Factor Analysis and Psychometric Properties Adaption of Chinese Version of the Decisional Engagement Scale (DES-10)

Introduction

Prostate cancer (PCa) was the second most common cancer,¹ which caused 6.6% of the total deaths of men, ranking fifth in leading cancer-related deaths in men worldwide.² The incidence of prostate cancer is increasing in most Asian countries especially in people with Chinese ancestry, and the mortality rates have also increased in China.³ Nowadays, alternative treatments are used in prostate cancer treatment,⁴ such as watchful waiting, active surveillance, surgery, chemotherapy, etc. Surgery and chemotherapy could help certain patients survive the disease, while with side effects like impaired urinary function,⁵ sexual dysfunction,⁶ and lower quality of life.⁷ Active surveillance might be a preferable choice for patients with low-risk disease,⁸ and watchful waiting could help patients with intermediate-risk disease to achieve satisfactory quality of life.⁹ However, because of the absence of an optimal treatment strategy regarding prostate cancer, patients have to make critical and highly preference-sensitive decisions involving trade-offs between benefits and potential risks.¹⁰ Shared decision-making (SDM) was recommended as a fundamental process in doctor-patient interaction in treatment decisions.¹¹

SDM was defined as a dynamic process in which patients and health care providers jointly make the best health decisions,¹² based on the aggregation of medical evidence and patient values, and highlights the ethical principle of respecting autonomy.¹³ The health professionals were encouraged to respect consumers’ preferences and share the decision-making responsibility with them in the process of SDM.¹⁴ Nowadays, SDM is being used in maternal-fetal care, stable ischemic heart disease, orthopedics surgery, and prostate cancer treatment.^15,16 SDM could increase disease-related knowledge and help patients make clinical decisions aligned with their self preferences.¹⁷ Related studies had demonstrated that medical information itself could not help patients improve their health, but patient engagement in their health decisions based on medical information would lead to better health care quality and efficiency.¹⁸ SDM could increase patients’ compliance with treatment,¹⁹ improve health care outcome,²⁰ and save on medical costs.²¹ The existing evidence has shown that the more engagement in the decision-making process, the less stress of decision-making for cancer patients.²² A study suggested that the desire to participate in health care decision-making is particularly strong in samples of cancer patients.²³ Another study conducted in a tumor hospital from Guangzhou reported that 71.85% of the investigated patients believed that SDM was essential.²⁴ Since the importance of the health care decision-making process in cancer therapy was well-documented, a tool to measure prostate cancer patients’ engagement in the health care decision-making process is necessary. Multiple measuring scales have been developed through easy-to-approach questionnaires for SDM. However, many of them are too general. The 10-item Decisional Engagement Scale (DES-10) had been confirmed by prostate cancer patients regarding its good reliability and validity, and it was able to measure the degree of health care decision-making for them.²⁵

The 10-item scale is a self-reporting instrument and has a stable and simple structure, excellent reliability and validity. The internal consistency reliability was 0.80 among 376 patients with prostate cancer.²⁵ Currently, SDM-Q-9 was translated and used among the in-patients in China and has been proven to be a reliable and eligible tool to assess the patient’s perspective in the process of SDM in a clinical situation, but the scale is not specifically for prostate cancer patients, and the correlation of test-retest reliability is relatively low.²⁶ Few generally accepted, culturally appropriate scales exist to accurately measure prostate cancer patients’, engagement in shared decision-making process in China, therefore, it is critical to identify such a tool. To the best of our knowledge, this is the first time in literature that the Chinese 10-item Decisional Engagement Scale (DES-10-Chinese) has been adapted among Chinese prostate cancer patients to demonstrate its reliability and validity.

Methods

Study Design

One cross-sectional study was conducted from December 2019 to January 2020 in four hospitals in Henan Province, the central part of People’s Republic of China.

