- Email: [email protected]
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine: survey and cluster analysis
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine: survey and cluster analysis M.F. Chan, E. Mok, Y.S. Wong, T.F. Tong, M.C. Day, C.K.Y. Tang, D.H.C. Wong School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, PR China
SUMMARY. Objectives: To determine whether within a cohort of Hong Kong out-patients definable subtypes exist based on their attitudes to traditional Chinese medicine and Western medicine. Design: Data were collected using a structured questionnaire. Setting: The sample of 503 subjects was recruited at two local outpatient clinics in Hong Kong. Main outcome measures: The study employs demographic variables, illness status, the Chinese-Western Medical Belief Scale, trust of physicians and subjects’ preferences on consultation fees, attitude of health care professional, efficacy of service and waiting time during consultation. Results: A cluster analysis yielded three clusters based on their attitudes towards traditional Chinese and Western medicine. One cluster, 24% of the sample, is noted for being older, poorer, more likely to be female and to have chronic conditions; they are sceptical of western physicians. The second cluster (63% of the sample) is younger and have considerably more belief in Western than traditional Chinese medicine. The third group (14%) is intermediate in age and is noted for a marked faith in both forms of medicine. Conclusions: A clear profile of these attitudes of Hong Kong Chinese may benefit health care professionals in making appropriate patient–doctor relationships and planning patient care. © 2003 Elsevier Science Ltd. All rights reserved.
INTRODUCTION
Esther Mok PhD, School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, PR China. Tel.: +852 2766 6410; Fax: +852 2364 9663; E-mail: hsemok@ inet.polyu.edu.hk
Traditional Chinese Medicine (TCM) has been practiced in China for over 2000 years. TCM has been widely used in Hong Kong and the Peoples’ Republic of China (PRC).1 Before the return of Hong Kong to PRC sovereignty in 1997, there was limited interest in developing the potential of TCM and very little government attention or support. However, after 1997, the Chief Executive of the Hong Kong Government Special Administrative Region (HKSAR) supported TCM research.2 He emphasised the potential of the HKSAR to develop into an international centre for the manufacture and trade of TCM, and for research, information and training in the use of
Complementary Therapies in Medicine (2003), 11, 103–109 doi:10.1016/S0965-2299(03)00044-X
© 2003 Elsevier Science Ltd. All rights reserved.
TCM.3 It has been important in the delivery of health care services.4 As a result, a consultation document on Health Care Reform proposed the introduction TCM provision into the public health care system. The Chinese Medicine Council of Hong Kong (CMCHK) was approved in 2001. Today, 7707 TCM physicians are registered as listed Chinese Medicine Practitioners (CMP) in Hong Kong, serving a population of almost 6.7 million.2 At the same time, many TCM clinics were established. This marked
103
Funding Central Research Grants (G-T328) from the Hong Kong Polytechnic University.
104
Complementary Therapies in Medicine
the beginning of a new era in Hong Kong with TCM an officially recognised profession. While TCM services may be widely used by Hong Kong people, Lau pointed out that for most patients, it is often some “trigger” that brings patients to their TCM healer.5 Lau and Yu showed that Hong Kong adolescents’ sought TCM for fall-related injuries, while those with skin disease used both TCM and Western medicine (WM).6 Lam suggested that the health attitudes of patients from different ethnic backgrounds might also affect their choice of TCM or WM.7 Understanding the factors that influence patients’ attitudes is helpful in identifying reasons for differences in seeking preference and utilisation patterns towards TCM and WM. The reasons are complex and little research was found locally to explore these relations. Therefore, it is pertinent to identify patterns of usage, attitudes and health perceptions towards TCM and WM, in order to understand patterns of TCM and WM utilisation. For the purpose of this research TCM refers to herbal medicine only, as this is the way it is generally practised in Hong Kong. The aims of the present study were to: (1) identify profiles of various Hong Kong Chinese based on their background variables, attitudes and health perceptions to TCM and WM. (2) investigate whether background variables, attitudes and health perceptions vary between Hong Kong Chinese in the different profiles groups.
