Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Mar 8.
Published in final edited form as: J Aging Soc Policy. 2018 Jan 8;30(2):93–108. doi: 10.1080/08959420.2017.1414538

Change in VA Community Living Centers 2004–2011: Shifting Long-Term Care to the Community

Kali S Thomas a,b, Danielle Cote a, Rajesh Makineni a,b, Orna Intrator c,d, Bruce Kinosian e,f, Ciaran S Phibbs g,h, Susan M Allen a,b
PMCID: PMC6407124  NIHMSID: NIHMS1006815  PMID: 29308990

Abstract

The United States Department of Veterans Affairs (VA) is facing pressures to rebalance its long-term care system. Using VA administrative data from 2004–2011, we describe changes in the VA’s nursing homes (called Community Living Centers [CLCs]) following enactment of directives intended to shift CLCs’ focus from providing long-term custodial care to short-term rehabilitative and post-acute care, with safe and timely discharge to the community. However, a concurrent VA hospice and palliative care expansion resulted in an increase in hospice stays, the most notable change in type of stay during this time period. Nevertheless, outcomes for Veterans with non-hospice short and long stays, such as successful discharge to the community, improved. We discuss the implications of our results for simultaneous implementation of two initiatives in VA CLCs.

Keywords: Veterans Health Administration, long-term care, Community Living Centers, Veterans

Introduction

The Veterans Health Administration (VHA) provides services to an increasingly aging population, especially Veterans older than 85, the age at which the need for long-term care (LTC) is greatest (Kinosian, Stallard, & Wieland, 2007). In 2014, almost 760,000 Veterans enrolled in the VHA were older than 85, and this number is projected to grow to more than 1.2 million by 2034 (U.S. Department of Veterans Affairs, 2015).

With the passage of the Veterans Millennium Health Care and Benefits Act of 1999 (Public Law 106–117), access to LTC was expanded for U.S. military Veterans. The Millennium Act mandated provision, once provided on a discretionary basis, of LTC to Veterans who are in need of LTC for a service-connected disability and to Veterans who need LTC and have a service-related disability rated at 70% or more (classification level Priority 1a). The Millennium Act also required that, when clinically appropriate, home- and community-based services become the preferred option for LTC, given the high cost of institutional care and Veterans’ preference to remain in the community, making such services a basic benefit for all Veterans. This mandate to increase access to LTC and to better balance the proportion of LTC costs between institutional and noninstitutional care reflects similar legislative pressures to “rebalance” Medicaid, the primary provider of long-term services and supports (LTSS) nationally (Wenzlow, Eiken, & Sredl, 2016).

Despite the passage of the Millennium Act almost 2 decades ago, expenditures for noninstitutional Veterans Affairs (VA) LTC remain unbalanced; in fiscal year 2014, the VA’s Office of Geriatrics and Extended Care (GEC) spent 74% of its budget on institutional care and 26% on noninstitutional LTSS (U.S. Department of Veterans Affairs’ Office of Budget, 2014). For comparison, the national average for state Medicaid programs, which are under a statutory obligation to provide institutional care, devoted the majority (53%) of its LTC spending to noninstitutional care in 2014 (Wenzlow et al., 2016). In general, both VA and Medicaid community-based LTSS recipients are less impaired than their institutionalized counterparts. In addition, there are differences in the VA and Medicaid LTSS populations, particularly the fact that VHA LTSS recipients are overwhelmingly men. Furthermore, Veterans receiving VA LTSS have higher rates of cognitive impairment than their peers receiving Medicaid (Shay & Burgess, 2008), and a larger proportion are younger than 65, reflecting the age of Veterans seriously wounded in wars in Iraq and Afghanistan (McCarthy, Blow, & Kales, 2004). Nevertheless, while these are not directly comparable populations, the difference in progress toward rebalancing LTSS spending between institutional and noninstitutional care is noteworthy.

The GEC has promoted policies and implemented diverse strategies to rebalance LTC (see Figure 1). Access to VA home care services that can help Veterans remain in the community was expanded, including homemaker and/or home health aide services, community hospice care, home-based primary care, respite care, and home telehealth (U.S. Department of Veterans Affairs, 2006). In addition, alternatives to nursing homes to maintain Veterans with various types and levels of need in the community were developed, such as Veteran-Directed Home and Community Based Services (Thomas & Allen, 2016), a participant-directed program for Veterans with high levels of need for assistance, and the Medical Foster Home program (Levy et al., 2016) for Veterans with cognitive as well as physical impairments. An additional strategy was to shift the service focus of VA nursing home care units, renamed Community Living Centers (CLCs; U.S. Department of Veterans Affairs, 2008), away from long-stay custodial care to short-stay rehabilitative care with the goal of safe and timely return to the community as soon as Veteran treatment goals were met.

Figure 1.

Figure 1.

Open in a new tab

Relevant Veterans Affairs (VA) initiatives, 1999–2011. VHA = Veterans Health Administration.

