access to nursing home hospice: perspectives of nursing home and hospice administrators
TRANSCRIPT
JOURNAL OF PALLIATIVE MEDICINEVolume 8, Number 6, 2005© Mary Ann Liebert, Inc.
Access to Nursing Home Hospice: Perspectives of Nursing Home and Hospice Administrators
LAURA C. HANSON, M.D., M.P.H.,1 SOHINI SENGUPTA, Ph.D.,2and MONICA SLUBICKI, M.D.3
ABSTRACT
Background and objectives: Hospice improves the quality of end of life care in nursing homesbut serves less than 10% of dying residents. For residents to elect hospice, nursing homesmust first contract for these services. We surveyed nursing home and hospice administratorsto describe facilitators and barriers to hospice in nursing homes, and to test whether nursinghome administrators’ attitudes correlate with hospice use.
Methods: In a mailed survey, all nursing home and hospice administrators in North Car-olina responded to items on hospice’s effect on quality of care, and on facilitators and barri-ers to its use in nursing homes. Among nursing home administrators, bivariate analyses wereused to test associations of attitudes with use of hospice.
Results: After 2 mailings, 241 (62%) nursing home administrators and 74 (85%) hospice ad-ministrators responded. Eighty-three percent of nursing homes had a hospice contract, witha median of 3 residents enrolled in the last 3 months. Nursing home administrators were lesslikely than hospice administrators to believe that hospice improves quality of care for pain,emotional and spiritual needs, and bereavement support. Nursing home administrators weremore likely to agree that, “Nursing homes provide good care without using hospice for dy-ing residents and their families,” (24% versus 1%, p � 0.001). Among nursing home adminis-trators with a hospice contract (n � 180), those who agreed that hospice improves quality ofcare had higher rates of hospice use in their facilities.
Conclusions: Nursing home administrators’ attitudes toward hospice may influence its avail-ability for nursing home residents.
1207
INTRODUCTION
MOST AMERICANS SAY they prefer terminalcare at home but a majority of deaths take
place in institutions. Nursing homes are an in-creasingly common site for terminal care. Nearlyhalf of United States citizens who reach the ageof 65 will spend some time in a nursing home.1
In 1997, 1 in 4 adult deaths took place in a nurs-
ing home, and terminal care in nursing homes islikely to increase as the population ages.2–4 Themortality rate among nursing home residents isapproximately 25% per year, and most long-stayresidents remain in the nursing home to dierather than transferring to the hospital.2,5
Nursing home residents do not consistently re-ceive high-quality palliative care. Staff trainingand incentive systems emphasize restorative care
1Division of Geriatric Medicine, School of Medicine, 2Department of Social Medicine, School of Medicine, 3Schoolof Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
rather than symptom control and quality of life.6,7
Surviving family members report greater dissat-isfaction with nursing homes than with any othercomponent of terminal care.8 Residents increas-ingly forego life-sustaining treatment and hospi-talization, but these decisions are not linked to ef-fective plans for palliative care.9,10 Usual nursinghome care results in high rates of untreated se-vere pain,11–14 unmet needs for emotional andspiritual care,15 and little or no support for be-reaved family members.16 Residents diagnosedwith advanced dementia receive treatments thatmay prolong life but also increase discomfort.17
Nursing staff in long-term care perceive the needto improve end of life care, including enhancedstaffing and improved training.18
The limitations of usual nursing home caremay increase the impact of hospice services. Baerand Hanson19 surveyed recently bereaved familyabout the added value of hospice in nursinghomes. Hospice increased favorable ratings ofsymptom management from 64% to 90%. Familyrespondents identified unique hospice services,and 53% believed it reduced the need for hospi-talization.19 In a case-control study of nursinghome residents who died with or without Hos-pice, Miller et al.20,21 found that Hospice resultedin improved pain management and reduced useof tube feeding, intravenous fluids and physicalrestraints during the dying experience.
Although hospice improves the quality of end-of-life care in nursing homes, decedents in thissetting are less likely to access hospice servicesthan people who die at home. Hospices vary intheir delivery of services to nursing home resi-dents. In 2000, North Carolina’s 87 hospice agen-cies served 2261 nursing home residents. Overhalf of these residents were served by 7 agencies,while 28 agencies did not serve any residents(Carolinas Center for Hospice and End of LifeCare, unpublished data). Use of hospice variesgeographically, from a national average of 5.6%to over 20% of nursing home decedents inFlorida.22,23 Nursing home organizational char-acteristics are associated with access to hos-pice.22,23 Variations in use suggest that providernorms, not patient preferences, determine accessto nursing home hospice.
