ED Transfers Less Likely Among Nursing Home Residents With Do-Not-Hospitalize Orders?

ED Transfers Less Likely Among Nursing Home Residents With Do-Not-Hospitalize Orders?

JAMDA xxx (2017) 1e4 JAMDA journal homepage: www.jamda.com Original Study Are Hospital/ED Transfers Less Likely Among Nursing Home Residents With D...

219KB Sizes 0 Downloads 76 Views

JAMDA xxx (2017) 1e4

JAMDA journal homepage: www.jamda.com

Original Study

Are Hospital/ED Transfers Less Likely Among Nursing Home Residents With Do-Not-Hospitalize Orders? Taeko Nakashima PhD, SW a, b, *, Yuchi Young DrPH a, Wan-Hsiang Hsu PhD c a

Department of Health Policy, Management, and Behavior, School of Public Health, State University of New York at Albany, Albany, NY Department of Economics, Rutgers University, Camden, NJ c New York State Department of Health, Bureau of Environmental & Occupational Epidemiology, Albany, NY b

a b s t r a c t Keywords: Advance directive do-not-hospitalize (DNH) orders hospital/ED transfer dementia nursing home

Objectives: This study aims to examine whether an advance directive “Do Not Hospitalize” (DNH) would be effective in reducing hospital/emergency department (ED) transfers. Similar effects in residents with dementia were also examined. Design: Cross-sectional study. Setting/subjects: New York State (NYS) nursing home residents (n ¼ 43,024). Measurements and analysis: The Minimum Data Set 2.0 was used to address the study aims. Advance directives with an indication of DNH and Alzheimer disease/dementia other than Alzheimer disease were coded (yes vs no). Logistic regression analyses were performed to quantify the relationship between DNH orders and hospital/ED transfers while adjusting for confounders. Results: Our results show that 61% of nursing home residents had do-not-resuscitate orders, 12% had feeding restrictions, and only 6% had DNH orders. Residents with DNH orders had significantly fewer hospital stays (3.0% vs 6.8%, P <.0001) and ED visits (2.8% vs 3.6%, P ¼ .03) in the last 90 days than those without DNH orders. Dementia residents with DNH orders had significantly fewer hospital stays (2.7% vs 6.3%, P < .0001) but not ED visits (2.8% vs 3.5%, P ¼ .11) than those without DNH orders. After adjusting for covariates in the model, the results show that for residents without DNH orders, the odds of being transferred to a hospital was significantly higher (odds ratio ¼ 2.23, 95% confidence interval ¼ 1.77e2.81) than those with DNH orders. Conclusion: Residents with DNH orders had significantly fewer transfers. This suggests that residents’ end-of-life care decisions were respected and honored. Efforts should be made to encourage nursing home residents to complete DNH orders to promote integration of the resident’s values and goals in guiding care provision toward the end of life. Ó 2016 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

Advance directives are intended to ensure that patients’ end-of-life care wishes are carried out when the individual lacks decision-making capacity. End-of-life care decisions are an ongoing process in which patients, their families, and their health care providers must answer difficult questions such as “which medical treatment is right for me near the end of my life?” Options for continuation or withdrawal of treatments include do-not-resuscitate orders, do-not-hospitalize (DNH), living wills, and durable power of attorney for health care, etc. Studies have found that advance care planning on selected outcomes is beneficial, especially when it is done early on following

The authors declare no conflicts of interest. * Address correspondence to Taeko Nakashima, PhD, SW, 330 Armitage Hall 311 North 5th Street, Camden, NJ 08102. E-mail address: [email protected] (T. Nakashima). http://dx.doi.org/10.1016/j.jamda.2016.12.004 1525-8610/Ó 2016 AMDA e The Society for Post-Acute and Long-Term Care Medicine.

admission to a nursing home or hospice.1,2 Programs to encourage advance care planning have shown positive results: patients are more likely to get palliative care referrals, improved end-of-life care, and reductions in the number of unnecessary hospitalizations,1,2 which are in turn associated with significant cost savings.2 Another study showed that 93% of people with living wills were more willing to use limited care, indicating the potential for advance care planning to reduce end-of-life health care costs.3 Patients who have an advance directive at the time of death have substantially higher comfort levels during the dying process, even after controlling for sociodemographic characteristics, and relatives also report less emotional distress.4 Dementia has become a global public health priority.5 In the United States, on average two-thirds of deaths of people with Alzheimer or other dementia-related diseases occur in nursing homes.6 And more than a quarter of elderly adults may require surrogate decision making at the end of life.3 It is difficult for caregivers,

