Minnesota Nursing Homes Research Articles

Association of Pay-for-Performance Reimbursement With Clinical Quality for Minnesota Nursing Homes Residents.

In 2016, Minnesota implemented a new pay-for-performance reimbursement scheme for Medicaid residents in nursing homes, known as Value-Based Reimbursement (VBR). This study seeks to understand whether there is an association between VBR and quality improvement. We use data from 2013 to 2019 including Centers for Medicare and Medicaid Services, Nursing Home Compare, and Long-term care Facts in the US. Using multivariate regression with commuting zone fixed effects, we compare five long-stay and two short-stay clinical quality metrics in Minnesota nursing homes to nursing homes bordering states, before and after VBR was implemented. We find minimal significant changes in quality in Minnesota nursing homes after VBR. Minnesota should reconsider its pay-for-performance efforts.

NOWHERE ELSE TO GO: EFFECTIVENESS OF THE PASRR PROGRAM TO MEET THE NEEDS OF RESIDENTS WITH SMI ADMITTED TO NURSING HOMES

Abstract The number of adults with serious mental illness (SMI) who receive care in nursing homes (NHs) continues to rise. The Preadmission Screening and Resident Review (PASRR) program requires screening for SMI prior to NH placement, in order to avoid inappropriate admission and unnecessary institutional care. We interviewed staff responsible for the processing of PASRR documentation at four NHs in Minnesota (n=15), and obtained and analyzed all completed PASRR-II assessments in Minnesota from 2019 (N=532). PASRR assessments overwhelmingly recommended 24-hour NH care (94.7%) with 94% of assessments indicating a need for mental health services while at the NH. Most NH staff interviewed noted that PASRR is not used in the care planning process and described PASRR as a regulatory hoop. Staff shared that PASRR assessments could provide insight into an individual’s mental health history, current and future needs, and can be helpful in assessing NH capacity to provide such services. Although mental health services provided while at the NH are supposed to be facilitated in partnership with the county, there is a lack of follow-up and NH staff are largely left to deal with SMI in isolation. PASRR assessments are supposed to be a tool for care coordination, but leave the NH as the sole responsible point of contact for residents with SMI. A more integrated PASRR program that better focuses on incorporating PASRR into care planning and mental health service delivery in NHs and the broader community is necessary to improve the lives of individuals with SMI.

Ownership Change and Care Quality: Lessons From Minnesota's Experience With Value-Based Purchasing.

Background and ObjectivesMinnesota’s implementation of a new nursing home value-based reimbursement (VBR) system in 2016 presented an opportunity to compare the response of nursing homes (NHs) to financial incentives to improve their quality and efficiency. The state substantially increased reimbursement for care-related costs and tied this rate increase to a composite quality score. Coinciding with rate increases of the new VBR system was an increase in ownership changes, with new owners being primarily for-profit entities from outside of Minnesota, including several private equity firms. Our objective was to examine NHs that underwent a change in ownership to determine their cost and quality response to the change.Research Design and MethodsOur sample consists of 342 Minnesota NHs that submitted Medicaid cost reports each year from 2013 to 2019. A time differential two-way fixed-effects difference-in-difference model is used to assess changes in quality metrics by comparing measures in years prior to and years following the sale for NHs that changed ownership versus NHs with consistent ownership. Nursing home characteristics, revenue, and spending patterns are examined to understand differences in performance.ResultsThose NHs with ownership change experienced a decline in quality scores with notable changes to expenditure patterns. They performed worse on Minnesota Department of Health inspection scores and had nonsignificant declines in measures of quality of life and clinical care. They had declining staff dental and medical benefits and occupancy rates, greater revenue growth from Medicare Part B, and larger increases in administrative management fees.Discussion and ImplicationsMinnesota like many other states has given wide latitude for nursing home ownership changes, without specific oversight for the quality of care and expenditure patterns of new owners. Recommendations include strict guidelines for the transparency of ownership structures, quality performance targets, rigorous financial auditing, and enhanced regulatory oversight.

