Information, deliberation, and decisional control preferences for participation in medical decision‐making and its influencing factors among Chinese cancer patients

Abstract Background Patient‐centred care should strive to respond to the individual patient's needs and preferences when possible. Yet, preferences of cancer patients for participation in different stages of the medical decision‐making process to increase matching of preferred and actual participation of the patients in decision‐making are not well known. Objective This study aimed to determine the level of information, deliberation and decisional control preferences in decision‐making practices among Chinese cancer patients and to explore their association with the sociodemographic and clinical variables of the patients. Methods A cross‐sectional study was conducted involving a convenience sample of 328 cancer patients from three public hospitals in Guangzhou, China. The Patient Expectation for Participation in Medical Decision‐making Scale (PEPMDS) was used to evaluate information, deliberation and decisional control preferences of the patients. Binary logistic regressions were conducted to determine the association between sociodemographic variables, clinical variables and preferences of the patients. Results Most patients reported a high preference for information (73.2%) and deliberation (73.8%), while a small number (37.2%) reported a high preference for decisional control. Younger patients and patients with higher levels of education were significantly more likely to have a high preference for information, deliberation and decisional control. Patients with low annual family incomes were significantly more likely to have a low preference for decisional control. Conclusion Preferences of patients for involvement in different stages of decision‐making practices could vary. The level of preferences appears to be related to the patient's age, education level, and financial status. Practice Implications Healthcare providers should tailor the participatory approaches of patients considering individual preferences for information, deliberation and decisional control during medical decision‐making. Patient Contribution Patients participated in the survey and filled in the questionnaires.


| INTRODUCTION
Patient involvement in the medical decision-making process has been considered as a core element in patient-centred care. 1 Research has shown benefits of patient involvement in decision-making practices for patients (e.g., increased satisfaction, improved treatment outcomes and health-related quality of life) and benefits related to the clinical encounter (e.g., reduced cost and improved quality of care). 2,3 In the context of cancer, in particular, decision-making is fraught with uncertainty, and some decisions are very difficult because of their preference-sensitive nature, and require taking into consideration the values and preferences of the patients. Therefore, decision-making practices should strive to respond to the preferences of the individual patients for involvement when possible. Yet, previous studies have shown discrepancies between preferred and actual involvement in the treatment decision-making process among cancer patients 4 as it has been shown that the preferred and actual roles of cancer patients in the treatment decision-making practice match in about 35%-65% of the cases. [5][6][7] Most studies found that patients desired more involvement initially than what actually occurred. 4 Also, it has recently been recommended that patients should be convinced to act as autonomous decision-makers 8 and most cancer patients are highly anxious when forced to participate beyond their expectations.
Hopmans and colleagues conclude that we should involve patients more often but also respect their preferences for participation in decision-making practices. 9 Individualizing the approach to decisionmaking will allow us to maximize the respect to the patients in our care. 10 As for decision-making practices, distinguishing preferences of patients in different stages of the medical decision-making process is an essential step to match their preferred and actual levels of involvement in medical decision-making.
Although many studies have been conducted to explore the expectation of patients to participate in decision-making practices, this topic is still underexplored, especially in Mainland China. A narrative review 11 has shown that previous studies had mainly explored the preferred role of patients in treatment decision-making and the Control Preferences Scale (CPS) was the most commonly used instrument.
Overall, the CPS is an easy-to-administer, valid and reliable measure of preferred and actual roles in medical decision-making among cancer patients. 12 It consists of five statements to elicit the role preferences of patients in decisional control over treatment decision-making and classifies the preferred roles of patients as 'active', 'collaborative' or 'passive'. However, the preferred roles are not a reliable indication of the preferred level of involvement of patients since they have different preferences for involvement during different stages of the decisionmaking process. Evidence has shown that not all stages of shared decision-making processes are equally desired by patients. 13 The conceptual framework of shared decision-making by Charles et al. 14 is the widely used model in the literature. 15 It describes the decisionmaking process in three analytic stages: information exchange, deliberation and control over the final decision. Although the Autonomy Preference Index (API) 16 and the Health Opinions Survey (HOS) 17 were validated instruments that can capture fundamental parts of the shared decision-making model by Charles et al., 14 both cannot measure all stages of the treatment decision-making process. The API includes two scales to distinguish between information seeking and decisional control, but has no explicit items about deliberation. The HOS includes two subscales to distinguish between preferences for information and behavioural involvement, but does not measure the decisional control. Besides, the Health Information Wants Questionnaire (HIWQ), developed and validated by Xie et al., 18,19 was also a widely used instrument, including two parallel scales, measuring preferences for seven specific aspects of health information and for participation in the corresponding types of decision-making. The content of HIWQ is relatively comprehensive, but it does not measure the preference for deliberation. To fully understand and distinguish patient preferences in decision-making practices, a validated instrument is essential to measure preferences for involvement in all stages of the decision-making process as conceptualized by Charles et al. 14 Guided by the above-mentioned decision-making model of Charles et al. 14 and on the basis of the items of API and HOS, Xu et al. 20 developed the Patient Expectation for Participation in Medical Decision-making Scale (PEPMDS) in relation to the Chinese cultural context. The Delphi method was used for the evaluation of items to ensure content validity. Its psychometric test was conducted in a population of 400 inpatients including cancer patients. Item discrimination, response analysis, t test, correlation analysis and factor analysis were used to screen items to ensure that the items have good sensitivity, independence and representativeness. 20 Exploratory factor analysis showed that the PEPMDS consists of three dimensions: information, deliberation and decisional control preferences, which was consistent with the theoretical model of Charles et al. 14 Discrepancies between the preferred and the actual level of participation in medical decision-making among cancer patients represent a common clinical phenomenon worldwide. Yet, preferences of cancer patients in terms of involvement in different stages of the decision-making process are not well known, especially in Mainland China. Besides, a review by Say et al. 11 demonstrates that various sociodemographic and clinical variables have been found to be related to the preferred roles of

