Refining a case-mix measure for nursing homes: Resource Utilization Groups (RUG-III)

Mental health quality and outcome measurement and improvement in Australia

PURPOSE OF REVIEW

This study outlines the broad approach taken in Australia to improve the quality of mental health services, including the use of performance and outcome measures along with a range of other linked quality initiatives.

RECENT FINDINGS

The investment of Australia in a strategic national approach to mental health reform with a range of nationally coordinated quality initiatives has provided a firm foundation for continuing to engage governments, healthcare organizations and clinicians in the ongoing quest for quality and outcome measurement and improvement.

SUMMARY

Australia has adopted a strategic national approach to quality and outcome measurement and improvement of mental health services despite having a federated health system involving nine governments and tiered care across public and private sectors. A range of interconnected initiatives including national plans, standards and guidelines, performance indicators and outcomes measures and benchmarking and reporting processes are the key components of this nationally coordinated system.

Use of physical restraints and antipsychotic medications in nursing homes: a cross-national study

OBJECTIVES

This study compares inter- and intra-country differences in the prevalence of physical restraints and antipsychotic medications in nursing homes, and examines aggregated resident conditions and organizational characteristics correlated with these treatments.

METHODS

Population-based, cross-sectional data were collected using a standardized Resident Assessment Instrument (RAI) from 14,504 long-term care facilities providing nursing home level services in five countries participating in the interRAI consortium, including Canada, Finland, Hong Kong (Special Administrative Region, China), Switzerland, and the United States. Facility-level prevalence rates of physical restraints and antipsychotic use were examined both between and within the study countries.

RESULTS

The prevalence of physical restraint use varied more than five-fold across the study countries, from an average 6% in Switzerland, 9% in the US, 20% in Hong Kong, 28% in Finland, and over 31% in Canada. The prevalence of antipsychotic use ranged from 11% in Hong Kong, between 26-27% in Canada and the US, 34% in Switzerland, and nearly 38% in Finland. Within each country, substantial variations existed across facilities in both physical restraint and antipsychotic use rates. In all countries, neither facility case mix nor organizational characteristics were particularly predictive of the prevalence of either treatment.

CONCLUSIONS

There exists large, unexplained variability in the prevalence of physical restraint and antipsychotic use in nursing home facilities both between and within countries. Since restraints and antipsychotics are associated with adverse outcomes, it is important to understand the idiosyncratic factors specific to each country that contribute to variation in use rates.

Equivalence of two healthcare costing methods: bottom-up and top-down

This paper compares two quite different approaches to estimating costs: a 'bottom-up' approach, represented by the US Department of Veterans Affairs' (VA) Decision Support System that uses local costs of specific inputs; and a 'top-down' approach, represented by the costing system created by the VA Health Economics Resource Center, which assigns the VA national healthcare budget to specific products using various weighting systems. Total annual costs per patient plus the cost for specific services (e.g. clinic visit, radiograph, laboratory, inpatient admission) were compared using scatterplots, correlations, mean difference, and standard deviation of individual differences. Analysis are based upon 2001 costs for 14 915 patients at 72 facilities. Correlations ranged from 0.24 for the cost of outpatient encounters to 0.77 for the cost of inpatient admissions, and 0.85 for total annual cost. The mean difference between costing methods was $707 ($4168 versus $3461) for total annual cost. The standard deviation of the individual differences was $5934. Overall, the agreement between the two costing systems varied by the specific cost being measured and increased with aggregation. Administrators and researchers conducting cost analyses need to carefully consider the purpose, methods, characteristics, strengths, and weaknesses when selecting a method for assessing cost.

Sensitivity of nursing home cost comparisons to method of dementia diagnosis ascertainment

This study compared the association of differing methods of dementia ascertainment, derived from multiple sources, with nursing home (NH) estimates of prevalence of dementia, length of stay, and costs an understudied issue. Subjects were 2050 new admissions to 59 Maryland NHs, from 1992 to 1995 followed longitudinally for 2 years. Dementia was ascertained at admission from charts, Medicare claims, and expert panel. Overall 59.5% of the sample had some indicator of dementia. The expert panel found a higher prevalence of dementia (48.0%) than chart review (36.9%) or Medicare claims (38.6%). Dementia cases had lower relative average per patient monthly costs, but longer NH length of stay compared to nondementia cases across all methods. The prevalence of dementia varied widely by method of ascertainment, and there was only moderate agreement across methods. However, lower costs for dementia among NH admissions are a robust finding across these methods.

Urinary incontinence among institutionalized oldest old Chinese men in Taiwan

AIMS

To explore prevalence and related factors for urinary incontinence (UI) among the oldest old institutionalized Chinese men in Taiwan.

METHODS

All residents living in Banciao Veterans Care Home were invited for study. UI was defined as urinary leakage at least once weekly. Additional data items from the Minimum Data Set (MDS Nursing Home Chinese Version 2.1) were used to explore impact associated with physical function, cognitive status and quality of life (social engagement, SocE). Depressive symptoms were screened by the Short Form Geriatric Depression Scale.

RESULTS

Data from 594 male residents (mean age: 80.9 +/- 5.3 years) were analyzed. Among all study subjects, 92.8% were functionally independent, 20.4% had certain cognitive impairment and 8.2% had depressive symptoms. The prevalence of UI in the Banciao Veterans Care Home was 10.1%. Compared with residents without UI, subjects with UI had poorer physical function, cognitive status, and more depressive symptoms. The mean SocE score was 1.5 +/- 1.3, and was similar between UI (+) and UI (-) subjects (1.4 +/- 1.2 vs. 1.6 +/- 1.3, P = 0.411). By multivariate logistic regression, poorer physical functional status, cognitive impairment and depressive symptoms were independent risk factors for UI (P < 0.05).

CONCLUSIONS

Poorer physical function, poorer cognitive status and depressive symptoms were all statistically significant independent risk factors for UI. However, SocE score (proxy indicator of quality of life) did not differ between subjects with and without UI. Further investigations are needed to evaluate the impact of UI on quality of life among oldest old institutionalized Chinese men in Taiwan.

