Healthcare Provider Knowledge, Attitudes, and Practices in Hospice Care and Their Influencing Factors: A Cross-sectional Study in Shanghai

Background: In 2017, the Chinese government launched a pilot project in hospice care (HC), in which Shanghai was a pioneer. Healthcare provider knowledge, attitudes, and practices in hospice care (KAPHC) may facilitate or hinder HC development. To determine how to design targeted training for healthcare providers and policies to improve their KAPHC, we conducted an original study based on an indigenized KAPHC scale to (a) comprehensively measure the KAPHC baseline of healthcare providers in Shanghai and (b) explore its influencing factors. Methods: A cross-sectional study was designed to evaluate healthcare providers’ KAPHC with the indigenized KAPHC scale. Descriptive analysis, linear regression, and Pearson’s (r) correlation analysis were performed to uncover providers’ KAPHC, its influencing factors, and their correlations. Results: The KAPHC scale proved applicable to the knowledge, attitudes, and practices of the large sample of providers we surveyed. The 7027 KAPHC scaling results revealed that 42.44% of participants had HC experience and 57.49% were willing to provide HC. The mean accuracy of the responses related to knowledge was 59.30%. Scoring rates for attitudes, confidence, and self-reported practices were 74.20%, 73.96%, and 75.55%, respectively. The linear regression revealed that higher KAPHC scores were associated with experience and willingness and varied with professional specializations. The Pearson’s (r) correlation evidenced that HC practices were strongly correlated with confidence (r = 0.648, P<.001) and moderately correlated with attitudes (r = 0.463, P<.001). Conclusion: We uncovered that targeted training for enhancing healthcare provider KAPHC in Shanghai should focus on psychological and spiritual care, the management of pain and other symptoms, and traditional Chinese medicine (TCM). Additionally, training should be scaled up for providers with different characteristics. Meanwhile, policy should encourage providers to work enthusiastically in HC—universal high-quality HC requires well-trained, supported, and motivated providers.


Background
Aging populations have become a tough challenge for many countries; they can yield complex and diversified demands for healthcare, 1 including end-of-life (EOL) hospice care (HC) and palliative care. China has the largest elderly population in the world. 2 According to the seventh census of China, people aged 60 and over comprised 18.70% of the total population in 2020. 3 Meanwhile, Shanghai is home to an even higher percentage of older adults, with people aged 60 and over comprising 23.38% of the city's population. 4 China's large aging population has given rise to an increasing demand for high-quality HC. The Economist Intelligence Unit's 2015 Quality of Death Index ranked mainland China 71 out of 80 countries 5 ; this indicates that HC in China is limited in terms of availability and quality. In addition, due to the miniaturization of family structure and traditional Chinese lifestyles, most EOL patients do not receive sufficient treatment for pain and other physical, psychological and spiritual issues. 6 The rapid global spread of coronavirus disease 2019 (COVID-19) and the great suffering it caused have emphasized the need to improve the functionality of HC-in May 2020, the World Health Organization (WHO) added palliative care to its Clinical Management of COVID-19 Interim Guidance. 7 Notably, HC can not only significantly reduce the costs of caring for terminal patients but also produce considerable economic benefits. 8 HC and palliative care are crucial parts of comprehensive, people-oriented health services; more specifically, they integrate physical, mental, spiritual, and social care for terminal patients by controlling their discomfort. 9 "HC" refers to the active holistic care given to individuals suffering from severe illness and EOL patients to improve their quality of life as well as that of their families. 10 HC involves treating pain and other symptoms to ensure the patient is as comfortable as possible. Typically, patients and their families are provided psychological and social support in step with a multidisciplinary collaborative model.
The development of HC and the integration of palliative medicine into health systems are among the most impactful parts of medical reform. 11 Since the New Medical Reform of China was published in 2009, the government has accelerated the construction of primary healthcare institutions and promoted the development of modern HC. The Chinese National Commission conducted modern experimental work in HC in 5 cities across China in 2017 12 and expanded this work to 87 cities in 2019 13 (in step with the principles of the Healthy China 2030 strategic plan). Shanghai is a pioneer of the HC movement in China; today, it is home to an early-stage HC service system with localized characteristics. 14 Healthcare provider awareness and services are often regarded as significant facilitators or barriers in the development of HC. Physicians and nurses play an important role in HC teams, which include multidisciplinary professionals, such as social workers, psychiatrists, psychologists, and volunteers. Healthcare providers are expected to deliver maximal benefits to their patients, and their knowledge, attitudes, and practices in hospice care (KAPHC) are of vital importance to the development of HC. While more attention has been paid to the development of HC of late, it remains crucial for healthcare providers to strengthen their knowledge of and attitudes toward HC. 15 Along these lines, training in HC is currently in high demand by most healthcare providers; in particular, there is a demand to learn knowledge and skills related to living wills and other legal concerns, techniques for stress relief, and social work. 16 The insufficient awareness about the processes of life and death and the lack of innovative teaching methods in China make the situation even worse. 17 Existing studies on HC were mostly conducted with physicians (especially oncologists) or nurses and often advise that more attention should be paid to educating healthcare providers in HC and EOL care to increase their self-efficacy in these areas. 18,19 Relatively few studies have assessed KAP in healthcare providers and effective strategies for barriers to developing HC in China.
Accordingly, we conducted a novel large-scale empirical study to comprehensively evaluate healthcare provider KAPHC based on an indigenized KAPHC scale published by our research team. 20 Then, we performed a linear regression and Pearson's (r) correlation analysis to uncover related influencing factors and their correlations. Ultimately, we sought to discover the problems in delivering HC and develop corresponding solutions rooted in improving healthcare provider KAPHC with targeted training. In particular, this study focused on the KAP of healthcare providers as a prerequisite to improve HC quality. Notably, this study provides an evidence-based scale and data that can be used to conduct similar inquiries in other areas of China. This study has theoretical and practical implications for improving the quality of life of terminal patients and their families and establishing a scientific basis for the development of highquality HC service.

