectorate

1994; 1995c), and leisure services (Greig 1997; McGrother, Hauck,
Bhaumik, et al. 1996). Each major group of persons with disabilities
may contain one or more subgroups that are characterized by particular
urgency (e.g., serious mental illness in addition to developmental dis-
ability; terminal and debilitating illness; or specific transitions, like
those from school to work and hospital to home). The relative impor-
tance of health and various other service systems can be expected to
change over the life cycle (Leutz 1999; Social Services Inspectorate 1995a).
2. Integration costs before it pays.

Saying that integration costs before it pays raises a caution sign when
promises of cost offsets are used as a selling point. The second law posits
that the savings from integration are hopes, whereas the costs cannot be
avoided if success is to be achieved. Whether the former outweigh the
latter to create net savings will depend on specific situations and must
be subjected to careful analysis. There are at least three types of costs to
integration: staff and support systems, services, and start-up.
Staff and Support Systems. Depending on the scope of integration,
staff should be involved at the clinical, management, and policy levels.
At each level it takes time to learn about the capabilities of the other
system, to decide how to work together (including changing to improve
the fit), and to communicate. There are up-front costs, ongoing opera-
tional costs, and future costs arising from evaluation and change. Com-
plaints about drains on staff time are common in evaluations of the
major UK integration efforts (Glendinning and Lloyd 1997; Mays 1997b);
in the United States, On Lok/PACE staffing appears to increase with the
complexity of administering larger sites (Branch, Coulam, Zimmerman,
et al. 1995). Successful integration requires any of a variety of support
systems, including consolidated assessment forms and clinical and man-
agement information systems. Failure to increase budgets to train staff
in new operations has hampered many integration efforts (Leutz 1994;
Leutz, Capitman, MacAdam, et al. 1992; Leutz, Greenlick, Ervin, et al.
1995; Social Services Inspectorate 1994).
Estimates of the magnitude of staff and support costs are difficult to
obtain because they are usually embedded in other costs. Efforts to
integrate medical care provide some guidance. As managerialist schemes
were implemented in the NHS, management costs increased from £1.2
90 Walter N. Leutz
to £2.1 billion between 1989–90 and 1993–94 (Ham 1995). Admin-

istrative costs of Medicare HMOs commonly run from 10 to 15 percent
of revenues (Wiener and Skaggs 1995). Larger TPP sites tended to have
dedicated managers, but lower per capita management costs, than smaller
sites, suggesting the addition of management tasks to clinical tasks for
physicians in smaller practices (see the Third Law) (Mays 1997b). The
most prominent ongoing administrative cost entailed in integrating
acute and long-term care is care management, which typically ranges
between $100 and $150 per month per client served in long-term-care
demonstrations and state programs (Leutz, Greenlick, and Capitman
1994; Phillips, Kemper, and Applebaum 1988; Thornton, Dunstan,
and Kemper 1988), and even higher in PACE (Branch, Coulam, and
Zimmerman 1995).
Service Costs. Experience shows that new service funding helps to
facilitate integration. These may be particularly difficult funds to obtain
from payers because promises are often made that integration initiatives
will self-fund new services through “savings” when an existing service
(ideally, lower-cost long-term care) is substituted for another. The prob-
lem with this tactic is that it is rare for people to give up control of
“their” services to substitute for someone else’s. In the United King-
dom, attempts to integrate, which generally rely on voluntary cooper-
ation, provide numerous examples of turf guarding by health authorities
(Glendinning and Lloyd 1997; Greig 1997), social services (Myles,
Popay, Wyke, et al. 1997), and even housing and educational authorities
( Jones 1996). The problem may even be worse at policy than at oper-
ational levels: at one TPP site, for example, even though line staff were
participating as members of a multidisciplinary geriatric team, policy-
level staff of the local authorities avoided planning meetings because
they were afraid of being saddled with health system costs (Myles,
Popay, Wyke, et al. 1997). This suggests that one ingredient of success
in integration experiments is the allocation of new service money des-
ignated for that purpose. There are precedents for such funding: Social
HMOs’ funds from member premiums and the “extra” money they
receive from Medicare for frail beneficiaries (Leutz, Greenlick, Ervin,
Feldman, et al. 1991); money restricted to community care to support
deinstitutionalization of persons with learning disabilities (Wadding-
ton 1995); generous Medicaid capitations to PACE and a Medicare
capitation concession to MSHO (Branch, Coulam, and Zimmerman
1995; Minnesota Department of Human Services 1997); Medicare dem-
onstration funds to the program, Channeling, and its predecessors (Weis-

sert, Cready, and Pawelak 1988); and new service funds for community-
care experiments in the county of Kent (Davies and Challis 1986).
Start-Up Costs. Successful integration projects generally have been
helped by significant start-up grants. In the United States, these include
the first round of Social HMOs (Leutz, Greenberg, Abrahams, et al.
1985), PACE (Kane, Kane, Haye, et al. 1996), Wisconsin Partnership
(Hamilton 1995), and MSHO (Minnesota Department of Human Ser-
vices 1997). In the United Kingdom they include special transitional
grants to implement Caring for People (Social Services Inspectorate
1995b) and TPP (Mays 1997b). In contrast to the Conservatives’ addi-
tions to administration for the purpose of creating an internal market in
the NHS, the new Labour plan to form primary groups of GPs is ac-
companied by a proposal to shift £1 billion from NHS administration to
services (Horton 1998), a policy that does not bode well for successful
integration.
As integration moves beyond the demonstration stage, it is harder to
find funds for replication. Failure to anticipate these costs, which could
legitimately be considered long-term investments (Davies and Challis
1986), is a typical shortcoming of public initiatives, which are often
strapped for cash and may not recognize the new management, training,
and supervision models that are required (Davies 1994; Leutz, Altman,
Capitman, et al. 1992).
In summary, unless these start-up and continuing costs are paid,
integration may not occur—either within or between organizations.
Staff may not participate in planning, smooth support systems will not
be developed, and inadequate training will hamper operations. If they
are not compelled by strong policy or financial controls or lured by new
service funding, subunits will hold on to the control of their services,
and thus, their jobs. Furthermore, if the deal is not attractive to private
providers and MCOs, they simply may choose not to participate.
3. Your integration is my fragmentation.

