Do Residents Want Automated External Defibrillators in Their Retirement Home?

ORIGINAL STUDIES

Do Residents Want Automated External Defibrillators in Their Retirement Home? Douglas C. Woolley, MD, MPH, CMD, Louis J. Medvene, PhD, Rick D. Kellerman, MD, Michelle Base, MA, and Victoria Mosack, MSN, RN, ARNP Purpose: The administration of a continuing care retirement community (CCRC), while weighing practical and ethical questions surrounding installation of automated external defibrillators (AEDs), wanted to consider resident opinions. No databased studies on this subject were found. Design and Methods: After an information session about AEDs, CCRC residents were surveyed concerning their opinions on AED installation, their beliefs and concerns regarding AEDs, their advance directive status, and their demographic characteristics. Correlations were sought between choices about AED installation and beliefs, advance directives, and demographics. Results: Seventy-eight percent of 107 eligible residents participated. Twenty-seven percent wanted AEDs installed, 37% were not sure, 23% were opposed, and 11% did not answer this question. Univariate analysis showed that women, the widowed or single, and those with a college degree were more likely to oppose AEDs. In the best logistic regression (LR) model

The placement of automated external defibrillators (AEDs) in public places is becoming routine, and even expected, where crowds gather in sometimes-stressful settings such as

Department of Family and Community Medicine, Kansas University School of Medicine, Wichita, KS (D.C.W., R.D.K.); Department of Psychology, Wichita State University, Wichita, KS (L.J.M., M.B., V.M.). Dr Woolley was supported with an Advanced Research Training Grant from the American Academy of Family Physicians Foundation. Address correspondence to Douglas C. Woolley, MD, MPH, CMD, Department of Family and Community Medicine, Kansas University School of Medicine, Wichita, 1010 N Kansas Street, Wichita, KS 67214-3199; E-mail: [email protected]

Copyright ©2006 American Medical Directors Association DOI: 10.1016/j.jamda.2005.08.004 ORIGINAL STUDIES

the hope that “AED use could be life saving” and the fears that “AED use might lead to a very poor quality of life” and that “AEDs might be misused” were more important than any demographic variables and only education remained in the model. Those opposing AEDs supplied powerful written comments to support their choice. Conclusions: There is no consensus and great indecision about AED installation among the residents of this CCRC. The subjects were somewhat older and more affluent than the typical retirement home population, pointing to the need for replicating the investigation with a larger and more diverse study population. However, these findings suggest that AED installation in a retirement home would be premature without engaging the entire community in discussions and education in a process considerate of the wishes of all residents, which are likely to be quite diverse. (J Am Med Dir Assoc 2006; 7: 135–140) Keywords: Defibrillation; resuscitation; sudden death; advance directives; continuing care retirement homes

airports, airplanes, casinos, fitness clubs, and sporting events.1–3 Modern AED devices may be successfully used with modest training.4,5 One of the critical factors influencing the prognosis for witnessed “out-of-hospital” cardiac arrest or “sudden death,” along with rapid response time by a trained resuscitation team, is the application of rapid external cardiac defibrillation.6,7 Defibrillation has been added to the classic “ABCs” of Advanced Cardiac Life Support to form the “ABCDs” of resuscitation.7–9 The epidemiology of “sudden death” shows a sharp increase in incidence in late middle age that continues into old age.10 –12 Outcomes of out-of-hospital cardiac resuscitation for aged people, though not as good as for younger victims, are not uniformly dismal.13–19 An emerging policy question for administrators of retirement homes is whether they too should install AEDs. The Woolley et al. 135