Sample

Four tertiary grade A hospitals were picked by convenience sampling method, which have more than 501 hospital beds, provide specialty medical and health services, and conduct advanced educational and scientific research tasks.²⁷ The total number of participants was determined by the Kendall criterion²⁸ (i.e., 5–10 fold the number of items). Since there are 24 items in total, including demographic information (5 items), 10-item Decisional Engagement Scale (DES-10) and 9-item SDM-Q-9, 120–240 participants were needed. Considering 20% convenience sampling error, 144–288 participants were essential in this study. Finally, 396 newly admitted prostate cancer in-patients were investigated from four of the previously mentioned hospitals, and 380 of them completed the survey with a response rate of 96.0%. After admission to hospitals, these patients may face watchful waiting, active surveillance, surgery, chemotherapy, etc. The study criteria were as follows, inclusion criteria: a) diagnosed with prostate cancer confirmed by clinicopathological examination; b) had known the diagnosis, and no barrier in communicating; c) voluntary participation in research. Exclusion criteria: a) patients with mental disorders, senile dementia, and consciousness disorders; b) with severe heart, brain or lung disease and tumor metastasis. Written informed consent was obtained from all participants before the data collection interview took place.

Measures

The questionnaire included three parts: the demographic information, 10-item Decisional Engagement Scale (DES-10), and the 9-item Shared Decision Making Questionnaire (SDM-Q-9).

The Demographic Information

The general information included demographic variables: age, marital status, educational level, family residence, and household monthly income.

Des-10

DES-10 was a 10-item, single-factor measure of cancer patients' engagement in the health care decision-making process, “which was able to well assess patients’ awareness of their diagnosis, sense of empowerment and involvement, and level of information-seeking and planning”.²⁵ The DES-10 used an 11-point Likert-type scale ranging from 0 (totally disagree) to 10 (totally agree) with a total score of 0~100. A higher total score indicated greater engagement in decision-making in cancer care.

Sdm-Q-9

SDM-Q-9 was a 9-item, single-factor measure.²⁶ A 6-point Likert-type scale ranging from 0 (totally disagree) to 5 (totally agree) with a total score of 0~45. SDM-Q-9 was able to be used to assess the patients’ perspective in the process of SDM in clinical situations. The author suggested converting the original score 0~100 by multiplying 20/9 for easy comparison. The Cronbach’s ɑ of SDM-Q-9 was 0.97 in this study.

Translation Procedure

The DES-10 scale was authorized by Doctor Michael Hoerger,²⁵ and the translation process was composed of several steps based on the Brislin translation model (Brislin, 1970).²⁹ Firstly, the DES-10 was translated from English into Chinese by two bilingual professional translators. A group of bilingual experts, including two psychology experts and two nursing experts, was organized to review the translated version concerning its content accuracy, semantic equivalence, and sentence structure. Secondly, the Chinese version was translated back into English by another two, who had not been exposed to the original scale, and they worked respectively between the forward and back translation process. Then experts including one sociology professor, two psychology experts, and two nursing experts were invited to evaluate each item of the scale one by one from the aspects of articulation, language habits, cultural and linguistic equivalence, to fully ensure the cultural relevance and content equivalence of the Chinese version of DES-10. Then we got the trial DES-10-Chinese, after making some minor revisions. Thirdly, the pilot study was carried out among 20 Chinese prostate cancer patients, the results showed the Cronbach’s ɑ was 0.86, reflecting an acceptable internal consistency, and both interviewees and interviewers were able to understand questions of the scale, and discrepancies were adjusted based on the participants’ feedback on the items. Lastly, we acquired the final Chinese version of DES-10, which was explicit, easy to understand, and had cultural equivalence.

Data Collection

The data were collected from December 2019 to January 2020 in in-patient wards in Henan Province by 10 wound and stoma therapists. Before conducting the one-on-one, face-to-face survey, they were all trained in a preliminary pilot study. 380 participants completed the survey anonymously. The longer the interval for measurement, the lower is the stability coefficient, so the recommended reliable interval is 1–2 weeks.³⁰ After data collection and encoding, 60 of them were randomly sampled after 1 week to test the test-retest reliability of the scale.

Ethics

This study was approved by the Ethical Review Board of Henan Provincial People’s Hospital (2019 Ethics No. 74), Zhengzhou, China, and was in accordance with the Declaration of Helsinki. All participants knew the purpose of the study, and all agreed to participate.