Evaluation of Minnesota’s Long-Stay Nursing Home Quality Indicators

Older adults with bipolar disorder are at increased risk of developing dementia. The literature suggests lithium treatment may reduce the incidence of dementia. This study sought to inform clinical practice in the Veterans Affairs (VA) health system by estimating the effect of past year lithium receipt on dementia incidence among Veterans with bipolar disorder. Divalproex receipt was used as a comparison. Using VA medical records, 121,094 Veterans aged 50 and older with a diagnosis of bipolar disorder but no dementia diagnosis were identified in fiscal years 2005-2019. Follow-up continued until dementia diagnosis, 36 months from the index date, death, or the end of fiscal year 2020, whichever came first. 4347 (3.6%) were diagnosed with dementia during follow-up. Time-varying indicators of receipt of lithium and divalproex in the prior 365 days were calculated for each day, categorized as 301-365, 61-300, 1-60, or 0 days of receipt. Unadjusted Cox proportional hazards regression analyses indicated reduced dementia incidence with 301-365 (HR=0.86, 95% Confidence Interval [95%CI] 0.75-0.99) and 61-300 (HR=0.75, 95%CI 0.65-0.87) days of lithium receipt, compared to 0 days. For divalproex, 301-365 (HR=1.34, 95%CI 1.23-1.47) and 61-300 (HR=1.13, 95%CI 1.03-1.23) days of receipt were each associated with increased dementia incidence. Lithium effects were not statistically significant after adjusting for age, sex, race, ethnicity, medical comorbidities, and antidepressant, antipsychotic, and anxiolytic medication receipt. Divalproex effects remained statistically significant. Past year divalproex, but not lithium, receipt was significantly associated with dementia incidence among VA patients with bipolar disorder when adjusting for demographics and medical comorbidities.

Quality-of-Life Scores for Nursing Home Residents Are Stable Over Time: Evidence From Minnesota

ObjectiveQuality of life (QoL) is a multidimensional construct that assesses the quality of lived experience in nursing homes (NHs). QoL is directly important to NH residents. However, QoL is only publicly reported in a few states, partly because of concerns regarding measure stability. To address these concerns, we tested the stability of Minnesota’s NH QoL measure over one year.Study DesignA pair of two-year cohorts of Minnesota NH residents who responded to the 2012-2013 (N = 4,448) or 2014-2015 (N = 4,644) QoL survey in consecutive years. Stability was measured using the intra-class correlation (ICC) from hierarchical linear models. Models were fit without any covariates, then individual and facility-level characteristics were added. Principal Findings: Overall QoL had ICCs of 0.602 and 0.614 in the earlier and later cohort respectively. Domain-level ICCs were lower, ranging from 0.374 (positive mood) to 0.571 (lack of negative mood) in the 2012-2013 cohort, with similar trends for the later cohort. Adjusting for important covariates reduces the ICCs slightly, but they remained at 0.565 or higher for the summary score.ConclusionsPerson-reported summary QoL has adequate stability over a period of one year. Our results provide impetus to assess and report NH QoL on a national level. Consumers can be confident that if an NH’s QoL scores improve from year to year, that represents a real improvement, and not just the scores varying due to which residents were sampled. Some caution, however, is warranted when presenting facility-level domain scores, as these are less stable.

Longitudinal Comparison of Stability and Sensitivity in Quality of Life Scores Among Nursing Home Residents With and Without Diagnoses of Alzheimer's Disease and Related Dementias.

Background and ObjectivesPrevalence of nursing home residents with Alzheimer’s disease and related dementias (ADRD) has increased along with a growing consensus that person-centered ADRD care in nursing homes should maximize quality of life (QoL). However, concerns about whether residents with ADRD can make appropriate QoL judgments persist. This study assesses the stability and sensitivity of a self-reported, multidomain well-being QoL measure for nursing home residents with and without ADRD.Research Design and MethodsThis study linked the 2012–2015 Minnesota Nursing Home Resident QoL and Satisfaction with Care Survey, Minimum Data Set 3.0 (nursing home assessments), and Minnesota Department of Human Services Cost Reports. The QoL survey included cohort–resident pairs who participated for 2 consecutive years (N = 12 949; 8 803 unique residents from 2012–2013, 2013–2014, and 2014–2015 cohorts). Change in QoL between 2 years was conceptualized as stable when within 1.5 SD of the sample average. We used linear probability models to estimate associations of ADRD/Cognitive Function Scale status with the stability of QoL summary and domain scores (eg, social engagement) and the absolute change in QoL summary score, controlling for resident and facility characteristics.ResultsMost (86.82%) residents had stable QoL summary scores. Residents with moderate to severe cognitive impairment, irrespective of ADRD, were less likely to have stable summary scores than cognitively capable residents without ADRD (p < .001), but associations varied by QoL domains. Among those with stable summary QoL scores, changes in health/functional status were associated with absolute changes in summary QoL score (p < .001), suggesting sensitivity of the QoL measure.Discussion and ImplicationsQoL scores were similarly stable over time for most residents with and without ADRD diagnoses and were sensitive to changes in health/functional status. This self-reported QoL measure may be appropriate for nursing home residents, regardless of ADRD diagnosis, and can efficaciously be recommended to other states.