administration, board of directors, health care staff, and fitness staff of a continuing care retirement community (CCRC), at which the lead author is the medical director, began an ongoing discussion about whether to install AED devices in the fitness center, dining room, and other congregate areas of the apartments for independently living residents. A few local retirement homes had already installed AEDs.20 The introduction of AEDs was the focus of a session of the institution’s ethics roundtable, where concerns were raised about residents’ rights and autonomy, particularly about whether AEDs might be considered by some residents as more of a threat than a source of security. Thoughtful observers who work with very aged people have questioned whether routine AED use is appropriate or desired by independently living people of advanced age.21,22 Although there are several studies of the attitudes and choices of the aged about advance directives and cardiopulmonary resuscitation,23–32 a search of the literature did not reveal relevant databased studies looking specifically at the question of residents’ attitudes toward the placement of AEDs in their retirement homes. The goal of this study was to understand the beliefs of the residents of one CCRC about whether AEDs would benefit them and to understand their opinions about whether AEDs should be installed in their facility. Also of interest was how residents’ opinions and beliefs about AEDs were related to their demographic characteristics, their perceived personal health status, and their preferences regarding “do not resuscitate” (DNR) and “other advance directive” choices. Based on studies of factors associated with cardiopulmonary resuscitation (CPR) preferences of seniors in other settings, we hypothesized that younger,24,32 married,24 male,23,24 less educated,24 and more highly functioning residents32 may be more likely to support AED use in the retirement home than their counterparts. However, these factors have not always been found to correlate with significant differences in choices about CPR among the aged.25,26 We did not hypothesize what percentage of our subjects would want AEDs installed in the retirement home because previous studies have found a wide range (from 25% to 74%) in the percentage of aged respondents who would wish to receive CPR.23,24 METHODS Study Design and Setting The study was based on a cross-sectional survey of independently living residents of a CCRC in the prosperous suburb of a mid-sized midwestern metropolitan area. The study was reviewed and approved by the University Human Subjects Committee. Eligible study subjects were all CCRC residents who were neither too sick nor too demented to complete the survey and who did not participate in the pilot study of the survey. It was deemed impractical to ask all CCRC residents to take a mental status test, and base eligibility on a specific cutoff score. Therefore, eligibility was determined subjectively by the home health nursing director, the activities director, and the medical director working together, answering the question, “Based on our work with this resident, can the resident attend a group informational session, under136 Woolley et al.

stand the information provided and discussed, and apply it in answering a questionnaire?” They did not know DNR preferences of the residents when determining eligibility. Survey Design and Administration A simply worded large-print survey was designed to allow correlation of the primary outcome variable, the participant’s opinion about whether AEDs should be installed in the CCRC, with their beliefs about AED use and its consequences and their knowledge and experience concerning AED use, as well as their health status, previous end-of-life planning, and demographic information. It concluded with a request for comments. In a pilot test, a group of 8 CCRC residents previewed a brief educational presentation about AEDs and their use, completed the survey, and then in a focus group, gave detailed critique of the survey and the presentation. The 8 residents were chosen by the activities director based on their active participation in resident association affairs. The activities director did not have knowledge of resident DNR choices or feelings about defibrillation prior to nominating the focus group participants. The survey went through several revisions based on the focus group outcome and reviews by psychology and gerontology colleagues. All eligible residents received a packet with introductory information about the study and an invitation to one of several scheduled meetings. The sessions were facilitated by trained research staff who showed an informational video (the same at each session), demonstrated an actual AED machine, answered basic factual questions about the AED according to a prepared script, and distributed the survey. The video, written information, and script for answering questions had been independently prescreened and edited to eliminate evident bias for or against the use of AEDs in this setting. Attendance was recorded but the surveys were returned anonymously. After the sessions were completed, a number of residents reported that they had wanted to participate but could not attend any of the sessions. In response, all eligible residents who did not attend sessions were sent a packet that included the survey, pictures of AED machines, and a bulleted summary of the video. Although individual anonymity was maintained, the introductory paragraph of these surveys was slightly changed to fit the mail-back format, and thus they were distinguishable at the time of data entry. Data Analysis The demographics of noneligible, eligible, and responding residents were compared to test for response biases (chi square, analysis of variance [ANOVA], t tests). The demographics of the residents as a whole and of study participants were compared with those of other CCRC populations and of same-aged community populations to aid in testing the generalizability of findings. Since almost all CCRC residents are older than the age of 75, we chose this segment of the community population for comparison.29 The main dependent variable of the survey, Q1 (“Are you for or against having AEDs placed here?”) had 3 responses: “for,” “not sure,” and “against.” For most analyses this could be treated as an ordered variable (see the Surgery Results: AssoJAMDA – March 2006

