Advance Care Planning to Improve End-of-Life Decision-Making Consistency between Older People and Their Surrogates in Taiwan

Abstract

Background:

Surrogates often do not accurately predict older people's preferences about end-of-life (EOL) care. Few studies have examined the impact of advance care planning (ACP) on EOL decision-making consistency between older people and their surrogates, and these studies have yielded conflicting results.

Objectives:

To evaluate the effectiveness of ACP in improving EOL decision-making consistency between older people and their surrogates.

Design:

The intervention in this pre–post quasi-experimental design included an informative video, a brochure about ACP, and a guided discussion about EOL wishes.

Setting:

Two geriatric wards in a medical center in northern Taiwan.

Subjects:

One hundred eight participants, as 54 pairs of older people and their surrogates, were randomly assigned to either the experimental or control group. The experimental group received an intervention, while the control group received usual care.

Measurements:

Life-Support Preferences Questionnaire.

Results:

The intervention did not improve decision-making consistency between older people and their surrogates. This was the first time that most pairs discussed specific EOL decisions, so additional preparation may improve comfort with this topic. This study also found that some older people had difficulty concentrating on the educational brochure or understanding the related terms.

Conclusions:

Preparation for ACP discussion is needed for older people and their surrogates. Longer-term effects of ACP should be monitored because ACP interventions may have enhanced empathy between older people and their surrogates. Additionally, a culturally sensitive illustrated questionnaire that explains life-support preferences and ACP topics may improve communication between older people and their surrogates.

Introduction

As adults age, their cognitive health may decline and they may lose the capacity to make health-related decisions.¹ When this happens, family members often serve as surrogate health care decision makers.² However, many studies have found that health care surrogates often do not accurately predict older people's preferences regarding medical treatments and end-of-life (EOL) care.^3–6 Therefore, improving communication between older people and their surrogates during advance care planning (ACP) is important to ensure appropriate decision making.

ACP addresses difficult problems related to medical and EOL care. ACP usually involves discussions with health professionals, family, and significant others about future health care preferences while the person has the capacity to do so.⁷ ACP often involves financial and funeral arrangements in addition to medical and quality-of-life decisions.⁸ However, personal or cultural stigma surrounding these discussions can complicate EOL decision making. In these circumstances, the substituted judgment standard is often used. This standard specifies that the designated decision makers (i.e., health care surrogates) make choices that align with and respect those of the patients.^9,10

Cultural differences between Western and Eastern countries affect patient care. Whereas Western culture focuses on individual freedom and autonomy, East Asian culture emphasizes family harmony and unity.^11–14 Thus, the concept of autonomy in Asia emphasizes family responsibilities over personal rights. In this cultural context, the balance between family harmony and patient autonomy is crucial. Enhancing EOL care communication between older people and their health care surrogates helps ensure this balance.

ACP can involve several approaches, such as motivational tools (e.g., advertising that promotes communication among family members), educational tools (e.g., brochures that target families with elderly members), professional guidance (e.g., discussions with doctors and other health care workers about options and expectations), informal discussions (e.g., family talks), and formal plans (e.g., legal documents that specify one's wishes). A few studies have examined how these approaches improved decision-making consistency between older people and their health care surrogates, yielding conflicting results.^15–18 Two of these studies showed that interventions did not improve surrogates' predictions,^15,16 whereas others reported that videos and discussions did improve decision making and communication among surrogates and the older people in their care.^17,18 Volandes et al.¹⁸ found that a two-minute video about an elderly patient's advanced dementia encouraged discussions about ACP and increased the congruence of EOL care decisions between elders and their surrogates. Matheis-Kraft and Roberto¹⁷ adopted 23 value indicators (including independence, control, dignity, and hope) to guide ACP that significantly improved the consistency of decision making between older women and their surrogates concerning EOL treatments.

Previous research is limited to Western countries, so it is unknown whether such interventions would be successful in different cultural settings. Asian individuals tend to discuss death and dying indirectly.¹⁹ For example, people prefer not to discuss death and dying in their daily life due to fears of inviting bad luck.^20,21 While older people might make funeral or financial arrangements for their death, they do not typically engage in ACP.^22–26 Thus, appropriate venues for discussing ACP should be culturally sensitive to those who prefer indirect discussions, such as a social event or commercial video appropriate to their situation. These might provide an opportunity for people to consider unforeseen events. Education is also crucial when discussing EOL care decisions. Sufficient knowledge of advance directives and EOL care, for example, could help people in their ACP and decrease misunderstanding concerning EOL issues.^19,27,28 Therefore, this study aimed to assess whether an ACP intervention (i.e., motivational video, educational brochure, and guided discussion about EOL wishes) increased decision-making consistency between older people and their health care surrogates regarding EOL preferences.

Methods

Design

In this pre–post quasi-experimental design conducted from March 2015 to March 2016, participants from two geriatric wards (25 beds each) at a medical center in northern Taiwan were randomly assigned to a control or experimental group.

Participants

Eligible participants were adult patients over 65 years of age with an appointed family member or friend over the age of 20 serving as their health care surrogate. The Civil Code in Taiwan defines people aged at least 20 years as adults with full capacity, thus defining the lower age limit for surrogates. Exclusion criteria included (1) severe hearing loss, (2) severe vision defects, (3) diagnosis of moderate to severe dementia, (4) critical status, (5) diagnosis of major depression, (6) current participation in clinical trials, and (7) difficulty communicating in Mandarin or Taiwanese.

To determine the adequate sample size, G*power 3.1.9.2²⁹ was used based on a two-tailed Wilcoxon–Mann–Whitney test with two groups. A sample size of 27 elder–surrogate pairs per group was required based on an alpha of 0.05, an effect size of 0.6, and power of 0.8. Predicting an 80% retention rate, 65 pairs were recruited. Of these 65 elder–surrogate pairs, 35 were randomly assigned to the experimental group and 30 to the control group. Eight pairs were excluded due to incomplete pretest or post-test data. Three additional pairs dropped out of the study because the older people felt uncomfortable about completing the questionnaire. They commented that it made them feel sad or that it was bad luck to discuss this issue. Therefore, 54 pairs completed the study (30 in the experimental group and 24 in the control; Fig. 1).