Questionnaires
This study was designed as a cross-sectional study with an anonymous questionnaire survey. Participants were informed that participation was voluntary and were required to give informed consent. We adopted the indigenized healthcare provider KAPHC scale, which was scientifically constructed by Shu et al 20 and which proved to have good reliability and validity. The scale comprised four parts: demographic characteristics (11 items), knowledge (15 items), attitudes (24 items with 5 sub-concepts), and practices (2 sub-concepts, confidence of practices and self-reported practices, which each comprised 11 items) in HC. Only healthcare providers with HC experience were asked to give responses to survey items on self-reported practices (see Supplementary file 1). Scores for each subcategory were calculated separately. Regarding the knowledge scale, each item was marked as "right" (1) or "wrong" (0), based on previous widely used scales and expert In particular, this study focused on the knowledge, attitudes, and practices (KAP) of healthcare providers as a prerequisite to improve hospice care (HC) quality and provides an evidence-based scale and data useful for measurement in other areas of China. • Health providers' self-reported practices in HC were strongly correlated with confidence and moderately correlated with attitudes. • There is an urgent need to enhance targeted training for healthcare providers across different skills. • Policy-makers can formulate policies based to the results of this study, which may include targeted training for healthcare providers to ensure high-quality HC for patients with advanced and incurable diseases.

Implications for the public
Hospice care (HC) is a crucial part of integrated, people-oriented healthcare services. Providing physical, mental, spiritual, and social care for older people and terminally ill patients by treating their discomfort are the basic skills of healthcare providers working in this context. By completing training in the knowledge, attitudes, and practices (KAP) related to HC, healthcare providers can better satisfy their patients' complex needs and improve their quality of life. Additionally, healthcare providers able to effectively apply their medical specializations in delivering integrated and people-oriented healthcare can benefit the public by enhancing the quality of life of terminal patients and helping them die peacefully with dignity. Moreover, improving healthcare providers' knowledge, attitudes, and practices in hospice care (KAPHC) will enhance relationships between healthcare providers and patients.

Key Messages
advice. Meanwhile, the attitudes section was measured using a five-point Likert scale, from totally agree (5) to totally disagree (1). The score of negative topics (Dimensions 1 and 4) was corrected in reverse before calculation. The other two sections-confidence and self-reported practices-were also measured using a five-point Likert scale, from rather confident (5) to rather non-confident (1) and always (5) to never (1).

