Introduction
As an essential component of health care, palliative care aims to improve the quality of life for patients facing life-limiting illnesses. It encompasses a wide range of supportive interventions that address physical, psychological, social, and spiritual needs (Radbruch et al. Reference Radbruch, De Lima and Knaul2020). With the increasing aging population and the rise in noncommunicable diseases, there is a rapidly growing global demand for palliative care services (World Health Organization 2021). In China, the aging population is growing at an unprecedented rate. The number of individuals aged 65 and above has reached 200 million by the end of 2021 in China, accounting for 14.2% of the total Chinese population (National Health Commission of the People’s Republic of China 2022). It is estimated that the number of people in this age group will reach 400 million in 2050 (Yang et al. Reference Yang, Wu and Tan2021), further driving the demand for palliative care.
However, the development of palliative care is still at the infant stage in China. The Chinese government initiated the promotion of palliative care in 2017, and the third batch has been launched, including 185 pilot cities in 2023. Additionally, palliative care education is currently not systematically integrated into medical education, and there is a scarcity of on-the-job training programs for palliative care professionals. Furthermore, the majority of palliative care professionals do not receive or only receive fragmented palliative care training as part of their clinical training programs (Lu et al. Reference Lu, Gu and Yu2018; Zheng et al. Reference Zheng, Guo and Chen2021). According to the 2021 Quality of Death Index, China ranked 53rd out of 81 countries, lagging far behind other countries, reflecting the poor quality of palliative care and limited availability (Finkelstein et al. Reference Finkelstein, Bhadelia and Goh2022).
Self-efficacy refers to an individual’s confidence in their ability to perform the activities required to achieve a desired goal (Bandura Reference Bandura1977). People with high self-efficacy are more likely to handle challenges, successfully accomplish tasks, and demonstrate greater proficiency in regulating their emotions and cognition (Almeida et al. Reference Almeida, Flores and Scheffer2013; Chan Reference Chan2021). The self-efficacy of palliative care providers in delivering care significantly influences the quality of care and patient satisfaction (Carey et al. Reference Carey, Zucca and Freund2019). Therefore, to ensure the quality of palliative care and enhance better patient experiences, particularly in the early stage of palliative care within the Chinese context, it is important to gain a deep understanding of the levels of self-efficacy among palliative care providers (Carey et al. Reference Carey, Zucca and Freund2019; Salins et al. Reference Salins, Ghoshal and Hughes2020). However, the palliative care self-efficacy tools are limited to nurses, and there is a lack of a reliable and valid Chinese version of self-efficacy tool for other health-care professionals involved in palliative care (Chen et al. Reference Chen, Chen and Bai2012; Hu Reference Hu2018).
The Palliative Care Self-Efficacy Scale (PCSS), developed by Eagar in 2003 (Eagar et al. Reference Eagar, Senior and Fildes2003) and validated by Phillips in 2011 (Phillips et al. Reference Phillips, Salamonson and Davidson2011), is a widely used tool for comprehensively assessing self-efficacy in palliative care, applicable not only to nurses, but also to doctors. It has demonstrated good reliability and validity, with the item-level content validity index (I-CVI) ranging from 0.88 to 1.00, and the Cronbach’s α for the scale and subscales ranging from 0.87 to 0.92 (Phillips et al. Reference Phillips, Salamonson and Davidson2011). Additionally, the PCSS has been translated, culturally adapted into multiple languages, including Arabic (Wazqar Reference Wazqar2023), Swedish (Andersson et al. Reference Andersson, Granat and Brännström2022), Mongolian (Kim et al. Reference Kim, Kim and Gelegjamts2020), and Persian (Dehghani et al. Reference Dehghani, Barkhordari-Sharifabad and Sedaghati-Kasbakhi2020). All these versions of the PCSS had been validated and demonstrated good reliability and validity. The PCSS includes 2 theoretically distinct subscales with a total of 12 items. Subscale 1 focuses on “Perceived capability to answer end-of-life care concerns” (6 items), and subscale 2 related to “Perceived capability to respond to patient’s end-of-life symptoms” (6 items). Respondents rate all the items on a 4-point Likert scale, with “4” indicating “confident of performing independently,” “3” indicating “confident of performing with minimal consultation,” “2” indicating “confident of performing with close supervision/coaching,” and “1” indicating “need further basic instruction.” The total score ranges from 12 to 48, with higher scores suggesting higher levels of palliative care self-efficacy.
