| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||
a Associate Professor, University of California at Los Angeles, School of Public Health and Borun Scholar, Borun Center for Gerontological Research, Los Angeles, CA
b SCAN Health Plan, Los Angeles, CA
c Borun Center for Gerontological Research, UCLA School of Medicine and UCLA School of Public Health, Reseda, CA
d University of Minnesota Institute for Health Services Research, Minneapolis, MN
e Formerly Jewish Home for the Aging, Reseda, California, and Borun Center for Gerontological Research. Currently Wesley Woods Center of Emory University and Atlanta V.A. Rehabilitation Research and Development Center, Atlanta, GA
Correspondence: Steven P. Wallace, PhD, Associate Professor, University of California at Los Angeles, School of Public Health and Borun Scholar, Borun Center for Gerontological Research, UCLA Box 951772, Los Angeles, CA 90095-1772. E-mail: swallace{at}ucla.edu.
Decision Editor: Vernon L. Greene, PhD
 |
Abstract |
|---|
 TOP Abstract MethodsResultsDiscussionReferences |
|---|
Key Words: Long-term care • Managed care • CCRC • Assisted living • Organizational change
The rapid growth of managed care in health services is a classic example of a changing environment that is forcing a number of different organizations to change in order to survive. The implications of managed care for hospitals, doctors, and other acute care providers and their patients are regularly reported in both the academic and popular press (Brodie, Brady, and Altman 1998
; Managed Care Evolution 1997). In contrast, the consequences of managed care for long-term care has received little attention except for studies of innovative programs such as Programs for the All-inclusive Care of the Elderly (PACE) and Social health maintenance organizations (S/HMOs; e.g., Branch, Coulam, and Zimmerman 1995; Harrington, Lynch, and Newcomer 1993). The growth of managed care is substantially changing the financing and referral environment for long-term care providers. The long-term care system developed under a mixture of free market incentives and regulatory constraints. Although fee-for-service reimbursements gave long-term care organizations flexibility in providing services, government regulations placed limits on the industry through certificate-of-need requirements, licensing and certification standards, and, since the late 1970s, restrictive public reimbursements. The end result is that most long-term care facilities have operated historically under a tradition of reaction rather than innovation (Kane 1996).
Organization theory points to both internal and external sources of change in complex institutions such as long-term care. Resource dependence theory views organizations as actively seeking opportunities for growth and avoiding threats to their survival. Changes in environmental resources are particularly important causes of organizational changes, according to this perspective (Scott 1998). A complementary perspective is the institutional model that emphasizes cultural controls both internal and external to the organization (Powell and Friedkin 1987). The growth of managed care provides an environmental challenge to long-term care institutions, whereas the existing framework for long-term care and the organizational cultures of its providers provide boundaries on what long-term care organizations may consider viable and legitimate changes.
Managed care has been growing rapidly in health care, although slowest among the Medicare population. Nevertheless, 6.5 million (over 16%) of all Medicare beneficiaries were in managed care plans nationally in 1998. In some counties, the Medicare managed care penetration is substantially higher. Five Oregon and two California counties have over 50% of their Medicare populations enrolled in managed care, and many more counties have over 40% enrolled (U.S. Health Care Financing Administration 1998). At the same time, hospitals are shifting increasing amounts of care to other types of facilities because of the reimbursement pressures of Medicare Diagnostic Related Group prospective reimbursements and/or competitively priced managed care contracts (DellaPenna, 1999). This is causing significant growth in subacute care, in which long-term care facilities are providing post–hospital care at increasingly acute levels (Lewin-VHI 1995), creating savings for hospitals and a new market for skilled nursing facilities (SNFs) to capture Medicare days.