The driving interest in integration comes from providers, managers, and
researchers who are intimately—perhaps exclusively—involved with a
particular group that “needs” integrated services: poor and frail elders
with chronic illnesses; infants with severe developmental disabilities;
teens with physical or developmental disabilities who are making the
92 Walter N. Leutz
transition from school to work; adults with physical disabilities who are
seeking control and autonomy. The would-be integrator’s focus for change
is on how quality, efficiency, access, user control, and the like break
down as a person needs help from multiple professionals and benefit
systems.
However, asking professionals and managers to integrate their
services—or even simply to cooperate—not only creates costs (as covered
in the Second Law), but also requires them to expand their knowledge,
perspectives, and interests. My job as a manager, provider, or professional
is much simpler if all I need to worry about is my own service. The more
special groups and procedures I must work with, the more I need to learn
and to accommodate. Even if I appreciate the potential of integration to
improve the quality of my service or to create savings to the system and
the service user, I may nevertheless still have a sense that my job is being
fragmented.
The situation of primary care physicians is the most critical case in
point. They have demanding, complex, and unique tasks. They already
deal with the needs of varied patients while experiencing increased
productivity demands. Asking them to add special information or ac-
tions for small subgroups of patients simply increases the already con-
siderable pressure of their practice routines.
Evaluators of integration efforts frequently seem disappointed with
the level and quality of physician participation or understanding (Myles,
Popay, Wyke, et al. 1997; Plant 1997; Social Services Inspectorate 1994;
Harrington, Lynch, Newcomer, et al. 1993). Whereas evaluators may or
may not be correct in their characterizations, a closer look at a few
examples of GP involvement in integration raises questions about the
reasonableness of expecting more from GPs in small practices. One total
purchasing site with seven physicians serving 16,000 patients added
continuing and community care to its agenda for total purchasing dis-
cussions. However, they also added emergency–accident, mental health,
maternity, and early discharge provisions. Not surprisingly, the most
frequent complaint about TPP concerned the demands it placed on
professionals’ time (Mays 1997b). Similarly, when the implementation
of Department of Health guidelines was evaluated, one finding was that
physicians were reacting to their experience of “consultation fatigue” by
seeking to limit the scope of discussions through categorizing continu-
ing care policy as a health authority rather than as a physician issue
(Glendinning and Lloyd 1997). Finally, in implementing the Social
HMO demonstration, one large HMO site decided not to train its
several hundred physicians to handle the new program features because

it had just trained them for a new Medicare HMO demonstration a few
years earlier (Leutz, Greenberg, Abrahams, et al. 1985). Instead, train-
ing focused on clinic nurses as coordinating points. The site was later
criticized by evaluators as not having integrated primary care and long-
term-care case management (Harrington, Lynch, Newcomer, et al. 1993).
The new Labour proposal to create primary groups of 50 GPs will mean
a larger base to support administrators in the handling of integration
issues, but delegating responsibilities to administrators will remove
planners from the everyday, operational issues that smaller, cross-
discipline teams have been able to address more successfully (Coulter
1996).
In summary, would-be integrators need to be sensitive to other de-
mands on providers and to consider carefully what level of integration is
appropriate to expect. Even providing information about how to iden-
tify and care for a special needs group and how to make appropriate
referrals (the basic linkage level of integration) requires time and effort
both to implement and to absorb. Coordination requires more attention,
training, and cooperation from more people. Full integration reduces
demands on the sections of the larger systems that are transferred to the
integrated team, but coordination, and even linkage (for referrals), are
required for functions that the team does not carry out.
4. You can’t integrate a square peg and a round hole.

Logical though integration seems as a concept, there are underlying
differences between acute and long-term care that have frustrated in-
tegration efforts. The sources of this square–round problem may de-
rive from the different entitlement bases for the two types of benefits.
First, in both the United States and the United Kingdom, health care
benefits (at least for elders) are universal national entitlements, and
even in the United States most persons under 65 have health insur-
ance. In contrast, long-term-care benefits are individuals’ out-of-
pocket responsibility up to the point when they have “spent down”
their means to a level where they are entitled to state or local author-
ity service benefits. Second, access to health care is based on medical
need as determined by physicians, whereas access to publicly funded
long-term-care benefits is based on functional status, which is usually
determined by nurses or social workers. Parallel clashes can occur
between medical care and other systems; one example would be the
94 Walter N. Leutz
developmental focus of early childhood educators working with chil-

dren who have learning disabilities. In summary, both long-term-care
and health funding systems pay for similar and overlapping nursing-
home, community, and home care services (the source of potential
savings from integration), but the conditions under which they do so
differ. Each of these differences creates targeting conflicts, manage-
ment challenges, and cultural clashes that frustrate integration.
First, differences between functional and health status approaches are
epitomized by the frequent conflicts between the front-line, community-
based staff in the two worlds: home health (community) nurses and
home care (home help) case managers. One government evaluation of
the UK program, Caring for People, diplomatically observed that the
program’s professionals lacked a “shared hypothesis for integrated ac-
tion” (Social Services Inspectorate 1995b, 45). Social services assess-
ments were characterized by description, eligibility criteria, negotiation,
social support systems, and long-term goals, whereas health assessments
were characterized by diagnosis, clinical judgment, expert advice, indi-
vidual health, and short-term goals. Social service staff tended to see
case management as a way to “drive organizational change,” whereas
health authority staff tended to see it as a “distraction from purely
health provision.”
Second, long-term-care benefits are granted only to those who are
financially eligible. This requires an integrated system to offer universal
access based on need for one set of its services while denying access to
many who need another set of services. This not only creates conflict and
frustration among professionals; it also can be confusing and time con-
suming when, to take one example, tracking financial eligibility. Such
a problem occurred in the United Kingdom when the NHS reduced the
number of its geriatric units and turned patients over to local authori-
ties (Lancet 1994). Most US demonstrations of integration have “solved”
this problem by designing the system for Medicaid eligibles only (as in
PACE, MSHO, and the Wisconsin Partnership), but even these segre-
gated systems have problems collecting and recording share of cost for
the many long-term-care users who “spend down”: that is, users who
become financially eligible for Medicaid because they use more health
and long-term-care services than they can afford. Difficulties also arise
when a system confronts differing Medicare and Medicaid regulations,
particularly the requirement to offer Medicare beneficiaries freedom of
choice among providers.
Third, long-term-care benefits, particularly for home-based and