To accurately evaluate the self-efficacy levels of palliative care providers in China, and establish baseline data for comparing these levels with those of palliative care professionals in other countries, therefore, the study aimed to translate, adapt, and validate the PCSS among Chinese palliative care professionals.
Methods
This methodological study consisted of 2 phases: (1) the translation and cross-cultural adaptation of the PCSS and (2) the evaluation of the psychometric properties of the Chinese version of PCSS (C-PCSS).
Phase 1: Translation and cross-cultural adaptation of the PCSS
After obtaining permission from the PCSS developer, we acquired the English version of the PCSS and proceeded to translated and culturally adapted it into Chinese, adhering to the corresponding guidelines (Brislin Reference Brislin1970; Sperber Reference Sperber2004). The detailed steps were as follows.
Step 1: Forward translation
The PCSS was independently translated into 2 versions (PCSS-1 and PCSS-2) by 2 nursing doctors who are native Chinese speakers and proficient in English.
Step 2: Reconciliation
Reconciliation was conducted through extensive discussions among the aforementioned 2 translators and an additional independent translator who had not participated in the initial translation process. A combined translation version was formed through this process (PCSS-3).
Step 3: Back translation
The back translation of the PCSS-3 was conducted by 2 bilingual translators who have completed at least 1 year of study in a native English-speaking country. Both of them had no prior involvement with the PCSS. After thorough discussions, the back-translated version was submitted to the scale developers for review, and the Chinese version of PCSS-4 was formed for expert review by combining the comments of the original developers.
Step 4: Cross-cultural adaptation and content validity
To enhance the clarity and readability of each item and assess the content validity of the C-PCSS, a modified Delphi consultation was conducted. This process comprised a single round of expert consultation followed by consensus discussions within the research group. A total of 21 experts (see Supplementary Table S1 for demographic characteristics), consisting of 15 palliative care professionals and 6 researchers with extensive experience in measures development, were invited to complete the online questionnaire. The questionnaire was distributed and collected via email or WeChat by researchers and consisted of 3 parts: (1) the brief introduction of the research program; (2) expert consultation form: experts were requested to rate the importance of each item in relation to palliative care self-efficacy on a Likert 5-point scale, score 1 indicates “not important,” and 5 “very important.” Additionally, an extra column was included for individual item comments or revisions; (3) the expert background information form: the demographic information of the experts, the judgment basis (practical experience, theoretical analysis, references, personal intuition), and their familiarity with items.
Phase 2: Evaluation of the psychometric properties of the C-PCSS
Participants and sample size
We employed a convenience sampling method to recruited participants from 11 March 2023 to 28 March 2023 across 32 medical institutions with palliative care centers in China. The inclusion criteria for participants were as follows: (1) age over 18 years and (2) current employment as health-care providers in palliative care centers. Participants who refused to participate in this study were excluded.
We utilized the Kendall sample estimation method to determine the sample size, following the recommendation of 5–10 times larger than the number of items in the scale (Wang Reference Wang2009). For the exploratory factor analysis (EFA) of the C-PCSS, given its 12 items and assuming a 20% rate of invalid responses, the estimated sample size ranged from 72 to 144 (Cheng et al. Reference Cheng, Zhao and Li2023; Lai et al. Reference Lai, Zhang and Hong2022). To perform the confirmatory factor analysis (CFA) for the C-PCSS, recognizing that the data from EFA cannot be reused, we followed the principle that the sample size for CFA should be no less than 200 and be greater than that of EFA (Barbara and Fidell Reference Barbara and Fidell2021; Keenan and James Reference Keenan and James2015). Therefore, a minimal of 272 participants were required in this study.