Nursing home residents have access to hospiceonly when the nursing home and hospice ad-ministrators have a contractual agreement. Onlyone prior study has examined the role of nursinghome administrators in facilitating or preventing
access to hospice. This study found that admin-istrators’ attitudes correlated with use of hos-pice.24 The study was limited to a small nursinghome sample, did not include hospice perspec-tives, and used a very brief survey. Our prior re-search suggests that long-term care staff do notalways perceive hospice as important to improvethe quality of end-of-life care.18 Experience andexpert opinion indicate that Medicare reimburse-ment rules, contractual constraints, and concernsover fraud may act as barriers to use of nursinghome hospice.7,25
To examine the potential effect of administra-tors’ attitudes on nursing home hospice use, wesurveyed all nursing home and hospice adminis-trators in the state of North Carolina. The study’sspecific aims were to compare nursing home andhospice administrators’ attitudes about facilita-tors and barriers to nursing home hospice and totest whether nursing home administrators’ atti-tudes are associated with greater or lesser actualuse of nursing home hospice.
METHODS
We mailed a structured survey to administra-tors in all 387 nursing homes and 87 hospices inNorth Carolina. Survey items included respon-dent demographic data, nursing home organiza-tional characteristics, hospice use, and attitudinalitems about nursing home hospice. To measurehospice use, respondents provided informationon presence or absence of a current nursing homehospice contract, current number of nursinghome hospice enrollees, and number of enrolleesduring the past 3 months. To measure attitudestoward hospice use, investigators developed sur-vey items based on literature review and inter-views with experts and clinicians familiar withnursing home and hospice care. Surveys includedparallel items for both types of administrators,with responses on a four-point Likert scale of“strongly agree, agree, disagree, or strongly dis-agree.” Item domains included effects of hospiceon quality of care, and contractual or financial fac-tors that might facilitate or limit the use of nurs-ing home hospice. A draft survey instrument waspiloted with 2 hospice and 2 nursing home ad-ministrators for clarity, relevance and face valid-ity of items. Investigators made revisions basedon their comments, and the final version of thesurvey was mailed to potential respondents.
HANSON ET AL.1208
Nonrespondents to the initial mailing receiveda second mailing of the survey. Nonrespondentsto the second mailing received a third mailingwith a brightly colored survey form. As an in-centive, all respondents were offered $10 for acompleted survey.
Survey data was entered in a secure database,and converted to a SAS database for analyses. In-vestigators first compared nursing home and hos-pice administrator responses to all parallel atti-tudinal items to examine differences between thetwo groups. To meet the second goal, investiga-tors compared responses for two groups of nurs-ing home administrators; those with less than orgreater than the median number of hospice en-rollees in the preceding 3 months. To ensure thathospice enrollment was possible, only responsesfrom administrators with a current hospice con-tract were included in this analysis. Comparisonsof proportions were done using chi-square test-ing, and comparison of continuous variables us-ing t tests. Initial analyses used data from the 4-point Likert scaled responses. The scale wassubsequently reduced to “agree/disagree” tosimplify communication of the findings. Thissimplification did not alter the findings of theanalyses.
All aspects of this study were reviewed and ap-proved by the University of North CarolinaSchool of Medicine Committee for the Protectionof Human Subjects.
RESULTS
Of the 387 potential nursing home administra-tors and 87 hospice administrators eligible for thestudy, 241 (62%) nursing home administrators
and 74 (85%) hospice administrators returnedcompleted surveys. Respondents in the twogroups did not differ by age, ethnicity, or reli-gious affiliation, although hospice administratorswere more likely to be women (88% versus 68%,p � 0.001) (Table 1).
Consistent with national data, only 15% of hos-pices were organized for profit, while 70% of nurs-ing homes were for profit according to the re-sponding administrators. Nearly all nursing homeadministrators reported current hospice contracts;83% of nursing home administrators had an ac-tive contract to provide hospice, and 63% of thosewith a contract had at least 1 resident currentlyenrolled in hospice care. Nursing home adminis-trators with a hospice contract reported a medianenrollment of 3 residents during the preceding 3months, with a range of 0 to 21.