FIG. 1.

Flowchart of the recruiting process.

Intervention

The study intervention included a motivational video, an educational brochure, and a guided discussion. The video was used to stimulate conversations about EOL care among the study participants. BNP Paribas Cardif TCB Life Insurance Co., Ltd., produced a 6-minute and 28-second video titled “A Humble Wish from My Father with Dementia” (https://www.youtube.com/watch?v=NqXYJ7HWtjg) that describes how a father nurtured and raised his son, but then became affected by dementia. As his father's caregiver, the son encountered many difficulties. Although this is a commercial video for mortgage life insurance, it could provide insight into EOL care. The company granted authorization to use the video through YouTube.

The educational brochure covered several topics: learning about ACP, choosing medical treatments, designating a surrogate for health care power of attorney, discussing wishes with family and health care teams, and writing down one's wishes. Dr. Ying-Wei Wang of the Hospice Foundation of Taiwan edited the Advance Care Planning brochure (https://health99.hpa.gov.tw/educZone/edu_detail.aspx?CatId=21853). The foundation granted permission to reprint the original A5 size (148 × 210 mm) as an A4 size (210 × 297 mm) with enlarged text to make it more readable for those with visual degeneration. Additionally, two hypothetical vignettes from the educational brochure were used to encourage participants to share their EOL wishes (Appendix A1).

Data collection

Medical records and interviews provided clinical and demographic data as previous studies show that patients' cognitive, physical, and psychological statuses influence whether they implement ACP or advance directives.^30,31 Elderly participants underwent routine assessments, including the Mini-Mental State Examination (MMSE), Activities of Daily Living (ADL) assessment, Instrumental Activities of Daily Living (IADL) score, and Geriatric Depression Scale (GDS), the results of which were included in their medical records. The MMSE screens for cognitive function using a 0–30-point score, with higher scores indicating better cognitive status.³² The ADL measures one's ability to perform daily tasks necessary for independent living at home, using a score of 0–100, with higher scores indicating better physical function.³³ The IADL assesses one's ability to live independently, using a score of 0–8, with higher scores demonstrating more independence.³⁴ The GDS measures depressive mood on a 0–5-point scale, with higher scores signifying more depressive symptoms.³⁵

The Life-Support Preferences Questionnaire (LSPQ) uses nine health-related scenarios to help people identify their preferences about EOL treatments.³⁶ The scenarios are (1) currently healthy, (2) severe dementia, (3) continuous dyspnea, (4) coma status with no chance of recovery, (5) coma status with a slight chance of recovery, (6) severe stroke with no chance of recovery, (7) severe stroke with a slight chance of recovery, (8) terminal cancer without pain, and (9) terminal cancer with pain. Each scenario includes four potential medical treatments (antibiotics, cardiopulmonary resuscitation, surgery, and artificial nutrition and hydration) for a total of 36 options. Participants used a 5-point Likert scale, ranging from absolutely unwanted (1 point) to absolutely wanted (5 points), to indicate their preferences for each option. Data were collected within three days of the elderly participants' admission dates and before their discharge dates. Regarding reliability of the LSPQ, Cronbach's alpha was 0.86–0.96.^36,37 In the study, Cronbach's alpha for older people was 0.98, and Cronbach's alpha for surrogates was 0.98.

Ethical considerations

The institutional review board of Taipei Veterans General Hospital approved this study (reference number: 2015-03-010B).

Procedures

After explaining the study aims to prospective participants who met the study criteria, informed consent was obtained from those who enrolled. Elderly participants then selected their health care surrogate. Researchers explained the purpose of the study to the appointed surrogates and obtained their informed consent. All participant pairs completed a pretest answering of the LSPQ separately, with the surrogates attempting to predict responses of the elder member of their pair. Researchers provided the following instructions to surrogates: “I would like you to respond to these questions as if you had to make decisions on behalf of your family member. Please base your decision on what you think your family member would want.” A question-and-answer approach was implemented to assist both older people and surrogates in completing the LSPQ. Researchers (L.-S.K., M.-J.C., and H.-C.C) collected the data.

After completing the LSPQ pretest, participant pairs in the experimental group watched the motivational video together on a laptop with a 13-inch screen. Subsequently, researchers (L.-S.K. and M.-J.C.) then read the educational brochure together with both members of the pair and used the hypothetical vignettes to encourage them to share their health care wishes. The family conference room or single-patient room was chosen for conducting the intervention. The average intervention time was 44 minutes (including watching the video). The time between the intervention and the post-test for the experimental group was 6.4 days. Participant pairs in the control group received usual care and received the educational brochure after completing the study.

Data analysis

Data analysis used the SPSS 20.0 software (IBM Corp., Armonk, NY), with p-values less than 0.05 indicating a significant effect. Descriptive statistics included mean, standard deviation, and percentage. We used the chi-square test to examine homogeneity of nominal variables and the Mann-Whitney (M-W) test to analyze homogeneity of continuous demographic variables. To answer the study question, differences between older people and their surrogates on the LSPQ were calculated to show the degree of agreement. Large gaps between participant pairs indicated poor agreement. Owing to the small sample size, nonparametric (i.e., M-W and Wilcoxon signed-rank) tests were employed.

Results

Before the intervention, control and experimental groups were tested with regard to homogeneity. Older people in the experimental and control groups did not differ significantly with regard to age, gender, educational level, religion, marital status, living condition, financial support, and number of children. In addition, there were no significant differences in self-reported current health; MMSE, ADL, IADL, and GDS scores; primary diagnosis; or the number of chronic diseases. Elderly participants were predominantly male, married, and living with family. The main diagnoses of older people were musculoskeletal and infectious diseases. In addition, most elderly participants had never heard of ACP, and over half had never discussed EOL-related issues (Table 1).

Table 1.