ciations of Choices With Opinions section). Therefore univariate analysis of the significant differences in Q1 responses based on each of the independent variables was done with the Kruskal-Wallis test for categorical independent variables, and with regression analysis for continuous independent variables.33 These results were confirmed using logistic regression (LR) with single independent variables. Logistic regression was further used to develop predictive models of Q1 by combining independent variables (SAS 8.02, SAS Institute, Cary, NC, 2001). Since the outcome variable Q1 is not dichotomous but 3-leveled and ordered, multinomial LR was used.34 Finally, the frequency, tone, and content of written comments were compared among the 3 groups as defined by responses to Q1. RESULTS Survey Participants Compared With Nonparticipants Of the 157 independent apartment residents living in the CCRC during the survey period, 42 were unable to participate because of dementia or severe illness, and 8 were ineligible because they were part of the focus group that piloted the survey. This CCRC does not have a separate assisted-living section, but has a very active home health department that supports impaired residents “aging in place” in their apartments who might be in assisted living in other institutions. The availability of this support explains the relatively large number of apartment residents ineligible for the study. Of the remaining 107 eligible residents, 83 completed usable surveys, representing a 77.6% rate of participation. The mean age of all apartment residents was 85.9 (SD ⫾ 5.8). The residents eligible to participate in the survey and the residents who did participate both averaged 84.8 years of age. The age range for participants was from 73 to 95 years. Ineligible residents were on average 3.9 years older than eligible residents (P ⫽ .0003). There was no difference in gender ratios between the ineligible and eligible residents (P ⫽ .66), or between the eligible residents and the survey participants (P ⫽ .70). The participant group was 71.6% female; 42.3% with living spouses; 9.8% with self-rated health “excellent,” 59.8% “good,” and 30.5% “fair” or “poor”; 67.9% with college or professional degrees; and 70.8% with income “above” or “much above” the average yearly retiree income of $18,000 to $25,000. There

were no significant differences in demographics or in responses to other survey items between those who attended the sessions and those who mailed in the survey. Population Comparisons The residents of this CCRC and the survey participants differ significantly in key demographics from residents of other CCRCs, and particularly from community-dwelling residents older than 75 (see Table 1).29,33,35 Our study participants were significantly older, more highly educated, and more likely to have living spouses than other CCRC residents. They were older, more likely to be female, better educated, more affluent, more likely to be white, and to have living spouses than community residents older than 75. They rated their own health higher than did age-matched community residents, but not as high as did other CCRC residents. Survey Results: Associations of Choices With Demographics The survey’s main outcome variable was the answer to Q1, “Are you for or against putting AEDs in your retirement home?” Twenty-two (26.5%) of respondents said “for,” 31 (37.3%) said “not sure,” 19 (22.9%) said “against,” and 11 did not answer (13.3%). See Table 2 for a summary of univariate analysis of the relationships between Q1 and the independent variables. There was no difference in age between those who were in favor and those who were against, but the 31 who were “not sure” were significantly older than those who made a definitive choice (P ⫽ .02). Better-educated residents, women, and widowed or unmarried residents were less likely to favor AED placement. There was no significant difference in responses to Q1 based on self-perceived health, heart health, income, or advanced directive planning. Of the 83 respondents, 46 left no survey items blank, 22 left 1 or 2 items blank, and 15 left more than 2 blank. There was no significant difference between these groups in age, gender, self-reported health, education, or opinion about AED placement. Survey Results: Associations of Choices With Opinions The survey items 2 through 6 were designed to test individual beliefs about the consequences of AED use (see Table 3). In each case there were highly significant differences in