METHODS Sample The study was conducted in two outpatient clinics (A and B) in Hong Kong that provides both TCM and WM services. A random sample of 503 patients (clinic A, n = 303; clinic B, n = 200) was recruited from July to October 2001. The inclusion criteria were: • ethnic Chinese, living in Hong Kong; • adult aged 18 or above; and • literate in Chinese and Cantonese. The characteristics of the subjects are detailed in Table 1.
Procedure and instrument A researcher approached patients waiting for consultation in the two clinics, briefly explained the purpose of the study and described the study procedures. Each participant was asked to complete a questionnaire while in the clinic. Demographic data from all participants were obtained. This information included age, gender, monthly family income (1 =HK$ 40,000; £1 = HK$ 12.4), education level (1 = never been to school, 5 = college/University), length of diagnosis (1 = <1 month, 6 = >36 months), travelling time to
Table 1 Characteristics of subjects recruited to survey of Hong Kong patients n (total = 503)
(%)
Gender Male Female
165 336
(32.93) (67.07)
Age (years) 18–19 20–29 30–39 40–49 50+
37.21 30 215 46 67 140
[18.33]a (6.02) (43.17) (9.24) (13.45) (28.11)
Education level Post-secondary or below College or above
223 280
(44.33) (55.67)
Family income (HK$/month) 5000 or below 5001–10,000 10,001–20,000 20,001–30,000 30,001–40,000 40,001+
65 77 151 107 39 52
(13.24) (15.68) (30.75) (21.79) (7.94) (10.69)
History of hospitalisation Yes No
219 284
(43.54) (56.46)
£1 = HK$ 12.3. a Mean [S.D.].
the clinic (1 = <30 min, 3 = >60 min), history of hospitalisation (yes or no), illness status including discomfort (1 = not at all, 5 = very much), severity (0 = no, 5 = very serious), and type (1 = chronic, 3 = acute) of present illness. Subjects’ attitudes towards TCM and WM were measured using the Chinese-Western Medical Beliefs (CWMB) Scale developed by Liang.8 The CWMB questionnaire contains 22 items measured with a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Four subscale scores related to TCM can be obtained from the questionnaire (12 items), restorative effects (3 items), side effects (3 items), therapeutic effects (3 items) and priority (3 items). These reflect an overall belief in the superiority of TCM. Two subscale scores related to WM can also be obtained from the questionnaire (10 items), efficiency of the treatment and therapeutic effects (3 items), and better quality and facilities (7 items). These reflect an overall belief in the superiority of WM. Two additional questions with a 5-point Likert scale were used. The first measured participants’ trust of TCM and WM physicians (1 = not at all, 5 = very much). The second measured their preference pattern (1 = lowest, 5 = highest) on four factors: consultation fees, attitude of health care professional, efficacy of service and waiting time during consultation.
STATISTICAL METHODS Factor analysis was used to identify number of factors of the CWMB Scale and Cronbach’s alpha
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
coefficients were used to examine the internal reliability of this scale. A cluster analysis was used to group subjects based on five main variables: (1) demographic data, (2) current illness status, (3) CWMB scores, (4) trust scores, and (5) preferences ranking. The K-means method, nearest centroid sorting, was used for this study.9,11 This is a conservative method suggested by Anderberg9 and SPSS10 if sample sizes are more than 200. The squared standardised Euclidian distance was chosen as the measure of similarity.7 The analysis used standard scores (z) based on 467 out of 503 subjects for the five main variables described previously. Differences in means and proportions between clusters on the above factors were tested for statistical significance by ANOVA and χ2 -test, respectively. In case of a significant F-statistic at P < 0.05, post hoc comparisons were performed with the Scheffé test, one of the most conservative with respect to Type I errors. Differences between cluster groups were delineated both descriptively and statistically.