community-based (HCB) services, are chronically underfunded in both
the United States and the United Kingdom. This is reflected in waiting
lists, individual spending caps, geographic differences, and service ex-
clusions for Medicaid and local authority HCB care (Bennet 1996; Bald-
win and Lund 1996; Kassner 1995; Liu, Hanson, and Coughlin 1995;
Wiener and Skaggs 1995). If these funding shortfalls are carried over
into the integrated program, it is then forced to adopt its own system of
rationing, which is contrary to standard medical practice and may or
may not be consistent with public policy on rationing. Also, most local
long-term-care systems have additional HCB services outside mainline
funding streams (e.g., Older Americans Act, social services block grants,
voluntary and self-help services, and, of course, informal care). The basis
for substitutability of these services with integrated system services
needs to be worked out to ensure that system members receive fair, but
not special, access to them.
On the surface, the integrated Social HMO and PACE models seem
to belie the Fourth Law. On closer examination, they do not. The first,
and more straightforward, factor is that Social HMOs’ greatest achieve-
ment in integration has been with the 90 to 95 percent of their mem-
bers who are private pay. Financing is integrated by adding a finite,
universal, privately funded long-term-care benefit to an HMO—thus
avoiding means testing. In contrast, Social HMO arrangements for en-
rollment of Medicaid eligibles, for transition of private-pay members to
Medicaid once they qualify by spending their assets, and for access to
Medicaid long-term-care benefits have been cumbersome at all sites
(Leutz and Hallfors 1993; Leutz, Greenberg, Abrahams, et al. 1985).
The PACE program does fully integrate Medicaid and Medicare fund-
ing and integrate management and clinical decisions in the interdisci-
plinary team operating in the day care setting, and its performance in
reducing hospital and nursing-home utilization (and perhaps morbidity
and mortality) is impressive (Gruenberg and Kaganova 1997; Burstein,
White, and Kidder 1996). However, PACE has supports and exclusions
that open the way for waste from a poor round–square fit (Branch,
Coulam, and Zimmerman 1995; Kane, Illston, and Miller 1992). First,
unlike Social HMOs, payment rates have not been adjusted downward
when the functional status of PACE members improves. (Rates will be
so adjusted under 1997 Balanced Budget Act provisions, granting PACE
permanent status.) Second, even though PACE enrollees reside in the
96 Walter N. Leutz
community on application, the Medicaid payment rates are based on

Medicaid’s higher average institutional spending—not on its HCB spend-
ing (reflecting the underfunding of HCB services). If these features lead
to payments that are higher than would have been the case had enrollees
stayed with the regular system, it means that scarce resources are being
taken away from beneficiaries who do not join PACE. A final shortcom-
ing is that few non-Medicaid members can pay the Medicaid rate ($2,000
to $4,000 per month, or more). Low-income elders can spend down to
Medicaid eligibility levels, but this requires them to give up nearly all
assets and income.
Several states are pushing the integration envelope even further than
Social HMO and PACE by seeking contracts with mega-MCOs to de-
liver both acute and long-term care on a prepaid basis. Like Ronald
Reagan’s Strategic Defense Initiative antimissile system, these inte-
grated systems may work out on paper, but in reality the chances for
malfunction are many. Voluntary enrollment and capitated financing
increase the likelihood of low enrollment, biased selection, and inaccu-
rate payment. Full risk raises the stakes, increasing the likelihood that
the bottom line will be a major factor for providers in choosing when to
participate, whom to target for enrollment, and when to quit if the
experience turns bad. During the first six months of Minnesota’s SHO,
for example, the largest HMO was not participating, and the next larg-
est was waiting to see if payment rates favored its strategy of enrolling
nursing-home patients only.
In contrast to American private initiatives, the United Kingdom has
taken largely administrative and joint planning approaches to these is-
sues, promoting TPP sites and encouraging joint commissioning efforts
to bridge the gap between acute and long-term care with joint planning,
purchasing, and care management strategies. Where these initiatives have
overcome the simple avoidance that originated from turf guarding, there
have been reports of culture clashes between medical and social service
personnel (Glendinning and Lloyd 1997; Myles, Popay, Wyke, et al. 1997;
Rummery and Glendinning, 1997). Operational, case-by-case solutions
to problems have been easier to find than resolution at the policy financ-
ing levels, where clashes arise over more fundamental issues. Integration
also has worked better with populations for which there is a longer tra-
dition of joint commissioning; persons with learning disabilities would
form one example (Social Services Inspectorate 1995b).
In summary, “successful” US integration models make compromises
on full integration in three ways:
1. by focusing on either private or welfare populations

2. by coordinating rather than integrating management and clinical
care
3. by filling gaps in long-term care through adding money rather
than (or perhaps in addition to) saving it
UK integration models, which spurn strong financial and enrollment

incentives to create integrated programs, are weaker, but perhaps more
practical. Although GP fundholder and joint commissioning initiatives
appear seldom to have made changes at the policy level, the value added
from linking and coordinating the systems operations so that they work
better for both employees and users on a daily basis should be recog-
nized. Until we dismantle the barriers between social insurance and
means testing and between national and local funding, perhaps we can-
not expect more.
5. The one who integrates calls the tune.

The first question about the Fifth Law will be, Why doesn’t it read as
follows?: “The one who pays the integrator calls the tune.” The answer is
that integration to date largely has been a provider–professional rather
than a payer–public initiative. For a number of reasons, which might
include the wish to foster innovation, a lack of consensus on policy, or
insufficient resources to direct and monitor implementation, payers gen-
erally have left the choice of tunes to the pipers. Some questions remain:
What tunes have been called, who has requested them, who has been left
out, and what impact have the characteristics of the payers had on the
nature of integrated programs?
The initiators of the major US demonstrations of integration (e.g.,
Social HMO, PACE, MSHO, Wisconsin Partnership) were long-term-
care providers, researchers, and bureaucrats, rather than physicians (An-
sak and Zawadski 1984; Leutz Greenberg, Abrahams, et al. 1985;
Minnesota Department of Human Services 1997; Hamilton 1995, re-
spectively). The priorities of these organizations are social and long-
term-care benefits, and they sought to bring the medical system around
to a social care orientation. In addition to their part in development,
nonphysicians also take the lead in day-to-day integration in some mod-
els. In the first-round Social HMOs, integration has been led by the care
management unit, staffed by a mix of social workers and nurses, and
98 Walter N. Leutz
supported by medical directors with a geriatric orientation. The in-

