A comparative study of dementia knowledge, attitudes and perception of care approach among Chinese nursing and medical students

Background Dementia care requires inter-disciplinary collaboration starting from formal health professional education. Yet, little is known about how undergraduate medical and nursing students perceive dementia care in China. The aim of this study was to investigate undergraduate medical and nursing students’ knowledge, attitudes and care approach of dementia in China. Methods A cross-sectional survey was conducted. Students enrolled in a 5-year Bachelor of Medicine Program and a 4-year Bachelor of Nursing Program from four universities with campuses across Eastern, Western, Southern and Northern China were recruited into the study. Three validated instruments, Alzheimer’s Disease Knowledge Scale (AKDS), Dementia Care Attitude Scale (DCAS) and Approach to Advanced Dementia Care Questionnaire (ADCQ), were used to measure students’ dementia knowledge, attitudes and perceived care approach. Data were collected using a self-administered survey. Results The number of medical and nursing students completing the survey was 526 and 467 respectively. The overall mean score of students’ dementia knowledge measured by the ADKS was 19.49 (SD=2.82) out of 30, students’ attitudes to dementia was 29.92(SD=3.35) out of 40, and students’ person-centred care approach of dementia was 5.42 (SD=2.20) out of 13. Students’ overall knowledge towards dementia was poor and the attitudes were generally positive. Medical students demonstrated higher dementia knowledge scores and showed less positive attitude scores than nursing students (p<0.05). Students would not apply a person-centered care approach. There were no statistically signicant differences in the mean scores of ADCQ between nursing students and medical students. Conclusions Study results highlight the urgent need to implement an inter-disciplinary approach to increasing dementia education among Chinese medical and nursing students, and ensuring that students have adequate knowledge, attitudes and experience in the care of people with dementia.

Often dementia education is delivered as a 'tick-box' exercise or an add-on, or there is an ad hoc approach. This kind of education has little value attached and does not positively impact on students' knowledge, attitudes and skills in providing effective care for people with dementia [11].
Numerous studies have found the lack of knowledge and inappropriate attitudes towards dementia among students in health and non-health areas [17][18][19][20]. In a systematic review, Ahmad et al. found that the majority of college and university students reported lower understanding of risk factors and causes related to dementia [20]. In a recent cross-sectional study involving 359 adolescent students, around 79% of them perceived that people with dementia demanded too much time on keeping them clean, healthy and safe [19]. Moreover, in another study by Tullo et al., they found the lack of person-centred approach in dementia care in medical students [21]. For example, most participants agreed that "it is acceptable to lie to people with dementia if you think the truth might be upsetting to hear " and "it is preferable to try to talk to a member of the family of people with dementia rst, before speaking to the patients" [21]. However, few studies were conducted to compare dementia care knowledge, attitudes and person-centred care among health professional students for the purpose of promoting inter-professional education in dementia care.
In recent years, globally some efforts have been made to improve dementia education for undergraduate medical and nursing students. The Higher Education Dementia Network (HEDN) in UK has developed dementia core skills education and training framework for undergraduate programs [22]. The Health and Social Care Board in Northern Ireland recently designed a Dementia Learning and Development Framework to facilitate development and provision of dementia education programs [22]. These initiatives are useful for supporting the development of dementia education in undergraduate programs. However, inter-disciplinary dementia education remains brief and inadequate [8,23,24]. The content, implementation and assessment of inter-disciplinary dementia education remain inherently challenging [8,23,24]. Identifying the current knowledge, attitudes and care approach of students from different disciplines is seen as a key component for the development of improved inter-disciplinary dementia education [8,24].
Most of the previous studies assessed students' dementia knowledge, attitudes and care approach were conducted in developed countries [25,26]. There is currently little understanding of how Chinese undergraduate medical and nursing students' perceived this important topic. To meet recommended clinical practice guidelines for inter-professional dementia care [8,10], it is imperative to evaluate and compare undergraduate medical and nursing students' dementia knowledge, attitudes and perception of care approach to provide evidence to inform reforms to inter-disciplinary dementia education undergraduate programs in China.