| Study design and participant
This study was designed as a descriptive cross-sectional study.
Cancer patients were consecutively recruited from three public/ teaching tertiary hospitals in Guangzhou, China, from May to July 2020. The eligibility criteria were as follows: (1) age ≥ 18 years, (2) diagnosis of either colorectal, breast or lung cancer, (3) able to read and understand the questionnaire, (4) no cognitive impairment or brain metastasis and (5) willing to participate in the study.

| Measures
Data on information, deliberation and decisional control preferences of the patients as well as the potentially influencing factors related to sociodemographic and clinical characteristics were obtained.

| Sociodemographic and clinical variables
Sociodemographic variables, such as age, gender, marital status, educational level, residence and annual family income, were obtained from the respondents using a self-prepared questionnaire and the information on illness-related variables, such as cancer type, cancer stage and duration of cancer, was collected from the medical records of the patients.

| Patient preferences for involvement in the decision-making process
The Patient Expectation for Participation in Medical Decision-making Scale (PEPMDS) is a 12-item self-reported scale that covers information, deliberation and decisional control preferences of patients. Each item is rated on a 5-point Likert-type scale: 1 = completely disagree, 2 = disagree, 3 = neutral, 4 = agree, and 5 = completely agree.
Nine items are worded positively and the other three items are worded negatively, which were reverse-scored so that all items indicated preferences for more participation in decision-making practices. The dimension of information preference consists of three items, related to the needs of illness information. The dimension of deliberation consists of six items, related to the needs of sharing treatment information, and expressing and discussing treatment preferences. The dimension of decisional control preference consists of three items, related to the needs of personal control over the decision. The total score of the above three dimensions ranges from 3 to 15, 6 to 30 and 3 to 15, respectively. The average score of the dimension item is more than 3, representing a high level of preference. For dimensions of information and decisional control preferences, a total score ≤9 is considered to indicate low information preference and low decisional control preference, while a total score >9 is considered to indicate high information preference and high decisional control preference. For the dimension of deliberation preference, a total score ≤18 is considered to indicate low deliberation preference, while a total score >18 is considered to indicate high deliberation preference. PEPMDS has good content and construct validity among Chinese patients. 20 Before our formal investigation, we conducted a pilot study among 30 cancer patients by convenience sampling to test the internal consistency, and good reliability was achieved (Cronbach's α = .89).