In 2005, to facilitate a move away from traditional nursing home care toward short-term rehabilitative care, the GEC required CLCs to conduct preadmission assessment screenings to ensure that Veterans were medically and psychiatrically stable before admitting them. In addition, the primary type of service required for each Veteran was to be documented using treating specialty codes for short-stay (≤ 90 days) and long-stay (> 90 days) service categories. Each code contains a description of the services that must be offered for Veterans admitted with that code “to provide admissions committees with a framework for admissions decisions based on services offered rather than demand alone” (U.S. Department of Veterans Affairs, 2005). Furthermore, the GEC mandated documentation of the anticipated length of stay and discharge destination for each Veteran during the required admission assessment, thus encouraging the initiation of discharge planning upon CLC entry (U.S. Department of Veterans Affairs, 2005). Although not specifically mandated, these directives combined with frequent and ongoing GEC communication with CLCs stressed a rebalancing agenda. To date, the extent to which VA CLCs have responded to the GEC’s efforts to shift the focus of VA CLCs away from long-term, custodial nursing home care has not been documented.

Nearly simultaneous with the implementation of policies to promote a shift away from long-stay custodial care to short-stay rehabilitative care in CLCs was the GEC’s introduction of policies to expand access to hospice and palliative care (see Figure 1). The Comprehensive End of Life Care (CELC) initiative of 2009, in particular, awarded up to one million dollars per year for 3 years for development or renovation of hospice and palliative care units in 54 VA medical centers. Forty-three of these units were in CLCs.

Our focus in this article is on the change in VA CLCs resulting from the GEC’s efforts to affect a shift from long-term custodial care to short-term rehabilitative, skilled nursing care. Specifically, we describe changes in the facility and resident characteristics of CLCs, as well as in Veteran outcomes, over the course of the study period (i.e., from 2004–2011). Outcomes are presented at the facility level. Specifically, we examine changes in the proportion of long stays, short stays, and hospice stays in CLCs. The outcomes length of stay, rates of discharge to the community, and rates of successful discharge to the community are reported for short and long non-hospice stays only, because neither increase nor decrease in these outcomes can be considered “desirable” for hospice care, nor are they indicators of CLCs’ response to the GEC’s rebalancing agenda. In addition, the implications of simultaneously implementing two CLC-related initiatives, one funded and one unfunded, are discussed.

Methods

Data sources

All data were assembled from existing administrative data sources available from the VA or the Center for Medicare and Medicaid Services. Data on CLC characteristics are from the VHA Support Services Center (VSSC). CLC staffing data were obtained from the GEC Data and Analyses Center (DAC), assembled from various intra-VA databases including personnel payments (Bartel, Beaulieu, & Phibbs, 2014; Uchida-Nakakoji, Stone, Schmitt, & Phibbs, 2015). Data used to describe the Veteran population included the Patient Treatment File and the VA CLC Minimum Data Set (MDS). These VA-specific data sources were merged with Medicare Part A and B claims data, including all Medicare inpatient, outpatient, skilled nursing, and hospice claims for all Veterans receiving care in a CLC based on patients’ dates of service, location, and care received. Together, this formed the VA Residential History File (RHF; Intrator, Hiris, Berg, Miller, & Mor, 2011).

Sample

We constructed an aggregate CLC facility data set to analyze the changes over time in average CLC characteristics from 2004 to 2011. CLC inclusion criteria required that a CLC be operating all 12 months of the calendar year with no fewer than two residents in the facility per month. The number of CLCs included in the sample ranged between 132 and 135 across the study years.

Variables

With the VA RHF, we calculated the types of stays, discharge destinations, and rates of successful discharge for each CLC. First, we constructed CLC episodes of care (stays) using the CLC entry and discharge dates. A CLC stay is any span of days in which there is evidence in the data that the Veteran was receiving services in a CLC. Multiple stays by the same Veteran were counted separately. Because of the unique structure of CLCs and the frequency of movement from a CLC unit to an inpatient unit and back to the CLC unit, we allowed for inpatient stays of fewer than 7 days to be considered part of the CLC stay so long as the Veteran returned to the originating CLC. However, hospitalizations with more than 7 consecutive inpatient days or a hospitalization of any length without return to the CLC were considered a CLC discharge and the end of a CLC stay. This was established to provide consistency among CLCs given varying bed hold practices across the country.

Short stays were calculated to include stays with lengths of 90 days or fewer, and long stays include stays of more than 90 days, consistent with VA’s ad hoc definitions of short and long stays. In addition, any stay including one or more days in hospice care was considered to be a hospice stay, regardless of actual number of days with hospice. Concurrent with the initiative to move toward short-term rehabilitative care in CLCs, the VA instituted an initiative to increase access to palliative and hospice care that in many VA medical centers was implemented utilizing CLC beds (Edes, Shreve, & Casarett, 2007). Because the goals of care and treatment differ for Veterans receiving hospice care, we created a separate category for these types of stays.