2

T. Nakashima et al. / JAMDA xxx (2017) 1e4

surrogates, and care providers who take care of people with dementia to make decisions without advance directives. Advance directives are also important for people with dementia and should be made when the person with dementia still has legal and cognitive capacity. Comfort is the goal of care for the majority of older adults as well as people with dementia; however, this goal is seldom achieved with hospitalization.7 Approximately 25% of nursing home residents with advanced dementia will have a hospital transfer in the last 6 months of life.8 The benefit of hospital transfers for people with advanced dementia is uncertain because the dementia patients often receive burdensome and costly interventions with limited benefit such as tube feeding. 9e13 Despite these benefits, advance care planning is often underutilized. One study of elderly patients in hospice found that only 27% had a living will.14 An even smaller percentage, specifically 21%, had a living will if they were not in hospice.14 A study in Maryland of nursing home residents with advanced dementia found that 34% of residents lacked advance directives in their medical charts.15 DNH orders appear to be rare, occurring in just 7.1% of nursing home residents with advanced dementia nationwide.16 Advance directives are more likely if the patient is older, white, and well educated.15 There are a number of factors that lead to underutilization of advance care planning. DNH orders are less likely to be given if physicians are not closely involved with the patient or do not understand DNH orders well.17 Health care providers have incentives not to implement advance care plans that reduce the number of treatments provided, because this will also reduce the amounts they are paid.18 Even when advance care plans are present, the directives often contain vague or inconsistent language that may lead nurses and physicians to fall back on a default pattern of always hospitalizing a patient when a problem arises.19 Few studies have examined the relationship between DNH orders and their relationship with hospitalizations and emergency department (ED) admissions, and few studies have compared DNH orders among nursing home residents with Alzheimer disease on hospital and ED admissions. The current study examines crosssectional data from the Minimum Data Set 2.0 (MDS2.0) to compare overall nursing home residents and residents with and without dementia with the presence of advance directives and DNH orders on hospital and ED transfers. These results will help health care providers determine which groups are most in need of advance care planning and suggest ways to increase the utilization of advance care planning in the future.

Study Variables MDS2.0 included information on age (in years), gender, activities of daily living (ADL), commodities, advance directives, number of hospital stays, and ED visits. All ADL functional disability measures in MDS2.0 are self-performance on a scale from 0 (independent) to 4 (total dependence). For purposes of this study, if a resident’s ADL item was recorded as “activity did not occur during entire 7 days,” that ADL was considered missing. The overall ADL was computed as bathing, toileting, dress, personal hygiene, transfer, and eating. The presence of the following comorbidities was coded using binary variables: dementia, hypertension, depression, arthritis, heart failure, and asthma/ chronic obstructive pulmonary disease (COPD). Dementia status was coded as yes vs no by Alzheimer disease or dementia other than Alzheimer disease. Advance directives (a binary variable) included living will, do-not-resuscitate orders, DNH orders, organ donation, autopsy request, feeding restrictions, medication restrictions, other treatment restrictions, or none of the above. Hospital stays were defined by whether a resident was admitted to a hospital 1 or more days with an overnight stay in the last 90 days. ED visits were defined by whether residents visited an ED without an overnight stay 1 or more times in the last 90 days. Data Analysis Frequency distributions and bivariate analyses with chi-square tests were used to organize and summarize the data. Logistic regressions were performed to quantify the effects of DNH orders for all residents and among residents with dementia status on hospital/ED transfers while adjusting for age, gender, overall ADL, and comorbidity (hypertension, depression, arthritis, heart failure, and asthma/COPD). Analyses were conducted using Statistical Analysis Software, version 9.3 (SAS Institute, Inc, Cary, NC). The significance level was set at alpha ¼ 0.05 (2-tailed). Results Table 1 presents sociodemographic and selected health conditions of the study population. Of the 43,024 residents, the mean age was 84.2 years, and 72.2% were female. The average overall ADL was 2.7, and 62.2% had dementia. The prevalence of chronic conditions were common, including hypertension (69.1%), depression (48.9%), arthritis (29.6%), heart failure (18.8%), and asthma/COPD (17.8%). The number of hospital stays was 2,814 and that of ED visits was 1,526 in the last 90 days. In comparison to nondementia residents, dementia residents

Methods Study Design This was a cross-sectional study. Inclusion criteria for the study population were older adults (aged 65 years) who received care from a nursing home in New York State and had an annual assessment in 2010. The study was approved by the Internal Review Board of the University at Albany.