The Relationships of Nursing Home Culture Change Practices With Resident Quality of Life and Family Satisfaction: Toward a More Nuanced Understanding.

Transforming nursing homes (NHs) from restrictive institutions to person-centered homes, referred to as NH culture change, is complex and multifaceted. This study, based on a survey of administrators in Minnesota NHs (n = 102), tested the domain-specific relationships of culture change practices with resident quality of life (QOL) and family satisfaction, and examined the moderating effect of small-home or household models on these relationships. The findings revealed that culture change operationalized through physical environment transformation, staff empowerment, staff leadership, and end-of-life care was positively associated with at least one domain of resident QOL and family satisfaction, while staff empowerment had the most extensive effects. Implementing small-home and household models had a buffering effect on the positive relationships between staff empowerment and the outcomes. The findings provide meaningful implications for designing and implementing NH culture change practices that best benefit residents’ QOL and improve family satisfaction.

Understanding Nursing Home Responses to Minnesota’s New Value Based Reimbursement System

Even though Value Based Reimbursement (VBR) systems for nursing homes (NH) continue to expand, we have little understanding of how NH respond to VBR. In 2016, Minnesota passed VBR legislation for NHs that increased care-related funding and tied increases to a facility’s composite quality score. While care-related expenditures increased with VBR, the incentive for quality did not work as intended. We investigated the differential responses of facilities in their care-related expenditures and quality scores. Data were derived from cost reports and quality measures for the years 2013-2017 from 300 free-standing Minnesota NHs. Latent Class Growth Analysis was used to cluster facilities by their joint care-related cost and quality score trajectories over the period. Three interpretable trajectory clusters emerged: medium-to-high cost and medium-to-high quality (n=172), low cost and medium-to-high quality (n=54), and low cost and low quality (n=74), all during the pre-VBR period. In all three clusters cost rose significantly with VBR, but only in the low cost and low quality cluster did quality also rise significantly. The quality improving cluster had the highest percentage of government-owned and rural facilities as well as the largest annual increase in care related spending. The medium-to-high cost and medium-to-high quality cluster had the highest concentration of urban facilities (Twin City Metro Area) and were the most likely to be non-profit and chain owned. Although the new VBR system appeared effective in achieving its goals for a subset of facilities with lowest cost and quality, the majority of facilities increased care-related costs without improved quality.

Unmet and Unimportant Preferences Among Nursing Home Residents: What Are Key Resident and Facility Factors?

ObjectivesThe Preferences Assessment Tool (PAT) in the Minimum Data Set (MDS) 3.0 assesses 16 resident preferences for daily routines and activities. Although integrating important preferences into care planning is essential to provide person-centered care in nursing homes (NHs), preferences rated as important but unmet or unimportant may not receive much attention. This study aims to (1) identify the prevalence of unmet preferences and unimportant preferences, and (2) examine their associations with resident and facility-level characteristics. DesignThis is a longitudinal study of residents in NHs. Settings and ParticipantsWe used data from 2012–2017 MDS assessments of long-stay residents aged 65 or older in 295 Minnesota NHs. In total, 51,859 assessments from 25,668 residents were included. MethodsGeneralized linear mixed models were used to analyze resident and facility-level characteristics associated with having any unmet preferences, and with the number of unimportant preferences. ResultsAcross all years for both daily routine preferences and activity preferences, 3.3% to 5.1% of residents reported that at least 1 or more preference was important but unmet, and 10.0% to 16.6% reported that 4 or more out of the 8 preferences were unimportant. Residents with higher depressive symptoms, and poorer physical and sensory function were more likely to report unmet preferences. Residents with poorer physical and sensory function, and living in rural facilities and facilities having fewer activity staff hours per resident day were more likely to report unimportant preferences. Conclusions and ImplicationsResidents with functional and sensory limitations and living in underresourced NHs are more likely to report that preferences are unimportant, or that they are important but unmet. It is important for staff to elicit preferences that truly matter for residents, and to enable residents to meet their preferences.