Current funding and financing issues in the Australian hospice and palliative care sector

This article overviews current funding and financing issues in the Australian hospice and palliative care sector. Within Australia, the major responsibilities for managing the health care system are shared between two levels of government. Funding arrangements vary according to the type of care. The delivery of palliative care services is a State/Territory responsibility. Recently, almost all States/Territories have developed overarching frameworks to guide the development of palliative care policies, including funding and service delivery structures. Palliative care services in Australia comprise a mix of specialist providers, generalist providers, and support services in the public, nongovernment, and private sectors. The National Palliative Care Strategy is a joint strategy of the Commonwealth and States that commenced in 2002 and includes a number of major issues. Following a national study in 1996, the Australian National Subacute and Nonacute Patient (AN-SNAP) system was endorsed as the national casemix classification for subacute and nonacute care. Funding for palliative care services varies depending on the type of service and the setting in which it is provided. There is no national model for funding inpatient or community services, which is a State/Territory responsibility. A summary of funding arrangements is provided in this article. Palliative care continues to evolve at a rapid rate in Australia. Increasingly flexible evidence-based models of care delivery are emerging. This article argues that it will be critical for equally flexible funding and financing models to be developed. Furthermore, it is critical that palliative care patients can be identified, classified, and costed. Casemix classifications such as AN-SNAP represent an important starting point but further work is required.

Exploring Factors Associated With Turnover Among Registered Nurse (RN) Supervisors in Nursing Homes

Because most turnover studies focus on certified nursing assistants (CNAs), licensed vocational nurses (LVNs), and administrators, little is known about registered nurses’ (RNs) higher turnover. This study builds on the current body of knowledge about turnover among RN supervisors in nursing homes. The article discusses a survey of RN nurse supervisors administered in more than 1,000 nursing homes that was merged with the 2003 Texas Medicaid cost report and the area resource file. Two 2-stage models are developed to predict RN turnover rates. RNs’ intent to leave predicts RN turnover through job satisfaction, perceived empowerment, and education level. High LVN and CNA turnover and high Medicare census are associated with higher turnover. Implications are that participation in management decisions and perceived wage competitiveness are an important determinant of RN retention in nursing homes. Future research should focus why RN supervisors with higher levels of education leave nursing homes.

Games policy makers and providers play: introducing case-mix-based payment to hospital chronic care units in Japan

Case-mix-based payment was developed for hospital chronic care units in Japan to replace the flat per diem rate and encourage the admission of patients with higher medical acuity and was part of a policy initiative to make the tariff more evidence based. However, although the criteria for grouping patients were developed from a statistical analysis of resource use, the tariff was subsequently set below costs, particularly for the groups with the lowest medical acuity, both because of the prime minister's decision to decrease total health expenditures and because of the health ministry's decision to target the reductions on chronic care units. Providers quickly adapted to the new payment system mainly by reclassifying their patients to higher medical acuity groups. Some hospitals reported high prevalence rates of urinary tract infections and pressure ulcers. The government responded by issuing directives to providers to calculate the prevalence rates and document the care that has been mandated for the patients at risk. However, in order to monitor compliance and to evaluate whether the patient is being billed for the appropriate case-mix group, the government must invest in developing a comprehensive patient-level database and in training staff for making on-site inspections.

Standardization of the continuing care activity measure: a multicenter study to assess reliability, validity, and ability to measure change

BACKGROUND

There is a lack of standardized mobility measures specific to the long-term care (LTC) population. Therefore, the Continuing Care Activity Measure (CCAM) was developed.

OBJECTIVE

This study determined levels of reliability, validity for clinical utilization, and sensitivity to change of this measure.

DESIGN

This was a prospective longitudinal cohort study among elderly people with primarily physical or medical impairments who were residing in LTC institutions that provide nursing home and more-complex care, with access to physical therapy services.

METHOD

The CCAM, the Clinical Outcome Variables Scale (COVS), the Social Engagement Scale (SES) of the Resident Assessment Instrument-Minimum Data Set (RAI-MDS) 2.0 instrument, and the Resource Utilization Groups, version 3, (RUG-III) were administered by clinical and research physical therapists, with timing dictated by the study purpose.

RESULTS

The participants were 136 residents of LTC institutions and 21 physical therapists. The CCAM interrater reliability (intraclass correlation coefficient [ICC]) was .97 (95% confidence interval=.91-1.00), and test-retest reliability (ICC) over a period of 1 week was .99 (95% confidence interval=.93-1.00). Over 6 months, the absolute change in total score was 5.88 for the CCAM and 4.26 for the COVS; the CCAM was 28% more responsive across all participants (n=105) and 68% more responsive for those scoring in the lower half (n=49). The minimal detectable difference of the CCAM was 8.6 across all participants. The CCAM correlated with the COVS, nursing care hours inferred from the RUG-III, and the SES.

LIMITATIONS

Some participants were lost to follow-up.

CONCLUSIONS

The CCAM is a reliable and valid tool to measure gross motor function and physical mobility for elderly people in LTC institutions. It discriminates among functional levels, measures individual functional change, and can contribute to clinical decision making.

Sharing clinical information across care settings: the birth of an integrated assessment system

Background

Population ageing, the emergence of chronic illness, and the shift away from institutional care challenge conventional approaches to assessment systems which traditionally are problem and setting specific.

Methods

From 2002, the interRAI research collaborative undertook development of a suite of assessment tools to support assessment and care planning of persons with chronic illness, frailty, disability, or mental health problems across care settings. The suite constitutes an early example of a "third generation" assessment system.

Results

The rationale and development strategy for the suite is described, together with a description of potential applications. To date, ten instruments comprise the suite, each comprising "core" items shared among the majority of instruments and "optional" items that are specific to particular care settings or situations.

Conclusion

This comprehensive suite offers the opportunity for integrated multi-domain assessment, enabling electronic clinical records, data transfer, ease of interpretation and streamlined training.

Helping nursing homes "at risk" for quality problems: a statewide evaluation

The Quality Improvement Program for Missouri (QIPMO), a state school of nursing project to improve quality of care and resident outcomes in nursing homes, has a special focus to help nursing homes identified as "at risk" for quality concerns. In fiscal year 2006, 92 of 492 Medicaid-certified facilities were identified as "at risk" using quality indicators (QIs) derived from Minimum Data Set (MDS) data. Sixty of the 92 facilities accepted offered on-site clinical consultations by gerontological expert nurses with graduate nursing education. Content of consultations include quality improvement, MDS, care planning, evidence-based practice, and effective teamwork. The 60 "at-risk" facilities improved scores 4%-41% for 5 QIs: pressure ulcers (overall and high risk), weight loss, bedfast residents, and falls; other facilities in the state did not. Estimated cost savings (based on prior cost research) for 444 residents who avoided developing these clinical problems in participating "at-risk" facilities was more than $1.5 million for fiscal year 2006. These are similar to estimated savings of $1.6 million for fiscal year 2005 when 439 residents in "at-risk" facilities avoided clinical problems. Estimated savings exceed the total program cost by more than $1 million annually. QI improvements demonstrate the clinical effectiveness of on-site clinical consultation by gerontological expert nurses with graduate nursing education.