Data Source The MDS2.0 from 2010 contains assessment forms for all patients residing in a Medicare- and Medicaid-certified nursing home in New York State. For residents with more than 1 annual assessment, we retain the latest record for the analysis. The data set was provided by the NYS Department of Health (NYSDOH); unique resident identifiers were removed before being given to the researchers.

Table 1 Selected Sociodemographic and Health-Related Characteristics of the Study Population (n ¼ 43,024) Variables Sociodemographic Age, mean (SD) Male Overall ADL, mean (SD) Health characteristics Dementia Hypertension Depression Arthritis Heart failure Asthma/COPD Hospital stays (1 d) ED visits (1 visit)

Total*

Dementia*

82.5 (9.1) 5,027 (11.7) 2.5 (1.0)

<.0001 <.0001 <.0001

26,756 (62.2) 16,268 (37.8) (69.1) 18,764 (43.6) 10,982 (25.5) (48.9) 13,379 (31.1) 7,649 (17.8) (29.6) 8,342 (19.4) 4,402 (10.2) (18.8) 4,636 (10.8) 3,464 (8.1) (17.8) 4,277 (9.9) 3,385 (7.9) (6.5) 1,612 (3.8) 1,202 (2.8) (3.6) 928 (2.2) 598 (1.4)

<.0001 <.0001 <.0001 <.0001 <.0001 <.0001 <.0001 .26

84.2 (8.5) 11,963 (27.8) 2.7 (1.0)

29,746 21,028 12,744 8,100 7,662 2,814 1,526

Nondementia* P Value

85.3 (7.9) 6,936 (16.1) 2.8 (0.9)

SD, standard deviation. *Values are n (%) unless otherwise noted. Components may not sum to totals because of rounding.

T. Nakashima et al. / JAMDA xxx (2017) 1e4

were significantly older (85.3 vs 82.5, P < .0001), had a higher proportion of men (16.1% vs 11.7%, P < .0001), and had higher ADL disability scores (2.8 vs 2.5, P < .0001). On selected health conditions, dementia residents had significantly higher prevalence of hypertension (43.6% vs 25.5%, P < .0001), depression (31.1% vs 17.8%, P < .0001), and arthritis (19.4% vs 10.2%, P < .0001). For hospital stays, dementia residents had a higher number of hospital stays (1,612 vs 1,202, P < .0001). Thirty-six percent of residents did not have an advance directive (Table 2). Of these, 64% had at least one document related to advance directives, the most prevalent forms being do-not-resuscitate orders (61%), feeding restrictions (12.3%), and living wills (11.2%), and only 6.1% had DNH orders. In general, the residents with dementia were significantly more likely to have living wills, do-not-resuscitate orders, DNH orders, feeding restrictions, medication restrictions, or other treatment restrictions. The bivariate results in Table 3 show that residents with DNH orders had significantly fewer hospital stays in the past 90 days than those without DNH orders (3.0% vs 6.8%, P < .0001). Likewise, residents with DNH orders also had significantly fewer ED visits (2.8% vs 3.6%, P ¼ .03). Dementia residents with DNH orders had significantly fewer hospital stays than those without DNH orders (2.7% vs 6.3%, P < .0001) but no difference in ED visits (2.8% vs 3.5%, P ¼ .11). The results of multivariate logistic regression analysis in Table 4 indicate that the increased odds of being admitted to a hospital were significantly associated with no DNH orders [odds ratio (OR) ¼ 2.23, 95% confidence interval (CI) ¼ 1.77e2.81]. Residents with dementia compared to their counterparts had higher odds of hospital transfers (OR ¼ 1.21, 95% CI ¼ 1.11e1.31). There were no statistically significant differences between ED admissions on DNH orders (OR ¼ 1.25, 95% CI ¼ 0.98e1.59) for the overall population or by dementia status (OR ¼ 1.03, 95% CI ¼ 0.92e1.15). Among residents with dementia, the increased odds of being admitted to hospital were significantly associated with no DNH orders (OR ¼ 2.36, 95% CI ¼ 1.78e3.14) but the odds of ED visits on DNH orders (OR ¼ 1.20, 95% CI ¼ 0.9e1.59) were not statistically significant.