RESULTS Cluster analysis Cluster analysis yielded three clusters, consisting of 23.8% (n = 111), 62.3% (n = 291) and 13.9% (n = 65) of the subjects, respectively (Table 2). The three subgroups of subjects (clusters) were formed on the basis of the similarity of the subjects’ responses to the questionnaire. The three subject profiles were compared with regard to five main variables. The variables included: (1) demographic data (i.e. age, gender, education level, family income, length of diagnosis, travelling time to the clinic, history of hospitalisation); (2) current illness status (i.e. discomfort, severity and type); (3) CWMB scores; (4) trust scores; and (5) preferences ranking order on four factors (i.e. consultation fees, efficacy of service, waiting time, attitude of health care professionals). Profiles are depicted in Figure 1a–f.
Demographic data Significant differences were noted for all demographic variables (Table 2). The subjects in cluster 1 were older (mean = 55.5 years) than subjects in the other two clusters (cluster 2: mean = 26.0; cluster 3: mean = 48.2). Further, the subjects in cluster 1 showed less favourable scores on education level (mean = 2.3) and family income (mean = 2.4), compared with subjects in the other two clusters. The subjects in cluster 1 had a longer length of diagnosis (mean = 2.9) than subjects in cluster 2 (mean = 1.1, P < 0.05) and cluster 3 (mean = 1.6, P < 0.05). More subjects in clusters 1 (60.4%) had a history of hospitalisation compared to subjects in cluster 2 (34.7%, P < 0.001). Moreover, the proportion of male subjects in cluster 1 (21.6%) was
105
smaller than in the other two subject profiles clusters (cluster 2: 37.8%; cluster 3: 38.5%) (Fig. 1a).
Illness status data The subjects in cluster 1 showed less favourable scores on chronicity (mean = 1.6) and less favourable scores on discomfort (mean = 3.3) and severity (mean = 3.0), compared with subjects in the others two clusters (Fig. 1b).
The Chinese-Western Medical Beliefs Scales Two factors were generated accounting for 32.1% of the variance (Table 3). Factor 1 accounted for 16.2% of the variance (factor loading ranged from 0.40 to 0.63) and reflected concerns to subjects’ beliefs towards TCM. Factor 2 contributed 15.9% of the variance (factor loading ranged from 0.48 to 0.68) and focused on subjects’ beliefs towards WM. Overall, the Cronbach’s alpha for 22 items, was 0.75. The internal reliability of the subscales for TCM and WM were 0.77 and 0.78, respectively. Reliability scores for the CWMB instruments were found to be moderate on average. On the CWMB question, significant differences among the three clusters were noted, and subjects in cluster 3 evidencing higher both on the TCM superior scores (mean = 4.1) and the WM superior scores (mean = 4.2) compared with subjects in the other two clusters (Table 2). For each TCM sub-scores, subjects in cluster 3 exhibited higher scores than subjects in other two clusters in all sub-scores (Fig. 1c). For each WM sub-scores, subjects in cluster 3 exhibited higher scores than subjects in other two clusters in all sub-scores (Fig. 1d).
Trust data In Table 2, the trust scores of TCM physicians, subjects in cluster 3 exhibited more favourable scores (mean = 3.8) than subjects in clusters 1 (mean = 3.4) and 2 (mean = 3.3). For the trust scores of WM physicians, subjects in cluster 1 exhibited lowest scores (mean = 3.5) than subjects in cluster 2 (mean = 3.9) and cluster 3 (mean = 3.9) (Fig. 1e).