volvement of primary care physicians with long-term care at the first-
round Social HMO sites has been primarily reactive (Abrahams, Macko,
and Grais 1992; Abrahams, VonStenberg, Zeps, et al. 1992). PACE
physicians are part of the team leadership, but PACE has had a difficult
time finding physicians who want to work in the multidisciplinary
team (Branch, Coulam, and Zimmerman 1995; Kane, Illston, and Miler
1992). The government chose a design for second-generation Social
HMOs that emphasizes geriatric care. A geriatrician has been in charge
of their development, and, true to the tenets of the Fifth Law, benefit
priority has been given to geriatric teams following geriatric protocols
(Finch, Kane, Kane, et al. 1992).
The TPP initiative in the United Kingdom sheds some light on
community physicians’ priorities for integration. The physicians who
were in charge of the most sites focused only on one or two areas for total
purchasing (Mays 1997b), and physicians showed no interest in broader
areas like housing, targeting policy, or medical–social care cultures (Myles,
Popay, Wyke, et al. 1997). No TPP prioritized at-risk children, phys-
ically disabled adults, or patients with learning disabilities (Mays 1997b).
Joint commissioning appears to be a more successful approach to
integration in the United Kingdom than GP fundholding, perhaps
because it relies on professional cooperation rather than on physician
leadership in managing the internal market. Donald Light contends
that cooperative planning soon won out over competition when the
NHS internal market was being created in the early 1990s:
Within the first two years of the reforms, the British started using
the term “commissioning” to connote the notion that purchasing
authorities should do more than purchase existing services “off the
shelf.” Rather, they should seek new, more cost-effective configura-
tions of services across specialty and organizational lines and move
upstream to prevention, health education, self-management of chronic
problems, and reduction in illnesses whose source lies in local envi-
ronments. (Light 1997a, 305)
A key to the success of joint commissioning is to recruit commissioners

for whom integrated services for a population group is a preeminent
issue. We should expect different views from district health authorities,
community trusts, GPs, and social and educational service staff (Social
Services Inspectorate 1995c; Light 1997b). Although physicians occa-
sionally take the lead in joint commissioning (Mays 1997a), Greig’s

(1997) finding that GPs were “almost totally absent” from joint com-
missioning for learning disabilities services should not be surprising.
It would be comforting to think that these local shortcomings could
be fixed by a dose of national policy guidelines. Apparently it is not so
simple. For example, studies of how localities carried out national man-
dates to promote better integrated care for those with learning disabil-
ities (Hatton and Emerson 1995), to define the limits of NHS coverage
of continuing and community care (Glendinning and Lloyd 1997), to
coordinate services for persons with physical disabilities, and to ensure
that TPPs implement continuing and community care policies (Myles,
Popay, Wyke, et al. 1997) revealed that knowledge of national policies
was limited (particularly among GPs), that there was not enough in-
formation on which to base decisions, and that compliance was poor.
Labour’s new primary groups of GPs should have better administrative
support than was provided to the smaller fundholding practices to im-
plement integration mandates, but they will still need planning models
that include other providers, guidance on priorities for different groups
of persons with disabilities, and resources to plan and implement change.
Before leaving the Fifth Law, it is worth considering its implications
for one of the most pressing issues in long-term care: autonomy and
control for users and carers. One hope for integration is that it will
transform complex and confusing systems of health and human services
so that they are more “user friendly.” Integration can do this by creating
single access points for multiple services, by empowering care managers
to control more services, and by strengthening their access to other
services they do not control directly. However, the more we integrate
the system in these ways, the more users must rely on professionals to
shape, maintain, and use it.
Of course professionals should take steps to empower users and carers,
but they will still be sharing their power not ceding it. Light (1997b)
contends that community and user involvement in planning is funda-
mentally threatening to professional values. Their professional and or-
ganizational perspectives cause them to value their own services more
highly than those the users or the community may perceive as impor-
tant. Nor is there much encouragement to be gained from the records of
other professionals and health service systems of their dealings with
service users and whether they provide good information about how to
use the system, ask about user preferences, give users and carers mean-
100 Walter N. Leutz
ingful control over care plan decisions, or involve users in planning and
policy (Clark, Dyer, and Hartman 1996; Leutz, Sciegaj, Capitman, et al.
1995; Myles, Popay, Wyke, et al. 1997; Rummery and Glendinning
1997; Social Services Inspectorate 1996). When user perspectives are
ignored, the priority assigned to special needs issues is also reduced
(Social Services Inspectorate 1996), often leading to other adverse con-
sequences. For example, in the rush to implement Caring for People,
service modalities that were comfortable to persons with learning dis-
abilities were eliminated without their consultation (Social Services In-
spectorate 1994).
One promising individual-level approach to empowerment in resource-
strained long-term-care systems is “contextual autonomy,” whereby care
managers clearly lay out the choices and limits for users (Capitman and
Sciegaj 1995). More systemic approaches bring users, carers, and advo-
cates into planning and commissioning efforts (Light 1997b; Morris
1995). Inclusion of service users and community-based providers allows
for a range of views to be expressed in planning as a counter to medical
views, leading to more resources for nonmedical service modalities,
leadership from other sectors (e.g., social service, education) (Ruta,
Donaldson, and Gilray 1996; Social Services Inspectorate 1995a; 1995c).
Unless users and carers become strategically involved, it is less likely
that individuals will find autonomy within the system (Social Services
Inspectorate 1996).
Lessons and Recommendations
The laws of integration perhaps have no scientific justification. Social

science lacks the tools to reveal and predict processes and outcomes in
complex areas like health, long-term care, education, and other related
systems. My intent in proposing laws is to draw attention to the fact
that a great deal of experience and knowledge about integration still
remains to be digested and applied. There is still plenty of room for
research and experimentation, but there is enough accumulated expe-
rience that policy makers, managers, and clinical leaders can now call
some of the tunes rather than having to encourage ad hoc initiatives
indefinitely. The three recommendations offered below are relevant, with
adaptations for the different system environments, both to the United
States and the United Kingdom.
1. Involve service users, carers, and community service providers in

planning and oversight.
Successful integration of medical and long-term-care services will occur
only if all parties participate in planning and implementation. This
review has made it clear that GP physicians or other mainstream acute
care providers will focus on their own issues, which understandably do
not single out relatively small, special-needs groups for attention. Plan-
ners and providers from all sectors must go out of their way to involve
users and carers. There are various ways to accomplish this: changing
the times, places, and formats of meetings to accommodate these groups;
organizing separate users’ and carers’ groups, which would hold their
own meetings and send empowered representatives to professional meet-
ings; including advocates and specialized community providers in plan-
ning and activities (Barnes and Walker 1996; Gerry and Mirsky 1992;
Tyne 1994).
2. Develop systems to integrate, coordinate, and link services for

