Abstract
The long-term services system has changed substantially since the mid-1970s, when the landmark book Last Home for the Aged argued that the move to the nursing home was the last move an older person would make until death. Using detailed nursing home utilization data from the Minimum Data Set, this study tracks three cohorts of first-time nursing home admissions in Ohio from 1994 through 2014. Each cohort was followed for a 3-year period. Study results report dramatic reductions in nursing home length of stay between the 1994 and 2011 cohorts. Reduction in length of stay has important implications for nursing home practice and quality monitoring. The article argues that administrative and regulatory practices have not kept pace with the dramatic changes in how nursing homes are now being used in the long-term services system.
Introduction
In the 1970s landmark book Last Home for the Aged, the authors describe the transition to a nursing facility as the final life move for many older people (Tobin & Lieberman, 1976). In the four decades since that book was published, the nursing home industry in America has experienced a dramatic change in both who is served and the type of services provided. This trend was initially identified two decades ago in a study done by the Centers for Disease Control examining the utilization patterns of nursing home residents. Comparing the national profile of the newly admitted nursing home residents age 65 years and older, findings showed that the average length of stay had dropped from 89 days in 1985 to 45 days in 1997. Newly admitted residents were also more likely to be discharged to the community, growing from 18% in 1985% to 30% in 1997 (Sahyoun et al., 2001).
Similar patterns have been observed in state studies. For example, the Minnesota Department of Health in 2012 reported that most of older nursing home residents had stays lasting from 1 to 3 months (Toles, Young, & Ouslander, 2013). Although there were still some very long-stay residents, many admissions were for shorter stays. This is reflected in the difference between the average length of stay (248 days), as compared with the median, marking the point at which half of the residents leave the facility (28 days). In Tennessee, from 2003-2012, the number of admissions to state nursing homes increased by 38%, while the average length of stay dropped by 10% (Tennessee Department of Health, 2014). Although these previous studies show a picture of the changing landscape of the nursing home industry, this work, based on 20 years of nursing home length of stay patterns, provides an opportunity for a more in-depth look at these dramatic nursing home industry changes by looking at three cohorts in the same state over time.
These significant changes in how nursing home care is delivered have occurred over a relatively short period of time and have been driven by several potential factors including changes at the federal, state, and industry levels. At the federal level, over this time period, there have been a series of changes in payment structure for long-term and short-term care programs and changes in hospital regulations that impact both payment and utilization. At the state level, there have been changes in reimbursement rates, a stronger emphasis on reimbursement incentives, and a substantial increase in Medicare beneficiaries enrolled in managed care. At the industry level, there have been changes in how nursing homes utilize long- and short-term stay revenue streams and greater competition for long-term services from a growing home- and community-based services industry. Nursing facilities have shifted from relying mainly on long-term care payments of Medicaid to relying more heavily on short-term rehabilitation care payments covered by Medicare. All these changes suggest that how we fund, monitor, regulate, and deliver nursing home services have evolved in response to system changes.
Method
This study examines length of stay patterns for nursing homes in Ohio over three time periods beginning in 1994 and ending in 2014. For the most recent cohort, we used the nursing home Minimum Data Set (MDS 3.0), available for all nursing homes in the state that accept Medicaid and/or Medicare reimbursement (about 99.7% of all beds in the state). For the earlier time periods, we used MDS 2.0. Each time we built a longitudinal data set by tracking all first-time nursing home admissions, regardless of age, entering between July and September of a cohort year (1994, 2001, 2011 selected to capture changes pre- and post-1997 Balanced Budget Act) and following each admission for a 3-year period. The MDS 2.0 tool included whether the nursing facility admission was the first one for an individual, but there was no such item in MDS 3.0. To determine the very first admission for everyone in that cohort, we took all of the 38,291 individuals newly admitted between July and September 2011 and matched them with previous MDS data going back 5 years. If any previous admission was found, the person was excluded from the study. There were 23,475 individuals who did not have a prior admission in 2011. We describe the changes in the nursing home stay patterns of the three groups. When comparing results of a single quarter with full-year results for select years, the overall length of stay patterns were unchanged.