Skilled nursing facilities are finding that an increasing number of their potential patients belong to HMOs that now actively manage their enrollees' Medicare SNF days. Multilevel facilities (MLFs) that provide skilled nursing as well as less intense forms of care such as assisted living and/or independent living have similar incentives as SNFs to capture Medicare SNF days. In addition, MLFs face challenges in coordinating acute care services for residents in their unskilled levels of care. This is likely to be especially true in those MLFs such as continuing care retirement communities (CCRCs), which have a relatively full range of acute medical services on-site (Golant 1992; U.S. General Accounting Office [GAO], 1997a). In addition, a number of states provide Medicaid waiver funds for care in assisted living settings as a way to reduce nursing home expenditures (U.S. GAO 1997b), providing further external pressure when Medicaid recipients move into managed care. The changing organizational environment brought about by the growth of Medicare managed care is therefore likely to be particularly relevant to MLFs.
Managed care companies generally prefer to use long-term care providers that have facilities in multiple locations, enabling the provider to offer services to a higher proportion of the HMO's patients. HMOs are also more interested in long-term care providers that have data systems that supply data on patient costs and outcomes, that can work cooperatively with the plan in patient management, and that provide consistent quality care (Buss 1994; Grant 1995). Many of these external pressures on MLFs are recent and require new and different organizational structures and technologies. On the other hand, many of the multilevel facilities that include full long-term care services for a single price are already experienced in managing risk across a continuum of care (U.S. GAO 1997a). This could serve to reduce the problems multilevel facilities face in contracting with managed care organizations (MCOs).
MLFs may be attracted to using the services of MCOs because managed care offers costs that are predictable and that are often cheaper than fee-for-service alternatives. Some LTC organizations, for example, have contracted out their pharmacy services to managed care companies that can both negotiate better volume discounts for medications and identify lower cost alternatives for commonly used drugs (Schulman, Rubenstein, Abernethy, Seils, and Sulmasy 1996). In addition, life care communities and other MLFs that provide acute medical services can limit their exposure to potentially high costs by covering residents under managed care contracts (Mollica 1998).
Given the changing health care environment, it is important to know how multilevel long-term care facilities are responding to the growth of managed care. The following analysis provides an early look at the field from a national survey. We examine both the external resource pressures and institutional ability to respond to those pressures, as well as the institutional cultural congruence, in predicting MLF contracting with MCOs.
|
Methods |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
). A random 50% sample was drawn from the remaining states. Respondents represented 43 states, with the largest number of respondents coming from Pennsylvania 
. In our bivariate analysis, we weighted the data to correct for the disproportionate sampling. After two mailings, a total of 605 facilities returned questionnaires (47%). Of the 605, 492 (81% of respondents) were usable, with the remainder indicating they did not meet the selection criteria (skilled nursing plus one unskilled level) or not completing the survey sufficiently. Respondents were almost all upper level management in organizational leadership positions, including the facility's president/CEO (11%), director/executive director (21%), administrator (61%), or "other," including a variety of vice president titles (7%). Using data from the directories, we compared the number of skilled nursing beds for respondents with a 10% sample of nonrespondents. Respondents reported a mean of 106 nursing-level beds, whereas nonrespondents had a mean of 95 beds, which is not statistically different (p > .05). Because the focus of the survey was on managed care, there may have been a higher nonresponse from organizations with no managed care experience. Managed care experience is a dependent variable in our analyses, reducing the likelihood that this would bias our results.
Our key dependent variable is constructed from several questions. First, we asked whether facilities had current contracts with managed care organizations (MCOs); 83% of those with contracts reported that subacute care and/or skilled nursing home care was one element of the contracts. Second, we asked what actions the facility is likely to take in the following 2 years with respect to managed care. On the basis of 9 questions offering different possible actions, we constructed a hierarchal variable of future managed care involvement:
(a) Planning contracts (if they answer "very likely" (rather than "somewhat" or "not likely") to one or more of 3 questions indicating intent to contract with managed care organizations);
(b) planning (if they did not answer "very likely" to the contracting questions but answered "very likely" to one or more of 4 questions involving information gathering or planning for managed care); and
(c) No action (if they did not answer "very likely" to any contracting or planning questions but did answer "very likely" to one or more of 2 items indicating no action or a wait-and-see approach).
Our final variable is a four-category hierarchical variable of (a) has contracts, (b) does not have contracts but anticipates contracting in the next 2 years, (c) does not anticipate contracting but anticipates planning, and (d) no anticipated MCO actions.