Hard times: the effects of financial strain on home care services use and participant outcomes in Michigan

PURPOSE

When government funding for long-term care is reduced, participant outcomes may be adversely affected. We investigated the effect of program resources on individuals enrolled in the Michigan Home- and Community-Based Services (HCBS) waiver program for elderly and disabled adults.

DESIGN AND METHODS

Using dates of major policy and budget changes, we defined 4 distinct time periods between October 2001 and December 2005. Minimum Data Set for Home Care assessment records for HCBS participants (n = 112,182) were used to examine temporal trends in formal care hours and 6 outcomes: emergency room (ER) use, hospitalization, caregiver burden, death, nursing facility (NF) use, and permanent NF placement. Controlling for demographics, functional status, and cognitive status, adjusted odds of outcomes were obtained using discrete-time survival analysis.

RESULTS

As resources diminished, mean formal care hours decreased, declining most for persons with moderate functional or cognitive impairment, for up to an approximately 30% decrease. In the most financially restricted period, 3 adverse outcomes increased relative to baseline: hospitalization (odds ratio [OR] = 1.10; 95% confidence interval [CI] = 1.03-1.18), ER use (OR = 1.13; 95% CI = 1.03-1.24), and permanent NF placement (OR = 1.20; 95% CI = 1.00-1.42).

IMPLICATIONS

Reductions in resources for home care were associated with increased probability of adverse outcomes. Cutting funds to home care programs can increase utilization of other more costly services, thus offsetting potential health care savings. Policymakers must consider all ways in which budget reductions and policy changes can affect participants.

Registered nurse staffing mix and quality of care in nursing homes: a longitudinal analysis

PURPOSE

To examine the relationship between registered nurse (RN) staffing mix and quality of nursing home care measured by regulatory violations.

DESIGN AND METHODS

A retrospective panel data study (1999-2003) of 2 groups of California freestanding nursing homes. One group was 201 nursing homes that consistently met the state's minimum standard for total nurse staffing level over the 5-year period. The other was 210 nursing homes that consistently failed to meet the standard over the period. All facility and market variables were drawn from California's cost report data and state licensing and certification data, as well as 3 other databases.

RESULTS

The RN to total nurse staffing ratio was negatively related to serious deficiencies in nursing homes that consistently met the staffing standard, whereas the ratio was negatively associated with total deficiencies in nursing homes that consistently failed to meet the standard over the 5-year period. As the RN to licensed vocational nurse ratios increased, total deficiencies and serious deficiencies decreased in both groups of nursing homes.

IMPLICATIONS

A higher RN mix is positively related to quality of care, but the relationship is affected by overall nurse staffing levels in nursing homes. Further studies are necessary for a better understanding of RNs' unique contributions to the quality of care in nursing homes.

A panel data analysis of the relationships of nursing home staffing levels and standards to regulatory deficiencies

OBJECTIVE

To examine the relationships between nursing staffing levels and nursing home deficiencies.

METHODS

This panel data analysis employed random-effect models that adjusted for unobserved, nursing home-specific heterogeneity over time. Data were obtained from California's long-term care annual cost report data and the Automated Certification and Licensing Administrative Information and Management Systems data from 1999 to 2003, linked with other secondary data sources.

RESULTS

Both total nursing staffing and registered nurse (RN) staffing levels were negatively related to total deficiencies, quality of care deficiencies, and serious deficiencies that may cause harm or jeopardy to nursing home residents. Nursing homes that met the state staffing standard received fewer total deficiencies and quality of care deficiencies than nursing homes that failed to meet the standard. Meeting the state staffing standard was not related to receiving serious deficiencies.

CONCLUSIONS

Total nursing staffing and RN staffing levels were predictors of nursing home quality. Further research is needed on the effectiveness of state minimum staffing standards.

Applying Resource Utilization Groups (RUG-III) in Hong Kong nursing homes

Resource Utilization Groups III (RUG-III) is a case-mix system developed in the United States for categorization of nursing home residents and the financing of residential care services. In Hong Kong, RUG-III is based on several board groups of residents. The aim of this study was to examine the reliability and validity of the RUG-III in Hong Kong nursing homes. A cross-sectional survey was conducted in seven residential facilities operated by one agency. Residents ( N = 1,127) were assessed by the Minimum Data Set (MDS) and nursing as well as auxiliary staff care times were recorded within 2 weeks before or after the completion of MDS assessment. Forty-five out 1,127 residents were re-interviewed by an independent assessor to assess the inter-rater reliability. The inter-rater reliability of MDS assessment was excellent (kappa = 0.76) and the original RUG-III accounted for about 30 per cent of nursing staff time. Results provide preliminary evidence to support that RUG-III is a reliable and valid case-mix system for Hong Kong nursing homes, but future studies must be explored to reduce the variance of resource use explained by this case-mix system.

Racial disparities in access to long-term care: the illusive pursuit of equity

While nursing homes were insulated from civil-rights enforcement at the time of the implementation of the Medicare program and lagged behind other parts of the health sector in providing comparable access to minorities, they are the only providers for which current reporting requirements make it possible to fully assess racial disparities in use and quality of care. We find that African Americans' use of nursing homes in 2000 in the United States was 14 percent higher than Caucasians' use. The largest relative African American use of nursing homes in 2000 took place in the South and West. Average nursing-home case-mix acuity for African Americans and Caucasians were essentially identical, suggesting that shifts in payment incentives have eliminated the selective admission of easy-care private-pay (predominantly Caucasian) patients and helped fuel the growth of private pay home care and assisted living for this segment of the population. While these shifts in incentives helped increase the use of nursing homes by African Americans, a high degree of segregation and disparity in the quality of the nursing homes used by African Americans persists. Parity in use is an illusive benchmark for measuring progress in assuring equity in treatment.

Maintaining continuity of care for nursing home residents: effect of states' Medicaid bed-hold policies and reimbursement rates

OBJECTIVE

Recent public concern in response to states' intended repeal of Medicaid bed-hold policies and report of their association with higher hospitalization rates prompts examination of these policies in ensuring continuity of care within the broader context of Medicaid policies.