persons with disabilities.
Integration of health care, long-term-care, and other human service
systems is needed, not just for the most severely ill and impaired, but
also for the moderately and newly impaired. In conjunction with de-
veloping fully integrated systems for the most frail and medically at risk
(the main focus of current policy and demonstration), we should be
reconsidering how mainline health care systems provide care for those
persons with disabilities who do not choose or do not qualify for separate
programs. Some possible approaches would be screening populations for
risk of disability, providing primary care for all types of persons with
disabilities, making appropriate referrals, and coordinating services with
closely related systems through case management.
We can begin to design these systems for various populations by
learning from tested models, by analyzing health and functional status
data and utilization patterns, and by inviting users and carers to par-
ticipate in design and evaluation. Two questions deserve special atten-
tion. First, what areas does each subgroup consider most critical for
integration? What systems should be linked and at what levels? Second,
who should oversee the integration effort? Who is most interested in
these areas, and whose clinical or management approach is oriented
102 Walter N. Leutz
most strongly toward cooperation and understanding rather than dom-

ination and conflict?
In the United States, both opportunities and risks increase as more
long-term-care system users join MCOs. HMOs have shown that they
can establish screening systems and specialty care units, and even that
they can coordinate continuing and community care benefits, but they
have not produced many examples of linkage and coordination with
public social care systems (Fama and Fox 1993; Kramer, Fox, and Mor-
genstern 1992; Wiener and Skaggs 1995). Even if some social care funds
(e.g., for long-term care) follow beneficiaries into MCOs in new inte-
grated models, other funds for the remaining members will stay in the
current systems.
The British situation is different, but it has some of the same ele-
ments. Labour’s plan to replace the Conservatives’ fragmented NHS
initiatives to empower primary care practices (fundholding, total fund-
holding, and total purchasing pilots alongside purchasing by district
health authorities for nonparticipating GPs) with single, large primary
care practices in each locality should lead to better integrated medical
care delivery. However, social, educational, and long-term-care services
lie outside the scope of the new practices. As in the United States, unless
all levels of integration for all types of persons with disabilities are
coordinated in all service organization approaches, the result may be
new types of fragmentation, frustration, and waste.
3. Clarify borders between medical and other systems.

One definition of the long-term-care system is that it picks up all the
pieces that the health care system does not cover (Leutz, Capitman,
MacAdam, et al. 1992). Only when we have clearly and consistently
defined the limits of medical care benefits (particularly skilled home
nursing, therapies, and nursing homes) can long-term-care and other
human service systems define where they begin. In both the United
States and the United Kingdom, national guidelines can become lost in
regional, local, and even private interpretations of coverage borders. In
the United States, private intermediaries interpret the boundaries for
Medicare fee-for-service beneficiaries, and private insurers and HMOs
perform this exercise for others. In the United Kingdom, when local
health authorities have been asked to define these borders, they have
found it difficult to assign workloads and are not oriented to think in
this way about the divisions (Glendinning and Lloyd 1997; Social Ser-
vices Inspectorate 1995c). It is naive to expect competing US providers
to cooperate in working out local agreements (Brown and McLaughlin
1990).
Instead, clear national guidance should set the standards and criteria
to be implemented locally. Clearer information, monitoring of within-
plan benefits, and stronger appeal paths for service users are also needed.
Opening up the debate over health benefit exclusions and limitations
would make it more difficult to cut health coverage, or at least would
strengthen the ability of social, educational, and other services to obtain
resources to pick up the pieces. The debate also would clarify for long-
term-care planners the exact point at which their coverage would begin
and would help to inform beneficiaries about what to expect from each
system.
Although clearer definitions of borders would help, the square–round
fit problem can be solved only by redrawing the borders for a better fit.
One approach would be to expand Medicare and the NHS to cover more
long-term care; proposals have been made in both countries to add home
care and short-term nursing-home care ( Joseph Rowntree Foundation
1996; Pepper Commission 1990). Another would be to expand Medic-
aid and local authority care by applying less stringent means testing for
long-term-care services. Means testing is seldom used for educational
services, which may be one reason that joint commissioning has often
achieved integration for the learning disabled (Greig 1997). Both en-
titlement expansions, however, quickly run up against cost constraints:
Medicare and the NHS at the national level, Medicaid and local author-
ities at the state and local level.
In the absence of public action, private insurers and MCOs could
commit themselves more aggressively to solving the square–round prob-
lem on their own. One way to expand borders would be to support
private long-term-care benefits financed in conjunction with acute care,
as occurs in the Social HMO and in the British Social Maintenance
Organization proposal for the United Kingdom (Davies and Challis
1986). The 1997 Balanced Budget Act contains language that could
make Social HMOs a permanent part of Medicare. Another example is
the Manifesto 2005 initiative by Kaiser Permanente, which is the larg-
est HMO in the United States. Based on its experience with the Social
HMO it has operated in Oregon since 1985, the entire Kaiser Perma-
nente system is conducting population screening and risk profiling, and
104 Walter N. Leutz
its Interregional Committee on Aging has adopted the following vision

