The informal discussion of advance care planning among Chinese older adults: Do education and social media use matter?

Abstract

This study aimed to examine and compare the associations between education, social media use, and advance care planning (ACP) discussion among Chinese older adults in mainland China (Wuhan), Taiwan (Taichung), and the United States (Honolulu). Community-dwelling older adults (≥ 55) were recruited from 2017 to 2018. The ACP discussion rate in Wuhan, Taichung, and Honolulu were 15.2%, 19.2%, and 31.3%, respectively. Logistic regression models revealed that education was positively associated with ACP discussion in Taichung and Honolulu. Social media use was positively associated with ACP discussions in Wuhan and Honolulu, and it attenuated the association between education and ACP discussion in Honolulu. The present study contributes to previous studies by comparing the associations between education, social media use, and ACP discussion in different settings within the same ethnicity. Policy and practice implications were also discussed.

Introduction

Advance care planning (ACP) is a process of understanding and discussing an individual’s values, goals, and preferences regarding future medical care, including end-of-life (EOL) care.¹ ACP is associated with improved EOL care and high quality of death.² However, ACP rates are low among Chinese older adults. In North America, compared with the general population, Chinese adults experience more intensive care interventions at the end of life (e.g., lower rate of hospice use and higher rate of receiving intensive treatments).³ Known barriers to ACP among Chinese older adults include individual (e.g., knowledge and attitudes toward ACP), health system level (e.g., lacking clinical training), and cultural (e.g., filial piety and family decision-making) factors. Considering the increasing importance and attention given to ACP due to the challenge of the population aging, how to promote ACP among Chinese older adults started to catch a lot of attention from scholars and policymakers.^2–4

ACP is a socially mediated health behavior, ^5,6 as it predicts good death and dying for patients and lowers stress for family members during and after the dying process.^7,8 In this case, Andersen’s behavioral health model can provide a useful theoretical framework to explain factors that are associated with ACP engagement among Chinese older adults.^8–10 The model postulates that need factors (i.e., psychological distress and chronic diseases), enabling or impeding factors (i.e., education, social media use, and financial strain), and predisposing factors (e.g., age, gender, and marital status) largely influence individuals’ health care seeking behaviors. ¹¹ The current investigation aimed to examine how education and social media use are associated with ACP engagement among Chinese older adults residing in different settings.

As with other socially mediated health behaviors (e.g., drinking, smoking, and physical activity), educational attainment and social media use are likely to be associated with ACP. Educational attainment, the key indicator of socioeconomic status when studying health disparities,¹² is associated with a wide range of health behaviors and social networks, especially among older adults. ^13,14 In some studies, educational attainment is also found to be associated with ACP.^12,15 Meanwhile, social media is often viewed as an effective tool to promote health behaviors and reduce health disparities.¹⁶ In addition, a growing body of literature suggested that social media use is positively associated with social interaction and network building, ¹⁷ which, in turn, are associated with ACP discussion.¹⁸ Therefore, social media usage may provide older adults with a platform to probe the ACP-related knowledge. Notably, substantial evidence points to the critical role of geographical and cultural contexts in modulating the impact of education and social media on health behaviors and outcomes.^19–21 Thus, this effect could potentially extend to ACP as well. While prior research largely analyzed the influence of education and social media usage on ACP within the context of a single country, we proposed a comparative study across different social settings to provide valuable insights.