Table 1. Comparing Subject Demographics With Other Senior Populations Demographic

Institution

Study Subjects

US CCRCs*

Community Dwelling

Age, average (⫾ SD) Female:male ratio Education, % with some college Income Health, self-related Race, % white Married, %

85.9 (⫾ 5.8) 100:45 Unavailable

84.8 (⫾ 5.0) 100:40 91.4

82.3 100:32.8 72.0

78.0 100:55.3 14.2

Unavailable Unavailable 100 42.6

71% ⬎ 25K 70% good or excellent 100 42.3

Median ⬃ 30K 75% good or excellent 98 36.6

Median ⬍ 20K 67% good or excellent 90 37.0

CCRS, continuing care retirement community. * All figures in columns 3 and 4 are from Sherwood. ORIGINAL STUDIES

Woolley et al. 137

Table 2. Association of Respondents’ Opinions About AED Installation With Survey Items Opinion About Installing AEDs Survey Item Demographics Age, y Sex Female, % Male, % Marital status Unmarried, % Married, % Educational acheivement No college, % College, % Health: self-defined* Had heart treatment No, % Yes, % Income Below average or average, % Above average, % Living will No, % Yes, % Do not resuscitate No, % Yes, %

For

Not Sure

Against

P Value

26% (n ⫽ 22) 84

37% (n ⫽ 31) 87

23% (n ⫽ 19) 84

.909

26 45

42 45

32 10

.047

21 45

48 38

31 17

.038

40 24 2.1

37 39 2.2

9 37 2.2

.047 .475

34 28

40 46

26 26

.79

22 34

50 45

28 20

.62

100 28

0 45

0 26

.94

54 25

38 44

8 31

.31

AED, automated external defibrillator. * Four-point scale where 1 is “excellent” and 4 is “poor.”

the answers to these items between those who were “for” and “against” AED use. Frequently there were significant differences between “for” and “not sure” responses, and between “not sure” and “against” responses as well. The average of the “not sure” respondents was always between that of the “for” and the “against” respondents in ratings of these items, supporting the treatment of Q1 as an ordered variable. Table 3 shows that agreement with items 2, 3, or 5 was associated with a choice against AED placement, and agreement with items 4 or 6 was associated with a choice for AED

placement. Before further LR modeling was done, a correlation table was constructed with all continuous and ordered variables to be used in model construction. This correlation table was important in guiding testing of LR models to predict Q1 choices, as inclusion in a LR model of independent variables with correlation coefficients approaching 0.45 or more can lead to model instability and failure.34 To avoid this, yet allow for inclusion of all 5 opinion variables in developing the LR model, new variables, made of linear combinations of Q2 through Q6, were created and tested for their effectiveness

Table 3. How Those For, Not Sure, and Against AEDs Answered Q2 Through Q6 Opinion Question

For AEDs Average

Not Sure Average

Against AEDs Average

P Value

Q2: “I am concerned that if I survived after use of an AED my quality of life might be very low.” Q3: “The presence of AEDs will make the home more like a hospital.” Q4: “Use of lifesaving devices such as AEDs is consistent with my advance directive preferences.” Q5: “I fear that AEDs might be used improperly and cause someone harm.” Q6: “I like the idea of AEDs because they might save my life.”

1.412

2.500

3.053

⬍.001

2.529

3.100

3.700

.004

3.813

3.318

2.611

.003

2.063

2.517

3.700

⬍.001

4.250

2.933

1.714

⬍.001

AED, automated external defibrillator. 1 ⫽ strongly disagree, 2 ⫽ disagree, 3 ⫽ neutral, 4 ⫽ agree, 5 ⫽ strongly agree 138 Woolley et al.