statement:
By the year 2005 Kaiser Permanente will have expanded its scope
of services to include a LTC system consisting of a broad range of
home- and community-based services that would be easily accessible
to persons with functional disabilities. This will be a multi-disciplinary
system, with providers and consumers of care working in collabora-
tion to maximize the independent function of persons with disabil-
ities of any age. (Nonnenkamp 1996)
Kaiser and other forward-looking MCOs will face limits in what they
can do without corresponding support from public payments, and they
could face competitive disadvantages if they overextend their benefits
and reputation for serving persons with disabilities (Schlesinger 1997).
There also are differences between what a group-model HMO like Kai-
ser (or a forward-looking local authority) can do compared with the
possible achievements of a network-model MCO or an independent
practice association.
Conclusion
I have recommended both more policy structure and more benefit cov-
erage than are politically popular—particularly in the United States.
My recommendations also may be interpreted as throwing cold water on
the popular movement that advocates full integration of acute and long-
term care for persons with severe disabilities. My intent is to place full
integration into the larger context of good human service practices by
integrating through linkage and coordination. The problem is that,
without specification, support, and enforcement, it cannot be assumed
that good human service practice will occur, particularly in the turbu-
lent worlds of emerging managed care in the United States and reorga-
nization in the United Kingdom. Full integration may be very effective,
and even efficient, for a few. The caution I have offered herein is that
while we are trying to identify who those few are, what they need, and
how to provide and pay for their care, we should ensure that the needs
of the many receive commensurate attention.
References
Abrahams, R., P. Macko, and M.J. Grais. 1992. Across the Great Di-
vide: Integrating Acute, Post-Acute, and Long-Term Care. Journal
of Case Management 1(4):124–34.
Abrahams, R., T. VonStenberg, D. Zeps, S. Dunn, and P. Macko. 1992.
Integrating Care for the Geriatric Patient: Examples from the Social
HMO. HMO Practice 6(4):14–19.
Ansak, M.L., and R.T. Zawadski. 1984. On Lok CCODA: A Consoli-
dated Model. Home Health Care Services Quarterly 5:147–70.
Baldwin, S., and N. Lund. 1996. Charging Ahead. Joseph Rowntree
Foundation. London.
Barnes, M., and A. Walker. 1996. Consumerism versus Empowerment:
A Principled Approach to the Involvement of Older Service Users.
Policy and Politics 24:375–93.
Bennet, F. 1996. Highly Charged: Policy Issues Surrounding Charging for
Non-Residential Care. Joseph Rowntree Foundation. London.
Branch, L., R. Coulam, and Y. Zimmerman. 1995. The PACE Evalua-
tion: Initial Findings. Gerontologist 35:349–59.
Brown, L.D., and C. McLaughlin. 1990. Constraining Costs at the Com-
munity Level. Health Affairs 9(4):5–28.
Burstein, N.R., A.J. White, and D. Kidder. 1996. Evaluation of the
PACE Demonstration: The Impact of PACE on Participant Outcomes.
Cambridge, Mass.: Abt Associates.
Cambridge, P. 1997. Joint Commissioning: Searching for Stability in an
Unstable World: Commentary. Tizard Learning Disability Review
2(1):26–30.
Capitman, J., and M. Sciegaj. 1995. A Contextual Approach for Under-
standing Individual Autonomy in Managed Community Long-
Term Care. Gerontologist 35(4):533–40.
Clark, H.S. Dyer, and S. Hartman. 1996. Going Home: Older People Leav-
ing the Hospital. Joseph Rowntree Foundation. London.
Coulter, A. 1996. Why Should Health Services be Primary Care-Led?
Journal of Health Services Research and Policy 1(2):122–4.
Davies, B. 1994. Divisions of Labor and Home Care in Britain. In Care
of the Chronically and Severely Ill, eds. J.R.H. & E.J. Hollingsworth,
107–44, New York: Aldine de Gruyter.
———. 1995. The Reform of Community and Long-Term Care of
Elderly Persons: An International Perspective. In International Per-
spectives on Community Care for Older People, eds. T. Scharf and C.C.
Wenger, 21–38. Brookfield, Vt.: Ashgate.
Davies, B., and D. Challis. 1986. Matching Resources to Need in Community
Care. Hants, England: Gower.
106 Walter N. Leutz
Department of Health. 1997. The New NHS: Modern, Dependable. Lon-

don: Her Majesty’s Stationery Office.
Fama, T., and P.D. Fox. 1993. An Extra Mile for the Elderly. HMO
Magazine (November/December): 22–9.
Finch, M., R. Kane, R. Kane, et al. 1992. Design of the 2nd Generation
S0HMO Demonstration: An Analysis of Multiple Incentives. Minneapo-
lis: University of Minnesota.
Gerry, M.H., and A.J. Mirsky. 1992. Guiding Principles for Public
Policy on Natural Supports. Natural Supports in School, at Work and
in the Community for People with Severe Disabilities, ed. J. Nisbet,
341–6, Baltimore: Brookes.
Glendinning, C., and B. Lloyd. 1997. Developing Local “Continuing Care”
Policies and Guidelines: The Implications for Primary and Community
Health Services (SR/DB/97). Manchester, U.K.: National Primary
Care Research and Development Centre, University of Manchester.
Greig, R. 1997. Joint Commissioning: Searching for Stability in an
Unstable World. Tizard Learning Disability Review2(1):19–25.
Gruenberg, L., and J. Kaganova. 1997. An Examination of the Cost-
Effectiveness of PACE in Relation to Medicare. Cambridge, Mass.:
DataChron Health Systems.
Ham, C. 1995. Population-Centered and Patient-Focused Purchasing:
The UK Experience. Milbank Quarterly 74:191–214.
Hamilton, T. 1995. The Wisconsin Partnership Program. Madison: Wis-
consin Department of Health and Social Services.
Harrington, C., M. Lynch, R. Newcomer, and N. Miller. 1993. Medical
Services in Social Health Maintenance Organizations. San Francisco:
Institute for Health and Aging, University of California.
Hatton, C., and E. Emerson. 1995. Services for Adults with Learning
Disabilities and Sensory Impairments. British Journal of Learning
Disabilities 23(1):11–17.
Horton, R. 1998. The Realpolitik of a New National Health Service for
the UK. Lancet 351:76–7.
Iglehart, J. 1992. The American Health Care System: Managed Care.
New England Journal of Medicine 327:742–7.
Jones, C. 1996. Public Expenditure on Services for People with Intel-
lectual Disabilities. Journal of Applied Research in Intellectual Disabil-
ities 9(4):289–306.
Joseph Rowntree Foundation. 1996. Meeting the Costs of Continuing Care.
London.
Kane, R., L. Illston, and N. Miller. 1992. Qualitative Analysis of the Pro-
gram for all-inclusive Care for the Elderly. Gerontologist 32:771–80.
Kane, R., R. Kane, N. Haye, et al. 1996. Managed Care Handbook for the
Aging Network. Minneapolis: University of Minnesota, National Long-
Term Care Resource Center.
Kassner, E. 1995. Long-Term Care: Measuring the Impact of a Medicaid Cap.