Data Collection
In order to establish a systematic and comprehensive understanding of the status of healthcare provider KAPHC in Shanghai, we designed a survey that covered all HC institutions between November 1 and December 31, 2019. For each investigated institution, 30 healthcare providers were chosen randomly, including physicians, nurses, administrators, and other health technicians, and the sample size was expanded by an extra 5% to avoid invalid data. The inclusion criteria of the study were as follows: healthcare providers who (1) were employed in the units at the time of the study; (2) were certified physicians, nurses, health technicians, and administrators; and (3) volunteered to participate in this investigation and gave informed consent. The exclusion criteria were as follows: healthcare providers who (1) were not employed or certified in the sampled units and (2) refused to participate or were unable to independently complete the survey. The anonymous, cross-sectional questionnaire survey was conducted through SO JUMP, a professional online questionnaire survey platform used by a large number of companies and individuals. To minimize the probable bias, strict quality control was conducted throughout the process. A concentrated training was held at the municipal level for all quality controllers before the investigation, and a backto-back quality control inspection on the questionnaire was conducted by the research group.
Statistical Methods Statistical analyses were performed using the commercial software IBM SPSS Statistics 24.0 (IBM Corporation, Armonk, NY). Descriptive statistics were used to analyze the participants' demographic characteristics and KAP scores. Continuity variables were described by mean ± standard deviation (SD) and categorical variables by frequency and proportion. A linear regression analysis was used to assess associated factors of healthcare provider KAP. A positive standardized regression coefficient (β) was used to indicate a protective factor and negative one to indicate a risk factor. A larger |β| suggested greater effects on the dependent variable (0<|β|<1). Pearson's (r) analysis was used to explore correlations between every two dimensions among knowledge, attitudes, confidence, and self-reported practices. A larger |r| indicated a stronger correlation and an |r| closer to 0 indicated a weaker correlation (0<|r|<1). The test level was α = 0.05 and all P values represented bilateral probability. Most participants were female (79.22%), 62.42% were between 30-50 years of age, 68.14% had a bachelor's degree or above, 38.17% were nurses, and 34.75% were physicians. More than half (79.18%) had witnessed the process of death of an EOL patient. In terms of participation, 42.44% of healthcare providers had provided HC and 57.49% of them were willing to provide HC. The most popular reason respondents were willing to provide HC was a sense of responsibility (75.50%); meanwhile, the most popular reason respondents were not willing to provide HC was pressure (73.52%). Table 1 details the respondents demographic characteristics.

Knowledge
The mean knowledge score of respondents was 8.90 ± 2.632; response accuracy was 59.30%. Table 2 details the questions and scores. Question 14, "HC should be provided by a multiprofessional HC team that includes general physicians, " had the highest rate of correct responses (93.88%), whereas Question 1, "The provision of HC requires emotional detachment, " turned out the lowest rate of correct responses (16.58%). The question with second lowest rate of correct responses (26.57%) occurred with Question 7, "During the terminal stages of an illness, respiratory depression medicine are appropriate for certain treatments of severe dyspnea. " In total, 8 items had a rate of correct responses below 60.00%.

Attitudes
The mean score of the attitudes scale was 89.04 ± 11.320 and the total scoring rate was 74.20%. Table 4 presents the attitudes scale items and their mean scores. Among the different dimensions, "Perception of the benefits for the life quality promotion" had the highest mean score (4.02 ± 0.635), whereas "Perception of the threats from the worsening conditions of advanced patients" had the lowest score (3.04 ± 0.841). Question 8, "Having care and accompany by medical team enhance quality of life" had the highest mean score (4.42 ± 0.728); by contrast, Question 19, "Advanced patients' complex symptoms are barriers to providing HC" had the lowest mean score (2.51 ± 1.068).
The linear regression analysis (Table 3) evidences that the following variables significantly correlated with higher attitudes scores: higher education level (β = 0.041, P = .001), experience in witnessing death (β = 0.054, P < .001), experience in providing HC (β = 0.044, P < .001), and willingness to provide HC (β = 0.362, P < .001). Physicians had lower attitudes scores (β = -0.045, P = .010) than others. Confidence and Self-reported Practices Among all participants, there were 2982 (42.44%) healthcare providers with self-reported HC experience; these providers completed the self-reported practices scale. The total scores for confidence and self-reported practices toward HC were 40.68 ± 7.822 and 41.55 ± 8.145, respectively; the total scoring rates were 73.96% for confidence and 75.55% for practices. Table 5 presents the mean scores of the confidence and selfreported practices scale items. Question 8, "Create good relationship between the medical staff and family members, " had the highest mean score (3.87±0.849, 3.99 ± 0.797), whereas Question 11, "Guide the management of afterwards and funeral preparation for families, " had the lowest (3.55 ± 0.996, 3.43 ± 1.048). The data analysis of Table 6 shows that correlations existed (P < .05) among two of the four dimensions (ie, knowledge, attitudes, confidence, and selfreported practices); more specifically, HC practices were strongly correlated with confidence (r = 0.648, P < .001) and moderately correlated with attitudes (r = 0.463, P < .001).