Discussion Our study results indicate that advance directives are significantly underutilized among older adults in New York State nursing homes. Approximately 36% of nursing home residents had no advance directives at all. Our finding is comparable to a previous study.20 Williams and associates found that in nursing homes in 2000 the prevalence of having any advance directive at admission was 56.7% and at 12 months after admission it was 64.9%.20

Table 2 Prevalence of Having Advance Directives Among Nursing Home Residents (n ¼ 43,024) Variables

Total*

Dementia*

Advance directives Living will 4,807 (11.2) 3,122 (7.3) Do not resuscitate 26,226 (61.0) 16,993 (39.5) Do not hospitalize 2,626 (6.1) 1,911 (4.4) Organ donation 242 (0.6) 144 (0.3) Autopsy request 31 (0.1) 18 (0.04) Feeding restrictions 5,269 (12.3) 3,557 (8.3) Medication 1,424 (3.3) 1,055 (2.5) restrictions Other treatment 5,542 (12.9) 3,737 (8.7) restrictions No advance directives 15,487 (36.0) 9,038 (21.0) (none of above)

Nondementia* P Value 1,685 9,233 715 98 13 1,712 369

(3.9) (21.5) (1.7) (0.2) (0.03) (4.0) (0.9)

<.0001 <.0001 <.0001 .39 .64 <.0001 <.0001

1,805 (4.2)

<.0001

6,449 (15.0)

<.0001

*Values are n (%). Components may not sum to totals because of rounding.

3

Table 3 Comparison of Nursing Home Residents With and Without DNH Orders (Unadjusted) Population

All residents (n ¼ 43,024) Dementia (n ¼ 26,756)

DNH Order

Hospital Stays

ED Visits

n (%)

P Value

n (%)

Yes No Yes No

78 2,736 51 1,561

<.0001

73 1,453 54 874

(3.0) (6.8) (2.7) (6.3)

<.0001

P Value (2.8) (3.6) (2.8) (3.5)

.03 .11

DNH orders were relatively uncommon, being found in only 6.1% in our study. Of those with dementia, 4.4% had DNH orders, which is below the national average of 7.1%. There is a wide range in DNH orders from 0.7% in Oklahoma to 25.7% in Rhode Island.16 Decisions to forgo hospitalizations are unconventional even for residents with advanced dementia. Mitchell and associates found that this could partially be explained by facility characteristics (eg, corporate chain, staffing ratio, and geographical location) and patients’ demographics and health status.16 Our study results indicate that having a DNH order seemed to be effective in reducing hospital stays from 6.8% to 3.0% and ED visits from 3.6% to 2.8% compared to those without DNH orders. The percentage of reduction may be small, but it has significant clinical and policy implications. The majority of hospitalizations/ED admissions are necessary but many are not. Unnecessary hospitalizations are a strong indicator of poor quality of care and a fragmented health care delivery system. Almost 50% of hospital admissions for nursing home residents in their last year of life were for potentially avoidable diagnoses, costing Medicare $1 billion.21 In 2012, there were 36.5 million hospital admissions, with an average length of stay of 4.5 days and an average cost of $10,400 per stay.22 In 2014, there were about 1.4 million nursing home residents with various levels of functional and mental disability.23 Our study results indicate that residents without DNH orders had a 3.8% greater number of hospital transfers than those with DNH orders. If we apply this to 1.4 million nursing home residents to 3.8%, this would equate to 53,200 hospital transfers that may have been averted per year had all residents filled out DNH orders and their wishes honored. This hospital transfer reduction would translate to about $553 million in cost savings (53,200  $10,400 ¼ 553,280,000). The implications of this potential cost savings from having DNH orders would include the following: (1) the patients’ wishes for DNH orders were respected and honored; (2) palliative or hospice care may be provided on site as directed by the residents’ advance directives; (3) unnecessary hospitalizations that might expose patient to iatrogenic complications are averted; and (4) costs saved can be used in other needed care areas. Table 4 Multivariate Logistic Regression Analysis Results on the Effect of DNH Orders and Hospital and ED Admissions* Variables