Data collection
we adopted an online data collection method by sending the website link or QR code of the electronic questionnaire to the potentially eligible participants. The electronic questionnaire includes 3 pages. On the first page of this survey, participants were directed to read the contents of the participant informed consent form, which included detailed information about the purposes of this study, the time that may be spent, the possible risks and benefits associated with participation, as well as avenues for raising complaints about this study. The second page is to fill in the demographic information of the participants including gender, age, marital status, educational background, hospital level, professional title, years of experience in palliative care, and experience of participating in palliative care training course. The final page is to fill the Chinese version of the PCSS.
Data analysis
Data were analyzed using IBM SPSS version 26.0 and SPSS AMOS version 26.0. The categorical data were described by calculating frequencies and percentages, the continuous data were described as the mean and standard deviation (SD). We assessed the item analysis, content validity, construct validity, known-groups validity, and reliability of C-PCSS using the following methods.
Item analysis
(1) Critical ratio (CR) method: the total score of C-PCSS was sorted from low to high, and those with scores in the top 27% were used as the low group and those in the bottom 27% were used as the high group. We conducted an independent sample t-test and considered a CR value was >3 and p < 0.05 as indicative of high discriminant validity (Kelley Reference Kelley1939). Entries that did not reach significance (p > 0.05) were removed. (2) Correlative coefficient method: the correlation of each item with the total score of the scale was analyzed. A higher correlation between the items and the total score indicated better homogeneity of the items with the scale, and items with low homogeneity (correlation coefficient <0.4) with the scale were removed (Tang et al. Reference Tang, Lu and Wu2022).
Validity
Content validity
The content validity index (CVI) was calculated at both the item level (I-CVI) and scale level (S-CVI). The I-CVI was derived from the proportion of experts who rated the item as 4 (important) or 5 (very important) on a Likert 5-point scale. The S-CVI was presented as the average value of I-CVIs. Adequate content validity was determined if I-CVI ≥ 0.78 and S-CVI ≥ 0.90 (Zhang and Zhou Reference Zhang and Zhou2020).
Construct validity
The EFA and CFA were used to analysis the construct validity (Heo et al. Reference Heo, Kim and You2022). Valid data were inputted into the computer which randomly divided the data in half. One half was used for EFA and the other for CFA.
EFA was conducted using principal component analysis with varimax rotation to establish the factor structure. Kaiser–Meyer–Olkin’s (KMO) measure and Bartlett’s sphericity were verified before the EFA; a KMO value >0.5 and significant Bartlett’s test p < 0.05 indicated that is suitable for factor analysis. The items with factor loadings <0.4 will be considered to remove (Zhang and Zhou Reference Zhang and Zhou2020).
CFA was performed by importing data into AMOS version 26.0 to establish a preliminary model and then fitting the model to further test the structure of C-PCSS. The following indices were used to evaluate the model fit: Chi-square freedom ratio (χ 2/df) < 3.00, root-mean-square error of approximation (RMSEA) < 0.08, goodness-of-fit index (GFI) > 0.90, comparative fit index (CFI) > 0.90, Tucker–Lewis index (TLI) > 0.90 (Hooper et al. Reference Hooper, Coughlan and Mullen2008).
Known-groups validity
We hypothesized that professionals with less than 1 year of palliative care experience, or without palliative care training would exhibit lower self-efficacy levels than their counterparts. To test the C-PCSS is sensitive to differentiate the levels of self-efficacy among different groups of professionals, known-groups validity was assessed by comparing mean scores on the total scale and subscales between participants with over 1 year of palliative care experience and those with 1 year or less, as well as between participants who received palliative care training and those who did not. We used the Mann–Whitney U test in a nonparametric test to analyze the known-groups validity. The result was considered statistically significant when the 2-tailed p-value was less than 0.05.
