In recent years, recovery has been a key component in psychiatric rehabilitation. Self-determination theory posits that people are motivated by a need to grow and gain fulfillment (
1,
2), and self-determination is viewed as an important ingredient of successful recovery. According to the 2003 report of the President’s New Freedom Commission on Mental Health, self-determination is important and of high value because it offers people with psychiatric disabilities realistic and meaningful choices that are considered an essential component of recovery. Similarly, research has shown that self-determination enhances patient motivation and treatment adherence and improves rehabilitation outcomes and quality of life (
2–
5).
Self-determination theory (
1,
2) proposes that all humans have three basic needs: autonomy (perceiving oneself as the volitional source of one’s actions), competence (perceived effectiveness in interactions with others), and relatedness (a sense of connection to and belonging with others). Cultures, however, differ in the amount to which they support the satisfaction of individuals’ basic needs. For example, autonomy is viewed as a leading principle of Western medicine. The autonomy need, however, is typically less valued in Asian societies, as evidenced by lower autonomy need satisfaction scores in those cultures (
6). Thus there has been concern that the concept of autonomy, a key component of self-determination, may be of less importance in Chinese cultures, whereby the individual’s autonomous choices may be in conflict with the expectations of the group or the family (
7).
In view of different autonomy needs across cultures, it is imperative to examine how Chinese consumers with psychiatric disabilities view autonomy in decision making. Most research on self-determination has been conducted with people who have intellectual disabilities (
8,
9), and little has been done with people who have psychiatric disabilities, not to mention from a cross-cultural perspective. Thus this study aimed to fill the void in the literature. Specifically, the purpose of the study was to develop an instrument, the Consumer and Family Decision Making Scale (CFDMS), to examine the opinions of consumers and their respective family members about autonomous decision making in various daily matters.
Methods
Sample
The sample consisted of 364 participants, 182 native Chinese consumers of psychiatric services and 182 of their family members, recruited from a local psychiatric hospital in Chengdu, China, from August 1, 2012, to July 31, 2014. Eligible participants had a diagnosis of schizophrenia on the basis of ICD-10 diagnostic criteria. People with organic brain dysfunction, psychosis induced by substance abuse, developmental disabilities, or severely impaired cognitive functions that would prohibit the completion of the questionnaire were excluded. The study received approval from the West China Hospital Institutional Review Board, and written informed consent was obtained from consumers and their family members.
Measures
The CFDMS was constructed specifically for the study because no published instruments were available. The development of the CFDMS involved several steps. First, a 90-minute focus group with consumers and family members generated 22 items. Second, a literature review generated ten additional items. Third, to establish the content validity, four of the five authors reviewed and rated the relevance of the items using a 5-point Likert scale (from 1, least relevant, to 5, most relevant). Items with average ratings <3 and raters’ intraclass correlation coefficients (ICCs) <.08 (
10,
11) were deleted. The mean±SD rating of the acceptable items was 4.1±.06, and the ICC was .86. On the basis of the cutoff criteria, five items were deleted mainly because of their cultural irrelevancy (for example, whether to obtain a driver’s license, whether to live in a group home or a half-way house, whether to register to vote, whether to purchase a car, and whether to practice birth control or safe sex). Thus the CFDMS consisted of 27 items. A 10-point Likert rating scale was used to assess CFDMS: 1, I [my son/daughter] should make no decision; 5, I [my son/daughter] should make this decision half of the time; and 10, I [my son/daughter] should make this decision all the time. In Chinese cultures, 10 is widely accepted as an anchor point for estimation, and thus it was adopted for the rating scale (scores range from 27 to 270). Therefore, a higher score represents a higher degree of autonomy in decision making. Fourth, to establish the test-retest reliability, two weeks after the initial assessment, 25 randomly selected consumers and 25 of their family members received another round of CFDMS assessment. The final step was an exploratory factor analysis (EFA), which examined the factorial structure and developed its psychometric properties.