Preferences ranking data Comparisons were made between the subjects in the three clusters of their preferences order on four factors: consultation fees, attitude of health care professionals, efficacy of service and waiting time (Table 2). For all subjects, efficacy of service ranked as the top priority, followed by consultation fees, then waiting time and attitudes of health care professionals ranked lowest. There were significant differences in the scores of subjects’ preference data between the three clusters for consultation fees (P = 0.002),
106
Table 2
Comparison between subjects in the three clusters
Demographic Age Education levela Family incomeb Traveling time to clinicc Length of diagnosisd Gender (number of males in percent) History of hospitalisation (number of yes in percent) Illness status Discomfort of present illnessf Severity of present illnessg Type of present illnessh Chinese-Western Medical Beliefs Scalei TCM restorative effect TCM fewer side effect TCM therapeutic effect TCM prioritised Sub-total of TCM superior scores WM quality and facilities WM efficiency of the treatment and therapeutic effects Sub-total of WM superior scores Trustj TCM physician WM physician Preferences ranking Consultation fees Attitude of the health care professional Efficacy of the service Waiting time
Cluster 1 (n = 111)
Cluster 2 (n = 291)
Cluster 3 (n = 65)
ANOVA significance
Mean
(S.D.)
Mean
(S.D.)
Mean
(S.D.)
55.48 2.26 2.35 2.15 2.86
(14.76) (0.96) (1.17) (0.62) (1.86)
25.98 4.73 3.82 2.43 1.43
(9.88) (0.73) (1.31) (0.66) (1.14)
48.20 2.77 2.54 2.31 2.17
(14.92) (1.18) (1.30) (0.64) (1.62)
P P P P P P P
21.62 60.36
37.80 34.71
38.46 53.85
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 = 0.007e < 0.001e
Scheffe´ test
Total (n = 467) Mean
(S.D.)
36.08 3.87 3.29 2.35 1.87
(17.79) (1.41) (1.44) (0.66) (1.53)
A, C
3.07 2.81 2.08
(0.95) (0.97) (0.98)
A, B, C B, C B, C A, B, C A, B, C B, C B, C B, C
3.66 3.78 3.40 2.89 3.45 3.68 3.77 3.71
(0.52) (0.58) (0.69) (0.79) (0.45) (0.59) (0.59) (0.49)
A, A, A, A A, A, A,
B, C B, C C B, C B C
3.30 2.97 1.56
(0.90) (1.13) (0.86)
3.04 2.80 2.33
(0.94) (0.86) (0.93)
2.86 2.62 1.86
(1.00) (1.13) (0.98)
P = 0.007 P = 0.060 P < 0.001
3.59 3.95 3.19 3.04 3.48 3.54 3.66 3.60
(0.52) (0.46) (0.68) (0.74) (0.35) (0.55) (0.55) (0.43)
3.56 3.59 3.35 2.66 3.31 3.63 3.67 3.65
(0.47) (0.54) (0.63) (0.67) (0.37) (0.57) (0.56) (0.46)
4.20 4.33 3.96 3.69 4.07 4.14 4.25 4.18
(0.43) (0.53) (0.67) (0.84) (0.41) (0.49) (0.57) (0.45)
P P P P P P P P
3.41 3.50
(0.91) (0.80)
3.30 3.90
(0.76) (0.65)
3.82 3.86
(0.93) (0.93)
P < 0.001 P < 0.001
B, C A, B
3.40 3.80
(0.84) (0.75)
3.50 2.59 4.50 2.95
(0.99) (0.87) (0.85) (1.00)
3.22 2.92 4.88 2.82
(0.88) (0.89) (0.40) (0.87)
3.57 2.86 4.72 2.85
(0.85) (0.95) (0.63) (0.87)
P P P P
= 0.002 = 0.004 < 0.001 = 0.443
A, C A A, B
3.33 2.84 4.76 2.85
(0.91) (0.90) (0.59) (0.90)
< < < < < < < <
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
A, B
A: significant difference of means between subjects in cluster 1 and cluster 2 (P < 0.05); B: significant difference of means between subjects in cluster 1 and cluster 3 (P < 0.05); C: significant difference of means between subjects in cluster 2 and cluster 3 (P < 0.05). a Scores are 1 (never been to school), 2 (primary), 3 (secondary), 4 (post-secondary/vocational), 5 (college/university). b 1 (60 min). d 1 (<1 month), 2 (1–3 months), 3 (4–6 months), 4 (7–12 months), 5 (13–36 months), 6 (>36 months). e Chi-square test. f Scores could range from 1 (not at all) to 5 (very much). g Scores could range from 0 (not at all) to 5 (very serious). h Scores could range from 1 (chronic) to 3 (acute). i Scores could range from 1 (very disagreed) to 5 (very agreed). j Scores could range from 1 (not at all) to 5 (very much).