Washington, D.C.:AARP Public Policy Institute.
Kerr, M., W. Fraser, and D. Felce. 1996. Primary Health Care for People
with a Learning Disability. British Journal of Learning Disabilities
24(1):2–8.
Klein, R. 1995. Big Bang Health Care Reform—Does it Work?: The
Case of Britain’s 1991 National Health Service Reforms. Milbank
Quarterly 73:299–337.
———. 1979. Control, Participation, and the British National Health
Service. Milbank Quarterly, 57:70–94.
Kramer, A.M., P.D. Fox, and N. Morgenstern. 1992. Geriatric Care
Approaches in Health Maintenance Organizations. Journal of the
American Geriatrics Association, 40:1055–67.
Lancet. 1994. Consuming Health Care. 344:763–65 (editorial).
Leutz, W. 1994. Integrated Systems of Care for Persons with Developmental
Disabilities and Mental Retardation: Not Easily Said and Not Easily
Done. Waltham, Mass: Brandeis University, Institute for Health
Policy.
———. 1999. Home Care Benefits for Persons with Disabilities. Amer-
ican Rehabilitation (forthcoming).
Leutz, W., R. Abrahams, M. Greenlick, R. Kane, and J. Prottas. 1988.
Targeting Expanded Care to the Aged: Early Social HMO Experi-
ence. Gerontologist 28:4–17.
Leutz, W., S. Altman, J. Capitman, et al. 1992. Final Report to HCFA on
Expansion Design for the Social HMO Demonstration. Waltham, Mass:
Brandeis University, Institute for Health Policy.
Leutz, W., J. Capitman, M. MacAdam, and R. Abrahams 1992. Care for
Frail Elders: Developing Community Solutions. New York: Auburn
House.
Leutz, W., J.N. Greenberg, R. Abrahams, J. Prottas, L.M. Diamond, and
L.Gruenberg. 1985. Changing Health Care for an Aging Society: Plan-
ning for the Social Health Maintenance Organization. Lexington, Mass.:
Lexington Books.
Leutz, W., M. Greenlick, and J. Capitman. 1994. Integrating Acute and
Long-Term Care. Health Affairs (fall):58–74.
Leutz, W., M. Greenlick, S. Ervin, E. Feldman, and J. Malone. 1991.
Adding Long-Term Care to Medicare in HMOs: Four Years of
Social HMO Experience. Journal of Aging and Social Policy 4:69–
88.
Leutz, W., M. Greenlick, S. Ervin, E. Feldman, and J. Ripley. 1995.
Medical Services in Social HMOs: A Reply to Harrington et al.
Gerontologist 35:6–8.
Leutz, W., and D. Hallfors. 1993. Lessons from Social HMO Marketing.
Waltham, Mass: Brandeis University, Institute for Health Policy.
108 Walter N. Leutz
Leutz, W., E. Sadowsky, and S. Pendleton. 1992. Cost Caps in State

Long-Term Care Programs. Waltham, Mass: Brandeis University, In-
stitute for Health Policy.
Leutz, W., M. Sciegaj, J. Capitman, and M.E. Henry. 1995. Service Plan-
ning Guidelines in Community Long-Term Care Programs. Waltham,
Mass: Brandeis University, National Resource Center for Diversity
and Long-Term Care.
Light, D.W. 1994. Strategic Challenges in Joint Commissioning. London:
National Health Service, NorthWest Thames Regional Health Au-
thority.
———. 1997a. From Managed Competition to Managed Cooperation:
Theory and Lessons from the British Experience. Milbank Quarterly
75:297–341.
———. 1997b. The Rhetoric and Realities of Community Health Care:
The Limits of Countervailing Powers to Meet the Health Care Needs
of the Twenty-First Century. Journal of Health Politics, Policy and
Law 22:105–45.
Liu, K., J. Hanson, and T. Coughlin. 1995. Characteristics and Out-
comes of Persons Screened into Connecticut’s 2176 Program. In
Persons with Disabilities: Issues in Health Care Financing and Service
Delivery, eds. J.M. Wiener, S.B. Clauser, and D.L. Kennell, 215–44.
Washington, D.C.: Brookings Institution.
Mays, N. 1997a. The Future of Locality Commissioning. British Medical
Journal, 314: 1212–13.
———. 1997b. Total Purchasing: A Profile of National Pilot Projects.
London: Kings Fund.
McCall, N., and J. Korb. 1994. Combining Acute and LTC in a Capitate
Medicaid Program: The Arizona LTC System. San Francisco, Calif.:
Laguna Research Associates.
McGrother, C., A. Hauck, S. Bhaumik, C. Thorp, and N. Taub. 1996.
Community Care for Adults with Learning Disability and Their
Carers: Needs and Outcomes from the Leicestershire Register. Jour-
nal of Intellectual Disability Research 40:183–90.
Minnesota Department of Human Services. 1997. Minnesota Senior Health
Options Program Summary. St. Paul, Minn.
Morris, J. 1995. The Power to Change: Commissioning Health and Social
Services with Disabled People. London: Kings Fund.
Myles, S., J. Popay, S. Wyke, J. Scott, A. Campbell, and J. Girling. 1997.
National Evaluation of Total Purchasing Pilot Schemes: Community and
Continuing Care for People with Complex Needs. London: Department
of Health.
National Health Service Management Executive. 1993. Guidance on the
Extension of the NCHS Element of the GP Fundholding Scheme for 1
April 1993. London: Health Care Directorate.
Nonnenkamp, L. 1996. Testimony to the Ways and Means Committee,

Subcommittee on Health. Washington, D.C.:U.S. Congress.
Pepper Commission. 1990. A Call for Action. Washington, D.C.: U.S.
Bipartisan Commission on Comprehensive Health Care.
Phillips, B.R., P. Kemper, and R.A. Applebaum. 1988. The Evaluation
of the National Long-Term Care Demonstration: 4. Case Manage-
ment under Channeling. Health Services Research 23:67–82.
Plant, M.. 1997. The Provision of Primary Health Care for Adults Who
Have Learning Disabilities: Summary. British Journal of Developmen-
tal Disabilities 43(84):75–8.
Prosser, H. 1997. The Future Care Plans of Older Adults with Intel-
lectual Disabilities Living at Home with Family Carers. Journal of
Applied Research in Intellectual Disabilities 10:15–32.
Rummery, K., and C. Glendinning. 1997. Working Together—Bringing
Primary Care into the Commissioning of Health and Social Care Services.
Manchester, U.K.: University of Manchester.
Ruta, D.A., C. Donaldson, and I. Gilray. 1996. Economics, Public Health
and Health Care Purchasing: The Tayside Experience of Programme
Budgeting and Marginal Analysis. Journal of Health Services Research
and Policy 1:185–93.
Schlesinger, M. 1997. Countervailing Agency: A Strategy of Princi-
paled Regulation Under Managed Competition. Milbank Quarterly
75:35–57.
Schulz, R., and S. Harrison. 1984. Consensus Management in the Brit-
ish National Health Service: Implications for the United States.
Milbank Quarterly, 62:657–81.
Small, N. 1989. Politics and Planning in the National Health Service. Phil-
adelphia: Open University Press.
Smith, R. 1997. The Future of Health Care Systems. British Medical
Journal 314:1495–6.
Social Services Inspectorate. 1994. Implementing Caring for People: Com-
munity Care for People with Learning Disabilities. London: Department
of Health.
———. 1995a. Growing Up and Moving On: Report of an SSI Project on
Transition Services for Disabled Young People. London: Department of
Health.
———. 1995b. Implementing Caring for People. London: Department of
Health.
———. 1995c. Searching for Service: An Inspection of Service Responses Made
to the Needs of Disabled Young Adults and their Carers. London: De-
partment of Health.
———. 1996. Progress in Services with Physically Disabled People. London:
Department of Health.
110 Walter N. Leutz
Thornton, C., S. Dunstan, and P. Kemper. 1988. The Effect of Chan-