Specifically speaking, the selection of Wuhan in mainland China, Taichung in Taiwan, and Honolulu in the United States offers a rich and comparative backdrop for our research inquiry. First, the rapid population aging in these places makes them face similar challenges in providing appropriate EOL care to older adults.^22,23 Second, people of Chinese ancestry are likely to be influenced by the traditional Chinese culture such as filial piety and family decision-making.²⁴ Third, the existence of varying degrees of palliative care development across these places further deepens their comparative appeal. According to a recent study on levels of palliative care provisions in 198 countries, China is still at the preliminary stage of integration, whereas Taiwan and the U.S. are already at the advanced stage of integration.²⁵ For example, in 1990, the United States passed the Patient Self-Determination ACT, emphasizing the significance of patient autonomy in health care decision-making. ²⁶ In 2019, Taiwan enforced the Patient Autonomy Act to promote ACP.²⁷ In China, however, such law does not exist. Consequently, these three cities provide an appropriate combination of settings to contextualize empirical findings on the association between education, social media use, and ACP engagement. This study, therefore, ventures to bridge a conspicuous gap in existing research. It aimed to underscore the critical role of social contexts - specifically the disparate stages of palliative care development - in shaping the relationship between educational attainment, social media usage, and ACP engagement.

ACP includes discussions of EOL wishes and documentation of these wishes such as a living will and advance directives.²⁸ Regarding guiding care, discussions with family members can be as vital as consultations with healthcare providers and the completion of advance directives.^29,30 In the context of Chinese culture, there is a pronounced preference for familial decision-making processes and the active participation of family members in making critical medical decisions.³¹ Meaningful discussions between older adults and their families are critical to ensuring that family members understand older adults’ wishes.⁶ This study aimed to examine and compare the associations among education, social media use, and ACP discussion in three cities located in mainland China, Taiwan, and the U.S.. By undertaking this comparative analysis, we aimed to generate insights that could contribute to the enhancement of ACP engagement among Chinese communities residing in varied sociocultural contexts.

In sum, the primary research question of this study was to examine whether education and social media use would be associated with ACP discussion and whether the relationships would hold true among Chinese older adults in Wuhan, Taichung, and Honolulu. Guided by Andersen’s behavioral health model, we hypothesized that both education and social media use would be positively associated with ACP discussion in these three regions. Finally, given that social media has been viewed as a platform that could promote healthy behaviors, we also explored whether social media use would weaken the association between education and ACP discussion.

Methods

Settings

Using survey data collected in Wuhan, Taichung, and Honolulu, this study was designed to compare how social settings shape health behaviors and health outcomes in three different settings. The study was approved by the institutional review board of University of Hawaiʻi at Mānoa and Tunghai University.

Participants

In Wuhan, data were primarily collected from the Jianghan district, which has a slightly higher percentage of aging population (60+) compared to the whole city (25% vs 22%). ³² Survey participants were residents aged 55 years and older. Convenient sampling was used to recruit participants. In the first step, researchers connected with key gatekeepers from 11 sub-districts (out of a total of 13 sub-districts) in the Jianghan district through local contacts. In the second step, after explaining the purpose of the research project and their expected responsibilities, these key gatekeepers distributed questionnaires to multiple communities that they administrated (e.g., each sub district includes 8–10 communities). Finally, the administrative staff of each selected community recruited participants from the community that they administrated. Using convenient sampling, communities within each sub-district and participants within each community were selected. From June to August 2017, a total of 850 questionnaires were collected, including 705 adults aged 55 and above. After deleting the missing values in the dependent variable (n = 45) and other analytical variables, our analytic sample included 534 respondents.

Taiwan data were mainly collected from Taichung, the second-largest city in Taiwan, which has an equivalent age structure to Taiwan as a whole, and provides representative respondents from diverse backgrounds.³³ Based on the official household registration provided by the Taichung City Government, the sampling frame targeted population aged between 56 and 85. A multistage stratified random sampling strategy (e,g., cities/countries, districts, administrative neighborhoods, individuals) and the “probability proportional to their size” method were used to recruit participants. Face-to-face interview was used to collect information from the recruited respondents. In 2017, a total of 645 questionnaires were collected. After listwise deletion of missing values in the dependent variable (n = 37) and other variables, 605 respondents were included in the analysis.