DATA SOURCES/STUDY DESIGN

Minimum Data Set assessments of long-stay nursing home residents in April-June 2000 linked to Medicare claims enabled tracking residents' hospitalizations during the ensuing 5 months and determining hospital discharge destination. Multinomial multilevel models estimated the effect of state policies on discharge destination controlling for resident, hospitalization, nursing home, and market characteristics.

RESULTS

Among 77,955 hospitalizations, 5,797 (7.4 percent) were not discharged back to the baseline nursing home. Bed-hold policies were associated with lower odds of transfer to another nursing home (AOR=0.55, 95 percent CI 0.52-0.58) and higher odds of hospitalization (AOR=1.36), translating to 9.5 fewer nursing home transfers and 77.9 more hospitalizations per 1,000 residents annually, and costing Medicaid programs about $201,311. Higher Medicaid reimbursement rates were associated with lower odds of transfer.

CONCLUSIONS

Bed-hold policies were associated with greater continuity of NH care; however, their high cost compared with their small impact on transfer but large impact on increased hospitalizations suggests that they may not be effective.

Maintaining continuity of care for nursing home residents: effect of states' Medicaid bed-hold policies and reimbursement rates

OBJECTIVE

Recent public concern in response to states' intended repeal of Medicaid bed-hold policies and report of their association with higher hospitalization rates prompts examination of these policies in ensuring continuity of care within the broader context of Medicaid policies.

DATA SOURCES/STUDY DESIGN

Minimum Data Set assessments of long-stay nursing home residents in April-June 2000 linked to Medicare claims enabled tracking residents' hospitalizations during the ensuing 5 months and determining hospital discharge destination. Multinomial multilevel models estimated the effect of state policies on discharge destination controlling for resident, hospitalization, nursing home, and market characteristics.

RESULTS

Among 77,955 hospitalizations, 5,797 (7.4 percent) were not discharged back to the baseline nursing home. Bed-hold policies were associated with lower odds of transfer to another nursing home (AOR=0.55, 95 percent CI 0.52-0.58) and higher odds of hospitalization (AOR=1.36), translating to 9.5 fewer nursing home transfers and 77.9 more hospitalizations per 1,000 residents annually, and costing Medicaid programs about $201,311. Higher Medicaid reimbursement rates were associated with lower odds of transfer.

CONCLUSIONS

Bed-hold policies were associated with greater continuity of NH care; however, their high cost compared with their small impact on transfer but large impact on increased hospitalizations suggests that they may not be effective.

Is it feasible and desirable to collect voluntarily quality and outcome data nationally in palliative oncology care?

Hospice/palliative care is a critical component of cancer care. In Australia, more than 85% of people referred to specialized hospice/palliative care services (SHPCS) have a primary diagnosis of cancer, and 60% of people who die from cancer will be referred to SHPCS. The Palliative Care Outcomes Collaboration (PCOC) is an Australian initiative that allows SHPCS to collect nationally agreed-upon measures to better understand quality, safety, and outcomes of care. This article describes data (October 2006 through September 2007) from the first 22 SHPCS, with more than 100 inpatient admissions annually. Data include phase of illness, place of discharge, and, at each transition in place of care, the person's functional status, dependency, and symptom scores. Data are available for 5,395 people for 6,379 admissions. After categorizing by phase of illness and dependency, there remain at the end of each admission 12-fold differences (mean, 26%; range, 4% to 52%) in the percentage of patients who became stable after an unstable phase; seven-fold differences (mean, 22%; range, 6% to 41%) in the percentage of patients with improved symptom scores, five-fold differences (mean, 25%; range, 12% to 64%) in discharge back to the community, four-fold differences (mean, 10%; range, 4% to 16%) in improved function, and three-fold differences in the length of stay (mean, 14 days; range, 6 to 19 days). PCOC shows it is feasible to collect quality national palliative care outcome data voluntarily. Variations in outcomes justify continued enrollment of services. Benchmarking should include all patients whose cancer will cause death and explore observed variations.

Relationship between state medicaid policies, nursing home racial composition, and the risk of hospitalization for black and white residents

OBJECTIVE

To examine racial differences in the risk of hospitalization for nursing home (NH) residents.

DATA SOURCES

National NH Minimum Data Set, Medicare claims, and Online Survey Certification and Reporting data from 2000 were merged with independently collected Medicaid policy data.

STUDY DESIGN

One hundred and fifty day follow-up of 516,082 long-stay residents.

PRINCIPLE FINDINGS

18.5 percent of white and 24.1 percent of black residents were hospitalized. Residents in NHs with high concentrations of blacks had 20 percent higher odds (95 percent confidence interval [CI]=1.15-1.25) of hospitalization than residents in NHs with no blacks. Ten-dollar increments in Medicaid rates reduced the odds of hospitalization by 4 percent (95 percent CI=0.93-1.00) for white residents and 22 percent (95 percent CI=0.69-0.87) for black residents.

CONCLUSIONS

Our findings illustrate the effect of contextual forces on racial disparities in NH care.

Validation of Resource Utilization Groups version III for Home Care (RUG-III/HC): evidence from a Canadian home care jurisdiction

BACKGROUND

The case-mix system Resource Utilization Groups version III for Home Care (RUG-III/HC) was derived using a modest data sample from Michigan, but to date no comprehensive large scale validation has been done.

OBJECTIVES

This work examines the performance of the RUG-III/HC classification using a large sample from Ontario, Canada.

METHODS

Cost episodes over a 13-week period were aggregated from individual level client billing records and matched to assessment information collected using the Resident Assessment Instrument for Home Care, from which classification rules for RUG-III/HC are drawn. The dependent variable, service cost, was constructed using formal services plus informal care valued at approximately one-half that of a replacement worker.

RESULTS

An analytic dataset of 29,921 episodes showed a skewed distribution with over 56% of cases falling into the lowest hierarchical level, reduced physical functions. Case-mix index values for formal and informal cost showed very close similarities to those found in the Michigan derivation. Explained variance for a function of combined formal and informal cost was 37.3% (20.5% for formal cost alone), with personal support services as well as informal care showing the strongest fit to the RUG-III/HC classification.

CONCLUSIONS

RUG-III/HC validates well compared with the Michigan derivation work. Potential enhancements to the present classification should consider the large numbers of undifferentiated cases in the reduced physical function group, and the low explained variance for professional disciplines.