Reliability
Reliability was evaluated by internal consistency reliability and split-half reliability. Internal consistency reliability was evaluated by calculating Cronbach’s α coefficients for each dimension and the total scale. Cronbach’s α coefficient 0.70 was considered an acceptable internal consistency level (Cronbach Reference Cronbach1951). The split-half reliability was assessed by calculating the correlation coefficient between the total scores of the first 6 items and the final 6 items of the C-PCSS.
Ethical statement
The study was approved by the Human Research Ethics Committee of Hunan Cancer Hospital (No. 2022-41). This study was conducted according to the principles of the Declaration of Helsinki and followed relevant guidelines and regulations. Informed consent was obtained from all participants prior to their participation in the survey.
Results
Phase 1: Translation and cross-cultural adaptation
In the translation phase, only minor grammatical and wording discrepancies were identified in the forward translations of the PCSS, and these were easily resolved through discussion by the 2 translators. In the cross-cultural adaptation section, to enhance clarity and comprehension in the Chinese context, for items 7–11, the phrase “Reacting to and coping with” was changed to “Assessing and managing.” Additionally, some experts suggested that the statement in item 12 needed to be adjusted. Therefore, for item 12, “Reacting to and coping with limited patient decision-making capacity” was revised to “Helping and supporting for patients with limited decision-making capacity” (see Supplementary Table S2 for C-PCSS). All the modifications were reviewed and confirmed with the developer.
Phase 2: Psychometric evaluation of C-PCSS
Demographic characteristics
In total, 493 valid questionnaires were included in our analyses. No missing data was observed in our study due to each item in our electronic questionnaire was required to be completed. The collected data were from 419 nurses, 44 doctors, and 30 administrators (nurse manager, doctor director). More details are presented in Table 1.
Demographic characteristics of the participants (n = 493)
Item analysis
The results showed the CR value of each item was >3 (p < 0.01), r value of each item was >0.4, which indicated that there was a good homogeneity of the items with the scale (Table 2).
Item analysis, factor loadings, and internal consistency reliability of the C-PCSS
Factor 1: Perceived capability to answer patient’s end-of-life care concerns; Factor 2: Perceived capability to respond to patient’s end-of-life symptoms.
** p < 0.001.
Validity
Content validity
The expert response rate was 100%, I-CVI ranged from 0.857 to 1.000, and S-CVI/Ave was 0.956.
Construct validity
The 493 questionnaires were randomly grouped into EFA (247 cases) and CFA (246 cases) groups using Excel (Microsoft, 2021) to explore the construct validity of the scale. EFA results showed that the KMO value of the C-PCSS was 0.909, Bartlett’s test was χ 2 = 2581.780, p < 0.001, indicating the fitness for EFA. The principal component analysis with varimax rotation extracted 2 common factors, which were the same as the original PCSS. These were “Perceived capability to answer patient’s end-of-life care concerns” and “Perceived capability to respond to patient’s end-of-life symptoms.” The cumulative variance contribution rate was 74.19%, which met the requirement that the cumulative variance contribution rate should be at least 40%. Details of factor loadings are presented in Table 2. CFA verified that the 2-factor model had a satisfactory model fit with χ 2/df = 2.724, RMSEA = 0.084, GFI = 0.916, CFI = 0.967, and TLI = 0.952. Detailed results are shown in Table 3 and Figure 1.
Confirmatory factor analysis of the C-PCSS.
The goodness-of-fit indices of the C-PCSS
χ 2/df = Chi-square freedom ratio; RMSEA = root-mean-square error of approximation; GFI = goodness-of-fit index; CFI = comparative fit index; TLI = Tucker–Lewis index; C-PCSS = Chinese version of palliative care self-efficacy scale.