Hawai’i data were primarily collected in Honolulu, where about 4.7% of the whole population is Chinese American, and 44% of the adult population are immigrants.³⁴ We used snowball and convenience sampling to recruit participants. Specifically, key informants were identified and recruited from local older adult groups, social organization, business, and faith-based agencies. Then relying on the connections of the selected key informants, researchers were able to connect to a variety of local Chinese communities in Honolulu. This is a common and effective strategy to recruit minority populations.³⁵ The inclusion criteria included Honolulu residents, aged 55 years and older, who self-identified as Chinese. From January 2018 to September 2018, a total of 430 participants were recruited. Given that US-born Chinese in general are inclined to accept the values of autonomy due to their acculturation to the American mainstream culture, we only included first-generation immigrants (n = 311). After deleting missing values in the dependent variable (n = 42) and other variables (n = 20), the final analytical sample included 249 participants.

Measures

Dependent variable- advance care planning

Our focal dependent variable was ACP discussions, a crucial component of ACP. Respondents were asked if they had ever discussed their EOL care plan with their family. We coded this variable as 0 (no) and 1 (yes).

Independent variables

We focused on two focal independent variables—education and social media use. Education was coded as 0 (lower than high school) and 1 (high school and above). Social media use was measured by asking participants how long they had interacted with their friends via social media. Because social media use was nonnormally distributed with high percentages of respondents reporting no interactions with friends via social media, it was dichotomized to contrast those using social media (1) with not using social media to interact with friends (0).

Control variables

We controlled for socio-demographics such as age (in years), gender (1 = female), marital status (1 = currently married), and financial strain (1 = a great deal/some). We also controlled for psychological distress and chronic conditions (1 = having at least one chronic condition). Psychological distress was measured by the Kessler Psychological Distress Scale.³⁶ Participants were asked whether they experienced symptoms of depression and anxiety (i.e., felt nervous, hopeless, restless, that everything takes effort, sad, and worthless) in the previous 30 days. We coded the frequency with which the participant had experienced each symptom as 1 = none of the time, 2 = a little of the time, 3 = some of the time, 4 = most of the time, and 5 = all of the time. We computed the mean value across these 6 items, which resulted in a psychological distress index ranging from 1 to 5, with a higher score indicating high levels of psychological distress. The reliability coefficient for the six items was above .85 in all three cities.

Analysis

In Table 1, we summarized descriptive statistics for study variables by region. We used multivariate logistic regressions to examine whether education and social media use are associated with ACP discussions in Chinese older adults for each city. We first examined the main effects, and then we added the interaction between education and social media use. Significance level was set at p < 0.05, and the tests were two-sided. The statistical analysis was performed in STATA version 15.0 (Stata Corp, College Station, Texas, USA)

Table 1:

Sample characteristics for Chinese older adults in Wuhan, Taiwan, and Honolulu.

Variables		%/M (SD)
	Wuhan (N=534)	Taiwan (N = 605)	Honolulu (N = 249)
End-of-life discussion
Have discussed end-of-life care plan with family	15.17%	19.17%	31.33%
Age	67.85 (6.90)	66.30 (7.42)	72.69 (9.28)
Gender
Female	60.86%	52.72%	57.83%
Marital Status
Married	83.71%	73.39%	73.09%
Psychological distress	1.98 (0.78)	1.72 (0.75)	1.77 (0.76)
Having at least one chronic disease	72.47%	71.74%	68.67%
Having some financial strain	20.60%	36.69%	26.91%
Education
≥High school	49.81%	36.53%	40.56%
Social media use
Yes	63.86%	48.10%	47.39%

Results

We showed sample characteristics in Table 1. Overall, the rates of ACP discussion in all three cities were lower than 33%, suggesting that Chinese older adults in our analytic samples had similarly low rates of ACP discussion rate. Importantly, the ACP discussion rate was 15.2% for Wuhan while the corresponding rates were 19.2% for Taichung and 31.3% for Honolulu.