An Empirical Typology of Nursing Home Culture Change Implementation.

Culture change in nursing homes (NHs) is a broad-based effort to transform NHs from impersonal institutions to genuine person-centered homes. Culture change practices have been implemented increasingly with varying levels of success. This study (a) generated an empirical typology of culture change implementation across Minnesota NHs using latent profile analysis based on the survey data from administrators in 102 NHs and (b) examined variations in NH characteristics and quality outcomes associated with the typology. Three types of culture change implementation were identified: high performers, average performers, and low performers. The distributions of culture change scores were distinct across the three types, with low performers lagging far behind others in family and community engagement, and end-of-life care. High performers were distinguished through demonstrating better resident quality of life and higher family satisfaction. The findings provide empirical support for policymakers, providers, and advocates to direct culture change expansion and resource allocation.

Occupational differences in workers' compensation indemnity claims among direct care workers in Minnesota nursing homes, 2005-2016.

Nursing assistants have one of the highest injury rates in the U.S., but few population-based studies assess differential injury risk by occupation in nursing homes. This statewide study assessed differences in musculoskeletal disorders (MSDs) and patient handling injuries among direct care workers in Minnesota nursing homes. Indemnity claims from the Minnesota workers' compensation database were matched to time at risk from the Minnesota Nursing Home Report Card to estimate 2005 to 2016 injury and illness claim rates for certified nursing assistants (CNAs), licensed practical nurses (LPNs), and registered nurses (RNs). Associations between occupation and claim characteristics were assessed using multivariable regression modeling. Indemnity claim rates were 3.68, 1.38, and 0.69 per 100 full-time equivalent workers for CNAs, LPNs, and RNs, respectively. Patient handling injuries comprised 62% of claims. Compared to RNs, CNAs had higher odds of an indemnity claim resulting from an MSD (odds ratio [OR] = 1.67; 95% confidence interval [CI], 1.31-2.14) or patient handling injury (OR = 1.89; 95% CI, 1.47-2.45) as opposed to another type of injury or illness. CNAs had lower odds of receiving temporary and permanent partial disability benefits and higher odds of receiving a stipulation settlement. CNAs in Minnesota nursing homes are at heightened risk for lost time MSDs and patient handling injuries. Claims filed by CNAs are more frequently settled outside the regular workers' compensation benefit structure, an indication that the workers' compensation system is not providing adequate and timely benefits to these workers.

AN EMPIRICAL TYPOLOGY OF CULTURE CHANGE IMPLEMENTATION IN NURSING HOMES

The implementation of culture change (CC) practices has been increasing in U.S. nursing homes (NHs) with varying success. No study has examined the typology of CC implementation across NHs using empirical data. This study aimed to identify an empirical typology of CC implementation and to examine NH organizational characteristics associated with types of CC implementation. An online CC survey using a validated instrument was sent to administrators of 363 NHs in Minnesota and 78 administrators completed the survey. No significant difference in NH characteristics was found between participants and non-participants. Using a hierarchical cluster analysis based on scores of five CC domains, this study identified three distinctive clusters. Overall, NHs most frequently engaged in resident-centered care and environment transformation and least frequently implemented family/community engagement practices. While NHs in cluster 1 (20.8%) performed best in all CC domains, they were distinguished from others by more frequency of family/community engagement, staff empowerment and staff leadership practices. NHs in cluster 3 (41.7%) were distinguished by overall lower engagement of CC practices in all domains. While NHs in cluster 2 (37.5%) and cluster 1 were comparable in environment transformation and resident-centered care, NHs in cluster 2 and cluster 3 were similar in their low family/community engagement. NHs in cluster 3 had significantly lower registered nurse retention, and lower mental health and social service staffing compared to NHs in cluster 2. This typology provides insight into the variations in implementing CC practices across NHs and pragmatic implications for promoting deep and extensive CC.