Our independent variables are all dichotomous. We dichotomized the number of nursing home beds at 125 to indicate a large facility that has at least the mean number of beds of facilities with existing MCO contracts. Eight states with the largest increase in Medicare managed care enrollees between 1987–1995 (Arizona, California, Florida, New York, Oregon, Pennsylvania, Texas, Washington) are coded as high-growth states. These eight states also accounted for 85% of all Medicare HMO enrollees in 1995 (U.S. GAO 1996).
Respondents were also asked to write unstructured responses to a few questions. A total of 58% of respondents wrote responses to the question, "What effect, if any, do you think the growth of managed care will have on your facility in the future?" and 13% answered the question, "What factors have deterred you [in MCO contracting] or concern you for the future?" An additional 7% wrote comments under "Other goals of your involvement with an MCO." We have used representative quotes from these responses to illustrate the meaning of some of our quantitative responses.
We first explored the bivariate relationships between different levels of planned MCO involvement, then estimated a multinomial logistic regression to examine the independent predictors of MCO involvement. None of the independent variables correlated with other variables above 0.5, suggesting that multicollinearity was not a problem. Multinomial regression was our choice for the multivariate analysis because we had a four-level categorical outcome variable (Hosmer and Lemeshow 1989). This procedure presented some difficulty in interpretation because it provided a two-way comparison across the independent variables and each level of the dependent variable. The regression provided the odds of the independent variables being associated with each level of the outcome (dependent) variable in comparison to the reference level of the outcome. In our analysis, we used "currently contracting" as the reference level of the dependent variable. This provided three odds ratios for each independent variable, corresponding to the odds of "no MCO actions" versus "current contact," "information only" versus "current contract," and "plan to contract" versus "current contract." We chose "currently contracting" as the reference variable because of its substantially larger sample size than "no MCO actions" and because that variable level maximized the contrasts with the other dependent variable levels. The advantage of this procedure over several dichotomous logistic regressions was that all levels of the dependent variable were simultaneously controlled. We calculated the variance explained (R2) using the method recommended by SAS (SAS Institute, Inc 1998).
|
Results |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
|
). Although our findings are therefore not generalizable beyond the nonprofit sector, the environmental and organizational structure forces that are driving MLF–MCO agreements may be similar across profit status.
In our sample, residents' MLFs were quite old: 48% of the average facility's residents were older than age 85. Facilities reported a mean of 39% of their nursing care level residents on Medicaid (Table 1 ), a much lower rate than the 67% of residents with Medicaid in all nursing homes nationally (Krauss et al. 1997). A substantial proportion of our sample (19.2%) reported no Medicaid residents in their nursing-level care, whereas 55% of those with Medicaid residents reported that those residents comprised half or more of their nursing-level census.
A modest proportion (29.9%) of multilevel long-term care facilities in our national sample reported having any type of contractual agreements with managed care. The most common type of agreement (61% of those with contracts) was a per-diem contract in which MCOs contracted for SNF beds. Few contracts (13%) involved full or partial capitation. Thus, most multilevel facilities contracted in rather traditional ways. To better understand what distinguished those facilities with current agreements from those without contracts but anticipating them, only planning, or doing nothing related to MCOs, we next examined a set of environmental, organizational structure, and organizational culture factors that the literature has suggested might influence the likelihood that organizations will change in response to environmental changes such as those presented by MCOs.
Environment
The managed care environment influences whether or not facilities have MCO contracts. Although the Medicare HMO market is relatively small, it continues to grow rapidly and has been particularly concentrated in a small number of states. Among our (weighted) respondents with formal MCO agreements in 1996, 62.6% were located in states with high Medicare growth rates, compared with only 14.3% of respondents that reported no MCO activities being in high MCO growth states (Table 2 ). The impact of managed care on the long-term care industry nationally is indicated by only 11.6% of all respondents, reporting that they intend to "wait and see" or do "nothing" in the near future regarding MCOs. Even facilities in low-HMO concentration states are looking to the future and gathering information on MCOs, many with trepidation. One respondent noted, "Managed care is still a little overwhelming to me and I'm not sure how we'll fit into the picture. Just hope I can keep up with the changes" (Respondent #293).