Comparison of nursing home and hospiceadministrators’ attitudes
Compared to hospice administrators, nursinghome administrators were less likely to agree thathospice improves the quality of care for dyingresidents (Table 2). Nearly all hospice adminis-trators, compared to smaller majorities of nurs-ing home administrators, believed that hospiceimproved the quality of pain management, emo-tional care, spiritual care, bereavement supportand care for personal cleanliness. The two groupsdiffered most on whether hospice improves per-sonal cleanliness care for dying nursing home res-idents; 68% of hospice versus 12% of nursinghome administrators agreed. Half of nursinghome administrators but 84% of hospice admin-istrators agreed with the statement, “Hospiceproviders avoid treatments that would cause
ACCESS TO NURSING HOME HOSPICE 1209
TABLE 1. CHARACTERISTICS OF NURSING HOME AND HOSPICE ADMINISTRATORS
Nursing home HospiceCharacteristic of administrator (n � 241) (n � 74)
Age (mean) 45 48Years in current job (median) 10 7Gender female 68% 88%a
RaceCaucasian 97% 97%African American 2% 1%
Religious affiliationProtestant 88% 90%Catholic 9% 8%
ap � 0.001.
death.” Only a small percentage of each groupagreed with the statement, “Hospice care cancause a nursing home resident to die faster” (3%versus 10%, p � NS).
In contrast to their differences on items mea-suring hospice’s effect on quality of care, the twogroups of administrators were similar in their be-liefs about facilitators and barriers to hospice use(Table 3). A high percentage of both nursinghome and hospice administrators agreed with thestatement, “Hospice team understands and re-spects the care provided by the nursing homestaff.” A majority of each group believedMedicare eligibility created a financial barrierand hindered cooperation. More nursing homeadministrators expressed concern about the po-tential for fraudulent billing practices (46% ver-
sus 10%, p � 0.001). Fewer nursing home admin-istrators perceived that hospice nurses and nurs-ing home staff found it difficult to coordinate aplan of care (26% versus 61%, p � 0.001).
Nursing home administrators’ attitudesassociated with hospice use
Nursing home administrators’ attitudes wereassociated with the actual use of hospice withintheir facilities. Of the 196 nursing home adminis-trators with a current hospice contract, 184 pro-vided complete answers to all items. For thisgroup of administrators, the median number ofhospice enrollees during the past 3 months was3. In bivariate comparisons, nursing home ad-ministrators who agreed with items measuring
HANSON ET AL.1210
TABLE 2. NURSING HOME AND HOSPICE ADMINISTRATORS’ PERCEPTIONS OF HOW HOSPICE AFFECTS QUALITY OF CARE
Agreement by Agreement bynursing home hospiceadministrators administrators
Attitude (n � 241) (n � 74)
Hospice improves quality of nursing home care 70% 99%a
Hospice necessary to improve care 54% 89%a
Nursing home provides good care without hospice 24% 1%a
Nursing homes do not need hospice 25% 1%a
Compared to nursing home, hospice does a better job at:pain management 60% 97%a
bereavement support 84% 99%a
emotional support 66% 97%a
spiritual counseling 78% 97%a
personal cleanliness care 12% 68%a
Hospice causes faster death 10% 3%b
Hospice avoids treatments causing death 49% 84%a
ap � 0.001.bp � 0.05.
TABLE 3. NURSING HOME AND HOSPICE ADMINISTRATORS’ PERCEPTIONS OF
CONTRACTUAL AND FINANCIAL BARRIERS TO NURSING HOME HOSPICE
Agreement by Agreement bynursing home hospiceadministrators administrators
Attitude (n � 241) (n � 74)
Hospice understands nursing home 84% 89%Nursing home loses if Medicare pays 65% 55%Medicare rules hinder cooperation 53% 61%Medicare & Medicaid can lead to fraud 46% 10%a
Nursing home loses if Medicaid pays 26% 20%Difficult to coordinate plan of care 26% 61%a
Too many nursing home residents are in hospice 8% 5%Too many nursing home residents in hospice not really dying 8% 5%
ap � 0.001.
hospice’s positive impact on quality of care weremore likely to have 3 or more enrollees in the past3 month (Table 4). Nursing home administratorswho reported greater hospice use were also morelikely to believe that hospice providers under-stand nursing home care, and less likely to be-lieve that care coordination was difficult. Concernabout hospice hastening death and attitudesabout financial or regulatory barriers were not as-sociated with actual use of hospice.