Data Analyses

Data analysis was conducted with SPSS21.0 and AMOS22.0. The internal consistency of the DES-10 was estimated by Cronbach’s α. The test-retest correlation coefficient (Intra-class Correlation Coefficient, ICC) was used to calculate the scale’s stability. The content validity was analyzed by item-total and inter-item correlations, and five experts were also invited to evaluate the content validity index (CVI) of DES-10, each of them marked each item based on the four grades: 4=very relevant; 3=strong correlation; 2=weak correlation; and 1=not related; very relevant and strong correlation were transformed to score of 1, and not related and weak correlation were transformed to score of 0.³¹ 50% of the data was randomly selected for exploratory factor analysis (EFA), and the remaining data were used for validity factor analysis (CFA) to test scale validity. Besides, Pearson’s correlation between DES-10-Chinese and SDM-Q-9 was calculated to estimate the concurrent validity, P<0.05 was defined as statistically significant.

Then, t-test and analysis of variance were used to test the demographic difference (marital status, homeplace, education level, and monthly household income) in DES-10. Pearson’s correlation was used to analyze the relationship between age and DES-10.

Results

The Sample

The age of the 380 participants ranged from 25 to 79, with a mean value of 59.77±11.42. The total scores of DES-10 and SDM-Q-9 were (71.16±17.14), (78.29±20.54), respectively (Table 1). The detailed demographic information was shown in Table 2.

Table 1 Item-to-Total Score Correlations for the DES-10

Table 2 Demographic Difference of DES-10 (n = 380)

Reliability

Cronbach’s ɑ was used to assess the internal consistency. The psychometric testing results showed a one-factor structure and high internal consistency in the final Chinese edition of DES-10 with a total Cronbach’s ɑ of 0.87. The inter-item correlations ranged from 0.18 to 0.70, the item-to-total correlations ranged between 0.40 and 0.82.

After 1 week of investigation, 20 patients who had participated in this study were randomly selected for the retest, and the retest reliability coefficient was 0.98 (95% CI, 0.94–0.99, P< 0.001) (Table 1).

Validity

Content validity

The average CVI in this study was 0.86, five experts provided the values of every item, ranging from 0.83 to 1.00.

Construct Validity

In EFA, the Kaiser-Meyer-Olkin (KMO) was 0.92, with a Bartlett spherical test value of 1658.43 (P<0.001), which showed that the data could be used for principal component analysis (PCA). PCA extracted one factor with eigenvalue >1.00, which explained 50.14% of the total variance, as was shown in Figure 1. In CFA, the indices of the structural equation model (SEM) of DES-10 were very good. SEM revealed CMIN/DF =2.92, RMSEA =0.07, CFI = 0.96, AGFI = 0.93, GFI=0.95, TLI=0.95. The SEM and standardized regression coefficients of one-factor model were shown in Figure 2.

Figure 1 Lithotripsy map of principal component analysis for Chinese version of the DES-10 (n =190).

Figure 2 Standardized one-factor structural model of the DES-10 (n = 190).

Concurrent Validity

With SDM-Q-9 as criterion, the correlation between SDM-Q-9 and DES-10-Chinese was high (r=0.88, P<0.001). The correlation coefficient between items of DES-10-Chinese and total score of SDM-Q-9 ranged between 0.33 to 0.73, item 6 (I would prefer to discuss treatment goals and options sooner rather than later) had the weakest correlation with SDM-Q-9 (r=0.33, P<0.01), and item 3 (I feel confident in my ability to make decisions about medical treatments) had the strongest correlation with SDM-Q-9 (r=0.73, P<0.01).

Demographic Difference

There was a correlation between age and DES-10-Chinese (r=−0.23, P<0.001). The engagement in shared decision-making varied by marital status, homeplace, and household monthly income, and the difference was statistically significant (P<0.05). The comparison of general demographic information was shown in Table 2.

Discussion

A reliable method to conduct the cross-cultural adaptation of the DES-10-Chinese is critical, and the scale must be proven to be reliable by statistical analysis before it could be used as a method for patients’ evaluation.³² The previously mentioned results showed a successful adaption process among Chinese prostatic cancer patients. This unidimensional DES-10-Chinese scale could evaluate the engagement in decision-making for prostate cancer patients specifically.

In this work, the Cronbach’s ɑ of DES-10 was 0.87, which is satisfactory, for the Cronbach’s ɑ coefficient between 0.70 and 0.90 was recommended when assessing an instrument,³³ and the results showed better internal consistency than the American version (0.80).²⁵ The Cronbach’s ɑ coefficient did not improve after deleting items of DES-10, except item 6. Item 6 “I would prefer to discuss treatment goals and options sooner rather than later” identified the earnest willingness of the prostatic cancer patients to receive quick treatment after admission. Studies have shown that “discuss treatment goals and options” was the primary concern for most Chinese patients, because of the fear of progression and recurrence.³⁴ Although there existed some limitations of item 6 concerning the reliability, we recommend not deleting it given its practical significance.