All residents (n ¼ 43,024) DNH order (ref ¼ DNH) Dementia status (ref ¼ dementia) Dementia (n ¼ 26,756) DNH order (ref ¼ DNH)

Hospital Stays

ED Visits

OR

95% CI

P Value

OR

95% CI

P Value

2.23

1.77e2.81

<.0001

1.25

0.98e1.59

.07

1.21

1.11e1.31

<.0001

1.03

0.92e1.15

.60

2.36

1.78e3.14

<.0001

1.20

0.9e1.59

.21

*Controlling for age, gender, overall ADL, hypertension, depression, arthritis, heart failure, asthma/COPD.

4

T. Nakashima et al. / JAMDA xxx (2017) 1e4

Results from a multivariate logistic regression analysis found that having DNH orders were associated with fewer hospital admissions. An association between DNH orders and fewer ED visits was detected but did not reach significance (P ¼ .07). For elderly nursing home residents without DNH orders, the odds of being admitted to hospital was 2.2 times higher than the odds for elderly residents with DNH orders, while adjusting for covariates in the model. This fact indicates that DNH orders were generally honored and were an effective means of reducing hospitalizations that might not have been necessary. This is consistent with studies indicating that programs to increase advance directives are effective at reducing the number of hospital stays.1,2 Among residents with dementia, DNH orders were associated with fewer hospital stays, although they made no difference for ED visits. This is consistent with another study that showed that DNH orders were the only identified modifiable factor associated with avoiding hospitalization for people with advanced dementia.24 Hospital transfer for people with dementia has a greater risk rather than others of receiving burdensome intervention such as restraints and intravenous lines,13 and these transfers can be anxiety-provoking because of the stress of unfamiliar surroundings. Reducing undesired hospitalization using advance directives such as DNH orders is important for nursing home residents with dementia. In addition to the reduction of unnecessary and undesired hospitalizations, written advance directives of the deceased was associated with the quality of dying reported by relatives/caregivers that relative/caregivers experienced lower levels of emotional distress at the patient’s end of life.4 The present study also compares the comfort assessment in dying with dementia as measured by the resident’s caregivers/relatives on DNH orders (yes vs no); the difference did not reach the level of significance.3 Given that advance directives and DNH orders improve quality of life of dying individuals and lessens their caregivers’/relatives’ emotional distress at the patient’s end of life,1,4 efforts should be made to encourage and facilitate residents to complete advance directives, when possible, to promote integration of the individual’s values and goals in guiding care toward the end of life. Advance care planning and DNH orders are not easy subjects to discuss for all parties concerned. Advance care planning is an ongoing process in which patients, caregivers, health care proxies, and health care providers reflect on the patient’s goals, beliefs, and wishes related to the patient’s future care, and it will become the official document to inform his or her future health care choice. Previous studies indicated that effective communication among patients, family members, caregivers, and physicians is a key factor in completing written advance directives.16,17 DNH orders are appropriate for patients whose goals are centered on comfort and maximizing quality of life without going to hospital. Physicians and health care providers’ involvement in the planning process related to medical care is critical. An open indepth discussion about prognosis and treatment options as their disease progresses and physical functions deteriorate may help patients and loved ones make informed decisions about advance directives and DNH orders. Study Limitations Several limitations should be noted. The study results may be generalizable only to states with similar sociodemographic and health characteristics. We could not include the residents’ information (race) in this study because the variable had a lot of missing data. There is a possibility that the variable may affect the results as it was not included in the multivariate model. MDS2.0 did not assess the reasons why the residents were transferred to hospital/ED even though they had DNH orders. Given the nature of secondary data analysis, we were not able to address this question in our study. In future research, we