Demographic Characteristics of Older People (N = 54)

Variables	Experimental group (n = 30)	Control group (n = 24)	Statistics	p
Clinic information
Current health status^a, mean ± SD	2.73 ± 0.87	2.71 ± 0.86	z = −0.286	0.775
MMSE, mean ± SD	22.97 ± 4.33	23.29 ± 3.77	z = −0.044	0.965
ADL, mean ± SD	74.50 ± 21.26	81.46 ± 18.62	z = −1.235	0.217
IADL, mean ± SD	3.00 ± 2.13	3.67 ± 2.18	z = −1.284	0.199
GDS, mean ± SD	1.90 ± 1.56	2.33 ± 1.52	z = −1.111	0.267
No. of chronic diseases, mean ± SD	4.77 ± 2.27	5.42 ± 1.84	z = −1.259	0.208
Main diagnosis, n			X² = 1.899	0.594
Musculoskeletal diseases	12	12
Infectious disease	9	4
Cancer	1	2
Others	8	6
Demographic information
Age, years, mean ± SD	85.2 ± 6.34	84.0 ± 5.37	z = −0.916	0.360
Gender, n
Male	23	19	X² = 0.048	0.826
Female	7	5
Educational level, n			X² = 0.761	0.859
Elementary school and below	10	10
High school	7	4
Senior high school	4	4
College degree and above	9	6
Marital status, n			X² = 0.20	0.887
Married	23	18
Widowed	7	6
No. of children, n			X² = 4.040	0.257
None	0	2
1–2 children	9	4
2–4 children	14	10
>5 children	7	8
Living condition, n			X² = 1.890	0.389
Living with family	28	20
Living in long-term care facilities	2	3
Living alone	0	1
Financial support, n			X² = 4.812	0.186
Pensions	14	11
Veterans subsidies	4	3
Savings	1	5
Family support	11	5
Religion, n			X² = 0.311	0.856
None	10	9
Buddhism/Taoism	14	9
Others	6	5
Have you ever heard of ACP? n			X² = 1.350	0.245
Yes	2	4
No	28	20
Have you ever discussed issues related to EOL with family or friends? n			X² = 1.148	0.284
Yes	13	7
No	17	17

The range of self-reported current health was 1–5 points, with higher scores indicating healthier status.

ACP, advance care planning; ADL, Activities of Daily Living; EOL, end-of-life; GDS, Geriatric Depression Scale; IADL, Instrumental Activities of Daily Living; MMSE, Mini-Mental State Examination; SD, standard deviation.

Surrogates were not significantly different between the experimental and control groups in terms of age, gender, educational level, religion, marital status, living condition, financial support, or number of children. Importantly, the relationships of surrogates with older people were not significantly different between the two groups (Table 2). Surrogates were predominantly adult children and spouses.

Table 2.

Demographic Characteristics of Surrogates (N = 54)

Variables	Experimental group (n = 30)	Control group (n = 24)	Statistics	p
Age, years, mean ± SD	58.5 ± 12.89	59.3 ± 10.60	z = −0.226	0.821
Relationships of surrogates with older people, n			X² = 0.464	0.793
Spouses	10	10
Adult children	19	13
Son-in-law/daughter-in-law	1	1
Gender, n			X² = 0.248	0.618
Male	8	5
Female	22	19
Educational level, n			X² = 1.331	0.722
Elementary school and below	4	5
High school	2	3
Senior high school	8	6
College degree and above	16	10
Marital status, n			X² = 0.672	0.715
Unmarried	5	4
Married	22	19
Others	3	1
No. of children, n			X² = 2.740	0.434
None	7	4
1–2 children	17	12
2–4 children	5	8
>5 children	1	0
Living condition, n			X² = 1.662	0.197
Living with family	28	24
Living alone	2	0
Financial support, n			X² = 6.667	0.083
Pension	6	12
Savings	7	2
Family support	6	2
Working	11	8
Religion, n			X² = 0.408	0.815
None	8	6
Buddhism/Taoism	18	16
Others	4	2

The LSPQ pretest revealed significant differences between pairs in the experimental and control groups for 9 of the 36 health-related scenarios (Table 3). Differences between pairs in the control group were higher than those in the experimental group, indicating that the experimental and control groups were unequal at the baseline.

Table 3.

Comparison of Differences between Older People and Surrogate Pairs from the Pretest to the Post-Test (Experimental Group n = 30 Pairs; Control Group n = 24 Pairs)