VALIDATING THE PREFERENCES ASSESSMENT TOOL FOR NURSING HOME RESIDENTS USING ITEM RESPONSE THEORY

The Preference Assessment Tool (PAT), part of the Minimum Data Set (MDS), assesses residents’ preferences to enable preference-based care in nursing homes (NHs). The two PAT sections including daily routine preferences and activity preferences are assumed to measure autonomy and meaningful activities as the underlying constructs associated with residents’ psychosocial needs. Yet, the validity of this assumption has not been tested. This study examines PAT’s construct validity using item response theory. We fitted graded response models to the two PAT sections using 2017 MDS annual assessments of 8,829 long-stay residents in 291 Minnesota NHs. Most items discriminated well between residents who had at a low versus high intensity of these two types of psychosocial needs (i.e. have discrimination parameters > 1). Two daily routine preference items (family’s involvement in care planning, and having snacks), and three activity preference items (having pets, keeping up with news, and reading) had poor discrimination in measuring autonomy and meaningful activities respectively. Overall, the PAT appeared to be more sensitive in measuring the lower middle range of the two constructs. We estimated the correlation between the two constructs as 0.65. In conclusion, the PAT performs fairly well in measuring the two types of psychosocial needs for NH residents. Nevertheless, more items are needed to capture a broader range of psychosocial needs beyond autonomy or basic leisure activities. The findings of this study brought attention to the utility of the PAT in measuring residents’ psychosocial needs and in guiding resident-centered care.

GENDER DIFFERENCES IN NURSING HOME RESIDENT QUALITY OF LIFE: WHY WOMEN DO BETTER

Despite research documenting gender differences in numerous outcomes in later life, we know little about gender differences in quality of life (QoL) for older adults who receive institutional long-term care. To address this gap, this study examines the relationship between gender and nursing home (NH) residents’ QoL, including possible reasons for differences observed. We used a mixed methods design including surveys with a random sample of Minnesota NH residents using a multidimensional measure of QoL (n=8,870), resident clinical data, facility-level characteristics, and qualitative interviews with NH residents (n=64). We used mixed models and thematic analysis of resident interviews to examine possible differences in resident QoL based on gender. After controlling for individual and facility characteristics, women reported higher overall QoL than men, with men reporting significantly lower QoL in 4 of 8 QoL domains. In interviews, men noted being especially dissatisfied with facility activities, whereas women more frequently described having friends in the facility and relying on family for support. Some women viewed the NH as a place of respite and described wanting to stay long-term, even though their families asked them to return home. In contrast, men more often described the NH as necessary due to physical needs, but undesirable for long-term living. Our findings provide preliminary evidence that men and women experience QoL differently, with men reporting lower QoL in several domains. Tailoring more activities for men and finding ways to strengthen relationships for men within the facility may help reduce the gender disparities in QoL we observed.

Relationship between nursing home quality indicators and potentially preventable hospitalisation

BackgroundHospitalisations are very common among nursing home residents and many of these are deemed inappropriate or preventable. Little is known about whether clinical care quality is related to hospitalisation, especially...

Racial Differences in Minnesota Nursing Home Residents' Quality of Life: The Importance of Looking Beyond Individual Predictors.

The aim of this study is to investigate racial differences in nursing home (NH) residents' quality of life (QOL) at the resident and facility levels. We used hierarchical linear modeling to identify significant resident- and facility-level predictors for racial differences in six resident-reported QOL domains. Data came from the following: (a) resident-reported QOL (n = 10,929), (b) the Minimum Data Set, and (c) facility-level characteristics from the Minnesota Department of Human Services (n = 376). White residents reported higher QOL in five of six domains, but in full models, individual-level racial differences remained only for food enjoyment. On the facility level, higher percentage of White residents was associated with better scores in three domains, even after adjusting for all characteristics. Racial differences in QOL exist on individual and aggregate levels. Individual differences are mainly explained by health status. The finding that facility racial composition predicts QOL more than individual race underscores the importance of examining NH structural characteristics and practices.

Longitudinal Changes in Nursing Home Resident-Reported Quality of Life: The Role of Facility Characteristics.

Improving quality of nursing homes (NHs) is a major social priority, yet few studies examine the role of facility characteristics for residents' quality of life (QOL). This study goes beyond cross-sectional analyses by examining the predictors of NH residents' QOL on the facility level over time. We used three data sources, namely resident interviews using a multidimensional measure of QOL collected in all Medicaid-certified NHs in Minnesota (N = 369), resident clinical data from the minimum data set, and facility-level characteristics. We examined change in six QOL domains from 2007 to 2010, using random coefficient models. Eighty-one facilities improved across most domains and 85 facilities declined. Size, staffing levels (especially activities staff), and resident case mix are some of the most salient predictors of QOL over time, but predictors differ by facility performance status. Understanding the predictors of facility QOL over time can help identify facility characteristics most appropriate for targeting with policy and programmatic interventions.