Less than 50% of older adults in the sample had high school education or higher. The rates of high school education or higher were highest in Wuhan (49.8%), and lowest in Taichung (36.5%) with Honolulu falling in between (40.6%). Also noteworthy was that about 50% of older adults used social media to chat with friends. In Wuhan, 63.9% of older adults used social media, while the corresponding figures were 48.1% for Taichung and 47.4% for Honolulu.

The regression results in Table 2 showed the associations between education, social media use, and ACP discussion for each city. In the Wuhan sample, social media use was positively related to ACP discussion (OR = 2.03, p<.05). In the Taichung sample, having high school or higher level of education was positively related to ACP discussion (OR = 2.38, p < 0.01). In the Honolulu sample, both social media use (OR = 1.96, p<0.05) and having high school or higher levels of education (OR = 4.13, p<0.01) were positively associated with ACP discussion. The interaction between education and social media use was significant (OR= 0.22, p<0.05) in Honolulu, indicating that social media use weakened the association between education and ACP discussion.

Table 2:

Logistic regressions examining the association between education, social media use, and ACP discussion.

	Wuhan (N = 534)		Taiwan (N = 605)		Honolulu (N = 249)
	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2
Variables
Age	1.02	1.02	0.98	0.98	1.03+	1.03
Female (ref = male)	1.17	1.18	1.41	1.41	1.29	1.35
Married (ref = unmarried)	0.71	0.70	0.65+	0.65+	0.79	0.71
Psychological distress	1.33 +	1.35+	1.58**	1.58**	1.52*	1.63*
Having at least one chronic disease (ref = 0)	1.19	1.20	1.04	1.04	1.75+	1.58
Financial strain (ref = none)	1.41	1.40	0.79	0.79	0.51+	0.45
≥ high school (ref = <high school)	1.13	1.69	2.38**	2.56**	1.84*	4.13**
Social media use (ref = no)	2.03*	2.56*	1.44	1.51	1.96*	3.87**
Interactions
≥ high school*		0.58		0.89		0.22*
*social media
Constant	0.02*	0.02*	0.23	0.22	0.01**	0.01**

Note:

⁺p < 0.1,

^*p < 0.05,

^**p < 0.01,

^***p< 0.001

In the Wuhan and Honolulu questionnaire, there were four categories for the EOL care discussion: “never and don’t want to discuss”, “never but want to discuss”, “reluctant to discuss”, and “have willingly discussed”. In the sensitive analysis, we used multinomial logistic regressions to compare the associations between education, social media use, and EOL care discussion in Wuhan and Honolulu. Similar to those in Table 2, we also found that social media use was positively associated with a higher likelihood of having an ACP discussion in Wuhan, whereas both education and social media use were positively associated with a higher likelihood of having an ACP discussion in Honolulu (see Appendix Table 1).

Discussion

This study revealed that, social media use was associated with ACP discussion in Wuhan, and education was associated with ACP discussion in Taichung. While education and social media use were associated with ACP discussion, and social media use attenuated the positive association between education and ACP discussion in Honolulu. Evidence from these three settings added much-needed nuances on how social context shape the association between education, social media use, and ACP discussion among Chinese older adults. This may pave the way for the development of culture-tailored ACP interventions among the Chinese population in different social contexts.

Education was found to be associated with ACP discussion among Chinese older adults in Taichung and Honolulu. These findings are consistent with previous studies conducted among Chinese and other populations, wherein individuals with a higher level of education were observed to be more inclined towards engaging in ACP discussions.^2,13,14,37 This correlation could be attributed to several factors. Firstly, older adults with higher educational attainment are likely to have adult children who are similarly highly educated.³⁸ This, in turn, could influence these older adults to engage in ACP discussions. Secondly, it is plausible that the benefits of ACP engagement initially become more evident to a more educated population, who consequently appear to be the early adopters of such practices.²⁰ Conversely, education was not associated with ACP discussion in Wuhan. The finding is similar to a previous study conducted in the same region, where educational levels did not seem to impact the ACP engagement, largely because over 95% of the older adult population had no prior awareness of ACP.³⁹