Length of stay was measured as the number of days from admission to the CLC and discharge. We calculated length of stay for both short and long stays. Lengths of stay are presented, over time, as stays lasting 1 to 14 days, 15 to 29 days, 30 to 59 days, 60 to 90 days, 91 to 119 days, 120 to 179 days, 180 to 364 days, and ≥ 365 days. It should be noted that although we allowed for inpatient stays of fewer than 7 days to be considered part of the CLC stay, those hospital days were not included in calculating length of stay.

Rates of discharge to the community were calculated as the proportion of Veterans whose stay ended with a discharge to the community (as opposed to a discharge to the hospital, discharge to another nursing home or another institution, or death) and were calculated separately for long stays and short stays. Rate of successful discharge was calculated among those short stays and long stays discharged to the community. Successful discharge was defined as any discharge to the community that did not have a hospitalization or admission to a CLC, community nursing home, or other institution within 30 days post–CLC discharge (Gozalo, Leland, Christian, Mor, & Teno, 2015). We conducted sensitivity analyses by lengthening the required length of time without institutionalization to 60 and 90 days post–CLC discharge. Rates of discharge to the community and successful discharge were not calculated for hospice stays because neither outcome is a goal of hospice care.

We examined changes in Veterans’ demographics using the MDS data. Specifically, we describe Veterans’ ages, gender, and race aggregated to the facility level. We also included measures of Veterans’ acuity using the MDS data. The average Nursing Case Mix Index is a measure of intensity of nursing resources required by a given Veteran’s condition and was calculated by applying Resource Utilization Groups Version III (RUG-III) residential classification system developed by the Centers for Medicare and Medicaid Services. Average activities of daily living (ADL) score across all admissions to the facility in the calendar year is based on an ADL measure that ranges from 0 (complete independence) to 28 (complete dependence), summing a score of 0 to 4 on seven ADLs (Morris, Fries, & Morris, 1999). Low cognitive impairment is defined as the proportion of residents with a Cognitive Performance Scale (CPS) score of 0, 1, or 2 (Morris et al., 1994). Severe cognitive impairment is defined as the proportion of residents with a CPS score of 5 or 6. The percentage of low-care residents is the proportion of residents who meet the low-care criteria, meaning that the Veteran does not require physical assistance in any of the four late-loss ADLs (i.e., bed mobility, transferring, toileting, and eating) and is not classified as either “special rehabilitation” or “clinically complex” in the RUG-III (Mor et al., 2007). Low-care individuals in nursing homes are likely to have their needs met equally well in community settings with adequate family or formal home care services.

Using the VSSC data, we estimated changes in a variety of CLC characteristics that measure workload and facility size. Average daily census is based upon the number of Veterans in the CLC on a given day of the year. The occupancy rate is the average daily census divided by operational beds. Operational beds are those that the facility considers available for Veterans’ CLC use and excludes beds that are temporarily unavailable. From the GEC DAC data, we included the monthly total number of nursing hours worked, and separately for each type of nurse, on each CLC ward; for those facilities with more than one ward, estimates for each ward were aggregated to provide one CLC estimate that was then divided by total patient days to create registered nurse (RN), licensed practical nurse, nursing aide, and total nursing hours per bed day.

Analysis

We conducted descriptive analyses and present the means and standard deviations of CLCs’ Veteran populations (i.e., demographics and acuity) and CLC characteristics (i.e., CLC size, workload, and staffing intensity) over time. CLC aggregate measures of Veterans’ lengths of stay are plotted, estimating the average proportion of Veterans in CLCs by length of stay. We also calculated change in the type of CLC stay, length of stay, and Veterans’ outcomes (i.e., discharge to the community and successful community discharge) for long- and short-stay Veterans over the study period. We tested for changes in CLC averages over the years with one-way analyses of variance. All analyses were conducted in SAS version 9.3 (SAS Institute, 2011).

Results

Over the study period, CLCs cared for a significantly more racially diverse patient population (from 11.56% of Black Veterans in 2004 to 17.17% in 2011, p < .001) and a growing proportion of females (from 2.63% in 2004 to 3.03% in 2011, p < .01); patients’ average age was relatively constant (see Table 1). We observed an increase in patient acuity. Specifically, patients admitted to the CLC had a higher average case-mix index (from .91 in 2004 to .95 in 2011, p < .001) and were more functionally impaired (from an average ADL score of 10.63 in 2004 to 11.95 in 2011, p < .001) over this time period, despite having negligible changes in cognition. The proportion of Veterans with low-care needs, whose needs could perhaps have been met in the community, peaked in 2006 at 8.84% but steadily declined to 6.39% in 2011. Concordant with the increases in patient complexity, there was an increase in total nurse staffing hours and nurse staffing levels, with the greatest proportional growth in RN staffing.

Table 1.