Known-groups validity
The results supported our hypothesis that professionals with less than 1 year of palliative care experience or without palliative care training exhibited lower self-efficacy levels than their counterparts. Participants with more than 1 year of palliative care experience scored higher on total C-PCSS score (M = 38.57, SD = 0.45) compared to those with 1 year or less experience (M = 34.80, SD = 0.78) (p < 0.001). Similarly, they also scored higher on subscale 1 and subscale 2 (Table 4). A significant difference was observed between participants who underwent palliative care training and those who did not. Those who participated the training (M = 37.88, SD = 0.41) had higher total C-PCSS scores than those who did not undertake any courses (M = 34.49, SD = 1.31) (p = 0.014). Additionally, they also scored higher on subscale 1 and subscale 2 (Table 4).
Known-groups validity of the C-PCSS
Subscale 1: Perceived capability to answer end-of-life care concerns.
Subscale 2: Perceived capability to respond to patient’s end-of-life symptoms.
** p < 0.01; *p < 0.05.
Reliability
Internal consistency reliability for the whole scale was 0.943, indicating high internal consistency or homogeneity for the scale. The Cronbach’s α coefficients for each subscale of the C-PCSS ranged from 0.904 to 0.933. Split-half reliability for the whole scale was 0.844.
Discussion
This study cross-culturally adapted and validated the PCSS within the Chinese context, it is a widely utilized measurement tool due to its reliability, validity, the advantage of being fast and simple to complete, and without professional-specific limitations (Andersson et al. Reference Andersson, Granat and Brännström2022; Dehghani et al. Reference Dehghani, Barkhordari-Sharifabad and Sedaghati-Kasbakhi2020; Kim et al. Reference Kim, Kim and Gelegjamts2020). The forward and backward translation processes were performed rigorously, adhering to established guidelines and recommendations (Brislin Reference Brislin1970; Sperber Reference Sperber2004). And the cultural adaptation procedure ensures that each individual item is clarify, readable in Chinese, and is equivalence to the original English version. Thus, we believe that the C-PCSS can be used by Chinese palliative care professionals without major difficulty.
In this study, all the item-total correlation coefficients were above 0.4, which indicated a high degree of item homogeneity within the scale. The internal consistency results for the C-PCSS exhibited an excellent level (raw coefficient alpha: 0.943), significantly surpassing the threshold of 0.70 recommended by Nunnally and aligning closely with the original version (raw coefficient alpha: 0.92) (Nunnally Reference Nunnally1978; Phillips et al. Reference Phillips, Salamonson and Davidson2011). The subscale 1 “Perceived capability to answer patient’s end-of-life care concerns,” and the subscale 2 “Perceived capability to respond to patient’s end-of-life symptoms” also showed an excellent internal consistency, with the Cronbach’s α of 0.904 and 0.933 respectively, The split-half reliability of the whole scale was 0.844, reinforcing the robustness of the C-PCSS as a reliable tool for assessing self-efficacy in palliative care.
In terms of the content validity, the majority of the experts in this study unanimously thought that each item in the PCSS could adequately reflect the purpose of the measurement. The findings from the I-CVI and S-CVI analysis established the excellent content validity of the C-PCSS. Meanwhile, the construct validity of the C-PCSS has been well supported by the results of the EFA and CFA in this study. The results of both EFA and CFA confirmed the 2-factor structure of this instrument, explaining 74.19% of the variance. In the original English version, the results of EFA also revealed the present 2 factors, accounting for a cumulative total variance of 65.3% (Phillips et al. Reference Phillips, Salamonson and Davidson2011). In addition, CFA verified that the 2-factor structure of the C-PCSS had a satisfactory model fit, and all the factor loadings in each domain were above 0.4, which further confirmed the stability of the 2-factor structure.
The findings of the known-groups comparison indicated that the C-PCSS exhibited good known-groups validity, showing sensitivity in differentiating levels of self-efficacy between professional with less than 1 year of palliative care experience and their counterparts (p < 0.001).