Knowledge
The mean knowledge score of healthcare providers regarding HC was 8.90 out of 15 (59.30%), indicating that the healthcare providers had insufficient knowledge of HC; this is generally consistent with other domestic studies 21,22 . This lack of knowledge may mainly be due to deficits in HC education. Physicians from different countries expressed that current undergraduate and postgraduate programs did not provide them with sufficient education on HC. 23  Our study found that healthcare providers had a basic level    Studies have shown that healthcare providers demonstrated better knowledge and attitudes in HC after being trained in the field. 29,30 Therefore, healthcare providers in different positions should complete professional HC and palliative care training to fill knowledge gaps, especially regarding psychological and spiritual care, the management of pain and other symptoms, and the application of TCM.

Attitudes
Our study revealed that healthcare providers had moderate attitudes toward HC; the score in our study was 89.04 out of 120 (74.20%), which is higher than the scores in other domestic studies. 31,32 Regarding international studies that used the Frommelt Attitude Toward Care of the Dying Scale, Lange et al 33 reported a score of 129 out of 150 (86.00%) for 355 oncology nurses in the United States, Abudari et al 25 reported a score of 111.66 out of 150 (74.44%) in Saudi Arabia, and another study reported a score of 71.67% for 190 nurses in Japan. 34 These differences are likely attributable to international differences in cultural approaches to the process of life and death and HC training. 34 Our study found that healthcare providers highly agreed that HC had many benefits and especially affirmed its ability to enable patients to access people-oriented care provided by multidisciplinary teams that allowed them to die with a relatively good quality of life and with dignity. However, healthcare providers only partly agreed with the idea that HC can help people die at home and avoid considering euthanasia; this negative attitude may impede the development of homebased HC to a certain degree. In addition, healthcare providers reported that the main barrier to providing HC was that advanced cancer patients had many complex symptoms. This echoes their lack of relevant knowledge and is consistent with studies by Liu et al, 35 Sun et al, 36 and Kuang et al. 37 Healthcare providers reported that they often felt sad and experienced a sense of loss when providing HC; sometimes they said they even felt guilty when patients died. Traditional medical education mainly focuses on the treatment of diseases rather than comforting or relieving the patient suffering; this may cause physicians to regard incurable diseases as "medical failures. " 35 The related negative attitudes of healthcare providers toward complex symptoms highlights a trend of poor medical skills in symptom control and management. A lack of professional training in HC will inevitably affect provider ability and attitude, which can cause psychological pressure and distress-key reasons why healthcare providers in our study were unwilling to participate in HC. Promoting professional training in HC for healthcare providers can improve their attitudes toward HC 38,39 (and thus help them deal calmly with death 40 ) and reduce or alleviate the barriers to providing HC services. Designing and offering professional training in HC will remain a worthwhile endeavor.