PANSS.
A Chinese version of the Positive and Negative Syndrome Scale (PANSS) was used to measure psychiatric symptom severity (
12). It is a structured clinical interview consisting of 30 items designed to assess severity of symptoms over the past week on a 7-point scale (1, absent, to 7, extreme); scores range from 30 to 210, with higher scores indicating greater symptom severity. PANSS subscales were used to measure negative symptoms and positive symptoms and general psychopathology (that is, depression, anxiety, guilt, and somatic concerns). The Chinese version of the PANSS (
12) shows acceptable Cronbach’s alphas of .73 on the positive scale, .83 on the negative scale, and .87 on the general psychopathology scale. The test-retest reliabilities for the positive, negative, and general pathology scales were .80, .68, and .60, respectively.
PSP.
A Chinese version of the Personal and Social Performance Scale (PSP-CHN) was used to assess participants’ social functioning (
13). The scale assesses four main areas of social functioning: socially useful activities, personal and social relationships, self-care, and disturbing and aggressive behaviors. Difficulty in each area is rated on a 6-point scale (absent, mild, manifest, marked, severe, and very severe), with lower ratings indicating better social functioning. A global item ranging from 1 to 100 in 10-point intervals is rated by the interviewer; lower scores indicate poorer functioning. Si and colleagues (
13) reported the psychometric properties of the PSP-CHN (Cronbach’s α=.84; interrater reliability κ=.82; ICC=.94 for the PSP-CHN total score). The test-retest reliability was .95.
FBS.
A Chinese version of the Family Burden Scale of Disease (FBS) was used to assess family burden (
14). The FBS scale has 24 items spread across six factors: economic burden, impact on daily activities, impact on social life, impact on free time, impact on physical health, and impact on mental health. Ratings of 24 items are made on a scale of 0 to 2 (scores range from 0 to 48), with the higher scores indicating more burdens. According to Zhang (
14), for the Chinese FBS, the Cronbach’s alpha was .87 and split-half reliability was .94.
Data Collection
The CFDMS has two versions with identical items, except the instructions were phrased to suit the specific group (either “I should” or “My son/daughter should”). Administration of the CFDMS was conducted on a one-to-one basis, separately with the consumer and their family member. The administration time ranged from 45 to 60 minutes for the consumers and from 30–45 minutes for the family members. Demographic and personal information was extracted from medical records, and other data collection was conducted during outpatient visits.
Data Analysis
For the EFA, several criteria were used to determine the number of factors retained, namely Eigenvalue >1.0 (Kaiser-Guttman rule), scree plot, total variance explained, and meaningfulness (that is, variables loading on the same factor make sense together). The number of factors with Eigenvalues >1.0 extracted for the sample was five. A scree plot (
15) of the Eigenvalues, however, suggested a range of three to five major factors. It was deemed desirable to review a range of solutions to ensure that no meaningful solution was overlooked. Accordingly, three-, four-, and five-factor solutions were rotated by Kaiser’s (
16) varimax method and then evaluated in terms of simple structure, parsimony, and clinical meaningfulness. A separate EFA was conducted for the consumer and family member samples, and sensitivity analyses were conducted to examine the factorial similarity of the two samples. Student’s t test and Pearson’s r were used to examine the consumers’ and family members’ opinions on the CFDMS. Correlation analyses were performed between other measures (PANSS, PSP, and FBS) and the CFDMS to explore their possible relationships.
Results
Table 1 presents data on the characteristics of the sample. The sample consisted of younger consumers (average age 29); the sample was younger than the samples in other similar studies (
17,
18). On average, the consumers had experienced two psychiatric episodes and two previous hospitalizations. Most consumers had a high school or college education, and they were mostly unemployed or working part-time during the study. Accordingly to the test manuals, scores on psychiatric symptoms (PANSS) and social functions (PSP) indicated that the consumers had satisfactory symptom control and social functioning.