Results
Nursing Home Length of Stay
Nursing home use patterns over the past two decades show a transformed industry. As shown in Table 1, after 3 months, nearly 57% of the 1994 cohort remained in the nursing facility, and after 6 months, 41% were still in the nursing home. After 12 months, about one third were still residing in a nursing home, and 2 years later, nearly one out of every four was still living in a nursing home. By 2001, the shift toward shorter stays was clear. After 3 months, 43% of all those admitted continued to reside in the nursing home, and after 6 months, less than one third (32%) remained. At 9 months, one fifth of all of those admitted were still there. The proportion residing in a nursing home after 1, 2, and 3 years were 16%, 9%, and 6%, respectively. The 2011 length of stay findings were more dramatic. More than four in five persons admitted were no longer in the nursing facility within 3 months (16% remained compared with 57% remaining on the 1994 cohort). After 6 months, 12% of the newly admitted residents remained in the facility.
Newly Admitted Nursing Home Residents and Changes in Their Stay Pattern Over a 3-Year Period (1994-2014).
Source. MDS 3.0 2011-2014; MDS 2.0 2001-2004; MDS 1994-1996.
Note. MDS = Minimum Data Set.
The time after the Balanced Budget Act of 1997 has been identified as a period of high nursing home Medicare expansion (Dalton & Howard, 2002). However, the 10-year period between the 2001 and 2011 cohorts showed a dramatically accelerated move toward shorter stays, with the proportion of residents staying 3 months or less decreasing from 43% in the 2001 sample to 16% in the 2011 group. Even Medicaid, the payment source thought of as a long-stay payer, showed three quarters of all admissions discharged in the first 3 months. It is also important to note that the sheer volume of first-time admissions for the 3-month period for each cohort has increased dramatically, growing from 5,800 in 1994 to 24,000 in 2011.
Profile of First-Time Nursing Home Residents Over Time
To gain a better understanding about how these changes impact the long-term services system, we examine the characteristics of those admitted across the three cohort time periods. As nursing homes have become a short-term stop for many between hospital and home, it is not surprising that the characteristics of those admitted have changed. The altered profile includes individuals who are younger, less likely to be female, and more likely to be married. For example, the average age of the new residents in the three cohorts dropped from 80 to 75; the percent female dropped from 66 to 60; and the proportion married increased from 27% to 37%. The average level of impairment as measured by the ability to perform activities of daily living remained constant (see Table 2).
Newly Admitted Nursing Home Residents Demographic and Functional Characteristics (1994, 2001, and 2011).
Source. MDS 3.0 2011; MDS 2.0 2001; MDS 1994.
Note. ADL = activities of daily living; MDS = Minimum Data Set.
Admission Pattern by Source of Payer
With 84% of the 2011 cohort leaving the nursing facility within the first 3 months, we take an in-depth look at these individuals to better understand how the system has changed. In 2011, 64% of all first-time admissions (14,977) were paid for by Medicare. One in five admissions were for individuals supported by Medicaid (4,844 persons) and 16% (3,654 persons) self/family pay, long-term care insurance, health care insurance, Veterans Administration (VA), Civilian Health and Medical Program of the Uniformed Services (CHAMPUS), and other sources together. This contrasts with the 1994 cohort, which included 45% of admissions supported by Medicare, and the 2001 group in which 61% of those admitted were paid for by Medicare.
Discussion
Our data show that it is clearly no longer your parent’s nursing home. Nursing homes in the United States have transitioned from primarily long-stay institutions, with a small amount of short-term care, to facilities that while continuing to provide long-stay care, serve a very high number of short-stayers. It is also clear that most of these short-stayers are funded by Medicare. In 1992, Ohio had 30,000 Medicare admissions, and by 2015, that number had increased to 147,000 (includes all admissions). The accompanying changes in length of stay documented in this article indicate that for the majority of those admitted, nursing home care is short-term rehabilitation care for individuals on their way home. We also see that nursing home admissions can be a much more commonly occurring event, with some individuals transitioning between home, hospital, and nursing home with frequency.