|
As a CCRC, we have had a form of managed care since 1965 because the physicians on staff actually manage the residents' care... [Our] plans are to contract with a MCO and continue managing the care of our residents as well as opening to all not-for-profit facilities in [the state] to increase the base. (#772)
The expansion of MCOs in a market can impair the ability of facilities to fill their nursing-level beds. Respondents with existing or planned MCO contracts were significantly more likely than others to cite maintaining or increasing their census as a goal of MCO contracts (Table 2 ). One respondent reported, "Currently our census is 100%. With the growth of managed care, this may reduce our market area, causing our census to drop" (419). Another elaborated:
Currently the majority of our residents have retained their Medicare benefits. We are seeing that change. We have a 50-bed SNF with much competition from other facilities in the surrounding community. We believe that we will need to establish managed care contracts in order to improve census and maintain a presence in the community. (#52)
Despite the concern over occupancy rates, our respondents reported high occupancy rates (averaging over 90%) across all levels of MCO involvement, suggesting that they perceived MCOs as potentially reducing rather than enhancing their resources.
In addition to external market forces that encourage MLFs to contract with MCOs to attract new residents, some MLFs were under pressure from existing residents who were also HMO members and who wanted to remain in the multilevel facility for as much of their care as possible. This reason was cited by 65% of facilities with existing MCO contracts, whereas only 26% of facilities reporting no MCO activities stated that such resident desires would encourage MCO relationships (Table 2 ). One respondent explained how this pressure works:
We have a closed system CCRC (Type A). We only take our own residents. So far, they have returned to us for care. However, MCOs are putting on a large campaign to enroll members. Our hope would be to negotiate with a MCO a package we could market to our residents. (#908)
Another respondent detailed a situation in which having MCO contracts would be important. "It [MCOs] may prevent some of our residential-level residents from having their subacute care covered in our SNF. They could be forced to go to another SNF. This is particularly difficult for long-time residents or residential couples where one may need subacute rehabilitation after a surgery and be unable to receive that at the facilities where they live" (#92).
Organizational Structure
A number of organizational characteristics could facilitate MLFs contracts with MCOs. Some organizational structures would be better than others at taking advantage of changing external resources. Facilities with distinct-part units are more likely to accept Medicare patients directly from hospitals and to have shorter and more expensive lengths of stay. The shorter stays increase the facility's need to market its services to a range of outside hospitals. Almost half of the respondents with MCO agreements reported having a distinct-part unit, compared with just over one quarter of those reporting no MCO activities (Table 2 ).
The presence of MCO contracts for subacute care had a number of effects on respondents. Many reported increasing levels of acuity in their skilled units and a shift of some care down to unreimbursed levels of care.
Managed care has already increased the acuity level in our SNF. We are seeing and will see a greater reduction in our Medicare revenue stream. Our managed care contract pays a good per-diem rate but the length of SNF coverage is about 50% of what Medicare allows. Since we are a full life-care provider, this is a problem. We are also seeing reduced utilization of therapies for managed care patients, which we believe is penny wise and pound foolish. (#39)
Accepting any Medicaid patients in the skilled facility is also associated with more current and planned involvement with managed care organizations (Table 2 ). Although relatively few states had Medicaid nursing home residents in mandatory managed care during the year of our survey, it is likely that facilities that have the organizational resources for Medicaid contracting standards are also more likely to be capable of undertaking MCO contracts. In addition, discussions in many states about moving Medicaid long-term care into managed care has some organizations planning for the future:
I believe that unless we eventually become part of a managed care system, we will have a census problem, as the state may go to a Medicaid managed care system. We have been a traditional nursing home with full census and a long waiting list for many years. (#226)
There is no bivariate association, however, with the percentage of nursing care residents covered by Medicaid and MCO contracting.