Variables	Pretest					Post-test					Post-test–pretest
Variables	Older people (O) Mean (SD)	Surrogates (S) Mean (SD)	Difference (D = O − S) Mean (SD)	z	p	Older people (O) Mean (SD)	Surrogates (S) Mean (SD)	Difference (D = O − S) Mean (SD)	z	p	Difference change (DC = post-test D − pretest D) Mean (SD)	z	p
Currently healthy
Antibiotics				−1.527	0.127				−1.055	0.292		−0.405	0.685
Experimental group	3.80 (1.38)	4.27 (1.05)	−0.47 (1.46)			3.30 (1.60)	3.67 (1.30)	−0.37 (1.81)	−0.122	0.903	0.10 (2.34)
Control group	4.17 (1.20)	4.12 (0.85)	0.04 (1.12)			4.04 (1.23)	4.04 (1.23)	0.00 (1.59)	−0.130	0.897	−0.04 (1.30)
CPR				−0.848	0.397				−0.809	0.419		−1.343	0.179
Experimental group	2.33 (1.56)	2.60 (1.63)	−0.27 (1.46)			2.10 (1.62)	2.60 (1.65)	−0.50 (1.74)	−0.824	0.410	−0.23 (1.79)
Control group	2.33 (1.55)	2.96 (1.55)	−0.63 (1.81)			2.33 (1.37)	2.67 (1.58)	−0.33 (1.79)	−0.723	0.469	0.29 (1.78)
Surgery				−0.697	0.486				−0.761	0.447		−1.462	0.144
Experimental group	2.77 (1.55)	3.47 (1.41)	−0.70 (1.95)			2.60 (1.57)	3.00 (1.26)	−0.40 (1.54)	−0.721	0.471	0.30 (1.93)
Control group	2.58 (1.64)	3.71 (1.27)	−1.13 (1.96)			3.25 (1.73)	3.29 (1.52)	−0.04 (2.01)	−2.198^a	0.028	1.08 (2.12)
ANH				−1.451	0.147				−0.009	0.993		−0.757	0.449
Experimental group	2.30 (1.24)	2.93 (1.41)	−0.63 (1.65)			2.07 (1.60)	2.87 (1.41)	−0.80 (1.97)	−0.480	0.631	−0.17 (1.70)
Control group	2.25 (1.65)	3.54 (1.35)	−1.29 (1.99)			2.21 (1.41)	3.08 (1.41)	−0.88 (1.83)	−0.809	0.418	0.42 (2.28)
Severe dementia
Antibiotics				−1.704	0.088				−0.402	0.688		−1.807	0.071
Experimental group	3.00 (1.49)	3.37 (1.30)	−0.37 (1.77)			2.33 (1.61)	3.10 (1.42)	−0.77 (1.85)	−0.861	0.389	−0.40 (2.25)
Control group	2.83 (1.63)	4.08 (1.10)	−1.25 (1.87)			3.04 (1.68)	3.54 (1.50)	−0.50 (1.96)	−1.768	0.077	0.75 (1.98)
CPR				−0.839	0.402				−1.307	0.191		−0.080	0.936
Experimental group	2.10 (1.56)	2.23 (1.31)	−0.13 (1.55)			1.90 (1.49)	1.83 (1.15)	0.07 (1.62)	−0.577	0.564	0.20 (2.16)
Control group	2.00 (1.47)	2.71 (1.37)	−0.71 (1.85)			2.00 (1.45)	2.63 (1.53)	−0.63 (1.74)	−0.261	0.794	0.08 (1.41)
Surgery				−1.981^a	0.048				−0.569	0.570		−1.304	0.192
Experimental group	2.67 (1.52)	2.93 (1.44)	−0.27 (1.70)			2.17 (1.51)	2.37 (1.25)	−0.20 (1.75)	−0.214	0.830	0.07 (1.86)
Control group	1.92 (1.38)	3.25 (1.36)	−1.33 (2.01)			2.46 (1.38)	3.08 (1.59)	−0.63 (1.86)	−1.522	0.128	0.71 (2.18)
ANH				−1.367	0.172				−1.135	0.256		−0.027	0.979
Experimental group	2.10 (1.35)	2.57 (1.36)	−0.47 (1.81)			1.80 (1.38)	2.27 (1.20)	−0.47 (1.74)	−0.209	0.834	0.00 (2.26)
Control group	1.75 (1.23)	2.92 (1.53)	−1.17 (1.81)			1.87 (1.39)	3.08 (1.44)	−1.21 (1.77)	−0.106	0.916	−0.04 (2.12)
Continuous dyspnea
Antibiotics				−0.565	0.572				−0.485	0.628		−0.276	0.783
Experimental group	3.07 (1.53)	3.30 (1.42)	−0.23 (1.91)			2.93 (1.57)	3.00 (1.49)	−0.07 (2.18)	−0.229	0.819	0.17 (2.55)
Control group	3.33 (1.52)	3.88 (1.26)	−0.54 (2.04)			3.25 (1.48)	3.67 (1.37)	−0.42 (2.10)	−0.438	0.662	0.13 (1.57)
CPR				−0.116	0.907				−1.279	0.201		−0.945	0.345
Experimental group	2.10 (1.49)	2.50 (1.38)	−0.40 (1.50)			1.90 (1.45)	1.73 (1.17)	0.17 (1.72)	−1.400	0.162	0.57 (2.34)
Control group	2.08 (1.44)	2.71 (1.52)	−0.63 (1.93)			2.04 (1.40)	2.50 (1.45)	−0.46 (1.82)	−0.477	0.633	0.17 (1.74)
Surgery				−1.414	0.157				−0.513	0.608		−0.952	0.341
Experimental group	2.57 (1.43)	2.87 (1.53)	−0.30 (1.74)			2.00 (1.46)	2.37 (1.19)	−0.37 (1.52)	−0.066	0.947	−0.07 (1.80)
Control group	2.17 (1.31)	3.13 (1.54)	−0.96 (2.05)			2.21 (1.35)	2.79 (1.56)	−0.58 (1.69)	−0.862	0.389	0.38 (2.04)
ANH				−1.656	0.098				−0.968	0.333		−0.666	0.505
Experimental group	2.17 (1.32)	2.57 (1.25)	−0.40 (1.54)			1.77 (1.31)	2.47 (1.28)	−0.70 (1.34)	−0.542	0.587	−0.30 (1.73)