While the data used in this study are comprehensive to the state of Ohio, they are limited to one state. As state-level changes in nursing home regulations and Medicaid regulations may vary from state to state, any generalization of the findings to other geographic locations should be made with caution. In addition, we focused on new nursing home admissions and excluded individuals with repeated admissions. This was done to help focus the data on new admissions and avoid any inflation of admission numbers caused by individuals who may frequently bounce back and forth between the nursing facility and hospital. This, however, may limit the generalizability of the findings to the entire nursing home population.
This dramatic shift in the nursing home industry is the product of an array of policy, program, and practice changes some designed and others not. Length of stay patterns are driven both by changes in medical practices, hospital interventions, outpatient therapy, and home health services, and state and federal policies that incentivize and encourage faster discharges from the hospital and ultimately the nursing home setting. For example, the expansion of home- and community-based services, the development of the new assisted living industry, and the effort to control the supply of nursing home beds represent policy changes designed to create better balance in the long-term services system. However, the changes experienced by nursing homes because of Medicare prospective payment reimbursement to hospitals were largely unanticipated by the nursing home industry. In total, the array of changes experienced over the last decades means that the nursing home industry is appreciably different than the earlier version.
However, it is not clear that the federal and state policies that guide and regulate nursing homes have accompanied these industry changes. This suggests that how we fund, monitor, regulate, and deliver nursing home services will need to continue to evolve in response to these considerable changes. For example, in response to concerns that some individuals were using nursing home care prematurely and inappropriately, many states implemented a nursing home pre-admission screening process. These typically involved an in-person or at least paper review of the nursing home applicant prior to admission. In Ohio, with 215,000 annual admissions, even a record review at admission that costs US$50 starts to add up and the in-person assessments could be several hundred dollars. When nursing home care was the default recommendation by professionals and there was a limited supply and coverage for home- and community-based services, this strategy made sense. But now that the clear majority of admissions stay for 3 months or less, this approach needs to be revised. It may make sense to use those resources to conduct a more comprehensive review for individuals who continue to be residents after 30 or 60 days as an example of an alternative approach. This has important cost implications as well, with 50% of those initially admitted under Medicare who become long-stayers, relying on Medicaid to fund their nursing home care within 1 year of admission.
The shift in use patterns has implications for other areas as well. For example, while some of the Centers for Medicare & Medicaid Services (CMS) star rating measures do attempt to recognize short- and long-stay residents, in general, the regulatory approach is still done from a long-term perspective. The state inspection survey, the staffing configurations, and even most of the quality indicators were developed within a long-stay context. Information available to consumers through such websites as Nursing Home Compare need to do a much better job of explaining the services provided in nursing facilities and how quality is monitored. Some states, like Ohio, have also been proactive in expanding consumer-based information by developing resident and family satisfaction measures, but again they are dominated by the long-stay data.
Another primary example of the lag between the change in practice and policy involves the Medicaid level of care review. In Ohio, Medicaid nursing home residents must meet state level of care eligibility criteria on admission, but once admitted they never receive any additional reviews. Individuals receiving in-home services through the Medicaid waiver program receive a re-determination assessment every 6 months. This issue has become much more important as even the Medicaid program is now using nursing homes for short-term rehabilitation care for the non-Medicare population. A sizable proportion of the Medicaid short-term admissions have behavioral health challenges and a review of their condition after admission is critical in many instances.
In addition to these major policy changes, practice and program modifications are also necessary. Nursing homes remain a long-stay option for a segment of the older population, particularly those living with dementia. It will be critical for providers and researchers to focus on innovative models of care to better serve the resident population of tomorrow’s facilities. Because nursing homes have been so heavily focused on shifting to the rehabilitation person, innovations for long-stay residents have not received as much attention in recent years.
These findings indicate that the nursing home industry of today is substantially different than even 10 years ago. With an array of individuals to serve, from short-term rehabilitation patients, to long-term older residents with dementia, to younger individuals with behavioral health needs, the nursing home of today has many goals and challenges. Figuring out what to emphasize and how to provide high quality services in these changing times will be a critical challenge for the industry as it prepares for the next generation of potential residents, patients, or whatever the boomers will call themselves in the last phases of life. Whatever the term, the nursing home and overall long-term services system of tomorrow will be so different that the funding, regulatory, and quality of care paradigms will need to be altered to keep up with the industry transformations that have and will continue to occur.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.