Chain ownership could be viewed as either a cultural variable (because chains are more likely to have a degree of nonlocal control and management that is more sensitive to broad market trends) or a structural variable (because MCOs are more likely to prefer contracts with multifacility nursing homes that can cover more of their service area). A higher proportion of facilities with contacts (or planning contracts) are parts of chains than are facilities without contracting plans (Table 2 ). One nonchain facility noted the importance of chain characteristics:
Our facility, as an independent stand-alone organization, recognizes that being part of an alliance of other homes will be necessary in dealing with MCOs. Being part of a larger negotiating group, we feel, will allow us better opportunities for fair contract and rates agreements favoring our home. As an independent, we may not carry enough size to negotiate with strength. (#622)
Another nonchain facility noted the disadvantages they face by being unaffiliated and therefore not having the resources and expertise of a central office to draw on:
We as an independent, church-related, not-for-profit, skilled nursing home will have difficulty competing with the larger profit chains. It will increase our costs as we work toward JCAHO accreditation, which is required by most managed care contracts. (#1108)
We hypothesized that skilled units combined with assisted living units (vs. only independent living units) would be more oriented organizationally to a higher level of care than those without assisted living and would therefore be more likely to seek out medically oriented MCO contracts. At the bivariate level, however, there was no statistically significant relationship (Table 2 ). Similarly, we did not find an anticipated association between MFLs with on-site primary care and MCO contracting that might have resulted from MLF interests in improving expenses and revenues and making them more predictable.
Finally, the level of MCO involvement was significantly related to the number of nursing beds in the multilevel institutions. Institutions with existing MCO contracts averaged 124 nursing beds, whereas those that planned no MCO activities averaged 78 nursing beds (Table 2 ). This is consistent with (a) MCOs' interest in contracting with larger facilities and with chains that can handle a substantial flow of MCO patients and (b) the administrative resources that larger facilities possess to seek out and make MCO contracts.
Organizational Culture
In addition to the environmental context and organizational structure, the culture of the organization may encourage or discourage multilevel long-term care organizations from contracting with MCOs. An organization's culture involves the set of values, beliefs, and understandings that are generally held by members of the organization (Rousseau 1990). There has been extensive public discussion about how managed care's business culture and cost-containment ethic conflict with traditional fee-for-service providers' philosophy of providing care with little attention to total costs (e.g., Davis, Schoen, and Sandman 1996; Southwick 1998). In addition, most (69%) of Medicare HMO enrollees were in for-profit HMOs in 1996 (U.S. Health Care Financing Administration 1996), whereas almost all of our MLFs respondents were nonprofit. Organizational cultural congruence facilitates interorganizational relationships (Oliver 1990), so we would expect resistance from traditionally nonprofit, fee-for-service MLFs to contracting with at-risk HMOs.
Among our multilevel, long-term care respondents, those with existing MCO contracts were mostly likely to report being comfortable with MCO cultures. Even among those with existing contracts, however, only two fifths agreed that their facility's culture was compatible with MCOs. The less experience and knowledge that facilities had with MCOs, the less compatible they felt their organizational culture would be with MCOs (Table 2 ). This suggests a high level of discomfort among multilevel, long-term care facilities with the way that MCOs operate or are thought to operate.
The different missions between MCOs and multilevel LTC organizations is summarized by this respondent from a high HMO penetration state.
Our major concern is that our historic 98-year Christian mission (and we live in the mission) will be lost by these MCOs who are concerned only for the bottom line. They may talk quality, consistency of care, focus on communities, etc., but it has not been proven. (#615)
Differing perspectives on what long-term care offers and how to care for patients can also cause conflict:
We have had a capitated contract with [HMO 1]—it was a mess! We lost money, $80,000. Residents are in and out in about 7 to 10 days. People who think their spouses will be here at least 20 days are shocked. The public doesn't know what it is buying. (#18)
Some facilities are also having to shift toward an orientation more suitable for higher acuity nursing home patients with shorter stays: "Our facility is undergoing rapid changes, education and awareness to managed care and subacute. Over the past few years, this had not been the culture" (#1252).
For-profit MLFs, although a small proportion of our respondents, would be expected to have a more business-oriented culture and be more oriented than nonprofits to these new marketing opportunities. MLFs that reported already having MCO agreements were significantly less likely to be nonprofit (Table 2 ). For-profit facilities were also significantly more likely to report that their organizational cultures were compatible with MCOs than were nonprofit facilities (48.7% vs. 29.6%, 
2(3, N = 492) = 8.4, p < .05. Unfortunately, the small number of for-profit facilities in our sample makes it impossible to use this variable in analyses beyond this bivariate level.