Fall-related hospitalization and facility costs among residents of institutions providing long-term care

PURPOSE

The purpose of this study was to estimate hospital and long-term-care costs resulting from falls in long-term-care facilities (LTCFs).

DESIGN AND METHODS

The study used a retrospective, pre/post with comparison group design. We used matching, based on propensity scores, to control for baseline differences between fallers and non-fallers. We estimated residents' propensity to fall from demographics, comorbidities, and reimbursement in the pre-period. The matched sample included 1,130 fallers and 1,130 non-fallers who were residents of a large, multifacility long-term-care chain. Cost estimates were based on information in the Minimum Data Set and were defined as hospital and LTCF reimbursements. We estimated fall-related costs as the difference between changes in costs for fallers and non-fallers from the pre- to post-period.

RESULTS

Fallers were substantially more likely to suffer fractures and hospitalizations in the post-period than were non-fallers. Fall-related LTCF and hospital costs were $6,259 (95% confidence interval = $2,034-$10,484) per resident per year. About 60% of this amount was attributable to higher hospitalization costs. Fallers were more likely to be discharged to hospitals or to die.

IMPLICATIONS

Falls in LTCFs are associated with costs of about $6,200 per resident per year. These results provide baseline estimates that one may use to estimate the cost-effectiveness of interventions to reduce fall rates.

Restructuring in response to case mix reimbursement in nursing homes: a contingency approach

BACKGROUND

Resident-based case mix reimbursement has become the dominant mechanism for publicly funded nursing home care. In 1998 skilled nursing facility reimbursement changed from cost-based to case mix adjusted payments under the Medicare Prospective Payment System for the costs of all skilled nursing facility care provided to Medicare recipients. In addition, as of 2004, 35 state Medicaid programs had implemented some form of case mix reimbursement.

PURPOSE

The purpose of the study is to determine if the implementation of Medicare and Medicaid case mix reimbursement increased the administrative burden on nursing homes, as evidenced by increased levels of nurses in administrative functions.

METHODOLOGY/APPROACH

The primary data for this study come from the Centers for Medicare and Medicaid Services Online Survey Certification and Reporting database from 1997 through 2004, a national nursing home database containing aggregated facility-level information, including staffing, organizational characteristics and resident conditions, on all Medicare/Medicaid certified nursing facilities in the country. We conducted multivariate regression analyses using a facility fixed-effects model to examine the effects of the implementation of Medicaid case mix reimbursement and Medicare Prospective Payment System on changes in the level of total administrative nurse staffing in nursing homes.

FINDINGS

Both Medicaid case mix reimbursement and Medicare Prospective Payment System increased the level of administrative nurse staffing, on average by 5.5% and 4.0% respectively. However, lack of evidence for a substitution effect suggests that any decline in direct care staffing after the introduction of case mix reimbursement is not attributable to a shift from clinical nursing resources to administrative functions.

PRACTICE IMPLICATIONS

Our findings indicate that the administrative burden posed by case mix reimbursement has resource implications for all freestanding facilities. At the margin, the increased administrative burden imposed by case mix may become a factor influencing a range of decisions, including resident admission and staff hiring.

Risk factors for falls among elderly men in a veterans home

BACKGROUND

Falls are major causes of mortality and morbidity in the elderly. Cognitive dysfunction, poor physical function and medical comorbidities are associated with many factors contributing to falls. The main purpose of this study was to explore the risk factors of falls among older institutionalized Chinese men in Taiwan.

METHODS

Residents aged over 65 in a veterans care home in northern Taiwan were enrolled for study after they gave their full consent. Falling was defined as a fall within the past 180 days as defined by the Minimum Data Set (MDS). Physical function was determined by MDS resource utilization group activity of daily living score (RUG ADL score). Cognitive status was measured by MDS cognition scale (MDS COGS).

RESULTS

In total, 585 residents (mean age, 80.9 +/- 5.4 years) were enrolled. Among all study subjects, 92.8% were physically independent and 20.2% were moderately cognitive impaired according to MDS COGS. By definition, 48 subjects (8.2%) had a past history of fall. Compared with non-fallers, fallers were significantly older (82.4 +/- 5.5 years vs. 80.7 +/- 5.4 years; p = 0.047) and had poorer functional status according to the MDS RUG ADL score (5.0 +/- 2.3 vs. 4.3 +/- 1.6; p = 0.044). Subjects with past history of anxiety disorder and cardiovascular disease were more prone to fall, and subjects who took hypnotics were also at a higher risk of falling. By using multivariate logistic regression, we found that higher RUG ADL score (odds ratio [OR], 1.18; 95% confidence interval [CI], 1.08-1.37; p = 0.017) and hypnotic use (OR, 2.0; 95% CI, 1.0-4.1; p = 0.048) were both independent risk factors for falls.

CONCLUSION

The prevalence of fall in the past 180 days was 8.2% among elderly residents in a veterans care home in northern Taiwan. The independent risk factors for falls in this setting included poorer functional status and hypnotic use.

Implementation of the resident assessment instrument/minimum data set in the nursing home as organization: implications for quality improvement in RN clinical assessment

The Resident Assessment Instrument/Minimum Data Set (RAI/MDS) used in nursing homes (NHs) participating in the Federal Medicare and Medicaid programs is a state-of-the-art, computerized clinical assessment instrument. RAI/MDS-derived data are essential, used for NH reimbursement, quality measurement, regulatory quality monitoring activities, and clinical care planning. Completing or coordinating the RAI/MDS, which may be conceived of as implementation, is a federally mandated responsibility of the RN involving clinical assessment, a core professional competency of any RN. How the RAI/MDS is implemented in each NH provides evidence of how each NH as an organization understands both the RAI/MDS process and its organizational level responsibility for promotion of RN competence in clinical assessment. Research literature related to RAI/MDS development, testing, and accuracy is used to identify what is known about organizational level implementation of the RAI/MDS. Evidence-based suggestions to enhance RN competence in RAI/MDS clinical assessments, given existing organizational barriers, are provided.

Nursing effort and quality of care for nursing home residents

PURPOSE

The purpose of this study was to determine the relationship between nursing home staffing level, care received by individual residents, and resident quality-related care processes and functional outcomes.