Complementary Therapies in Medicine
Variables
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
Fig. 1 Cluster analysis of subjects’ demographic data and attitudes.
107
108
Complementary Therapies in Medicine Table 3
Factor analysis of Chinese-Western belief itemsa
Items
Factor loading
Factor 1b : Chinese medicine superior beliefsc Compared with WM, the therapeutic effect of TCM for chronic diseases is greater Compared with WM, the therapeutic effect of TCM for medical diseases is greater TCM treats the causative agent, while WM only relieves symptomatic distress Compared with WM, TCM have fewer side effects Compared with WM, TCM cause less gastrointestinal side effects Compared with WM, TCM are slower and less potent in their action WM is used only when all other treatments in TCM have failed While relying mainly on TCM, when necessary WM may be used as an adjunct TCM should be prioritised over WM when they are equally effective TCM can cure diseases and also promote health TCM can have a restorative effect which can promote health over time TCM can improve one’s constitution
0.44 0.46 0.56 0.59 0.63 0.61 0.53 0.60 0.60 0.40 0.46 0.45
Factor 2d : Western medicine superior beliefse Compared with TCM, WM can diagnose illness more accurately Compared with TCM, WM has better facilities available Compared with TCM, the overall quality of WM is better Compared with TCM, the process of education and training in WM is more rigorous Compared with TCM, WM has a more scientific approach The labelling of content of WM is more accurate than TCM Compared with TCM, WM is specialised towards particular diseases Compared with TCM, WM acts more quickly Compared with TCM, the therapeutic effect of WM for acute disease is greater Compared with TCM, the therapeutic effect of WM is more apparent
0.62 0.56 0.59 0.65 0.68 0.54 0.48 0.49 0.50 0.64
Reliability for all 22 items, α = 0.75. Variance percentage: 16.2. c Chinese medicine belief items, α = 0.77. d Variance percentage: 15.9. e Western medicine belief items, α = 0.78. a
b
attitude of health care professionals (P = 0.004) and efficacy of service (P < 0.001), but no significant differences in scores for waiting time (P = 0.44) (Fig. 1f ).
DISCUSSION This study aimed to explore some major health beliefs of a group of Hong Kong Chinese. This research identified three clusters of attitudes among Hong Kong Chinese on the basis of five aspects related to health care (i.e. demographics variables, illness status, the CWMB Scale, trust scale and preferences ranking). People in cluster 1 tend to be older (mean 55 years), are predominantly female, and not particularly well educated or wealthy. They have chronic conditions which they perceive as more serious with greater discomfort. They are more likely to have spent time in hospital. Their belief in Western medicine physicians is lower than other groups. Their chronic illness may have made them somewhat sceptical. They represent about a quarter of the sample. People in cluster 2 are clearly different: they are more likely to be younger (mean 26 years), and are likely to be male. They have had longer education, they earn more, and they have more short-lived conditions on the whole. This group have the greatest faith in Western physicians and the least belief in TCM. They represent more than half of this sample.