Aims
The aims of this study were to measure medical and nursing students' dementia knowledge, attitudes and perceived care approach, and explore associated factors to inform inter-disciplinary dementia education curriculum design in China.

Study design
From January to May 2018, a cross-sectional research survey of Chinese undergraduate nursing and medical students was completed to gather self-reported data of dementia knowledge, attitudes and perception of care approach at four universities with campuses across Eastern, Western, Southern and Northern China.

Participants
Using convenience sampling, all students enrolled in a 5-year Bachelor of Medicine Program and a 4-year Bachelor of Nursing Program were invited to participate in the study. Students were included if they meet the following inclusion criteria: currently enrolling in 5-year Bachelor of Medicine Program or a 4-year Bachelor of Nursing Program in the participating universities.

Sample size analysis
The sample size was based on the primary outcome, the medical and nursing students' dementia knowledge using the Alzheimer's Disease Knowledge Scale (ADKS). An early study examined sample of students using the same instruments and reported mean knowledge scores of 20.19±3.59 [27]. Using this mean score as a reference for sample size calculation with a 95% con dence interval, a beta error of 10% and an alpha of 0.05, the required sample size was calculated using this formula: A sample size of 400 was required based on the calculated. Considering a 50% return rate in survey studies, 600 participants were needed in the study.

Procedure and Participants
Prior to participant recruitment, permission to contact nursing and medical students was obtained from each of the four universities. Two trained research assistants were responsible for participant recruitment and questionnaire distribution at each campus.
Interested students were given detailed information about the study by trained research assistant. Students were informed about the aim, data collection, procedures, bene ts, risks and con dentiality of the study via written information. Returning completed questionnaire survey was the evidence that students consented to participate in the study and written consent was not required [28].
Completion of the questionnaire took about 25 minutes. Students were requested to complete questionnaire independently and anonymously. To guarantee the anonymity and con dentiality, students submitted the completed questionnaire in a closed envelope and put it inside a closed survey collection box.

Measurements
The students' demographic information included age, gender, academic year, informal caregiving experience for people with dementia, clinical practicum experience in geriatrics, dementia education or training experience and interest in learning more about dementia was collected.
The Alzheimer's Disease Knowledge Scale (ADKS) contains 30 true/false items to assess knowledge about dementia. The items cover seven domains of dementia knowledge: symptoms, risk factors, disease progression, assessment and diagnosis, caregiving, life impact, treatment and management [27]. The total score ranges from 1 to 30 with higher score indicating better dementia knowledge. The Chinese version of the ADKS has adequate internal consistency with a Cronbach's alpha coe cient of 0.72 [29].
The Dementia Care Attitude Scale (DCAS) consists of 8 items that measure students' attitudes towards dementia [29]. Each item is rated on a 5-point Likert scale from 1 (strong disagree) to 5 (strongly agree). The instrument includes two factors (labeled as 'Heartfelt' and 'Heartsink') and each factor consists of four items. The items of Heartfelt factor are positively worded and the items of Heartsink factor are negatively worded. A total score is calculated by reverse coding the items of Heartsink factor and summing the 8 items scored in the two factors. The minimum score value for the total score is 8 and the maximum score value is 40. A higher total score indicates more positive attitudes towards dementia. The Chinese version DCAS shows adequate internal consistency with a Cronbach's alpha of 0.71 [29].
The Approach to Advanced Dementia Care Questionnaire (ADCQ) was used to evaluate students' perception of care approach for people with dementia [30]. The choice of answers indicates either a person-centred approach or a reality-oriented approach. Personcentred approaches are scored 1, and reality-oriented approaches are scored 0. A maximum score of 13 is possible. The reliability of the Chinese version of the ADCQ was established by Lin et al with a Cronbach's alpha of 0.60 [31].