Health-care professionals with over 1 year of palliative care experience scored significantly higher on the C-PCSS in comparison to their counterparts who had less than or equal to 1 year of experience, as well as they reported higher scores on both subscale 1 and subscale 2 of the C-PCSS. This noteworthy difference highlights the positive impact of long-term engagement in palliative care services on professionals’ self-efficacy. The findings were supported by studies conducted by Peng and Van (Peng et al. Reference Peng, Chen and Liu2019; Van Dyk et al. Reference Van Dyk, Siedlecki and Fitzpatrick2016), health-care professionals with more years of experience have more contact with end-of-life patients, and can effectively perform palliative care clinical work and cope with emergencies, and exhibit lower levels of fear regarding mortality compared to their less experienced counterparts (Peng et al. Reference Peng, Chen and Liu2019; Van Dyk et al. Reference Van Dyk, Siedlecki and Fitzpatrick2016). These factors collectively play a substantial role in elevating their self-efficacy when caring for patients. Additionally, the study demonstrated a marked and statistically significant difference between participants who had completed palliative care training courses and those who had not, consistent with the findings of Phillips et al. (Reference Phillips, Salamonson and Davidson2011). In our study, those who had taken part in palliative care training courses achieved higher total scores on the C-PCSS in comparison to their counterparts who had not undergone training, which further identified that the palliative care education and training programs contribute significantly to the development and improvement of self-efficacy in the realm of palliative care (Li et al. Reference Li, Liu and Xu2021). Health-care educational administrators can develop culturally appropriate palliative care curriculum training programs in the future to improve practitioners’ perceived competence in palliative care symptom management and in solving patient-related concerns.
Limitations
This study makes a valuable contribution to the development of palliative care in China. However, it is important to acknowledge some limitations that should be considered in our study. First, there is a professional disproportion in the phase of evaluating the psychometric properties for the PCSS, with the majority of participants being nurses, this may cause bias into the study results. Second, the criterion-related validity of the C-PCSS was not assessed in our study due to the unavailability of suitable criterion tools. Future research should address this aspect to confirm the instrument’s applicability in China. Finally, this study primarily focused on content validity, structure validity, and known-groups validity, the face validity assessment was not conducted in our study. It is advisable for future researchers to incorporate an appraisal of face validity to further ensure the clarity, relevance, and appropriateness of the C-PCSS.
Conclusions
The findings from this study affirmed the good validity and excellent reliability of the C-PCSS. The scale’s strong internal consistency, content and construct validity, and known-groups validity collectively demonstrated its utility in assessing the self-efficacy of Chinese palliative care professionals. Therefore, the C-PCSS can be emerged as a valuable and reliable instrument for quantifying palliative care professionals’ self-efficacy levels in China. Considering the professional disproportion among the participants in this study, future studies need to further validate the C-PCSS in a more diverse range of medical personnel cohorts with larger sample sizes.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1478951524000518.
Data availability statement
The data used during this study are available from the corresponding authors on reasonable request.
Acknowledgments
Thanks to all the health-care professionals who participated in this study.
Author contributions
Conceptualization, Y.D. and Y.C.; methodology, Y.D., B.S., L.W., and W.P.; software, J.G.; validation, J.G., Y.D., and Y.C.; formal analysis, J.G.; investigation, Y.D.; resources, Y.D. and Y.C.; data curation, J.G. and Y.D.; writing – original draft preparation, J.G.; writing – review & editing, Y.D.; visualization, Y.C.; supervision, Y.C. All authors read and approved the final manuscript.
Funding
This research was funded by the Research and Development Project of Hunan Provincial Department of Science and Technology (grant number: 2020SK2121).
Competing interests
The authors declare no conflict of interest.
Ethical approval
The study was approved by the Human Research Ethics Committee of Hunan Cancer Hospital (No. 2022-41). This study was conducted according to the principles of the Declaration of Helsinki and followed relevant guidelines and regulations. Informed consent was obtained from all participants before they participated in this survey.