Confidence and Self-reported Practices
The mean score of the confidence scale was 40.68 out of 55 (73.96%), which indicated that healthcare provider confidence levels leaned toward "confident" (3.70 ± 0.711). Regarding the self-reported practices of the providers with HC experience, the mean score was 41.55 out of 55 (75.55%) and the frequency leaned toward "usually" (3.78 ± 0.740). The results showed that most healthcare providers needed support in providing HC; this is consistent with some previous studies, such as a cross-sectional descriptive study on nurses in Mongolia 19 and another physician-related study in Vietnam. 29 Domestically, our result was higher than that of Gai et al, 41 who reported that 128 oncology nurses in Shenzhen scored 64.81 out of 100 (64.18%). However, our result was lower than that of Wang et al, 21 who reported that 1037 oncology physicians and nurses in Shandong scored 3.92 out of 5 (78.40%). This was mainly due to different questionnaire design models and participants. We also found that the frequency of self-reported practices was consistent and positively correlated with the outcomes of the confidence scale (r = 0.648, P < .001). Healthcare providers had better confidence and practices in maintaining stronger doctor-patient relationships, understanding and responding to family members' wishes and concerns, assessing patients' pain, and satisfying patients' physical and mental needs. Participants usually alleviated the pain and discomfort of dying patients, but they were not confident about their abilities, mainly because they lacked professional skills in pain control. Moreover, healthcare providers had low confidence about guiding death and funeral preparations and their related scores for practices were also low. The traditional doctrines of Confucianism, Taoism, and Buddhism have deeply impacted  the Chinese view of life and death and continue to affect contemporary life, 42 such as the taboos of discussing death and inappropriate view of life and death. Without a professional education in death, some healthcare providers feel helpless in regard to discussing patient deaths with family members. 43 Comprehensive education on the process of life and death should be given not only to patients and their families but also their caregivers. Healthcare providers who had a scientific outlook on death were able to help EOL patients and their families face death more positively and with dignity. 44 The experiences of developed countries like the United Kingdom have proved that life-and-death-related research can make HC more acceptable to healthcare providers, patients, and family members. Additionally, education on death has brought great improvements to the development of HC. 45 Influencing Factors of KAPHC There were positive correlations among the dimensions of KAP. Healthcare providers with higher scores in knowledge and attitudes usually expressed more confidence and practiced more frequently, which indicated a strong correlation between practices and confidence. The linear regression analysis suggested that healthcare providers with experience in witnessing death or a high willingness to deliver HC scored significantly higher on the KAPHC scale; this illustrates that work experience and good professional identity promote HC practices. While most healthcare providers fully recognize the need to promote quality of life, 42.51% of respondents were unwilling to participate in HC delivery because of the high level of pressure it involved (reason cited by 73.52% of respondents unwilling to deliver HC). Thus, relieving work stress is an urgent issue that should be solved to encourage healthcare providers to deliver HC. In addition, the government and HC institutions should introduce and implement innovative payment and incentive policies to encourage providers to work actively and enthusiastically in HC. The study revealed that the KAPHC scores of healthcare providers were significantly influenced by their work specialties. Physicians demonstrated better practices and knowledge but worse attitudes. In comparison, nurses reported better attitudes but poorer practices and professional knowledge. Additionally, gender, age, marital status, education level, and professional title were also influencing factors, as other studies have evidenced. 46,47 Healthcare providers at all levels need professional HC training tailored to their characteristics, which can improve their own perceptions of their levels of KAPHC as well as their real relative levels of KAPHC.

Limitations
This study had several limitations. Firstly, Shanghai is a developed region that is pioneering HC in China; accordingly, the KAP levels of the healthcare providers in Shanghai may be better than those of providers in other areas of China. The results should be verified by work conducted across different areas of China. Secondly, some bias may appear due to the format of the question design: the questions for the attitudes scale were mainly subjective items and the confidence and practices items were self-reported. Lastly, the study did not specify how the influencing factors affected the KAPHC scores of the healthcare providers; further research is needed to explore this in depth.

Conclusion
This study comprehensively investigated healthcare providers' knowledge, attitudes, confidence regarding practices, and self-reported practices in HC and the correlations between these different dimensions. We found that the KAPHC scale was applicable to the KAP of the large sample of healthcare providers we surveyed, which spoke to China's current state of HC. This study accordingly offers insights on how to conduct similar measurements in other regions of China. Healthcare providers demonstrated insufficient knowledge in HC, especially related to psychological care, the management of pain and other symptoms, and the application of TCM. Meanwhile, the outcomes of attitudes, confidence, and selfreported practices were moderate. The KAPHC of healthcare providers varied significantly with gender, education level, professional title, work specialty, HC experience, and willingness to provide HC. Additionally, there were close correlations among KAP dimensions. Ultimately, the study evidences that the government and HC organizations would do well to scale up targeted HC training for healthcare providers with different characteristics to improve their KAPHC; moreover, they should also introduce innovative payment and incentive policies to encourage providers to work enthusiastically in HC. Universal access to high-quality HC can only be achieved through the participation of welltrained, supported, and motivated healthcare providers.