need to figure out the reasons for the disconnection between DNH orders and hospital/ED transfers. Conclusion This study examines whether DNH orders would be effective in reducing hospital/ED transfers among nursing home residents and among residents with dementia. Our findings indicate that residents with DNH orders had significantly fewer hospital admissions. This suggests that residents’ end-of-life care decisions were respected and honored. Efforts should be made to encourage nursing home residents to complete advance directives to promote integration of individual’s values and goals in guiding care provision toward the end of life. References 1. Levy C, Morris M, Kramer A. Improving end-of-life outcomes in nursing homes by targeting residents at high risk of mortality for palliative care: Program description and evaluation. J Palliat Med 2008;11:217e225. 2. O’Sullivan R, Murphy A, O’Caoimh R, et al. Economic (gross cost) analysis of systematically implementing a programme of advance care planning in three Irish nursing homes. BMC Res Notes 2016;9:237. 3. Silveira MJ, Kim SYH, Langa KM. Advance directives and outcomes of surrogate decision making before death. N Engl J Med 2010;362:1211e1218. 4. Vandervoort A, Houttekier D, Vander Stichele R, et al. Quality of dying in nursing home residents dying with dementia: Does advanced care planning matter? A nationwide postmortem study. PLoS One 2014;9:e91130. 5. World Health Organization. Dementia: A public health priority. Geneva: World Health Organization; 2012. 6. Mitchell SL, Teno JM, Miller SC, et al. A national study of the location of death for older persons with dementia. J Am Geriatr Soc 2005;53:299e305. 7. Mitchell SL, Teno JM, Kiely DK, et al. The clinical course of advanced dementia. N Engl J Med 2009;361:1529e1538. 8. Lamberg JL, Person CJ, Kiely DK, et al. Decisions to hospitalize nursing home residents dying with advanced dementia. J Am Geriatr Soc 2005;53: 1396e1401. 9. Fillenbaum G, Heyman A, Peterson BL, et al. Use and cost of hospitalization of patients with AD by stage and living arrangement: CERAD XXI. Neurology 2001;56:201e206. 10. Bynum JP, Rabins PV, Weller W, et al. The relationship between a dementia diagnosis, chronic illness, medicare expenditures, and hospital use. J Am Geriatr Soc 2004;52:187e194. 11. Goldfeld KS, Stevenson DG, Hamel MB, et al. Medicare expenditures among nursing home residents with advanced dementia. Arch Intern Med 2011;171: 824e830. 12. Teno JM, Mitchell SL, Gozalo PL, et al. Hospital characteristics associated with feeding tube placement in nursing home residents with advanced cognitive impairment. JAMA 2010;303:544e550. 13. Morrison RS, Siu AL. Mortality from pneumonia and hip fractures in patients with advanced dementia. JAMA 2000;284:2447e2448. 14. Buchanan RJ, Choi M, Wang S, Ju H. End-of-life care in nursing homes: Residents in hospice compared to other end-stage residents. Palliat Med 2004;7: 221e232. 15. Triplett P, Black BS, Phillips H, et al. Content of advance directives for individuals with advanced dementia. J Aging Health 2008;20:583e596. 16. Mitchell SL, Teno JM, Intrator O, et al. Decisions to forgo hospitalization in advanced dementia: A nationwide study: Hospitalization decision in advanced dementia. J Am Geriatr Soc 2007;55:432e438. 17. Mann E, Goff SL, Colon-Cartagena W, et al. Do-not-hospitalize orders for individuals with advanced dementia: Healthcare proxies’ perspectives. J Am Geriatr Soc 2013;61:1568e1573. 18. Lynn J, Goldstein NE. Advance care planning for fatal chronic illness: Avoiding commonplace errors and unwarranted suffering. Ann Intern Med 2003;138: 812e818. 19. Happ MB, Capezuti E, Strumpf NE, et al. Advance care planning and end-of-life care for hospitalized nursing home residents. J Am Geriatr Soc 2002;50: 829e835. 20. McAuley WJ, Buchanan RJ, Travis SS, et al. Recent trends in advance directives at nursing home admission and one year after admission. Gerontologist 2006; 46:377e381. 21. Xing J, Mukamel DB, Temkin-Greener H. Hospitalizations among nursing home residents in the last year of life: Nursing home characteristics and variation in potentially avoidable hospitalizations. J Am Geriatr Soc 2013;61:1e16. 22. Weiss AJ, Elixhauser A. Overview of hospital stays in the United States, 2012. Rockville, MD: Agency for Healthcare Research and Quality; 2014. p. 1e14. 23. US Department of Health and Human Services. Long-term care providers and services users in the United States: Data from the National Study of Long-term Care Providers, 2013e2014. Vital Health Stat 3 2016;(38). xexii 1e105. 24. Givens JL, Selby K, Goldfeld KS, et al. Hospital transfers among nursing home residents with advanced dementia. J Am Geriatr Soc 2012;60:905e909.