JAMDA – March 2006

in competition with use of these same independent variables placed directly into the model in various combinations. Model effectiveness was judged by simplicity of interpretation and large size of the Homer Lemeshow “c” statistic.36 The simplest strong model (c ⫽ 0.93) involves a linear combination of Q2 and the inverse of Q6 (Cronbach’s alpha for Q2 and Q6 is 0.76), then Q5, and finally educational level, Q17— education ordered from most to least important. The new Q2Q6 variable can be interpreted as, “I like the idea of AEDs because they might save my life and I am not so concerned that it would be a low quality of life,” or conversely, “I don’t like idea of AEDs because I am not interested in resuscitation, and I am concerned that if I did live after AED use my quality of life might be very low.” Those who were “against” AED placement were significantly less likely to want more information on AED use than were those who were “for” placement. Those who were “not sure” were significantly more likely to want more information than either of the other groups. All but one of those “against” AED placement were also against having the board of directors of the CCRC explore the issue further; all but one of those “for” AED placement wanted the board to pursue the issue further; 78% of those “not sure” also wanted the board to further explore the issue. Comparing the written comments of those “for,” “not sure,” and “against” AED placement, no one group made significantly more comments than any other. However, those “against” AED placement made particularly pointed comments to emphasize the strength of their feelings, such as the following: ● “How do you propose to keep ‘do-gooders’ from violating my DNR instructions?” ● “Personally I would welcome a heart attack!” ● “We are old and ready for death. There are many bad ways to die. Cardiac sudden death is one of the easiest.” DISCUSSION This work addresses a timely but previously unstudied issue that directly impacts the approximately 2 million Americans living independently in retirement homes or apartments.37 The respondents’ average age of 85 is remarkably high for an opinion survey. The response rate of 78% of potential participants minimizes the likelihood of significant response bias. The internal consistency of participants’ responses provides strong evidence for the validity of the survey. For example, participants who were “not sure” about AED placement fell between those “for” and those “against” AED placement in all of the questions concerning beliefs about the consequences of using AEDs. “Unsure” participants were most likely to want further information about AEDs and very likely to want the board of directors to consider the issue further. A key finding from this study is that participants’ preferences about AED installation in their retirement home are strongly associated with their beliefs, hopes, and fears about the effectiveness and outcomes of cardiac resuscitation and are not generally associated with health or demographic variables. Based on previous work, we hypothesized there would be an association between AED preferences with health and ORIGINAL STUDIES

such demographics as gender, age, health, and marital status; hypothesizing that females, those without living spouses, the very aged, and the more chronically ill being less likely to want AEDs available.23,24,26 In this well-educated and moderately healthy population ranging in age from the mid 70s to the mid 90s, gender and education were the only demographic variables that showed significant univariate association with choice about AED placement. Only education remained a significant part of the best LR models of AED choice, but even then it was the weakest contributor. It may be that in our study age does not correlate with AED choices because the majority of our respondents were in their mid 80s or older, creating an age-related ceiling effect. The median age was 85 years; two thirds were 83 or older, and one third were 88 or older. In our study, education was negatively associated with a wish for AED availability. Other studies suggest that people of all ages, but particularly the aged, overestimate the effectiveness of CPR.24 It may be that better educated and informed populations are more aware of the limitations of AED use, and that our well-educated subjects are less interested in AED use than other aged populations, such as those attending community senior centers or living in subsided housing for seniors. One plausible interpretation of our findings is that participants’ definition of quality of life influenced their opinions about placement of AEDs. In other words, participants who had higher quality-of-life expectations were opposed to the placement of AEDs. This may be explored with further research. Those who were “unsure” or did not answer the question about AED placement were significantly older than those with firm opinions either way. Thus, regarding this question, decisional self-efficacy may drop with very advanced age. Important findings from this study for those involved with policy and planning for retirement homes are these: (1) only a quarter of respondents wanted AEDs available; (2) there was much ambivalence about the issue, with more than half of respondents “not sure” about AED placement and feeling they needed more information and discussion; and (3) almost a quarter of respondents did not want AEDs available and among this group are people with strong feelings about the issue. Therefore it would seem unwise and ethically tenuous to install AEDs without engaging the resident community in a process of education, discussion, and thoughtful attempt to accommodate everyone’s wishes. Boards of directors of CCRCs are ethically and probably legally obligated to respect residents’ preferences regarding AEDs.38 On the one hand, boards will need to develop administrative procedures that enable residents to have the benefit of AEDs, if this is their preference. On the other hand, boards are likewise obligated to enable residents the right to refuse this treatment, if this is their preference. Developing administrative policies that comply with these ethical and legal obligations will be challenging. Possible policies might include bracelets, or pins, or some other symbolic way for residents to express their preference. The members of each retirement community, including its residents and administration, must work together to decide how best they might comfortably and effectively achieve this end. Woolley et al. 139