Characteristics of Community Living Centers (CLCs) and Veterans Admitted to CLCs (2004–2011).

2004
 2005
 2006
 2007
 2008
 2009
 2010
 2011
N = 133
 N = 135
 N = 135
 N = 134
 N = 133
 N = 132
 N = 132
 N = 132
Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD) Mean (SD)
Patient demographics
 Black (%) 11.56 (11.24) 12.78 (13.27) 14.86 (15.14) 15.65 (15.58) 15.92 (15.40) 16.55 (15.95) 18.06 (16.78) 17.17 (16.51)
 White (%) 76.97 (14.77) 63.38 (19.83) 71.58 (17.96) 75.01 (18.32) 75.54 (18.42) 74.60 (18.06) 73.91 (18.92) 75.54 (18.19)
 Female (%) 2.63 (1.51) 2.46 (1.42) 2.64 (1.62) 2.70 (1.63) 2.90 (1.71) 2.81 (1.70) 2.94 (1.80) 3.03 (1.65)
 Average age (years) 71.49 (2.58) 71.31 (2.67) 71.49 (2.54) 71.50 c(2.63) 71.55 (2.67) 71.7 (2.83) 71.69 (2.68) 71.84 (2.63)
Patient acuity
 Average RUG Case Mix Index 0.91(0.07) 0.92(007) 0.91(0.08) 0.92(0.08) 0.92(0.09) 0.94(0.10) 0.96(0.10) 0.95
 Low CPS score (%) 74.31 (12.58) 73.89 (13.14) 72.49 (12.84) 73.35 (12.50) 72.80 (12.94) 73.12 (11.17) 72.32 (11.55) (11.41)
 High CPS score (%) 10.60 (8.56) 10.81 (9.25) 11.44 (7.97) 10.10 (6.99) 10.09 (7.15) 10.25 (6.74) 10.21 (6.60) 9.87 (6.95)
 Average ADL score 10.63 (2.58) 10.73 (2.68) 11.00 (2.59) 11.18 (2.68) 11.27 (2.72) 11.75 (2.74) 12.19 (2.90) 11.95 (2.74)
 Low-care needs (%) 7.06 (6.18) 6.84 (6.01) 8.84 (7.48) 8.00 (6.21) 7.47 (6.03) 7.29 (7.29) 6.30 (5.98) 6.39 (5.44)
CLC characteristics
 RN HPBD 1.5 (0.55) 1.53 (0.58) 1.57 (0.60) 1.57 (0.59) 1.8 (0.67) 1.95 (0.69) 1.99 (0.66) 2.03 (0.69)
 LPN/LVN HPBD 1.21 (0.57) 1.28 (0.65) 1.3 (0.62) 1.26 (0.58) 1.43 (0.7) 1.51 (0.72) 1.49 (0.65) 1.5 (0.7)
 Aide HPBD 1.77 (0.77) 1.8 (0.76) 1.88 (0.80) 1.86 (0.77) 2.06 (0.83) 2.08 (0.86) 2.1 (0.89) 2.14 (0.81)
 Total nursing HPBD 4.48 (1.07) 4.6 (1.08) 4.76 (1.13) 4.7 (1.06) 5.28 (1.16) 5.54 (1.25) 5.59 (1.15) 5.68 (1.1)
 Total nurse staffing hours* 1361.85 (815.95) 1377.69(825.01) 1399.48 (829.68) 1458.86(844.29) 1535.06(889.41) 1518.79 (887.38) 1558.96 (886.98) 1561.53 (875.93)
 Average daily census 93.47 (58.40) 88.77 (54.76) 83.97(53.26) 83.30 (52.60) 80.93 (51.26) 79.17 (48.30) 77.97 (47.53) 76.91 (47.00)
 Occupancy rate (%) 83.67 (19.51) 78.66 (14.83) 75.37 (16.22) 75.33 (16.22) 73.83 (15.95) 72.83 (16.13) 74.15 (18.69) 73.98 (17.27)
Open in a new tab

Note. Means represent the average of the facility-level patient demographic, patient acuity, or CLC characteristics.

*

Hours presented per 100 hours.

RUG = resource utilization group; CPS = Cognitive Performance Scale (low indicates little to no cognitive impairment; high indicates severe cognitive impairment); ADL = Activities of Daily Living Scale (score of 0–28, with higher scores suggesting more impairment); low-care needs = Veterans who did not require assistance with any of the late loss activities of daily living or fall into the two lowest-functioning RUG categories (i.e., clinically complex or special rehabilitation); RN = registered nurse; LPN/LVN = licensed practical nurse/licensed vocational nurse; HPBD = hours per bed day.

There was also a decrease in the average daily census during the period, from 93 Veterans in 2004 to 77 Veterans in 2011. Occupancy rates also declined, in the aggregate, from 83.7% to 74%. The drop in occupancy was in part the result of decreases in the average daily census and also of decreases in some CLCs’ operating bed capacity (data not shown).