The sample of family members was composed mainly of parents. Most had a high school or college education. Twenty-five percent were retirees, and the rest were working in various occupations. The mean FBS score of 17.24 indicated a moderate degree of family burden (
14).
On the basis of the aforementioned criteria, the EFA for the CFDMS identified four factors that accounted for 57.39% of the variance: factor 1, psychiatric care and treatment (nine items accounted for 17.16% of the variance); factor 2, personal and social function (seven items accounted for 16.06% of the variance); factor 3, community and daily living (seven items accounted for 15.06% of the variance); and factor 4, money management (four items accounted for 9.12% of the variance).
Table 2 shows the factorial structure of the CFDMS. [A table in an
online supplement to this article presents the factorial structure of consumer and family member samples.] Visual examination of the two samples’ factorial structures and Tucker's congruence coefficient test (.94) showed that both samples shared comparable factorial structures.
To establish the internal consistency of the CFDMS, Cronbach’s alphas were obtained: factor 1, .86; factor 2, .89; factor 3, .87; and factor 4, .76. Overall, the CFDMS possesses sufficient internal consistency. The test-retest reliabilities of the CFDMS were as follows: total score r=.87; factor 1, r=.81; factor 2, r=.89; factor 3, r=.80; and factor 4, r=.88. With an established content validity, acceptable test-retest reliability, meaningful factorial structure, and good internal consistency, the CFDMS appears to be an acceptable instrument to measure consumers’ and family members’ views of daily decision making.
The consumers’ CFDMS mean total score was 124.73±45.6, and the mean item score was 4.62±1.69. The family members’ CFDMS mean total score was 147.42±43.52, and the mean item score was 5.64±1.61. Consumers and family members had different views about decision making in the four areas. As shown on
Table 3, factor 1 (psychiatric care and treatment) was viewed as an area in which consumers should have the least autonomous decision making, followed by factor 4 (money management). Factor 3 (community and daily living) was viewed as the area in which consumers should have the highest degree of autonomous decision making, followed by factor 2 (personal and social function). Further analyses of the correlation between consumers and family members on CFDMS indicated that the groups had similar viewpoints on decision making. In addition, family members expected consumers to have more say in decision making than did the consumers themselves (t=6.17, df=179, p<.001) (
Table 3).
Table 4 shows the relationships between PANSS, PSP, FBS, and CFDMS. For the consumers, significant correlations were found between the CFDMS total score, personal and social function, and the community and daily living and the PANSS and PSP. Psychiatric care and treatment was not correlated with any measures. No significant correlations were found between FBS and CFDMS.
For the family members, PANSS was negatively correlated with psychiatric care and treatment and with personal and social function, and PSP was positively correlated with the CFDMS total score, personal and social function, and money management. No significant correlations were found between FBS and CFDMS.
Discussion
The results, while preliminary, shed light on Chinese consumers’ day-to-day life decision making. The CFDMS was found to possess good psychometric properties and appears to be a reliable and valid instrument for assessment of consumers’ views of decision making. Although there is no basis for comparison, the results seem to indicate that the consumers had a low tendency for making autonomous decisions. This unwillingness was more marked regarding decisions about psychiatric care and treatment.
Several factors may account for consumers’ low autonomous decision making. First, individuals with psychiatric disabilities may have less autonomy needs than individuals without psychiatric disabilities. In recent studies of self-determination among people with schizophrenia, Breitborde and colleagues (
19) and Gard and colleagues (
20) reported that individuals with schizophrenia showed less autonomy needs than general populations. Some authors have suggested that deficits in autonomy, competence, and relatedness may already be present early in the course of a psychiatric illness (
19). A second factor, as reported by Chambers and colleagues (
21) is that people with disabilities have lower self-efficacy regarding decision making, tend to perceive themselves as less capable than they actually are, and are afraid of making poor decisions. Third, as mentioned by Sheldon (
6), Chinese culture values autonomy less than most Western cultures do. Finally, Bao and Lam (
22) found that it is possible for Chinese persons to feel autonomous when they follow a choice made by others as long as they concur fully with and endorse this choice. In that sense, if consumers have internalized the choices made by trusted others (for example, family members), they might experience autonomy even if they do not actually make the decision.