DESIGN AND METHODS

Nurses recorded resident care time for 5,314 residents on 156 units in 105 facilities in four states (Colorado, Indiana, Minnesota, and Mississippi). We linked residents' care times to their measures of health and functioning from Minimum Data Set assessments. Major variables were unit- and resident-specific minutes of care per day, process measures (physical restraints, range of motion, toileting program, and training in activities of daily living [ADLs]), outcome measures (ADL decline, mobility decline, and worsening behavior between the time study and 90-day follow-up), and covariates such as unit type and resident health status. We used multilevel analysis to examine staffing and quality relationships.

RESULTS

Residents with toileting programs, range of motion or ADL training, and restraints received significantly more care from unlicensed but not from licensed staff. However, functional outcomes were not significantly related to care received from licensed or unlicensed staff, except for ADL decline, which was greatest for residents receiving more unlicensed minutes of care. Unit staffing level (licensed and unlicensed) was unrelated to any of the care processes or outcome measures, although higher overall staffing was associated with more time devoted to direct resident care.

IMPLICATIONS

Future research into nursing home quality should focus on organization and delivery rather than simply the amount of care available.

Comparing staffing levels in the Online Survey Certification and Reporting (OSCAR) system with the Medicaid Cost Report data: are differences systematic?

PURPOSE

This study had two goals: (a) to assess the validity of the Online Survey Certification and Reporting (OSCAR) staffing data by comparing them to staffing measures from audited Medicaid Cost Reports and (b) to identify systematic differences between facilities that over-report or underreport staffing in the OSCAR.

DESIGN AND METHODS

We merged the 2002 Texas Nursing Facility Cost Report, the OSCAR for Texas facilities surveyed in 2002, and the 2003 Area Resource File. We eliminated outliers in the OSCAR using three decision rules, resulting in a final sample size of 941 of the total of 1,017 non-hospital-based facilities. We compared OSCAR and Medicaid Cost Report staffing measures for three staff types. We examined differences between facilities that over-reported or underreported staffing levels in the OSCAR by using logistic regression.

RESULTS

Average staffing levels were higher in the OSCAR than in the Medicaid Cost Report data. The two sets of measures exhibited correlations ranging between 0.5 and 0.6. For-profit and larger facilities consistently over-reported registered nurse staffing levels. Factors associated with increased odds of over-reporting licensed vocational nursing or certified nursing assistant staffing were lower Medicare or Medicaid censuses and less market competition. Facility characteristics associated with over-reporting were consistent across different levels of over-reporting. Underreporting was much less prevalent.

IMPLICATIONS

Certain types of facilities consistently over-report staffing levels. These reporting errors will affect the validity of consumer information systems, regulatory activities, and health services research results, particularly research using OSCAR data to examine the relationship between staffing and quality. Results call for a more accurate reporting system.

Predicting nursing facility transition candidates using AID: a case study

PURPOSE

Although the nursing facility transition literature is growing, little research has analyzed the characteristics of individuals so assisted or compared participants to those who remain institutionalized. This article describes an analytic method that researchers can apply to address these knowledge gaps, using the Arkansas Passages nursing facility transition program as a case study.

DESIGN AND METHODS

This study employed Arkansas Minimum Data Set 2.0 data for 111 transitioned individuals, a derivation sample of 1,000 other residents, and a validation sample of all residents from the transitioned individuals' nursing facilities. Tree classification techniques identified distinct groups of transitioned and nontransitioned residents.

RESULTS

Nearly two thirds of transitionees were part of a group comprising only 1.5% of all Arkansas nursing facility residents. Five characteristics identified this group: age, day of stay (i.e., current day of stay at the time of the assessment), having hemiplegia/paraplegia, cognitive impairment level, and classification into one of eight Resource Utilization Groups (RUG-III) case-mix groups associated with the least nursing staff time. Another group containing 92% of the transitionees comprised 22% of all residents. Two characteristics identified this group: being younger than age 65 or being in the eight low-resource RUG-III groups.

IMPLICATIONS

Given that the majority of individuals assisted by this pilot represent a small and unusual nursing facility subpopulation, policy makers may wish to exercise caution in utilizing these data to forecast future transition populations, costs, or outcomes. Replicating this analysis using additional states' data could increase understanding about the characteristics of those assisted across transition programs and could help construct a more robust definition of what constitutes a transition success.

Prevalence and treatment of diagnosed depression among elderly nursing home residents in Ohio

OBJECTIVES

To examine the prevalence and treatment of diagnosed depression among elderly nursing home residents and determine the resident and facility characteristics associated with diagnosis and treatment.

DESIGN, SETTING, AND PARTICIPANTS

Documented depression, pharmacotherapy, psychotherapy, sociodemographics, and medical characteristics were obtained from Ohio's Minimum Data Set for 76 735 residents in 921 nursing homes. The data were merged with Online Survey Certification and Reporting System data to study the impact of facility characteristics. Chi-squared statistics were used to test group differences in depression diagnosis and treatment. Multiple logistic regressions were used to examine the prevalence of diagnosed depression, and among those diagnosed, of receiving any treatment.

RESULTS

There were 48% of residents who had an active depression diagnosis; among those diagnosed, 23% received no treatment; 74% received antidepressants; 0.5% received psychotherapy; and 2% received both. African Americans, the severely cognitively impaired, and those in government facilities were less likely to be diagnosed. Residents aged 85 and older, African Americans, individuals with severe mental illness, those with severe ADL or cognitive impairment, and individuals living in a facility with 4 or more deficiencies were less likely to receive treatment.

CONCLUSION

Significant disparities exist both in diagnosis and treatment of depression among elderly residents. Disadvantaged groups such as African Americans and residents with physical and cognitive impairments are less likely to be diagnosed and treated. Our results indicate that work needs to be done in the nursing home environment to improve the quality of depression care for all residents.

Cost and Quality

BACKGROUND

Although both quality and cost are important concerns for long term care (LTC) facility management and policy, the relationship between cost and quality is poorly understood. Such knowledge is necessary to guide facility management and policy action.

OBJECTIVE

We sought to determine the net effect of quality on cost in LTC hospital settings.

STUDY SAMPLE

A 4-year panel dataset from April 1997 through March 2002 comprising observations from 99 LTC hospitals were included in this analysis.

METHODS

We examined the relationship between direct resident costs and 7 indicators of quality for long-stay residents. We used panel data methods to control for unobserved facility-level characteristics.

RESULTS

We found that increases in restraint use and incident pressure/skin ulcers were associated with lower per diem costs, whereas incontinence prevalence was associated with higher per diem costs.