The average proportion of Veterans with short stays, excluding hospice stays, across CLCs decreased slightly over the study period (from a mean of 68.2% in 2004 to 61% in 2011, p < .001; see Figure 2). The proportion of long stays also decreased (from 21.9% in 2004 to 16.4% in 2011, p < .001), while the proportion of Veterans classified as having a hospice stay (regardless of length of stay) increased during the study period (from 9.8% in 2004 to 22.4% in 2011, p < .001). We did not observe any significant change in the length of stay among Veterans with short and long stays in CLCs during this time period (see Figure 3). The most prevalent length of stay was 2 weeks or less, with approximately 30% discharged within 15 days of admission. Among long stays, the greatest proportion had stays lasting more than 1 year.

Figure 2.

Figure 2.

Open in a new tab

Average facility-level proportion of types of stays admitted to Community Living Centers (2004–2011). Note. Red diamonds indicate Community Living Center mean.

Figure 3.

Figure 3.

Open in a new tab

Average facility-level lengths of stay among short and long stays from 2004 to 2011. Note. Average proportions across Community Living Centers are displayed in the figure. Hospice stays are excluded from this figure.

We observed a statistically significant increase in the proportions of patients who were discharged to the community, from 63% in 2004 to 74% in 2011 for short-stay Veterans and from 38% in 2004 to 60% in 2011 for long-stay Veterans (p < .01, see Figure 4). In addition, among those who were discharged to the community, the proportion who remained in the community successfully (i.e., without additional institutionalization) rose for the short- and long-stay population from 76% to 82% for short stays and 84% to 88% for long stays over the time period (p < .001). These increases in rates of successful discharge for short and long stays were robust using 60- and 90-day post–CLC discharge time frames.

Figure 4.

Figure 4.

Open in a new tab

Facility-level average rates of discharge to the community and successful discharge for Veterans admitted to Community Living Centers by type of stay (2004–2011). Note. Discharge to the community represents those who were discharged to a setting other than the hospital, a community nursing home, or another Community Living Center or died during their stay. Successful discharge is the proportion of those who were discharged to the community who were not readmitted to any Community Living Center, a community nursing home, or the hospital within 30 days of their discharge.

Discussion

Overall, this study suggests a mixed picture of VA CLCs’ responsiveness to GEC policy intended to rebalance VA LTC. There were moderate reductions in the proportion of admissions for both long stays (> 90 days) and short stays (< 90 days) that can be attributed largely to the increase in hospice stays, although long stays decreased by one-quarter relative to their baseline versus a 10% decrease in short stays, demonstrating CLCs’ effort to move away from custodial LTC. In addition, the GEC’s objective of decreasing length of stay for short-stay and long-stay patients was not realized to a significant degree. However, the proportions of Veterans with both long stays and short stays who were discharged to the community increased substantially from 2004 to 2011. These increases in successful discharge rates occurred despite increases in residents with functional limitations and decreases in the proportion of residents with low-care needs.

Amid this period of change, CLCs witnessed a decrease in their average daily census and, relatedly, their occupancy rate. The proportion of long stays admitted to CLCs over the study period was reduced by one-quarter, making more beds available to meet the demand for short-stay rehabilitative and skilled nursing care as well as for hospice and palliative care. In other words, beds that were previously dedicated to long-stay patients were now repurposed as hospice or short-stay beds and no longer continuously occupied.

It is likely that the directive (U.S. Department of Veterans Affairs, 2005) to estimate anticipated length of stay and discharge destination as the basis for initiation of discharge planning at admission was at least partly responsible for the increase in the proportion of Veterans discharged to the community over the study period. Families who agreed to support Veterans’ return to the community at admission may have been committed to that goal, remaining engaged in discharge planning and their Veterans’ care. CLC nursing staff and rehabilitation therapists may have worked with Veterans in the context of a target discharge date, and Veterans may have been motivated to work toward a timely return home. Outcomes of care, operationalized for this study as discharge to the community without bounce back to the CLC or rehospitalization within 30 days postdischarge, may have been enhanced by families’ expectation of their loved ones’ discharge within a specific time frame and by working with CLC staff to understand their needs and how to meet them. An anticipated discharge date and destination may have given discharge planners ample time to make arrangements for services necessary for Veterans returning to the community, further enhancing the quality of care received at home postdischarge. Indeed, qualitative data from our site visit interviews with staff in eight CLCs support this interpretation (Mills et al., in press).

However, documenting Veterans’ anticipated length of stay at admission did not achieve a secondary goal of reducing length of stay for short- and long-stay Veterans. Exacerbation of illness during CLC stays may have made hoped-for discharge dates unfeasible for some Veterans. Lack of financial and social resources in the community is also likely to be a primary reason for some stays extending past their target discharge date. Finally, dementia, mental illness, and behavioral issues associated with both conditions may eliminate some Veterans as candidates for transition to community nursing homes or traditional community-based LTSS, thereby preventing significant reductions in average length of stay. In fact, the lack of community-based options for difficult-to-place Veterans was the motivation for the GEC’s investment in the development of new models of community-based care described in the introduction to this article. With greater dissemination, these innovative models and others like them may greatly facilitate further progress toward rebalancing.