Much work is yet to be done on this subject. The results of our study are being shared with residents and administrators of the CCRC in conjunction with more sessions devoted to education and discussion. A representative working group of residents and staff will be convened to explore possible systems of AED implementation designed to address all resident concerns. Our study population is small, from a single CCRC, and rather privileged. Therefore the investigation should be expanded to include a variety of other congregate settings for seniors, including retirement communities and community senior citizens’ centers. Its expansion should also encompass rural and urban settings and include key ethnic and racial groups. REFERENCES 1. Page RL, Joglar JA, Kowal, RC, et al. Use of automated external defibrillators by a U.S. airline. N Engl J Med 2000;343:1210 –1216. 2. Valenzuela TD, Roe D J, Nichol G, et al. Outcomes of rapid defibrillation by security officers after cardiac arrest in casinos. N Engl J Med 2000; 343:1206 –1209. 3. American College of Sports Medicine and American Heart Association joint position statement: automated external defibrillators in health/ fitness facilities. Med Sci Sports Exerc 2002;34:561–564. 4. Caffrey SL, Willoughby PJ, Pepe PE, Becker LB. Public use of automated external defibrillators. N Engl J Med 2002;347:1242–1247. 5. Hoffman CE, Marenco J, Wang PJ, et al. Public access defibrillation programs: the role of the automated external defibrillator. Cardiovasc Rev Rep 2002;23:286 –291. 6. Weisfeldt ML, Kerber RE, McGoldrick RP, et al. Public access defibrillation. A statement for healthcare professionals from the American Heart Association Task Force on Automatic External Defibrillation. Circulation 1995;92:2763. 7. Wik L, Hansen TB, Fylling F, et al. Delaying defibrillation to give basic cardiopulmonary resuscitation to patients with out-of-hospital ventricular fibrillation: A randomized trial. JAMA 2003;289:1389 –1395. 8. Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Part 6: Advanced cardiovascular life support: Section 7: Algorithm approach to ACLS emergencies: Section 7A: Principles and practice of ACLS. The American Heart Association in collaboration with the International Liaison Committee on Resuscitation. Circulation 2000;102(8 Suppl):I136 –139. 9. Marenco JP, Wang PJ, Link MS, et al. Improving survival from sudden cardiac arrest: The role of the automated external defibrillator. JAMA 2001;285:1193–1200. 10. Demirovic J, Myerburg RJ. Epidemiology of sudden coronary death: An overview. Prog Cardiovasc Dis 1994;37:39 – 48. 11. Myerburg RJ, Kessler KM, Castellanos A. Sudden cardiac death: Epidemiology, transient risk, and intervention assessment. Ann Intern Med 1993;119:1187–1197. 12. Nicolas G, LeComte D. La mort subite d’origine cardiaque. Épidémiologie. Bulletin Académique Nationale de Médicine 1999;183:1573–1580. 13. Carlen PL, Gordon M. Cardiopulmonary resuscitation and neurological complications in the elderly. Lancet 1995;345:1253–1254. 14. Ghusn HF, Teasdale TA, Pepe PE, Ginger VF. Older nursing home residents have a cardiac arrest survival rate similar to that of older persons living in the community. J Am Geriatr Soc 1995;43:520 –527.