Control group	2.04 (1.50)	3.29 (1.57)	−1.25 (2.07)			1.79 (1.25)	2.96 (1.49)	−1.17 (1.76)	−0.309	0.757	0.08 (2.34)
Coma status with no chance of recovery
Antibiotics				−0.622	0.534				−1.829	0.067		−0.957	0.339
Experimental group	2.23 (1.50)	2.67 (1.45)	−0.43 (1.63)			1.90 (1.37)	2.33 (1.37)	−0.43 (1.72)	−0.230	0.818	0.00 (2.05)
Control group	2.29 (1.50)	3.21 (1.62)	−0.92 (1.86)			2.00 (1.41)	3.33 (1.58)	−1.33 (2.22)	−1.254	0.210	−0.42 (1.82)
CPR				−1.218	0.223				−2.194^a	0.028		−0.063	0.949
Experimental group	2.00 (1.46)	1.80 (1.06)	0.20 (1.45)			1.67 (1.32)	1.47 (0.86)	0.20 (1.54)	−0.311	0.756	0.00 (2.27)
Control group	1.87 (1.39)	2.54 (1.44)	−0.67 (1.76)			1.83 (1.20)	2.46 (1.41)	−0.63 (1.84)	0.000	1.000	0.04 (1.57)
Surgery				−1.065	0.287				−1.388	0.165		−0.295	0.768
Experimental group	1.93 (1.34)	2.33 (1.47)	−0.40 (0.97)			1.73 (1.26)	1.83 (1.09)	−0.10 (1.40)	−1.411	0.158	0.30 (1.47)
Control group	1.71 (1.23)	2.67 (1.55)	−0.96 (2.01)			1.92 (1.35)	2.67 (1.44)	−0.75 (1.70)	−0.264	0.792	0.21 (1.93)
ANH				−2.005^a	0.045				−2.262^a	0.024		−0.116	0.908
Experimental group	2.03 (1.30)	2.20 (1.30)	−0.17 (1.34)			1.80 (1.38)	1.80 (1.10)	0.00 (1.89)	−0.400	0.689	0.17 (1.95)
Control group	1.71 (1.27)	3.00 (1.59)	−1.29 (1.94)			1.79 (1.32)	3.00 (1.50)	−1.21 (1.64)	−0.627	0.530	0.08 (1.95)
Coma status with a slight chance of recovery
Antibiotics				−1.854	0.064				−0.655	0.512		−1.321	0.187
Experimental group	2.20 (1.38)	2.70 (1.54)	−0.50 (1.61)			1.87 (1.31)	2.33 (1.37)	−0.47 (1.72)	−0.296	0.767	0.03 (1.92)
Control group	1.92 (1.35)	3.42 (1.59)	−1.50 (1.74)			2.50 (1.69)	3.25 (1.51)	−0.75 (1.92)	−1.790	0.073	0.75 (2.01)
CPR				−2.086^a	0.037				−1.803	0.071		−1.028	0.304
Experimental group	1.90 (1.27)	1.70 (1.06)	0.20 (1.21)			1.60 (1.22)	1.37 (0.72)	0.23 (1.38)	−0.024	0.981	0.03 (1.83)
Control group	1.79 (1.38)	2.75 (1.51)	−0.96 (1.73)			1.75 (1.40)	2.33 (1.44)	−0.58 (1.64)	−1.168	0.243	0.38 (1.56)
Surgery				−1.296	0.195				−1.409	0.159		−0.205	0.838
Experimental group	2.00 (1.44)	2.37 (1.47)	−0.37 (1.33)			1.63 (1.25)	1.73 (1.02)	−0.10 (1.32)	−0.902	0.367	0.27 (1.62)
Control group	1.67 (1.20)	2.75 (1.51)	−1.08 (1.74)			1.87 (1.33)	2.67 (1.49)	−0.79 (1.69)	−0.708	0.479	0.29 (2.07)
ANH				−2.254^a	0.024				−1.637	0.102		−0.471	0.638
Experimental group	2.00 (1.41)	2.13 (1.20)	−0.13 (1.53)			1.73 (1.39)	1.83 (1.15)	−0.10 (1.71)	−0.093	0.926	0.03 (2.13)
Control group	1.83 (1.34)	3.13 (1.48)	−1.29 (1.92)			1.92 (1.41)	3.00 (1.53)	−1.08 (1.82)	−0.816	0.415	0.21 (1.74)
Severe stroke with no chance of recovery
Antibiotics				−1.718	0.086				−1.343	0.179		−0.348	0.728
Experimental group	1.80 (1.24)	2.90 (1.40)	−1.10 (1.40)			1.83 (1.29)	2.37 (1.38)	−0.53 (1.80)	−2.194^a	0.028	0.57 (1.52)
Control group	1.62 (1.17)	3.50 (1.47)	−1.88 (1.85)			1.96 (1.43)	3.17 (1.55)	−1.21 (2.13)	−1.365	0.172	0.67 (2.22)
CPR				−1.497	0.134				−1.083	0.279		−1.387	0.165
Experimental group	1.77 (1.25)	1.83 (1.09)	−0.07 (1.08)			1.57 (1.25)	1.63 (1.07)	−0.07 (1.64)	−0.019	0.984	0.00 (2.07)
Control group	1.71 (1.27)	2.75 (1.60)	−1.04 (1.92)			1.75 (1.23)	2.25 (1.39)	−0.50 (1.82)	−1.331	0.183	0.54 (1.79)
Surgery				−1.670	0.095				−1.361	0.174		−0.758	0.448
Experimental group	1.90 (1.27)	2.20 (1.30)	−0.30 (1.18)			1.70 (1.26)	1.90 (1.09)	−0.20 (1.52)	−0.362	0.718	0.33 (1.74)
Control group	1.62 (1.10)	2.83 (1.55)	−1.21 (1.91)			1.87 (1.19)	2.75 (1.51)	−0.88 (1.54)	−0.868	0.386	0.10 (1.18)
ANH				−2.637^b	0.008				−1.892	0.058		−0.447	0.655
Experimental group	1.93 (1.26)	2.10 (1.16)	−0.17 (1.21)			1.77 (1.38)	1.93 (1.17)	−1.67 (1.78)	−0.323	0.747	0.00 (1.78)