In addition to early initiation of discharge planning, care quality during this time period may also have been enhanced by the introduction of culture change into CLCs in 2005, an initiative intended to transform care culture from a medical model to one in which “care is driven by the needs of the individual, as impacted by medical conditions” (U.S. Department of Veterans Affairs, 2008). In fact, recent studies in community nursing homes have demonstrated an increase in the quality of care processes and outcomes in nursing homes with a high level of culture change implementation (Miller, Lepore, Lima, Shield, & Tyler, 2014), as well as reduction in health-related survey deficiencies (Grabowski et al., 2014).

While the average daily census did decrease over this time period, the numbers of full-time equivalent nursing staff and staffing hours in the CLC unit increased. The consequence was a significant increase in the nurse-to-patient staffing ratios over this time period, with an additional hour of nursing per bed day from 2004 to 2011. These rates are much higher than in community nursing homes, whose residents in 2011 received on average 4 hours of nurse staffing per day with less than 2.5 hours of those consisting of licensed nurse time (Harrington, Carrillo, & Garfield, 2015). The high staffing levels and their increase during this time period were in large part made possible through funding available from the CELC initiative to enable the observed dramatic growth of hospice care provided in CLCs. With the large body of literature documenting the positive relationship between nurse staffing levels and patient outcomes in community nursing homes (e.g., Hyer et al., 2011; Konetzka, Stearns, & Park, 2008), it is reasonable to assume that the impact of the increase in CLCs’ nurse staffing levels spilled over to non-hospice stays, thus contributing to the increase in rates of discharge to the community and rates of successful discharge.

We expected to see an increase in short-stay rehabilitative care and a decrease in long stays admitted to CLCs over the course of the study period as part of GEC efforts. However, we observed a decrease in the proportions of both types of stay, largely attributable to the CELC-facilitated increase in hospice stays in CLCs. While some medical centers established dedicated hospice/palliative care wards in acute care facilities, CLCs became the locus of hospice and palliative care in the majority of VA medical centers. An article recently published by Miller et al. (2017) found that male Veterans aged 66 and older had increases in their rates of hospice use within the last year of life following the CELC initiative. Therefore, the findings from these two studies, taken together, suggest that the CELC initiative was successful in increasing the number of Veterans who had access to hospice services. Whether or not we would have observed a larger increase in short-term rehabilitative and skilled care in the absence of the CELC initiative is unknown.

There was variation by CLC in all outcomes, some of it extreme. We tested in additional analyses whether the proportion of short (or long) stays, as a proxy for specializing in the care of short- (or long-) stay Veterans, was partly responsible for this variation, as it has been demonstrated for other outcomes, for example, successful community discharge of patients with hip fracture from nursing homes (Gozalo et al., 2015). However, we did not find a relationship between the proportion of types of stays in a facility and any outcome presented here, nor did we find a relationship between outcomes and other CLC characteristics known to be important in the health care organizational literature, such as total number of beds and rural versus urban location.

There was also variation in patterns of CLC change in types of stays over the study period, although the national pattern of increase in hospice stays and decrease in non-hospice short and long stays was dominant. Although the rebalancing and hospice expansion initiatives were not competing, certainly the funding for physical infrastructure and increased staffing that powered the CELC initiative made expansion of hospice care an attractive strategy for CLCs. Furthermore, all CLCs were mandated to have sustainable Palliative Care Consult Teams, with seed funding provided by CELC. In contrast, the shift in CLC focus from traditional long-stay custodial care to short-stay rehabilitative care promoted by the GEC was neither mandated nor accompanied by special funding, although it was encouraged by frequent conversations with GEC leadership and supported by change in admission policies. Prior studies have documented the inherent difficulties in concurrent implementation of multiple change initiatives (Nystrom, Garvare, Westerlund, & Weinehall, 2014). In the case of VA CLCs, both initiatives appeared to be successful, although the greater effect of CELC was obvious.

Limitations

Our study was limited by the lack of availability of more recent non-VA data used to operationalize outcomes. Clearly it would be of interest to continue to monitor these indicators of change to the present. More rapid receipt of essential data including Medicaid and Medicare claims should be a priority for the VA, especially for studies related to the rapidly growing population receiving LTC so that programs and policy can be better informed. In addition, due to data limitations, we could not test the changes that occurred in the years immediately following the Millennium Act of 1999 and rather our analyses begin in 2004. Furthermore, this study was designed to be descriptive in nature and present a macro picture of the changes in CLCs that occurred during this time period. Therefore, we do not account for intra-CLC variation or test for any causal relationships. Future work should investigate Veteran and facility characteristics that are predictive of positive care outcomes at the Veteran level. This will allow for a better understanding of the person-level and facility-level mechanisms behind the changes we observed in the aggregate in this study.