140 Woolley et al.

15. Juchems R, Wahlig G, Frese W. Influence of age on the survival rate of out-of-hospital and in-hospital resuscitation. Resuscitation 1993;26:23– 29. 16. Kim C, Becker L, Eisenberg MS. Out-of-hospital cardiac arrest in octogenarians and nonagenarians. Arch Intern Med 2000;160:3439 –3443. 17. Rogove HJ, Safar P, Sutton-Tyrrell K, Abramson N. Old age does not negate good cerebral outcome after cardiopulmonary resuscitation: Analyses from the brain resuscitation clinical trials. The Brain Resuscitation Clinical Trial I and II Study Groups. Crit Care Med 1995;23:18 –25. 18. Swor RA, Jackson RE, Tintinalli JE, Pirrallo RG. Does advanced age matter in outcomes after out-of-hospital cardiac arrest in communitydwelling adults? Acad Emerg Med 2000;7:762–768. 19. Tresch DD, Thakur RK. Cardiopulmonary resuscitation in the elderly. Beneficial or an exercise in futility? Emerg Med Clin North Am 1998; 16:649 – 663. 20. Anonymous. Defibrillators for at-risk locations. Newton Kansan. May 31, 2003:5. 21. Kolata G. Extending life, defibrillators can prolong misery. NY Times (Print). March 25, 2002;A1, 18. 22. Scheetz A. Informed consent for public automated external defibrillation. JAMA 2001;286:47. 23. Hansdottir H, Gruman C, Curry L, Judge JO. Preferences for CPR among the elderly: The influence of attitudes and values. Conn Med 2000;64: 625– 630. 24. Hui E, Ho SC, Tsang J, et al. Attitudes toward life-sustaining treatment of older persons in Hong Kong. J Am Geriatr Soc 1997;45:1232–1236. 25. Liddle J, Gilleard C, Neil A. Elderly patients’ and their relatives’ views on CPR. Lancet 1993;342:1055. 26. Mead GE, Turnbull CJ. Cardiopulmonary resuscitation in the elderly: Patients’ and relatives’ views. J Med Ethics 1995;21:39 – 44. 27. Morgan R, King D, Prajapati C, Rowe J. Views of elderly patients and their relatives on cardiopulmonary resuscitation. BMJ 1994;308:1677– 1678. 28. Phillips K, Woodward V. The decision to resuscitate: Older people’s views. J Clin Nurs 1999;8:753–761. 29. Sherwood S, Ruchlin HS, Sherwood CC, Morris SA. Continuing Care Retirement Communities. Baltimore, MD: Johns Hopkins University Press, 1997. 30. Walker RM, Schonwetter RS, Kramer DR, Robinson BE. Living wills and resuscitation preferences in an elderly population. Arch Intern Med 1995;155:171–175. 31. Sayers GM, Schofield I, Aziz M. An analysis of CPR decision-making by elderly patients. J Med Ethics 1997;23:207–212. 32. Mead GE, O’Keefe ST, Jack CI, et al. What factors influence patient preferences regarding cardiopulmonary resuscitation? J R Coll Physicians Lond 1995;29:295–298. 33. Rosner B. Fundamentals of Biostatistics, 4th ed. Belmont, CA: Wadsworth, 1995. 34. Tabachnik BG, Fidell LS. Using Multivariate Statistics, 4th ed. Boston, MA: Allyn and Bacon, 2001. 35. American Association of Homes and Services for the Aged. The Continuing Care Retirement Community (CCRC) Industry Profile—1997. Washington, DC: AAHSA, 1997. 36. Hosmer DW, Lemeshow S. Applied Logistic Regression. New York, NY: Wiley, 2000. 37. Department of Health and Human Services. Profile of Older Americans: 2003, Figure 3: Living Arrangements of Persons 65⫹: 2002. Available at: http://www.aoa.gov/prof/Statistics/profile/2003/6.asp#figure3. Accessed November 8, 2005. 38. Beauchamp TL, Walters L. Contemporary Issues in Bioethics. Belmount, CA: Thomson-Wadsworth, 2003.

JAMDA – March 2006