Control group	1.67 (1.09)	3.08 (1.53)	−1.42 (1.91)			1.67 (1.17)	2.92 (1.59)	−1.25 (1.78)	−0.819	0.413	0.17 (1.63)
Severe stroke with a slight chance of recovery
Antibiotics				−1.826	0.068				−1.010	0.313		0.000	1.000
Experimental group	2.27 (1.44)	2.87 (1.48)	−0.60 (1.90)			2.03 (1.45)	2.37 (1.33)	−0.33 (1.94)	−0.752	0.452	0.27 (2.03)
Control group	1.87 (1.33)	3.50 (1.47)	−1.63 (1.97)			2.33 (1.52)	3.25 (1.45)	−0.92 (2.00)	−1.147	0.251	0.71 (2.61)
CPR				−1.436	0.151				−1.648	0.099		−0.260	0.795
Experimental group	1.70 (1.18)	1.80 (1.10)	−0.10 (1.06)			1.77 (1.36)	1.63 (1.07)	0.13 (1.66)	−0.621	0.534	0.23 (1.92)
Control group	1.54 (1.06)	2.58 (1.56)	−1.04 (1.83)			1.71 (1.20)	2.37 (1.38)	−0.67 (1.79)	−0.814	0.416	0.38 (2.04)
Surgery				−1.209	0.227				−1.745	0.081		−0.143	0.887
Experimental group	1.83 (1.29)	2.33 (1.45)	−0.50 (1.25)			1.67 (1.27)	1.83 (1.02)	−0.17 (1.44)	−1.021	0.307	0.33 (1.56)
Control group	1.67 (1.05)	2.92 (1.59)	−1.25 (1.67)			1.92 (1.41)	2.79 (1.47)	−0.88 (1.42)	−0.838	0.402	0.38 (1.79)
ANH				−2.831^b	0.005				−1.997^a	0.046		−0.708	0.479
Experimental group	2.97 (1.35)	2.10 (1.19)	−0.13 (1.50)			1.87 (1.43)	1.93 (1.14)	−0.07 (1.86)	−0.015	0.988	0.07 (2.21)
Control group	1.67 (1.09)	3.25 (1.48)	−1.58 (1.86)			1.83 (1.31)	3.00 (1.47)	−1.17 (1.55)	−1.316	0.188	0.42 (1.56)
Terminal cancer without pain
Antibiotics				−1.218	0.223				−0.619	0.536		−0.283	0.777
Experimental group	2.07 (1.29)	3.03 (1.61)	−0.97 (1.61)			1.97 (1.45)	2.60 (1.38)	−0.63 (2.16)	−0.880	0.379	0.33 (2.07)
Control group	2.08 (1.38)	3.62 (1.38)	−1.54 (1.82)			2.25 (1.51)	3.21 (1.44)	−0.96 (1.99)	−1.043	0.297	0.58 (2.26)
CPR				−1.207	0.227				−1.929	0.054		−0.200	0.842
Experimental group	1.73 (1.26)	1.97 (1.25)	−0.23 (1.59)			1.53 (1.14)	1.53 (1.04)	0.00 (1.39)	−0.726	0.468	0.23 (1.79)
Control group	1.67 (1.27)	2.58 (1.41)	−0.92 (1.64)			1.79 (1.29)	2.50 (1.38)	−0.71 (1.71)	−0.573	0.566	0.21 (1.50)
Surgery				−2.410^a	0.016				−0.979	0.328		−1.415	0.157
Experimental group	1.83 (1.23)	2.00 (1.20)	−0.17 (1.53)			1.53 (1.07)	1.70 (0.88)	−0.17 (1.39)	−0.124	0.902	0.83 (1.81)
Control group	1.38 (0.82)	2.75 (1.54)	−1.38 (1.50)			1.79 (1.29)	2.33 (1.31)	−0.54 (1.84)	−2.010^a	0.044	0.00 (2.00)
ANH				−2.272^a	0.023				−1.932	0.053		−0.205	0.838
Experimental group	1.97 (1.35)	2.17 (1.23)	−0.20 (1.69)			1.63 (1.22)	1.77 (1.10)	−0.13 (1.55)	−0.149	0.881	0.07 (1.91)
Control group	1.75 (1.19)	3.17 (1.44)	−1.42 (2.00)			1.71 (1.23)	2.92 (1.50)	−1.21 (1.61)	−0.522	0.601	0.21 (2.02)
Terminal cancer with pain
Antibiotics				−1.404	0.160				−0.891	0.373		−0.660	0.509
Experimental group	1.70 (1.12)	2.57 (1.52)	−0.87 (1.70)			1.70 (1.32)	2.13 (1.25)	−0.43 (1.96)	−0.670	0.503	0.43 (2.40)
Control group	1.87 (1.23)	3.46 (1.41)	−1.58 (1.77)			2.04 (1.49)	3.00 (1.47)	−0.96 (2.26)	−1.324	0.186	0.63 (2.16)
CPR				−1.704	0.088				−0.693	0.488		−1.377	0.169
Experimental group	1.67 (1.12)	1.67 (1.21)	0.00 (1.39)			1.37 (0.96)	1.40 (0.62)	−0.03 (1.13)	−0.185	0.853	−0.03 (1.45)
Control group	1.58 (1.10)	2.38 (1.25)	−0.79 (1.67)			1.79 (1.32)	2.17 (1.31)	−0.38 (1.79)	−1.215	0.224	0.42 (1.50)
Surgery				−1.827	0.068				−0.468	0.640		−0.934	0.350
Experimental group	1.73 (1.20)	1.80 (1.16)	−0.07 (1.66)			1.33 (0.96)	1.57 (0.68)	−0.23 (1.30)	−0.533	0.594	−0.17 (2.00)
Control group	1.46 (0.88)	2.46 (1.47)	−1.00 (1.56)			1.71 (1.27)	2.25 (1.36)	−0.54 (1.91)	−1.005	0.315	0.46 (1.69)
ANH				−2.261^a	0.024				−0.127	0.899		−1.838	0.066
Experimental group	1.80 (1.24)	2.07 (1.41)	−0.27 (1.76)			1.37 (0.96)	1.90 (1.16)	−0.53 (1.46)	−0.531	0.595	−0.27 (1.95)
Control group	1.62 (1.06)	3.13 (1.42)	−1.50 (1.82)			2.04 (1.60)	2.79 (1.59)	−0.75 (1.96)	−2.044^a	0.041	0.75 (1.65)