Statistical analysis
The Statistics Product and Service Solutions (SPSS) version 22.0 statistical package (IBM Corp., Armonk, New York, USA) was used for data analysis. Two research assistants checked the surveys for any missing data and entered the data into SPSS for analysis. The Kolmogorov-Smirnov test was used to determine normality of distribution of all variables. The ADKS, DCAS and ADCQ scores showed normality of distribution. An independent-sample t test, one-way ANOVA and Pearson correlation analysis were used to compare mean scores among groups. Multivariate analysis of variance (MANOVA) was used to compare scores in each content domain and total scores between medical and nursing students. Bonferroni correction was used to adjust the alpha level used to judge statistical signi cance. Multivariate regression models were conducted to examine independent variables (demographic characteristics of participants) affecting ADKS, DCAS and ADCQ scores (dependent variables). Potential factors affecting ADKS, DCAS and ADCQ scores identi ed in the literature and in the bivariate analysis were entered as independent variables in three separate multivariate regression models. Preliminary analysis was conducted to ensure no violation of the assumptions of normality, linearity, multi collinearity and homoscedasticity were present in the multivariate regression analysis. Statistical signi cance was set at p-value<0.05 in 2-tailed tests.

Demographics of medical and nursing students
Of the 1200 questionnaires distributed, 85% (n=1020) were returned. Data from participants with more than 10% missing responses were excluded (n=27); the remaining 993 (526 medical students and 467 nursing students) completed questionnaires yielded a response rate of 82.3%. The mean age of the sample was 19.9±1.8 years (range: 16-25 years). Most were female (74.4%), without informal caregiving experience for people with dementia (82.1%) and had interest in learning more about dementia (85.5%). A higher proportion of nursing students had geriatric clinical practice experience and dementia education or training than medical students (p<0.05). The demographics of medical and nursing students are summarized in Table 1.

Knowledge
The overall mean knowledge score was 19.49±2.82 (ranging from 10 to 30 out of 30) and equivalent to 65% of answers correct.
Medical students demonstrated higher mean scores in the 'symptoms' and 'life impact' content domains than nursing students (p<0.05). The lowest overall scores were achieved on the 'symptoms' content domain (53%). From analyzing the four items of 'symptoms', a majority of students (72.3% n=718) mistakenly believed that 'tremor or shaking of the hands or arms was a common symptom in people with Alzheimer's disease'. In addition, only 27.8% (n=276) of students responded correctly to the item that 'in rare cases, people have recovered from Alzheimer's disease' and 31% (n=308) gave the correct answer to the item that 'if trouble with memory and confused thinking appears suddenly, it is likely due to Alzheimer's disease'. Higher correct rates were obtained in the content domains of 'treatment and management' (72%) and 'life impact' (74%). Most students knew that 'people whose Alzheimer's disease was not yet severe could bene t from psychotherapy for depression and anxiety'. In the 'life impact' domain, most students knew the statement 'it was safe for people with Alzheimer's disease to drive, as long as they had a companion in the car at all times' was wrong. Moreover, in the 'course of disease' domain, most students knew that 'a person with Alzheimer's disease became increasing likely to fall down as the disease got worse' (Table 2).
Informal caregiving experience for people with dementia was not found to be signi cantly related to knowledge scores (p>0.05).
Knowledge scores were signi cantly different when related to gender, academic year, clinical practicum experience in geriatrics, dementia education or training, and interest in learning more about dementia (p<0.05). In particular, students in year three had the highest knowledge scores. The total score range was 0-30 with higher scores indicating better knowledge. p-value was based on Multivariate analysis of variance(MANOVA). Bonferroni correction for multiple comparisons was applied. With 8 group comparisons conducted, a corrected p-value of 0.0062 was required. *Significant after Bonferroni correction (p-value <0.0062). Table 3 shows the total mean attitude score (mean=29.92, SD=3.35) and the mean scores of two factors ('heartfelt' and 'heartsink'). There were statistically signi cant differences in the total mean attitudes score (F=18.235, p<0.01) and mean score of the 'heartfelt' factor (F=7.409, p<0.01) between nursing and medical students. In the 'heartfelt' factor, most students agreed that that 'much could be done to improve the quality of life of people with dementia (mean=4.33, SD=0.63) and caregivers of people with dementia' (mean=4.20, SD=0.72). In the 'heartsink' factor, most students disagreed that 'patients with dementia could be a drain on resources with little positive outcome'. However, lowest scores were achieved in the items 'the primary care team had a limited role to play in the care of people with dementia' (mean=2.34, SD=0.87) and 'managing dementia was more often frustrating than rewarding' (mean=3.52, SD=0.99).