DISCUSSION
In a statewide survey, nursing home adminis-trators differed from hospice administrators intheir perception of hospice’s impact on quality ofcare for dying nursing home residents. Nursinghome and hospice administrators agreed on therelative importance of contractual and financialbarriers, with a majority endorsing Medicarerules as important barriers. Among nursing homeadministrators, more positive attitudes were as-sociated with greater use of hospice by residentsin their facilities. Nursing home administratorsmay influence access to hospice services for resi-dents within their facilities.
Our findings may be interpreted to mean thatnursing home administrators who believe hos-pice improves the quality of end-of-life care mayfacilitate access to hospice services. Alternatively,greater experience with hospice may lead ad-ministrators to develop more positive percep-
tions of its value. We found an association be-tween attitudes and practice for items related toquality of care, but no association for items mea-suring regulatory or financial barriers. Althougha majority of nursing home administrators be-lieve that Medicare rules hinder cooperation andcreate a financial disincentive, these beliefs werenot associated with actual use of the service oncea contract was established.
The results of this study must be considered inlight of its limitations. The cross-sectional designmeans that associations are not directional, andcannot state cause and effect relationships. Ad-ministrators’ attitudes may influence hospice re-ferral patterns, or simply be the result of greateror lesser exposure to hospice care. The study islimited to the experience of administrators in onestate, and may not generalize to other geographicregions. Although response rates are good, non-response bias may affect results. Investigatorscould not obtain data on nonrespondents to as-certain differences. However, this sample is moregeneralizable than any other published study onthis topic, and North Carolina end of life prac-tices are similar to national averages.4 Finally, itis important to acknowledge that administrators’attitudes are likely to explain only a portion ofthe variance in hospice use among nursinghomes. Other factors such as local patient char-acteristics, physician practices, and willingness ofhospices to serve patients with dementia are alsolikely to influence access.26 Individual hospice or-ganizations variably restrict access to care based
ACCESS TO NURSING HOME HOSPICE 1211
TABLE 4. NURSING HOME ADMINISTRATORS’ ATTITUDES ASSOCIATED WITH HOSPICE USE IN PAST THREE MONTHS
Agreement, Agreement,administrators administrators
with 0–2 enrollees with �3 enrolleesAttitude (n � 84) (n � 97)
Hospice necessary to improve care 42% 73%a
Nursing home provides good care without hospice 37% 11%a
Hospice improves quality of care 59% 87%a
Compared to nursing home, hospice does a better job at:pain management 49% 70%b
bereavement support 80% 92%c
emotional support 52% 78%a
spiritual counseling 70% 89%d
Difficult to coordinate plan of care 36% 15%Hospice understands nursing home 77% 96%a
ap � 0.001.bp � 0.005.cp � 0.02.dp � 0.002.
on patient requests for types of treatment, anddifficulty in prognosis related to diagnoses suchas dementia.27,28
The results of this study have implications forclinical practice, for policy, and for future re-search. Dying nursing home residents who wishto access hospice services must overcome barri-ers unique to this setting. Physicians and familycaregivers may need to convince reluctant ad-ministrators, or select nursing homes based onhospice access. Given that a majority of nursinghome and hospice administrators agree thatMedicare rules are a major barrier, reformingthese rules may be an important strategy to fa-cilitate collaborative care. Future research shouldtest new models of hospice services tailored tothe needs and conditions of the nursing home en-vironment. Stronger evidence that hospice im-proves the quality of care in nursing homes maybe needed to convince administrators to facilitateaccess to this service.
ACKNOWLEDGMENTS
Supported by grants from the Duke Instituteon Care at the End of Life and from the Holder-ness Foundation.
REFERENCES
1. Kemper P, Murtaugh CM: Lifetime use of nursinghome care. N Engl J Med 1991;324:595–600.
2. Hanson LC: Creating excellent palliative care in nurs-ing homes. J Palliat Med 2003;6:7–9.
3. Sager MA, Easterline DV, Kindig DA, Anderson OW:Changes in the location of death after the passage ofMedicare’s prospective payment system. N Engl JMed 1989;320:433–439.
4. Teno JM: Facts on Dying. �www.chcr.brown.edu/dy-ing/factsondying.htm� (Last accessed June 2003).
5. Holtzman J, Lurie N: Causes of increasing mortalityin a nursing home population. J Am Geriatr Soc1996;44:258–264.
6. Keay TJ, Schonwetter RS: Hospice care in the nursinghome. Am Fam Physician 1998;57:491–494.
7. Zerzan J, Stearns S, Hanson LC: Access to palliativecare and hospice in nursing homes. JAMA2000;284:2489–2494.
8. Hanson LC, Danis M, Garrett J: What is wrong withend of life care? Opinions of bereaved family mem-bers. J Am Geriatr Soc 1997;45:1339–1344.