Another limitation is that CLCs received two co-occurring directives during this study period. It is therefore difficult to untangle their impact and evaluate the rebalancing of CLCs. Given data limitations, we were unable to determine the reasons that Veterans were admitted to CLCs (e.g., short-term rehabilitative care, hospice care, long-stay custodial care) prior to 2007. Therefore, our indicator of hospice stays includes Veterans who received any hospice care, regardless of their length of stay, in an effort to describe the outreach of hospice to Veterans in CLCs during this time period. Because the goals of hospice are not consistent with those that were the focus of this study (i.e., discharge to the community and reducing length of stay), it was important to separate Veterans receiving hospice care in CLCs from those receiving other types of care. Finally, this is an ad hoc evaluation of the impact of the GEC’s efforts to shift care focus ina manner consistent with their broader agenda of rebalancing VA LTC.

Conclusions

In conclusion, we find mixed responsiveness to the GEC’s initiatives to decrease length of stay and increase the focus of CLCs on providing more short-term postacute, rehabilitative care. However, the outcomes achieved by CLCs during this time period are notable and suggest that care delivery, discharge planning, and coordination of services between the CLC and community settings have improved and are consistent with a rebalancing agenda. As the GEC prepares to serve the impending wave of older Veterans who will require LTSS, it is imperative that attention to these and similar efforts continue in order to shift the delivery of LTC for VHA-enrolled Veterans from care provided in institutional settings to less restrictive environments in community-based settings.

Acknowledgments

Funding

This work was supported by the U.S. Department of Veterans Affairs [CRE12-036]; U.S. Department of Veterans Affairs [CDA14-422].

Footnotes

Disclosure Statement

The views expressed in this article are those of the authors and do not necessarily reflect the position or policy of the Department of Veterans Affairs or the US government.