A final organizational culture issue is the extent to which respondents were concerned about potential quality-of-care problems resulting from MCO contracts. A majority of all respondents reported that quality-of-care concerns were a deterrent to their entering into contracts with MCOs. There was no statistically different rate between those with different levels of MCO experience (Table 2 ). The concern was expressed by facilities in high HMO penetration states that have contracts: "The issue that most concerns us revolves around the quality of care. We believe MCOs do not provide as high a quality of care as we have previously provided" (#52). The concern was also expressed by facilities in states with low HMO penetration and limited MCO experience: "Our nursing home is considered one of the top homes in the area, and I will resist any attempts by an MCO to reduce the quality or quantity of services to get the cost down" (#459), and "Our excellent care may have to be lowered to stay competitive and keep the facility full" (#441).
A few respondents saw culturally valued effects coming from MCO pressures to contain costs. Some viewed MCOs as delaying or reducing the need for nursing home care. One respondent with existing MCO contracts predicted that MCOs "will affect outreach and residential services in a positive way, as we believe that wellness programs will be funded in a capitated market place" (#943). Another respondent thought that MCOs would provide the "opportunity for funding of community-based services to avoid more costly services" (#1047). And a few others believed that their quality would be rewarded under managed care: "[The effects of MCO are an] ability to be rewarded for our positive outcomes... [and] [a]llow the best providers to become even better" (#860).
Multivariate analysis
We next estimated a multinomial logistic regression that entered all of the variables that were significant in the bivariate analysis, except for the for-profit variable because of the small sample size. A multinomial logistic regression compares the odds of each of the predictor variables on each of the outcome levels of the dependent variable in comparison to the reference level. Because of the distribution across the four levels of our dependent variable, MCO involvement, we used "has formal MCO agreement" as the reference group. Thus, we were attempting to predict how far away the organization was from MCO contracts. To make the resulting odds ratios more intuitive, we reversed the wording of the independent variables to make the odds greater than one.
The strongest effect was state-level Medicare HMO growth. A multilevel facility in a low HMO growth state was about 3.6 times more likely to have "no MCO actions" as to "having a current MCO agreement" or contract, controlling for the other independent variables and controlling for other levels of the dependent variable (Table 3 ). Being in a low HMO growth state increased the odds of "information only" and "planning to contract" versus the reference level of "current agreement" by somewhat over and just under two times, respectively (Table 3 ). This shows that being outside the high HMO growth states independently increases the odds of limited or no MCO involvement.
|
Among the facility structure variables, there was a mixed pattern of significant results. "No MCO actions" were predicted only by the absence of Medicaid residents. The odds of "information only" versus "currently contracting" was increased by having a smaller number of nursing beds and (at the p < 0.1 significance level) being independent (i.e., not part of a chain). The odds of "planning to contract" versus "currently contracting" were increased by having a smaller number of nursing beds and (at the p < 0.1 significance level) not having a Medicare distinct-part unit. Overall, it appears that the multilevel facility structure had a moderate independent influence on the extent of MCO contracting.
Finally, the single organizational culture variable was associated with "no MCO actions" and "information only" compared with "currently contracting." Across almost all of the independent variables, there was an ordinal-like effect, with the odds ratios shrinking as the extent of MCO involvement increased from "no MCO actions" towards "currently contracting" (the reference). In this model, the state HMO variable accounted for 22% of the variance, the other environmental variables an additional 13%, the organizational structure variables an additional 5%, and organizational culture 1% additional for a total R2 of 0.41.
|
Discussion |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
Structural characteristics of MLFs had a weaker but independent association with MCO contracting. Having a large number of skilled nursing beds, having a Medicare-certified distinct-part unit, having any Medicaid patients, and being part of a chain were all related in different ways to managed care involvement. Each of these facility characteristics were attractive to managed care organizations. MCOs were more likely to want to contract with long-term care organizations that can handle a significant number of patients in geographically dispersed locations. Facilities with distinct-part units were more likely to have the subacute beds many MCOs now use to reduce hospital stays. And Medicaid managed care is being extended to long-term care in some states.