The outstanding feature of patients in the smallest group, cluster 3 is their trust and belief in medicine— both Western medicine and TCM. In other respects they are rather similar in sociodemographic status and aspects of their illness to cluster 1, though they are rather younger (mean 48 years) and more likely to be male. This study shows the CWMB scores of some Hong Kong Chinese, on TCM (range: 3.31–4.07) and WM (range: 3.60–4.18) range from moderate to great, while their trust scores on TCM (range: 3.30–3.82) and WM (range: 3.50–3.90) physicians were moderate. These might partly explain why Chinese choose TCM or WM for a specific type of illness that they are suffering from.5–8,12,13 By understanding patients’ views on WM and TCM, physicians may begin discussions about TCM or WM and consider how they should be advising their patients. It may also improve the patient–doctor relationship and patient care if physicians can show an understanding of patient attitudes and beliefs while being proactive in discussing whether their patients are using TCM or WM. Some caution must be exercised in generalising the results of the present study. One potential limitation is that data for this study were obtained from a self-report questionnaire and open to a possible response bias.14 The results are limited due to the use of a rather small sample collected from only two clinics. The sample contains an overrepresentation of subjects between the ages of 20–29 (43.17%)
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
and 50+ (28.11%). No adolescents (below 18) were included in the sample. Future research should evaluate process and outcome of TCM and WM focusing on longitudinal designs and larger samples of patients.
REFERENCES 1. Wang DIC, Leung JC, Wu IC, Gao N. Biotechnology and Hong Kong. In: Berger S, Lester R, eds. Made in Hong Kong. Hong Kong: Oxford University Press, 1997. p. 249. 2. CMCHK http://www.cmchk.org.hk. The Chinese Medicine Council of Hong Kong, 2001. 3. Tung CH. Building Hong Kong for a new era. Policy address for Hong Kong Government Special Administrative Region, the PRC, 8 October 1997. 4. Tsang JC, Lo YL. Biotechnology development in Hong Kong: a renewed scientific interest in traditional Chinese medicine. Biochem Educ 1998; 26: 301–303. 5. Lau BWK. When and how patients seek medical help? An exploration in Hong Kong. Hong Kong Pract 1996; 18: 109–115.
109
6. Lau JTF, Yu A. The choice between Chinese medicine and Western medicine practitioners by Hong Kong adolescents. Am J Chin Med 2000; 28(1): 131–139. 7. Lam TP. Strengths and weaknesses of traditional Chinese medicine and Western medicine in the eyes of some Hong Kong Chinese. J Epidemiol Commun Health 2001; 55: 762–765. 8. Liang C. The development and examination of the Chinese-Western Medical Beliefs Scale. Nurs Res (in Chinese version) 1999; 7: 445–458. 9. Anderberg MR. Cluster analysis for applications. New York: Academic Press, 1973. 10. SPSS Base 7.5 Applications guide, SPSS Inc., 1997. 11. Morris R, Blashfield R, Satz P. Neuropsychology and cluster analysis: potentials and problems. J Clin Neuropsychol 1981; 3: 79–99. 12. Wong TW, Wong SL, Donnan SPB. Traditional Chinese medicine and Western medicine in Hong Kong: a comparison of the consultation processes and side effects. J Hong Kong Med Assoc 1993; 45(4): 278–284. 13. Wong TW, Wong SL, Donnan SPB. A study of traditional Chinese medicine practitioners in Hong Kong. J Hong Kong Med Assoc 1993; 45(4): 285–290. 14. Polit DF, Hungler BP. Nursing research: Principles and methods, 5th edn. Philadelphia: Lippincott, 1995.