9. Mor V, Intrator O, Fries BE, Phillips C, Teno J, Hiris
J, Hawes C, Morris J: Changes in hospitalization as-sociated with introducing the Resident AssessmentInstrument. J Am Geriatr Soc 1997;45:1002–1010.
10. Holtzman J, Pheley AM, Lurie N: Changes in orderslimiting care and the use of less aggressive care in anursing home population. J Am Geriatr Soc1994;42:275–279.
11. Bernabei R, Gambassi G, Lapane K, Landi F, Gatse-nis C, Dunlop R, Lipsitz L, Steel K, Mor V: Manage-ment of pain in elderly patients with cancer. JAMA1998;279:1877–1882.
12. Won A, Lapane K, Gambassi G, Bernabei R, Mor V,Lipsitz LA: Correlates and management of nonma-lignant pain in the nursing home. J Am Geriatr Soc1999;47:936–942.
13. Ferrell BA: Pain evaluation and management in thenursing home. Ann Intern Med 1995;123:681–687.
14. Sengstaken EA, King SA: The problems of pain andits detection among geriatric nursing home residents.J Am Geriatr Soc. 1993;41:541–544.
15. Hanson LC, Henderson M, Schulman A, Reynolds K:Needs of the dying in nursing homes. J Gen InternMed 1999;14:37.
16. Murphy K, Hanrahan P, Luchins D: A survey of griefand bereavement in nursing homes: The importanceof hospice grief and bereavement for the end-stageAlzheimer’s disease patient and family. J Am GeriatrSoc 1997;45:1104–1107.
17. Mitchell SL, Kiely DK, Hamel MB: Dying with ad-vanced dementia in the nursing home. Arch InternMed 2004;164:321–326.
18. Zimmerman S, Sloane PD, Hanson L, Mitchell CM,Shy A: Staff perceptions of end-of-life care in long-term care. J Am Dir Assoc 2003;4:23–26.
19. Baer WM, Hanson LC: Families’ perceptions of theadded value of hospice in the nursing home. J AmGeriatr Soc 2000;48:879–882.
20. Miller SC, Gozalo P, Mor V: Outcomes and utilizationfor Hospice and non-Hospice nursing facility dece-dents. HHS contract #100-97-0010. �www./aspe.hhs.gov/reports/oututil.htm�
21. Miller SC, Mor V, Teno J: Hospice enrollment andpain assessment and management in nursing homes.J Pain and Symptom Manage 2003;26:791–799.
22. Petrisek AC, Mor V: Hospice in nursing homes: A fa-cility-level analysis of the distribution of Hospice ben-eficiaries. Gerontologist 1999;39:279–290.
23. Miller SC, Intrator O, Gozalo P, Roy J, Barber J, MorV: Government expenditures at the end of life forshort- and long-stay nursing home residents: Differ-ences by hospice enrollment status. J Am Geriatr Soc2004;52:1284–1292.
24. Jones B, Nackerud L, Boyle D: Differential utilizationof Hospice services in nursing homes. Hospice J1997;12:41–57.
25. Miller SC, Mor V, Coppola K, Teno J, Laliberte L,Petrisek AC: The Medicare Hospice benefit’s influ-ence on dying in nursing homes. J Palliat Med1998;1:367–376.
HANSON ET AL.1212
26. Friedman BT, Harwood MK, Shields M: Barriers andenablers to hospice referrals: an expert overview. JPall Med 2002;5:73–84.
27. Lorenz KA, Asch SM, Rosenfeld KE, Liu H, Ettner SL:Hospice admission practices: where does hospice fitin the continuum of care? J Am Geriatr Soc2004;52:725–730.
28. Hanrahan P, Luchins DJ: Access to hospice programsin end Bstage dementia: a national survey of hospiceprograms. J Am Geriatr Soc 1995;43:56–59.
Address reprint requests to:Laura C. Hanson, M.D., M.P.H.
Division of Geriatric Medicine and Program on Aging
CB 7550, 258 MacNiderUniversity of North Carolina
Chapel Hill, NC 27599
E-mail: [email protected]
ACCESS TO NURSING HOME HOSPICE 1213