neling on Health and Long-Term Care Costs. Health Services Research
23(1):131–42.
Tyne, A. 1994. Taking Responsibility and Giving Power. Disability and
Society 9:249–54.
U.S. General Accounting Office. 1996. Medicaid Managed Care: Serving
the Disabled Challenges State Programs. GAO/HEHS-96-136. Wash-
ington, D.C.
Valk, H.M.F. van Schrojenstein Lantman-de, M. van den Akker, M.A.
Maskaant, et al. 1997. Prevalence and Incidence of Health Prob-
lems in People with Intellectual Disability. Journal of Intellectual
Disability Research 41(1):42–51.
Vladeck, B.C. 1995. Overall Perspectives. Persons with Disabilities: Issues
in Health Care Financing and Service Delivery, eds. J. M. Wiener, S. B.
Clauser, and D. L. Kennell, 19–28. Washington, D.C.:Brookings
Institution.
Waddington, P. 1995. Joint Commissioning of Services for People with
Learning Disabilities: a Review of the Principles and the Practice.
British Journal of Learning Disabilities 23:2–10.
Weiner, J., and D. Ferriss. 1990. GP Budget Holding in the UK: Lessons
from America. Report no. 7. London: King’s Fund Institute.
Weissert, W.G., C.M. Cready, and J.E. Pawelak. 1988. The Past and
Future of Home and Community-Based Long-term Care. Milbank
Quarterly 66:309–88.
Welch, P. 1996. Growth in HMO Share of the Medicare Market, 1989–
1994. Health Affairs 15(3):201–14.
Wiener, J., and J. Skaggs. 1995. Current Approaches to Integrating Acute
and LTC Financing and Services. Washington, D.C.: AARP Public
Policy Institute.
Wilkin, D., T. Butler, and A. Coulter. 1997. New Models of Primary Care:
Developing the Future. National Primary Care Research and Devel-
opment Centre. Manchester, U.K.: University of Manchester.
Acknowledgment: This paper was supported by a grant from the Fulbright Com-
mission and was completed with assistance from the University of Kent at
Canterbury, where the research and writing were done. Special thanks are given
to Professor Bleddyn Davies, director of the Personal Social Services Research
Unit at Kent University, for his hospitality and for many thoughtful dialogues
on the issues covered here.
Address correspondence to: Walter N. Leutz, PhD, Associate Professor, Institute for
Health Policy, Heller Graduate School, Brandeis University, 415 South Street,
MS035, Waltham, MA 02254 (e-mail: Leutz@Brandeis.edu).

Milq 125

Uploaded by

Copyright:

Available Formats

Milq 125

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Milq 125

Uploaded by

Copyright:

Available Formats

Five Laws for Integrating Medical and Social

Services: Lessons from the United States and

T he integration of acute and long-term care

The Milbank Quarterly, Vol. 77, No. 1, 1999

1. demonstrating which approaches have worked and which have not

Although there is clearly much left to learn, innovators should build on

Health Care Systems and Integration in the

UK Systems and Initiatives

• The Community Care Act (1990): strengthened care management

• Total Purchasing Pilots (TPPs) (1995): expanded the range of health

In December, 1997, the Blair government released a blueprint for

US Systems and Initiatives

• Medicare HMOs: These organizations receive a capitation for ben-

both state workers and Medicaid beneficiaries to join contracted

Initiatives designed to consolidate long-term care include:

• Gatekeeping systems: A single agency—sometimes public and

• Social HMOs: These demonstration MCOs were launched in 1985

Examples of carve-out models include the following demonstration

• On Lok/Program for All-inclusive Care for the Elderly (PACE)

This brief summary of the US and UK systems and integration ini-

Maskaant, et al. 1997). Personnel from nonmedical systems also estab-

Operations Linkage Coordination Full integration

teams set up to facilitate deinstitutionalization of dually diagnosed pa-

and supports for vocational, educational (Social Services Inspectorate

2. Integration costs before it pays.

to £2.1 billion between 1989–90 and 1993–94 (Ham 1995). Admin-

onstration funds to the program, Channeling, and its predecessors (Weis-

3. Your integration is my fragmentation.

several hundred physicians to handle the new program features because

4. You can’t integrate a square peg and a round hole.

developmental focus of early childhood educators working with chil-

Third, long-term-care benefits, particularly for home-based and

community on application, the Medicaid payment rates are based on

1. by focusing on either private or welfare populations

UK integration models, which spurn strong financial and enrollment

5. The one who integrates calls the tune.

supported by medical directors with a geriatric orientation. The in-

A key to the success of joint commissioning is to recruit commissioners

sionally take the lead in joint commissioning (Mays 1997a), Greig’s

Lessons and Recommendations

The laws of integration perhaps have no scientific justification. Social

1. Involve service users, carers, and community service providers in

2. Develop systems to integrate, coordinate, and link services for

most strongly toward cooperation and understanding rather than dom-

3. Clarify borders between medical and other systems.

its Interregional Committee on Aging has adopted the following vision

Department of Health. 1997. The New NHS: Modern, Dependable. Lon-

Kassner, E. 1995. Long-Term Care: Measuring the Impact of a Medicaid Cap.

Leutz, W., E. Sadowsky, and S. Pendleton. 1992. Cost Caps in State

Nonnenkamp, L. 1996. Testimony to the Ways and Means Committee,

Thornton, C., S. Dunstan, and P. Kemper. 1988. The Effect of Chan-

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.