CONCLUSIONS

Our results point to different implications regarding cost and quality for different quality indicators. Although facilities have a strong internal business case to improve quality in incontinence, policy-makers may need to provide financial incentives to encourage reductions in restraint use and incident skin ulcers so as to defray potential higher costs associated with improving quality in these areas.

Long-term care market competition and nursing home dementia special care units

BACKGROUND

The proliferation of noninstitutional long-term care (LTC) alternatives, particularly assisted living (AL), raises questions about how the nursing home (NH) industry is affected. Dementia special care units (SCUs), which provided NHs a competitive edge in the past, are of interest because they have traditionally served residents who now have other options.

OBJECTIVES

To quantify the effect of LTC competition on the odds of SCU presence in a NH; among those NHs with an SCU, to quantify the effect of competition on overall acuity.

RESEARCH DESIGN

Cross-sectional with LTC market defined as the NH's county.

SUBJECTS

Nine hundred forty-two NHs, 88 with an SCU, in 122 Texas counties during 2004.

MEASURES

LTC competition was measured by a NH's share of beds, the presence of another NH SCU, the presence of AL, the presence of dementia care in AL, and the presence of home health in the market. The outcomes were the presence of an SCU in a NH and the average level of resident acuity on those SCUs.

RESULTS

Competition from another NH-based SCU was associated with the presence of an SCU in a NH [adjusted odds ratio, 3.3; 95% confidence interval (CI), 1.6-6.6] but there were no associations with other measures. Within NH-based SCUs, mean acuity was higher when there was AL in the market (beta, 1.6; 95% CI, 0.4-2.7), but there was no additional effect of AL dementia beds (beta, 1.6; 95% CI, 0.2-3.0).

CONCLUSIONS

A NH's investment in specialized dementia care is influenced by the behavior of nearby NHs but not yet by other forms of LTC; however, the profile of residents served by SCUs is affected by AL competition.

Hospitalization of nursing home residents: the effects of states' Medicaid payment and bed-hold policies

OBJECTIVE

Hospitalizations of nursing home residents are costly and expose residents to iatrogenic disease and social and psychological harm. Economic constraints imposed by payers of care, predominantly Medicaid policies, are hypothesized to impact hospitalizations.

DATA SOURCES/STUDY SETTING

Federally mandated resident assessments were merged with Medicare claims and eligibility files to determine hospitalizations and death within 150 days of baseline assessment. Nursing home and market characteristics were obtained from the Online Survey Certification and Reporting, and the Area Resource File, respectively. States' average daily Medicaid nursing home payments and bed-hold policies were obtained independently.

STUDY DESIGN

Prospective cohort study of 570,614 older (> or =65-year-old), non-MCO (Medicare Managed Care), long-stay (> or =90 days) residents in 8,997 urban, freestanding nursing homes assessed between April and June 2000, using multilevel models to test the impact of state policies on hospitalizations controlling for resident, nursing home, and market characteristics.

PRINCIPAL FINDINGS

Overall, 99,379 (17.4 percent) residents were hospitalized with rates varying from 8.4 percent in Utah to 24.9 percent in Louisiana. Higher Medicaid per diem was associated with lower odds of hospitalizations (5 percent lower for each $10 above average $103.5, confidence intervals [CI] 0.91-0.99). Hospitalization odds were higher by 36 percent in states with bed-hold policies (CI: 1.12-1.63).

CONCLUSIONS

State Medicaid bed-hold policy and per-diem payment have important implications for nursing home hospitalizations, which are predominantly financed by Medicare. This study emphasizes the importance of properly aligning state Medicaid and federal Medicare policies in regards to the subsidy of acute, maintenance, and preventive care in the nursing home setting.

Hospitalization of nursing home residents: the effects of states' Medicaid payment and bed-hold policies

OBJECTIVE

Hospitalizations of nursing home residents are costly and expose residents to iatrogenic disease and social and psychological harm. Economic constraints imposed by payers of care, predominantly Medicaid policies, are hypothesized to impact hospitalizations.

DATA SOURCES/STUDY SETTING

Federally mandated resident assessments were merged with Medicare claims and eligibility files to determine hospitalizations and death within 150 days of baseline assessment. Nursing home and market characteristics were obtained from the Online Survey Certification and Reporting, and the Area Resource File, respectively. States' average daily Medicaid nursing home payments and bed-hold policies were obtained independently.

STUDY DESIGN

Prospective cohort study of 570,614 older (> or =65-year-old), non-MCO (Medicare Managed Care), long-stay (> or =90 days) residents in 8,997 urban, freestanding nursing homes assessed between April and June 2000, using multilevel models to test the impact of state policies on hospitalizations controlling for resident, nursing home, and market characteristics.

PRINCIPAL FINDINGS

Overall, 99,379 (17.4 percent) residents were hospitalized with rates varying from 8.4 percent in Utah to 24.9 percent in Louisiana. Higher Medicaid per diem was associated with lower odds of hospitalizations (5 percent lower for each $10 above average $103.5, confidence intervals [CI] 0.91-0.99). Hospitalization odds were higher by 36 percent in states with bed-hold policies (CI: 1.12-1.63).

CONCLUSIONS

State Medicaid bed-hold policy and per-diem payment have important implications for nursing home hospitalizations, which are predominantly financed by Medicare. This study emphasizes the importance of properly aligning state Medicaid and federal Medicare policies in regards to the subsidy of acute, maintenance, and preventive care in the nursing home setting.

Effect of staff turnover on staffing: A closer look at registered nurses, licensed vocational nurses, and certified nursing assistants

PURPOSE

We examined the effects of facility and market-level characteristics on staffing levels and turnover rates for direct care staff, and we examined the effect of staff turnover on staffing levels.

DESIGN AND METHODS

We analyzed cross-sectional data from 1,014 Texas nursing homes. Data were from the 2002 Texas Nursing Facility Medicaid Cost Report and the Area Resource File for 2003. After examining factors associated with staff turnover, we tested the significance and impact of staff turnover on staffing levels for registered nurses (RNs), licensed vocational nurses (LVNs) and certified nursing assistants (CNAs).

RESULTS

All three staff types showed strong dependency on resources, such as reimbursement rates and facility payor mix. The ratio of contracted to employed nursing staff as well as RN turnover increased LVN turnover rates. CNA turnover was reduced by higher administrative expenditures and higher CNA wages. Turnover rates significantly reduced staffing levels for RNs and CNAs. LVN staffing levels were not affected by LVN turnover but were influenced by market factors such as availability of LVNs in the county and women in the labor force.