These facility characteristics are also likely to provide the multilevel facility with the administrative and financial capacity necessary to enter into contracting arrangements. However, facility characteristics that might have led MLFs to approach MCOs as a new resource, such as providing primary care on-site, were not related to managed care involvement. Some of the influence of facility structure may have been reduced by unmeasured actions that affected the facilities' characteristics. A number of free-standing facilities (i.e., not in a chain), for example, reported that they were forming alliances that served as the structural equivalent of a chain for the purposes of MCO contracting, probably weakening the effect of the chain variable.
Finally, a general question about the compatibility of MCO and MLF cultures is an organizational culture variable that was a minor but independent predictor of MCO involvement. Open-ended responses showed that many of the mission-driven facilities were uncomfortable with the demands made by more financially oriented managed care companies. There was also a widespread concern over the quality of care they would be able to supply under the constraints of MCO contracts, although this concern was not associated with the extent of MCO involvement. The finding, that concerns about quality of care did not significantly change as knowledge and experience with managed care increased, suggests that this concern was not exaggerated by those without MCO experience.
Conclusion
The resource constraints that managed care organizations (MCOs) are perceived to impose or threaten were the strongest independent correlates of MCO contracts by MLFs. Both quantitative and open-ended questions indicated that MLFs were more likely to have or seek MCO contracts when the MLF management was feeling pressured by the external forces of current residents' desires and threats to a high occupancy rate and market share. MLFs should be in an ideal position to shift toward managed care because they already supply and/or coordinate health care that includes both chronic and acute care. Yet most MLFs appear to be reluctantly responding to the changing environment rather than trying to take advantage of new resources in that environment.
It may be that Medicare policies have forced this position on MLFs. With the exception of the PACE and S/HMO demonstrations, Medicare has not yet risk-adjusted capitation payments in a way that would account for the higher acuity of the average MLF resident. If Medicare payments to risk HMOs reflected the levels of disability of chronic care residents in the different levels of MLF care, those facilities might be encouraged to design care systems in collaboration with MLFs. Alternatively, the increasing consolidation of Medicare HMO plans in some regions of the country may encourage MLFs to seek out MCOs. When an MLF finds a high proportion of their entering residents in one or two HMO plans, it may be to the advantage of both the MLF and HMO to develop specific contractual arrangements to maximize the coordination of needed care for the resident.
Model programs exist that MLFs could follow to integrate managed care into their skilled nursing units. Several HMOs around the country have special programs for their members in nursing homes that increase their primary care oversight through the use of nurse practitioners and/or physician's assistants. The additional primary care, beyond what would be reimbursable under Medicare fee-for-service, can significantly reduce emergency room use, hospitalizations, and hospital days (Reuben et al. 1999). In a capitated system, the savings from a reduction of two hospital days per year more than pays for the additional primary care services. A range of additional interventions at both the SNF and residential levels have been suggested that could reduce hospital and other medical costs, more than compensating for the additional services (Kane 1998; Siu, Beers, and Morgenstern 1993). These innovations require new ways of organizing and financing some services but are not revolutionary.
In an organizational field in which resource dependence drives relationships with MCOs, resource changes can be either constraining or enriching. Current patterns of MLF–MCO contracting appear to be primarily the result of actual or anticipated constraints. The push by MCOs to reduce hospital days by shifting care into subacute units is one outcome that many respondents observed, but the resulting increased acuity of the MLF's skilled unit is driven primarily by the benefits for the MCO rather than for the benefit of the MLF residents. If MLFs want managed care to benefit their residents and institutions, more MLFs will need to identify where they can benefit from MCO resources and seek out MCO contracts that build, rather than simply sustain, MLF programs and assets.
|
Acknowledgments |
|---|
Received for publication October 16, 1998. Accepted for publication December 7, 1999.
|
References |
|---|
|
TOPAbstractMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