The test-retest ICC of DES-10-Chinese was 0.98, which met the recommended criterion of 0.90, representing good stability, for ICC ranging from 0.00–0.20 indicated slight agreement, 0.21–0.40 for fair agreement, 0.41–0.60 for moderate agreement, 0.61–0.80 for substantial agreement, and 0.81–1.00 for almost perfect agreement.³⁵ And the results supplemented the test-retest reliability for the original English version of DES-10. The CVI was 0.86, indicating good content validity.³⁶ After repeated cultural adjustment, the DES-10-Chinese had good content validity and achieved the equivalence of the original scale in terms of concept, semantics, and content.

EFA was conducted to estimate the construct validity of DES-10-Chinese, item loading (≧0.40) on one of the common factors was suggested as an ideal factor, and the cumulative variance contribution ratio of the common factor was ≧40%.³⁷ In EFA one factor was explored, which contributed 50.14% of the total variance, and the loading value of each item ranged from 0.40–0.84, indicating acceptable construct validity. Then the CFA was constructed to verify the validity of the scale, fit indexes showed the explained variances strongly agreed with the EFA, and the DES-10-Chinese was demonstrated to be a one-factor structure. Pearson’s correlations were utilized to verify the concurrent validity of DES-10-Chinese and there was a correlation between SDM-Q-9 and DES-10-Chinese (r=0.88, P<0.001), which indicated good concurrent validity in the expected direction.³⁸ These results showed good validity and reliability to assess the patients' engagement in shared decision-making.

Table 2 showed that several sociodemographic variables such as age, marital status, homeplace, and household monthly income were correlated with DES-10. Age was negatively correlated with DES-10, the older, the less participation in SDM, which was in accordance with the previous study.²⁵ The study of Haesebaert et al³⁹ showed that age could affect patients’ decision participation, and participants aged ≧65 years experienced the least shared decision-making. DES-10 scores of patients with current spouses were higher than those of patients with deceased spouses, which was consistent with the study of Rose et al,⁴⁰ which stated that family support was able to encourage frail elderly patients to make shared decisions on rehabilitation options. In the study of Zhai et al,⁴¹ Chinese colorectal cancer surgery patients tended to rely more on close family members to make decisions. Patients from rural areas showed less participation in SDM than those in urban areas. The study of Shi et al⁴² found that urban patients had easier access to information than rural patients, more thorough understanding of tumor-related knowledge, and stronger desire to know about their condition. Higher income could promote SDM participation. Cancer patients usually suffer from high medical expenses, which would affect the treatment choice and hinder their access to relevant medical services. The lower the family’s annual income, the more likely they were to adopt simple surgical treatment. For high-income cancer patients, their financial status and social networks enabled them to seek more treatment options.⁴³ These results suggest that we should pay more attention to the older, lower-income, widowed, and rural patients, when making shared decisions.

This research is limited in several aspects. Firstly, in the measuring process of the retest reliability, only a small group of participants were included, and the telephone combined with WeChat replaced the face-to-face interviews the second time. Therefore, some divergence might exist between the two investigations. Secondly, the patients who were included in this study were all in-patients at large academic hospitals in China, which could have potential selection bias given the high likelihood that these patients are more educated and have more resources than patients treated in the smaller community hospitals. Thirdly, the study did not verify the sensitivity and specificity, for picking an appropriate cut-off value for identifying lower SDM prostate cancer patients. A future study could recruit a larger number of participants to explore the cut-off value and further confirm the reliability and validity of DES-10-Chinese for prostatic cancer patients in other settings in China. Moreover, we may need to adapt this scale in a different population of patients or in a setting outside of the academic hospitals to ensure its external validity. Furthermore, other related factors should be added to improve the engagement of SDM for Chinese prostatic cancer patients.

Conclusion

It is the first time the DES-10 has been tested with respect to homogeneity, stability, content, and construct validity among Chinese prostatic cancer patients. The result of this research showed adequate evidence of stability and homogeneity, as well as of content and construct validity to assess the prostatic cancer patients' engagement in health care decision-making.