p < 0.05.

p < 0.01.

ANH, artificial nutrition and hydration; CPR, cardiopulmonary resuscitation.

After the intervention, we observed significant differences in LSPQ responses between participant pairs in the experimental and control groups for 3 of 36 health-related scenarios. In the control group, the LSPQ post-test revealed significant differences between older people and their surrogates for three scenarios. For the experimental group, significant differences were observed in only one scenario. We also compared each individual pair to determine if the size of the difference changed from the pretest to the post-test and found no significant differences between the experimental and control groups (Table 3).

The results indicated that the intervention did not improve decision-making consistency between older people and their health care surrogates. These results may be related to the complex familial relationships between older people and their children as 59.3% of surrogates in the study were the adult children of the patients. For example, during the video, many of the children had tears in their eyes while watching with their parents. The adult children seemed eager to learn their parents' wishes, but older people seemed uneasy sharing their health care wishes. For most participant pairs, it was their first discussion of specific EOL decisions. Perhaps more preparation would increase comfort with the topic.

Additionally, this study found that some older people had difficulty concentrating on the educational brochure or understanding the related terms. For instance, they did not understand the difference between intubation with a breathing tube and insertion of a nasogastric tube.

Discussion

The study intervention had no effect on agreement about EOL care between older people and their surrogates. Bravo and her colleagues' study facilitated ACP through an educational program and a structural approach, but was unsuccessful in improving surrogates' ability to predict older people's choices.¹⁵ Another study used two kinds of instructional advance directives to guide ACP and found that neither produced significant improvements in the accuracy of surrogates' substituted judgment for any illness scenario or medical treatment.¹⁶ Our study found that the ACP intervention may have enhanced empathy between older people and their surrogates, which may have led to each member of the pair adjusting their answers on the LSPQ out of concern for each other's feelings. However, this study further confirms the importance of communication between health care surrogates and the people in their care to improve EOL decision making.

Most surrogates are spouses or adult children with strong, complex emotional bonds with the older people in their care. When predicting health care wishes, surrogates are not only concerned with older people's wishes but also consider what the best choices are for the older person and the family.³⁸ Even surrogates' prediction errors represent their personal wishes for their beloved.³⁹ It is not surprising that one intervention lasting less than an hour did not improve surrogates' ability to predict older people's preferences. Thus, a qualitative study that explores how surrogates predict older people's EOL preferences might identify specific approaches to improve decision-making accuracy.

Cultural adaption of the LSPQ also needs to be noted. In Taiwan, when people are sick and have a poor appetite, they or their family members ask health professionals to administer nutritional injections.⁴⁰ It is a paradox that some participants refused nasogastric tube feeding, but accepted intravenous nutrition when considering artificial nutrition and hydration on the LSPQ. Future studies should account for local cultural aspects of health care decisions when using the LSPQ.

Increasing understanding of health care knowledge among participants is necessary. It is possible that the educational brochure was not suitable for this experiment despite enlarging the text for easier readability. Older people often have degenerated physical function, diminished abstract thinking, and memory deterioration. These adults might need specialized materials for ACP and EOL decision making. For example, to compensate for visual degeneration, fonts must be of sufficient size and strong contrasting colors such as black and white rather than blue, green, and purple must be used.⁴¹ To account for diminished abstract thinking and memory, teaching materials should be simple, short, specific, and practical. Visual aids, such as pictures and models, may also be helpful.⁴¹ Avoiding complex language can aid comprehension.²¹

Furthermore, studies involving older people can be challenging. The average age among the elderly participants in this study was 84 years. During the question-and-answer approach, when the researcher assisted older people in completing the LSPQ, they struggled to understand the terms. Thus, an illustrated version of the LSPQ might help increase understanding among elderly participants.

Limitations

The researchers acknowledge several limitations in the study. First, the method of matching older people and their surrogates restricts the numbers of participants able to be recruited; however, it is appropriate for conducting such research in a familial society. Second, the study did not focus on surrogates' personal views, but rather the surrogates' ability to predict older people's wishes as a measure of decision-making consistency regarding EOL care preferences, which should be noted when using these results. Third, talking about death and dying is taboo. Three pairs dropped out of this study because the elderly participants felt uncomfortable completing the LSPQ. Thus, it follows that the people who willingly participated in the study also were more willing to discuss these topics, which could have led to selection bias. Additionally, the experimental and control groups were unequal at baseline; therefore, a future, randomized controlled trial might be needed to address these concerns.

Conclusions

The intervention used in this study did not improve consistency of EOL decision making between older people and their surrogates. For most participant pairs, this was the first time they discussed specific EOL decisions, possibly requiring more preparation to become comfortable with the topic. ACP interventions may have enhanced empathy between older people and their surrogates, so potential long-term effects of ACP should be followed. Additionally, an illustrated culturally sensitive questionnaire that explains life-support preferences and other ACP topics may help improve decision-making communication between older people and their surrogates.

Footnotes

Acknowledgment

Parts of this research have been presented at the ICN Conference, Barcelona, Spain, 2017.

Funding Information

This research was supported by Taipei Veterans General Hospital (V104A-036).

Author Disclosure Statement

No competing financial interests exist.

Appendix

Your True Wishes

In the following scenario, what kind of medical care would you choose?

【Situation 1】

Suppose you have a serious brain injury (perhaps dementia or are in a postcoma unresponsive state). Two specialists judge your clinical chances of recovery to be small, and there is no treatment to reawaken you. Your expectations for medical treatment in this situation would be to:

□ Use all methods to extend my life span.

□ Provide treatment and continue evaluating. If the treatment is ineffective, then stop.

□ Only provide less invasive treatment.

□ Only provide palliative care (treatment to improve comfort only).

□ Other (please specify)____________________________.

【Situation 2】

Suppose that you are in the end stages of life, with only a few weeks to live. Sometimes you may be able to respond to external stimulus, but most of the time you may not respond to it. At this point, your expectations for medical treatment are to:

References

Brown

: The history of advance directives. A literature review. J Gerontol Nurs, 2003; 29:4–14.

Sauler

, Siegel

: The past, present, and future of advance directives as a guide to end-of-life decision making. Chest, 2012; 141:9–10.

Coppolino

, Ackerson

: Do surrogate decision makers provide accurate consent for intensive care research?. Chest, 2001; 119:603–612.

Foo

, Lee

, Soh

: Discrepancies in end-of-life decisions between elderly patients and their named surrogates. Ann Acad Med Singapore, 2012; 41:141–153.

Shalowitz

, Garrett-Mayer

, Wendler

: The accuracy of surrogate decision makers: A systematic review. Arch Intern Med, 2006; 166:493–497.