| Statistical analysis
After data were double-checked during the data input phase, statistical analyses were performed using SPSS for Windows version 23.0, with the level of significance set two-sided at p < .05. Descriptive statistics were used to describe sociodemographic and clinical variables of the respondents and their preferences for involvement in decision-making practices. The information, deliberation and decisional control preferences were categorized into two separate XIAO ET AL.
| 1727 groups, respectively, and we first performed univariate analyses to test for an individual predictor with a dependent variable. Consequently, the potential predictors were included in the multivariate analysis if the criterion (p ≤ .10) was fulfilled. The binary forced-entry logistic regressions were used to determine the factors predicting information, deliberation and decisional control preferences. The sample size of our study was considered adequate for a logistic regression model of 10-15 times per covariate in binary outcomes according to the rule of thumb. 23

| Preferences for involvement in medical decision-making
The overwhelming majority of respondents reported high preferences for information (73.2%, n = 240, total score >9) and In our study, almost three-quarters of the cancer patients expressed high preferences for information and deliberation, and nearly two-thirds reported a low preference for decisional control. This was consistent with the findings of a qualitative study, 25 which showed that most patients expressed a desire to participate in decisionmaking with their doctors but also desired that the doctors make the treatment decision. In general, cancer patients require detailed information about their cancer and its treatment, and they desire information exchange. 26,27 Meeting information needs of the patients can yield several benefits, including increased patient satisfaction in decision-making, 28 reduced mood disturbances 29 and better psychological well-being. 30 On the one hand, the medical staff should provide patients with their preferred information and provide more opportunities for patients to participate in the discussion of treatment options, so as to help them make the best treatment decision.
On the other hand, we should respect the wishes of those who prefer their doctors to retain decisional control.
Our findings suggest that the majority of patients wanted as much information as possible, good or bad, and patient preference for information varies according to their age and education level.
Younger patients were significantly more likely to express a stronger desire for information and patients with a higher level of education were more likely to have a high preference for information, which was consistent with the findings obtained from non-Western ethnic minority cancer patients, as reported in a systematic review. 31 34 although other studies found a relationship. 33,39 As far as we are aware, our study preliminarily explored patient preferences for involvement in different stages of medical decisionmaking and provided new insights into decision-making practices from a Chinese perspective. However, some limitations should be kept in mind when interpreting our results. First, this research was carried out in a developed city, which could limit the generalization of our findings. Second, we could not rule out potential selection bias because of convenience sampling. Third, most data were self-reported, with a risk of response bias. Fourth, the PEPMDS is a Chinese version scale, developed in the culture context of China, and its use limits the comparison of the findings with those of previous research.

| Conclusion
Our results confirm that not everyone desires to be involved in the same way during treatment decision-making practices and even the same patient desires to be involved at different levels during different stages of decision-making. Most patients desired a high preference for information and/or discussion of treatment options with their doctors, but a low preference for personal control over the decisions.
The sociodemographic characteristics like age, education level and economic status could be used to predict preferences of patients for participation in medical decision-making practices. The clinical factors play a relatively small role in predicting patient preferences for all stages of the decision-making process, which should be explored further.

| Practice implications
As a practical implication of our results, healthcare providers should tailor the involvement approaches of patients by considering their preferences for information, deliberation and T A B L E 3 (Continued) .914

Note:
Bold p values are statistically significant.
Abbreviations: 95% CI, 95% confidence interval; OR, odds ratio. decisional control, to decrease the mismatch between the preferred and actual levels of involvement of the patients and improve patient satisfaction with decision-making practices.
Existing strategies focused on increasing the quality of treatment decisions for cancer patients need to deal with the preferences of patients for involvement, particularly for specific cancer populations, such as younger patients, those with higher levels of education and those with better economic conditions. To strive for the development of patient-centred care, we should assess the preferences of individual patients and respond to their needs and preferences when possible.