Attitudes
In addition, students who were female, had dementia education or training, and showed interest in learning more about dementia showed statistically signi cant higher attitudes scores (p<0.05). There were statistically signi cant decreases in attitudes scores between students in year 1 and other academic years (p<0.05). Items of the DCAS were rated on a 5-point Likert scale with higher scores indicating better attitudes. p-value was based on Multivariate analysis of variance(MANOVA). Bonferroni correction for multiple comparisons was applied. With 3 group comparisons conducted, a corrected pvalue of 0.0167 was required. *Significant after Bonferroni correction (p-value <0.0167).

Care approach
The total mean score of care approach was 5.42±2.20, indicating most students would not apply a person-centered approach in dementia care. Over 85% (n=848) of students realized that people with dementia should be allowed to freely express themselves even in seemingly meaningless behavior. However, more than 88% of students thought that when people with dementia participated in group activities they should be told what to do and that the aim of communicating should be to help them to better handle their life situation. Moreover, in all content domains a student's major was not related to their dementia care approach (Table 4). There were no statistically signi cant differences in care approach scores between student groups. The person-centred approach answer earned a score of 1 point, whereas the reality-oriented answer earned no points. The total score range was 0-13. p-value was based on Multivariate analysis of variance(MANOVA). Bonferroni correction for multiple comparisons was applied. With 6 group comparisons conducted, a corrected pvalue of 0.0083 was required.

Factors affecting knowledge, attitudes and care approach
The results revealed that gender, major, clinical practicum experience in geriatrics, dementia education or training, and interest in learning more about dementia were factors affecting dementia knowledge (F=6.614, P=0.000), explaining 5% of the variance in knowledge. The multiple regression model using attitudes as the dependent variable was statistically signi cant and three factors were identi ed to be associated with attitudes (F=10.956, P=0.000), explaining 7% of the variance. These factors were gender, dementia education or training, and interest in learning more. Moreover, only gender and interest in learning more about dementia were factors affecting the care approach (F=2.234, P=0.023), accounting for 1% of the total variance (Table 5).