Sharma

, Hughes

, Nolan

, et al.: Family understanding of seriously-ill patient preferences for family involvement in healthcare decision making. J Gen Intern Med, 2011; 26:881–886.

Seymour

, Almack

, Kennedy

: Implementing advance care planning: A qualitative study of community nurses' views and experiences. BMC Palliat Care, 2010; 9:4.

, Huang

, Hu

, et al.: Experiences and perspectives of older people regarding advance care planning: A meta-synthesis of qualitative studies. Palliat Med, 2017; 31:394–405.

Suhl

, Simons

, Reedy

, et al.: Myth of substituted judgment. Surrogate decision making regarding life support is unreliable. Arch Intern Med, 1994; 154:90–96.

10.

Torke

, Alexander

, Lantos

: Substituted judgment: The limitations of autonomy in surrogate decision making. J Gen Intern Med, 2008; 23:1514–1517.

11.

Khizar

, Harwood

: Making difficult decisions with older patients on medical wards. Clin Med (Lond), 2017; 17:353–356.

12.

Ling

, Yu

, Guo

: Truth-telling, decision-making, and ethics among cancer patients in nursing practice in China. Nurs Ethics, 2019; 26:1000–1008.

13.

Nakagawa

, Cho

, Gondo

, et al.: Subjective well-being in centenarians: A comparison of Japan and the United States. Aging Ment Health, 2017:1–8.

14.

Rothbaum

, Morelli

, Pott

, et al.: Immigrant-Chinese and Euro-American parents' physical closeness with young children: Themes of family relatedness. J Fam Psychol, 2000; 14:334–348.

15.

Bravo

, Trottier

, Arcand

, et al.: Promoting advance care planning among community-based older adults: A randomized controlled trial. Patient Educ Couns, 2016; 99:1785–1795.

16.

Ditto

, Danks

, Smucker

, et al.: Advance directives as acts of communication: A randomized controlled trial. Arch Intern Med, 2001; 161:421–430.

17.

Matheis-Kraft

, Roberto

: Influence of a values discussion on congruence between elderly women and their families on critical health care decisions. J Women Aging, 1997; 9:5–22.

18.

Volandes

, Mitchell

, Gillick

, et al.: Using video images to improve the accuracy of surrogate decision-making: A randomized controlled trial. J Am Med Dir Assoc, 2009; 10:575–580.

19.

Kwak

, Salmon

: Attitudes and preferences of Korean-American older adults and caregivers on end-of-life care. J Am Geriatr Soc, 2007; 55:1867–1872.

20.

Bowman

, Singer

: Chinese seniors' perspectives on end-of-life decisions. Soc Sci Med, 2001; 53:455–464.

21.

Venkatasalu

, Arthur

, Seymour

: Talking about end-of-life care: The perspectives of older South Asians living in East London. J Res Nurs, 2013; 18:394–406.

22.

Carrese

, Mullaney

, Faden

, et al.: Planning for death but not serious future illness: Qualitative study of household elderly patients. BMJ, 2002; 325:125–130.

23.

Glass

, Nahapetyan

: Discussions by elders and adult children about end-of-life preparation and preferences. Prev Chronic Dis, 2008; 5:A08.

24.

Malcomson

, Bisbee

: Perspectives of healthy elders on advance care planning. J Am Acad Nurse Pract, 2009; 21:18–23.

25.

Piers

, van Eechoud

, Van Camp

, et al.: Advance care planning in terminally ill and frail older persons. Patient Educ Couns, 2013; 90:323–329.

26.

Samsi

, Manthorpe

: ‘I live for today’: A qualitative study investigating older people's attitudes to advance planning. Health Soc Care Community, 2011; 19:52–59.

27.

, Berkman

: Advance directives among Korean American older adults: Knowledge, attitudes, and behavior. J Gerontol Soc Work, 2012; 55:484–502.

28.

Winland-Brown

: Death, denial, and defeat: Older patients and advance directives. Adv Pract Nurs Q, 1998; 4:36–40.

29.

Faul

, Erdfelder

, Lang

, et al.: G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods, 2007; 39:175–191.

30.

Lee

, Shin

, Son

, et al.: Factors influencing attitudes toward advance directives in Korean older adults. Arch Gerontol Geriatr, 2018; 74:155–161.

31.

Menon

, Campbell

, Ruskin

, et al.: Depression, hopelessness, and the desire for life-saving treatments among elderly medically ill veterans. Am J Geriatr Psychiatry, 2000; 8:333–342.

32.

Folstein

, Folstein

, McHugh

: “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res, 1975; 12:189–198.

33.

Mahoney

, Barthel

: Functional evaluation: The Barthel index. Md State Med J, 1965; 14:61–65.

34.

Lawton

, Brody

: Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist, 1969; 9:179–186.

35.

Hoyl

, Alessi

, Harker

, et al.: Development and testing of a five-item version of the geriatric depression scale. J Am Geriatr Soc, 1999; 47:873–878.

36.

Coppola

, Bookwala

, Ditto

, et al.: Elderly adults' preferences for life-sustaining treatments: The role of impairment, prognosis, and pain. Death Stud, 1999; 23:617–634.

37.

Suen

, Lee

, Morris

: Life-sustaining treatment: A comparison of the preferences of Taiwanese older adults and their family caregiver. J Nurs Res, 2013; 21:261–269.

38.

, Nelson

, Hanson

, et al.: How surrogate decision-makers for patients with chronic critical illness perceive and carry out their role. Crit Care Med, 2018; 46:699–704.

39.

Marks

, Arkes

: Patient and surrogate disagreement in end-of-life decisions: Can surrogates accurately predict patients' preferences?. Med Decis Making, 2008; 28:524–531.

40.

, Chiu

, Lo

, et al.: Knowledge, attitudes, and behavioral intentions of nurses toward providing artificial nutrition and hydration for terminal cancer patients in Taiwan. Cancer Nurs, 2008; 31:67–76.

41.

Zurakowski

, Taylor

, Bradway

: Effective teaching strategies for the older adult with urologic concerns. Urol Nurs, 2006; 26:355–360.