References

  1. Bartel AP, Beaulieu N, & Phibbs CS (2014). Human capital and productivity in a team environment: Evidence from the healthcare sector. American Economic Journal: Applied Economics, 6 (2), 231–259. [Google Scholar]
  2. Edes T, Shreve S, & Casarett D (2007). Increasing access and quality in department of Veterans affairs care at the end of life: A lesson in change. Journal of the American Geriatrics Society, 55 (10), 1645–1649. doi: 10.1111/(ISSN)1532-5415 [DOI] [PubMed] [Google Scholar]
  3. Gozalo P, Leland NE, Christian TJ, Mor V, & Teno JM (2015). Volume matters: Returning home after hip fracture. Journal of the American Geriatrics Society, 63 (10), 2043–2051. doi: 10.1111/jgs.2015.63.issue-10 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Grabowski DC, O’Malley AJ, Afendulis CC, Caudry DJ, Elliot A, & Zimmerman S (2014). Culture change and nursing home quality of care. The Gerontologist, 54 (Suppl 1), S35–S45. doi: 10.1093/geront/gnt143 [DOI] [PubMed] [Google Scholar]
  5. Harrington C, Carrillo H, & Garfield R (2015). Nursing facilities, staffing, residents and facility deficiencies, 2009 through 2014 Retrieved from https://kaiserfamilyfoundation.files.wordpress.com/2015/08/8761-nursing-facilities-staffing-residents-and-facility-deficiencies.pdf
  6. Hyer K, Thomas KS, Branch LG, Harman JS, Johnson CE, & Weech-Maldonado R (2011). The influence of nurse staffing levels on quality of care in nursing homes. The Gerontologist, 51 (5), 610–616. doi: 10.1093/geront/gnr050 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Intrator O, Hiris J, Berg K, Miller SC, & Mor V (2011). The residential history file: Studying nursing home residents’ long-term care histories(*). Health Services Research, 46 (1 pt 1), 120–137. doi: 10.1111/j.1475-6773.2010.01194.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Kinosian B, Stallard E, & Wieland D (2007). Projected use of long-term-care services by enrolled Veterans. The Gerontologist, 47 (3), 356–364. doi: 10.1093/geront/47.3.356 [DOI] [PubMed] [Google Scholar]
  9. Konetzka RT, Stearns SC, & Park J (2008). The staffing-outcomes relationship in nursing homes. Health Services Research, 43 (3), 1025–1042. doi: 10.1111/j.1475-6773.2007.00803.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Levy CR, Alemi F, Williams AE, Williams AR, Wojtusiak J, Sutton B, … Argyros L (2016). Shared homes as an alternative to nursing home care: Impact of VA’s medical foster home program on hospitalization. The Gerontologist, 56 (1), 62–71. doi: 10.1093/geront/gnv092 [DOI] [PubMed] [Google Scholar]
  11. McCarthy JF, Blow FC, & Kales HC (2004). Disruptive behaviors in Veterans affairs nursing home residents: How different are residents with serious mental illness? Journal of the American Geriatrics Society, 52 (12), 2031–2038. doi: 10.1111/jgs.2004.52.issue-12 [DOI] [PubMed] [Google Scholar]
  12. Miller SC, Intrator O, Scott W, Shreve ST, Phibbs CS, Kinosian B, … Edes TE (2017). Increasing Veterans’ hospice use: The Veterans health administration’s focus on improving end-of-life care. Health Affairs (Millwood), 36 (7), 1274–1282. doi: 10.1377/hlthaff.2017.0173 [DOI] [PubMed] [Google Scholar]
  13. Miller SC, Lepore M, Lima JC, Shield R, & Tyler DA (2014). Does the introduction of nursing home culture change practices improve quality? Journal of the American Geriatrics Society, 62 (9), 1675–1682. doi: 10.1111/jgs.12987 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Mills WL, Tyler DA, Shield RR, Harrison J, Morgan KE, Cutty ME, … Allen SM (in press). Preparing Veterans for a timely and safe discharge from post-acute care through early initiation of discharge planning in VA Community Living Centers. The Gerontologist
  15. Mor V, Zinn J, Gozalo P, Feng Z, Intrator O, & Grabowski DC (2007). Prospects for transferring nursing home residents to the community. Health Affairs (Millwood), 26 (6), 1762–1771. doi: 10.1377/hlthaff.26.6.1762 [DOI] [PubMed] [Google Scholar]
  16. Morris JN, Fries BE, Mehr DR, Hawes C, Phillips C, Mor V, & Lipsitz LA (1994). MDS cognitive performance scale. Journal of Gerontology: Medical Science, 49 (4), M174–M182. doi: 10.1093/geronj/49.4.M174 [DOI] [PubMed] [Google Scholar]
  17. Morris JN, Fries BE, & Morris SA (1999). Scaling ADLs within the MDS. The Journals of Gerontology Series A, 54 (11), M546–M553. doi: 10.1093/gerona/54.11.M546 [DOI] [PubMed] [Google Scholar]
  18. Nystrom ME, Garvare R, Westerlund A, & Weinehall L (2014). Concurrent implementation of quality improvement programs. International Journal of Health Care Quality Assurance, 27 (3), 190–208. doi: 10.1108/IJHCQA-08-2012-0085 [DOI] [PubMed] [Google Scholar]
  19. SAS Institute. (2011). SAS version 9.3 Cary, NC: SAS Institute. [Google Scholar]
  20. Shay K, & Burgess JF (2008). Setting the stage for a new strategic plan for geriatrics and extended care in the Veterans health administration: Summary of the 2008 VA state of the art conference, “The changing faces of geriatrics and extended care: Meeting the needs of Veterans in the next decade. Journal of the American Geriatrics Society, 56 (12), 2330–2339. [DOI] [PubMed] [Google Scholar]
  21. Thomas KS, & Allen SM (2016). An interagency partnership to deliver Veteran-directed home and community-based services: Interviews with aging and disability network agency personnel regarding their experience with partner Veteran affairs medical center. Journal of Rehabilitation Research and Development, 53 (5), 611–618. doi: 10.1682/JRRD.2015.02.0019 [DOI] [PubMed] [Google Scholar]
  22. Uchida-Nakakoji M, Stone PW, Schmitt SK, & Phibbs CS (2015). Nurse workforce characteristics and infection risk in VA Community living centers: A longitudinal analysis. Medical Care, 53 (3), 261–267. doi: 10.1097/MLR.0000000000000316 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. U.S. Department of Veterans Affairs. (2005). VA Nursing Home Care Unit (NHCU) admission criteria, service codes, and discharge criteria Washington, DC: Department of Veterans Affairs; (vol. VHA directive 2005–061). [Google Scholar]
  24. U.S. Department of Veterans Affairs. (2006). Purchased home health care services procedures Washington, DC: Department of Veterans Affairs; (vol. VHA Handbook 1140.6). [Google Scholar]
  25. U.S. Department of Veterans Affairs. (2008). Criteria and standards for VA Community Living Centers (CLC) Washington, DC: Department of Veterans Affairs; (vol. VHA Handbook 1142.01). [Google Scholar]
  26. U.S. Department of Veterans Affairs. (2015). Enrollee health care projection model Planning Retrieved from http://www.va.gov/HEALTHPOLICYPLANNING/planning.asp
  27. U.S. Department of Veterans Affairs’ Office of Budget. (2014). FY2015 Congressional submission Washington, DC: Veterans Affairs. [Google Scholar]
  28. Wenzlow A, Eiken S, & Sredl K (2016). Improving the balance: The evolution of medicaid expenditures for LTSS, FY 1981–2014 Retrieved from https://www.medicaid.gov/medicaid/ltss/downloads/evolution-ltss-expenditures.pdf

RESOURCES