Effect of Urinary Incontinence on Older Nursing Home Residents’ Self‐Reported Quality of Life

To evaluate the effect of urinary incontinence (UI) on elderly nursing home (NH) residents' self-reported quality of life (QOL), especially on the specific QOL domains most closely associated with UI. Retrospective cross-sectional study using the Minimum Data Set (MDS) and Minnesota Nursing Home Report Card data in 2010 to compare overall QOL and its domains of residents with and without UI using propensity scores and inverse probability weighting (IPW) adjustments to reduce selection bias. All Medicare- or Medicaid-licensed NHs in Minnesota. All residents aged 65 and older except those with conditions that could readily overwhelm the effect of UI on QOL, such as coma and cerebral palsy. Urinary incontinence (UI) was defined as leakage two or more times a week (score ≥ 2 on MDS 5-point scale), and continence was defined as continent or usually continent (score 0 or 1 on MDS 5-point scale). QOL was assessed using a self-reported QOL questionnaire that measured general QOL, not QOL specific to UI symptoms. Urinary incontinence (UI) prevalence was 65.8% in 10,683 older NH residents. Self-reported QOL was good (>0.7 on a scale from 0 to 1) in 8,620 eligible residents in 371 NHs. Mood and meaningful activity domains had lower scores; dignity had the highest score. UI was associated with being older and female, ADL dependence, impaired cognitive ability, Alzheimer's disease, non-Alzheimer's dementia, bowel incontinence, diabetes mellitus, and long-term NH stay. Bivariate analysis found that residents with UI had lower QOL than those without. Using IPW to reduce selection bias, it was found that, although UI was not associated with overall QOL, it decreased the QOL domains of dignity, autonomy, and mood. To improve the QOL of residents with UI, attention should be paid to dignity, autonomy, and mood.

The Influence of Cognitive Impairment, Special Care Unit Placement, and Nursing Facility Characteristics on Resident Quality of Life

We examined the (a) influence of nursing facility characteristics on resident quality of life and (b) the impact of cognitive impairment and residence on a dementia special care unit(SCU) on QOL after controlling for resident and facility characteristics. Multilevel models (resident and facility) were estimated for residents with and without cognitive impairment on conventional units and dementia SCU. Data came from the 2007 Minnesota Nursing Home Resident Quality of Life and Consumer Satisfaction Survey (N = 13,983). Level of resident CI was negatively related to QOL, although residing on a dementia SCU was positively related to QOL. Certified Nursing Assistant and activity personnel hours per resident day had a positive relationship with resident QOL. Our results highlight the need to ensure adequate levels of paraprofessional direct care staff and the availability of dementia-focused (SCU)s despite current constraints on long-term care funding.

Falls and Nursing Home Residents With Cognitive Impairment: New Insights into Quality Measures and Interventions

Prevention and public reporting of falls have suffered due to inadequate attention given to the association of falls and cognitive impairment (CI) among nursing home (NH) residents. This study examines the relationship between CI, residence on dementia special care units (SCUs) and other resident characteristics and likelihood of residents experiencing new falls in NHs. Retrospective cohort study. A total of 21,587 residents from 381 Minnesota NHs. The NH Minimum Data Set (MDS) for 21,587 residents from 381 Minnesota NHs in the first calendar quarter of 2008 were analyzed. New falls, (fall noted on a current MDS assessment but not on a prior assessment); cognitive status, (as defined by Cognitive Performance Scale); residence on an SCU, and health and functional status covariates were recorded. A random effects logistic regression model was used to examine relationships between new falls and the resident's cognitive status, type of unit, and covariates. The likelihood of a new fall had a nonlinear association with CI. Compared with residents with normal or mild CI, the likelihood of a new fall was significantly higher among residents with moderate CI (OR= 1.43). The risk decreased slightly (OR= 1.34) for residents with more advanced CI, whereas the presence of severe CI was not significantly associated with new falls. Overall the likelihood of new falls was significantly higher for residents on SCUs compared with those on conventional units (OR= 1.27). Severity of CI and residence on SCU impact fall incidence and should be accounted for in future fall- prevention interventions and quality-reporting indicators and measures.