Both consumers and family members in this study believed that consumers should not engage in decision making related to psychiatric care and treatment. Deference to authority is a common Chinese decision-making style, in which mental health professionals are expected to make decisions in the best interest of the greatest number of people involved with the consumer (
23). In addition, the professional can serve as a buffer so that the consumer’s family members do not have to be blamed for making poor decisions.
Chinese mental health laws to some degree may contribute to consumers’ low autonomy in psychiatric treatment. For many years, the mental health laws have taken away consumers’ rights to decide on treatment and care regarding their psychiatric illness (
24). The laws assume that people with psychiatric illness are “incompetent” to make any treatment decisions, and the right to make decisions about psychiatric hospitalization and discharge from a hospital was granted only to family members and health care professionals. Thus consumers may learn to be submissive and yield medical decision making to health care professionals and families.
Although both consumers and family members had similar views of decision making, family members expressed higher expectations for consumers to make autonomous decisions, especially on daily living, personal, and social matters. As in the study by Balaji and colleagues (
25), family members expected consumers to take charge in these areas and viewed social functioning, fulfillment of personal responsibilities, and independent daily living as important ingredients in recovery. Consumers’ psychiatric symptoms and personal and social functioning were related to their participation in decision making. It stands to reason that consumers who show less symptomology and higher social and personal functioning would be more willing to make and capable of making autonomous decisions. The finding that decisions about psychiatric care and treatment were not related to consumers’ psychiatric and social functioning suggest that these decisions were made mainly on the basis of a familial-cultural belief (
23) rather than on a functional basis. The finding that the FBS was not correlated with the CFDMS may suggest that these two constructs were independent of each other.
Implications for Cross-Cultural Clinical Practice
Although our results did not point to the source of decision making about psychiatric treatment, Chinese family members typically assume that it is the responsibility of the psychiatrist and family to make such decisions (
24). Thus Chinese families have a strong influence on deciding the course of treatment. Effective communication and service engagement between mental health practitioners and Chinese consumers rely on an understanding of Chinese cultural traditions and on the aforementioned decision-making styles. Mental health providers must respect this Chinese decision-making style and remain nonjudgmental when dealing with situations or decisions that may be contradictory to their own culture and values. Until there is more research evidence and understanding about self-determination among Chinese consumers, pushing for autonomous decision making in psychiatric treatment may be questionable.
Limitations and Implications for Future Studies
In this first attempt to examine autonomous decision making among Chinese consumers with psychiatric disabilities, several limitations should be noted. First, our sample consisted primarily of younger consumers who were living with their families. It is very common for Chinese consumers to live with their families; it is culturally acceptable even for individuals without these disabilities to do so (
26). Thus our sample is representative of the population studied, and the findings may not generalize to other populations, especially to consumers who are older and living independently. Future studies should employ a larger sample that includes consumers from multiple sites and cities and consumers who are older and from other living environments. To further understand the concepts of family decision making and self-determination, other samples could be considered, such as a control group with no psychiatric disabilities and individuals with other medical conditions or chronic illnesses. In addition, a comparable sample from a Western society would aid cross-cultural comparisons of self-determination.
Although the CFDMS was found to possess good psychometric properties and appears to be a reliable and valid instrument, the CFDMS items are by no means exhaustive in representing actual day-to-day decision making. It is quite possible that other areas of decision making are not included in the CFDMS. Future participatory research in this area should include more inputs from consumers and family members. In addition, the CFDMS measures only opinions about autonomous decision making, not actual behaviors.