IMPLICATIONS

Staffing levels are not always associated with staff turnover. We conclude that staff turnover is a predictor of RN and CNA staffing levels but that LVN staffing levels are associated with market factors rather than turnover. Therefore, it is important to focus on management initiatives that help reduce CNA and RN turnover and ultimately result in higher nurse staffing levels in nursing homes.

Prospects and Pitfalls: Use of the RAI-HC Assessment by the Department of Veterans Affairs for Home Care Clients

PURPOSE

The U.S. Department of Veterans Affairs has adopted two functional assessment systems that guide care planning: one for nursing home residents (the Resident Assessment Instrument [RAI]) and a compatible one for home care clients (RAI-HC). The purpose of this article is to describe the RAI-HC (often referred to as the Minimum Data Set-Home Care or MDS-HC) and its uses and offer lessons learned from implementation experiences in other settings.

DESIGN AND METHODS

We reviewed implementation challenges associated both with the RAI and the RAI-HC in the United States, Canada, and other adopter countries, and drew on these to suggest lessons for the Department of Veterans Affairs as well as other entities implementing the RAI-HC.

RESULTS

Beyond its clinical utility, there are a number of evidence-based uses for the assessment system. The resident-level data can be aggregated and analyzed, and scales identify clinical conditions and risk for various types of negative outcomes. In addition, the data can be used for other programmatic and research purposes, such as determining eligibility, setting payment rates for contract care, and evaluating clinical interventions. At the same time, there are a number of implementation challenges the Department of Veterans Affairs and other organizations may face.

IMPLICATIONS

Policy makers and program managers in any setting, including state long-term-care programs, who wish to implement an assessment system must anticipate and address a variety of implementation problems with a clear and consistent message from key leadership, adequate training and clinical support for assessors, and appropriate planning and resources for data systems.

Nurse staffing levels and Medicaid reimbursement rates in nursing facilities

OBJECTIVE

To examine the relationship between nursing staffing levels in U.S. nursing homes and state Medicaid reimbursement rates.

DATA SOURCES

Facility staffing, characteristics, and case-mix data were from the federal On-Line Survey Certification and Reporting (OSCAR) system and other data were from public sources.

STUDY DESIGN

Ordinary least squares and two-stage least squares regression analyses were used to separately examine the relationship between registered nurse (RN) and total nursing hours in all U.S. nursing homes in 2002, with two endogenous variables: Medicaid reimbursement rates and resident case mix.

PRINCIPAL FINDINGS

RN hours and total nursing hours were endogenous with Medicaid reimbursement rates and resident case mix. As expected, Medicaid nursing home reimbursement rates were positively related to both RN and total nursing hours. Resident case mix was a positive predictor of RN hours and a negative predictor of total nursing hours. Higher state minimum RN staffing standards was a positive predictor of RN and total nursing hours while for-profit facilities and the percent of Medicaid residents were negative predictors.

CONCLUSIONS

To increase staffing levels, average Medicaid reimbursement rates would need to be substantially increased while higher state minimum RN staffing standards is a stronger positive predictor of RN and total nursing hours.

Explaining direct care resource use of nursing home residents: findings from time studies in four states

OBJECTIVE

To explain variation in direct care resource use (RU) of nursing home residents based on the Resource Utilization Groups III (RUG-III) classification system and other resident- and unit-level explanatory variables.

DATA SOURCES/STUDY SETTING

Primary data were collected on 5,314 nursing home residents in 156 nursing units in 105 facilities from four states (CO, IN, MN, MS) from 1998 to 2004. Study Design. Nurses and other direct care staff recorded resident-specific and other time caring for all residents on sampled nursing units. Care time was linked to resident data from the Minimum Data Set assessment instrument. Major variables were: RUG-III group (34-group), other health and functional conditions, licensed and other professional minutes per day, unlicensed minutes per day, and direct care RU (wage-weighted minutes). Resident- and unit-level relationships were examined through hierarchical linear modeling.

DATA COLLECTION/EXTRACTION METHODS

Time study data were recorded with hand-held computers, verified for accuracy by project staff at the data collection sites and then merged into resident and unit-level data sets.

PRINCIPAL FINDINGS

Resident care time and RU varied between and within nursing units. RUG-III group was related to RU; variables such as length of stay and unit percentage of high acuity residents also were significantly related. Case-mix indices (CMIs) constructed from study data displayed much less variation across RUG-III groups than CMIs from earlier time studies.

CONCLUSIONS

Results from earlier time studies may not be representative of care patterns of Medicaid and private pay residents. New RUG-III CMIs should be developed to better reflect the relative costs of caring for these residents.

Hospice enrollment and evaluation of its causal effect on hospitalization of dying nursing home patients

OBJECTIVE

To examine the patient, nursing home (NH), hospice provider, and local market factors associated with the selection of the Medicare hospice benefit by eligible NH residents, and evaluate the causal effect of hospice on end-of-life hospitalization rates.

DATA SOURCES/STUDY SETTING

Secondary data for 1995-1997 for NH residents.

STUDY DESIGN

This retrospective cohort study includes NH residents in five states (Kansa, Maine, New York, Ohio, South Dakota) who died in the years 1995-1997. Medicare claims identified hospice enrollment and hospitalizations. Geocoding of NHs, hospice providers, and hospitals was used to identify local markets. The two outcome measures are hospice enrollment and hospitalization of NH residents in their last 30 days of life.

DATA COLLECTION/EXTRACTION METHOD

A file was constructed linking MDS assessments to Medicare claims and denominator files, NH provider files (OSCAR), hospice provider of service files, and the area resource file.

PRINCIPAL FINDINGS

Twenty-six percent of hospice and 44 percent of nonhospice residents were hospitalized in their last 30 days of life (odds ratio [OR] 0.45; 95 percent confidence interval [CI]: 0.42-0.48). Adjusting for confounders, hospice patients were less likely than nonhospice residents to be hospitalized (OR 0.47; 95 percent CI: 0.45-0.50). Adding inverse propensity score weighting, hospice patients were still less likely than nonhospice residents to be hospitalized (OR 0.56; 95 percent CI: 0.53-0.61).

CONCLUSIONS

Hospice selection introduces some bias in the evaluation of the causal effect of hospice on end-of-life hospitalization rates. However, even after adjusting for selection bias, hospice does have a powerful effect in reducing end-of-life hospitalization rates.