Discussion
A body of evidence con rms that the most effective dementia care is provided by health professionals through inter-professional collaboration [32][33][34]. However, few inter-disciplinary studies comparing knowledge, attitudes and care approach of dementia between medical and nursing students are available. In this study, self-reported knowledge, attitudes and care approach of dementia among medical and nursing students were explored to determine whether students were well prepared with core dementia care competence during their formal education years. The present study is signi cant in contributing to a broader understanding of the topic, providing insight into current inter-disciplinary curriculum reform of medicine and nursing undergraduate programs, and may be used to improve preparation of the future health workforce.
Our study found that the knowledge scores of nursing and medical students in China are signi cantly lower than their counterparts in Malta and US using the ADKS [26,27]. Possible reasons for this difference may be that dementia education is not included in the curricula of bachelor of medicine and bachelor of nursing [35]. In China, education sessions in gerontology for undergraduate medical and nursing students are generally covered during the third year with our results suggesting that such education has a positive in uence on students' knowledge [36]. A study by A Scerri and C Scerri [26] also found that education sessions had a positive in uence on students' knowledge about dementia. However, it should be noted that knowledge of dementia decreases after students' clinical placements in the present study. This may suggest that theory-based 'gerontology' classroom study alone is insu cient and students need to be better supported to apply theory to practice [36]. The lack of suitable clinical placements for medical and nursing students is one important barrier to dementia care education [10,25]. Strategies used to support clinical preceptors to develop dementia care competence need to be considered in the Chinese context. Clinical preceptors have a responsibility to provide a supportive clinical learning environment and involve students in a team approach to improve their care practices. Effective teaching strategies, learning resources and meaningful feedback are also vitally important to improve students' knowledge of dementia care.
The data presented here is also indicative of the lack of knowledge about dementia symptoms (correct rate 53%) and care giving issues (correct rate 62%). A similar nding was recently observed among Chinese primary health care professionals [37]. Moreover, in a survey involving 280 nursing students, questions on dementia care giving had a similar correct rate (65%) [26]. More education and training is needed to increase knowledge about dementia care giving, especially person-centered care, which is internationally recognized as best-practice. This present study also identi ed that nursing students have less knowledge about dementia symptoms and life impact compared to medical students. Yet, nurses are in an ideal position to coordinate early detection and dementia management as they constitute the majority of health professionals and have more contact and closer relationships with health service recipients. A limited knowledge of symptoms and life impact among nursing students will limit their professional roles in providing dementia care services upon graduation.
Overall, medical and nursing students expressed moderately positive attitudes towards people with dementia and their family caregivers. This result is consistent with a previous study which showed nursing students demonstrate more positive attitudes scores than medical students [26]. However, nursing students in our study had a low score in person-centered dementia care. These results con rm ndings from previous studies in China [31,37] and may indicate that person-centered dementia care is not well known in China. Much work needs to be done throughout China to promote the person-centered care approach as the gold standard in dementia care and to integrate this into government policies, dementia care guidelines, curricula of health professional education and clinical practice [31,38]. It is recommended that person-centered care education be fostered by creative, experimental and re ective processes [39]. Educators need to facilitate students' learning of dementia care by creating respectful dialogue through critical thinking, self-awareness, personal knowing and re ection using case scenarios.
The present study identi ed that clinical practicum experience in geriatrics, dementia education and training, and interest in learning dementia care were associated with better knowledge, attitudes and care approach. This nding supports previous studies in developed countries [26,40,41]. University academics, physicians, registered nurses and clinical preceptors need to work closely to involve students in dementia care to produce a better quali ed healthcare workforce. The dementia care curricula need to address the full range of appropriate knowledge and skills, and to be embedded across topics. For example, ethics, communication, evidencebased science and psychology topics [11]. One-off and short-term sessions, lectures or assessments that have few activities to engage students have limitations in improving students' dementia knowledge, attitudes and care approach. Moreover, collaborative practice and teamwork among medical and nursing students are core to developing their competence in inter-disciplinary dementia care and need to be incorporated in practicum topics throughout the medical and nursing curricula in undergraduate programs [8,42].

Strengths and limitations of the study
To our knowledge, this study is the rst to evaluate dementia knowledge, attitudes and care approach among undergraduate medical and nursing students in China. The main strengths of the present study are the participation of four universities and the large sample size, allowing a fairly good generalizability and representativeness of the ndings. The high response rate also represents students' interest in dementia. There are several limitations to this study that should be considered. As with cross-sectional design, data are collected only once over a relative short period of time, limiting the ability to draw conclusions of possible causal relationships. Scores from self-report measures can be in uenced by social desirability bias. Therefore, attitudes scores should be interpreted cautiously. In any future study, the multivariate regression models should be improved to better explain variance.

Conclusion
The present study identi ed that medical students enrolled in a 5-year undergraduate program and nursing students enrolled in a 4- year undergraduate program in China demonstrate low scores for dementia knowledge and person-centered care approach, while showing moderately positive attitudes towards dementia. Findings have implications for curriculum intervention and further research to strengthen dementia education for medical and nursing students considering the rapidly ageing population and demand for highquality care for people with dementia. First, comprehensive undergraduate dementia care education needs to be a key component in nursing and medical curricula that include standalone dementia care topics and/or integrated dementia care content. Second, interprofessional dementia education needs to be established in the nursing and medical curricula to enable nursing and medical students to learn and share dementia knowledge, skills and experiences. This kind of education will prepare them to provide interdisciplinary care for people with dementia. Third, studies on effective teaching and learning strategies to improve inter-professional learning and inter-disciplinary collaboration between nursing and medical students in dementia care education are needed.