Given that the levels of palliative care provision in China is still at the preliminary stage of integration,²⁵ it is possible that the information about ACP has not sufficiently permeated even among more educated older adults. This suggests that the association between education and ACP discussion is heavily contingent on the sociocultural contexts within which older adults reside. For instance, in places like China, where palliative care is limited and seldom discussed by medical practitioners, educational disparities in ACP discussions are virtually non-existent.⁴⁰ Future studies are needed to examine how contextual factors, such as policies on palliative care development, could influence the ACP engagement with varying educational levels over time.

Social media use was found to be associated with ACP discussion among Chinese, except for those in Taichung. Previous studies suggested that social media is an effective tool of ACP promotion,^41–43 as the social media platforms are often used to promote information on ACP and discuss the preferences of medical treatment. The nature of social media enables frequent interactions among its users, making older adults have the opportunity to absorb knowledge pertaining to ACP, which is frequently disseminated through these digital platforms. Consequently, this demographic, when engaged in social media interactions, displays a higher likelihood of ACP discussions as compared to their non-user counterparts.

Our findings extended previous research by demonstrating that social media use measured by whether having interactions with friends via social media increased ACP engagement among Chinese and that the use of social media attenuated educational disparities in the U.S.. The finding is consistent with previous studies that social media intervention is more effective in promoting health equity among disadvantaged populations, such as older adults, and those with low levels of socioeconomic status.¹⁶ Despite the U.S. being at an comparatively more advanced stage of palliative care integration, Chinese older adults are often subjected to intensive care interventions towards the end of their life.³ Social media serves as a valuable resource for information dissemination, potentially increasing ACP awareness among Chinese older adults. Therefore, it can be concluded that social media interventions could potentially enhance ACP engagement among Chinese older adults both in China and among the immigrant population in the United States.

Limitations

This study has several limitations. First, it is difficult to establish causal relationships from the cross-sectional design of this study. Especially, ACP became particularly important during the COVID Pandemic.⁴⁴ The association that we found in this study may not capture the associations between education, social media use and ACP discussion after the Covid −19 Pandemic. Longitudinal studies are needed to compare the differences in the association between education, social media use and ACP discussion before the COVID-19 and after the COVID 19 Pandemic. However, due to the difficulties of collecting the data from different contexts, this comparative study provides insights on how to improve ACP discussion among Chinese across different settings. Second, our samples may involve certain levels of selection bias—older adults who completed the survey may be more comfortable to talk about EOL care. Third, except for the Taichung sample, the samples were small and convenience samples derived from geographically restricted regions. Therefore, our findings may be not generalizable to Chinese people in other regions. Fourth, despite the similarities in the culture regarding EOL care among Chinese in three areas, the difference in the degree of Westernization as well as the palliative development may also account for the differences in the associations between education, social media use and ACP discussion. A deeper understanding of these factors needs to be considered to promote ACP discussion among Chinese.

Conclusions and Implications

This study sheds light on how social context such as the levels of palliative care development shape the associations between education, social media use, and ACP discussion among Chinese population in different regions. We found that education is associated with ACP discussion in Taichung and Honolulu. Whereas social media use was found to be associated with ACP discussion in Wuhan and Honolulu, and it can attenuate the association between education and ACP discussion in Honolulu. Education is not always associated with ACP engagement among Chinese older adults. Future studies need to explore the mechanism of the association between education and ACP engagement. Social media ACP interventions have potential to increase the awareness of ACP and reduce educational disparities in settings, such as China, and the United States, where social medias are the main channel for the delivery of ACP information.

Highlights.

• Discussion with families about-end-of-life care is important in Chinese families.

• We considered the settings differences in the association between education, social media use, and ACP discussion.

• Education and social media use are not always associated with ACP discussion.

• Social media use weakened the educational disparities in ACP discussion in Honolulu.