SUMMARY. Objectives: To determine whether within a cohort of Hong Kong out-patients definable subtypes exist based on their attitudes to traditional Chinese medicine and Western medicine. Design: Data were collected using a structured questionnaire. Setting: The sample of 503 subjects was recruited at two local outpatient clinics in Hong Kong. Main outcome measures: The study employs demographic variables, illness status, the Chinese-Western Medical Belief Scale, trust of physicians and subjects’ preferences on consultation fees, attitude of health care professional, efficacy of service and waiting time during consultation. Results: A cluster analysis yielded three clusters based on their attitudes towards traditional Chinese and Western medicine. One cluster, 24% of the sample, is noted for being older, poorer, more likely to be female and to have chronic conditions; they are sceptical of western physicians. The second cluster (63% of the sample) is younger and have considerably more belief in Western than traditional Chinese medicine. The third group (14%) is intermediate in age and is noted for a marked faith in both forms of medicine. Conclusions: A clear profile of these attitudes of Hong Kong Chinese may benefit health care professionals in making appropriate patient–doctor relationships and planning patient care. © 2003 Elsevier Science Ltd. All rights reserved.
INTRODUCTION
Esther Mok PhD, School of Nursing, The Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, PR China. Tel.: +852 2766 6410; Fax: +852 2364 9663; E-mail: hsemok@ inet.polyu.edu.hk
Traditional Chinese Medicine (TCM) has been practiced in China for over 2000 years. TCM has been widely used in Hong Kong and the Peoples’ Republic of China (PRC).1 Before the return of Hong Kong to PRC sovereignty in 1997, there was limited interest in developing the potential of TCM and very little government attention or support. However, after 1997, the Chief Executive of the Hong Kong Government Special Administrative Region (HKSAR) supported TCM research.2 He emphasised the potential of the HKSAR to develop into an international centre for the manufacture and trade of TCM, and for research, information and training in the use of
Complementary Therapies in Medicine (2003), 11, 103–109 doi:10.1016/S0965-2299(03)00044-X
© 2003 Elsevier Science Ltd. All rights reserved.
TCM.3 It has been important in the delivery of health care services.4 As a result, a consultation document on Health Care Reform proposed the introduction TCM provision into the public health care system. The Chinese Medicine Council of Hong Kong (CMCHK) was approved in 2001. Today, 7707 TCM physicians are registered as listed Chinese Medicine Practitioners (CMP) in Hong Kong, serving a population of almost 6.7 million.2 At the same time, many TCM clinics were established. This marked
103
Funding Central Research Grants (G-T328) from the Hong Kong Polytechnic University.
104
Complementary Therapies in Medicine
the beginning of a new era in Hong Kong with TCM an officially recognised profession. While TCM services may be widely used by Hong Kong people, Lau pointed out that for most patients, it is often some “trigger” that brings patients to their TCM healer.5 Lau and Yu showed that Hong Kong adolescents’ sought TCM for fall-related injuries, while those with skin disease used both TCM and Western medicine (WM).6 Lam suggested that the health attitudes of patients from different ethnic backgrounds might also affect their choice of TCM or WM.7 Understanding the factors that influence patients’ attitudes is helpful in identifying reasons for differences in seeking preference and utilisation patterns towards TCM and WM. The reasons are complex and little research was found locally to explore these relations. Therefore, it is pertinent to identify patterns of usage, attitudes and health perceptions towards TCM and WM, in order to understand patterns of TCM and WM utilisation. For the purpose of this research TCM refers to herbal medicine only, as this is the way it is generally practised in Hong Kong. The aims of the present study were to: (1) identify profiles of various Hong Kong Chinese based on their background variables, attitudes and health perceptions to TCM and WM. (2) investigate whether background variables, attitudes and health perceptions vary between Hong Kong Chinese in the different profiles groups.
METHODS Sample The study was conducted in two outpatient clinics (A and B) in Hong Kong that provides both TCM and WM services. A random sample of 503 patients (clinic A, n = 303; clinic B, n = 200) was recruited from July to October 2001. The inclusion criteria were: • ethnic Chinese, living in Hong Kong; • adult aged 18 or above; and • literate in Chinese and Cantonese. The characteristics of the subjects are detailed in Table 1.
Procedure and instrument A researcher approached patients waiting for consultation in the two clinics, briefly explained the purpose of the study and described the study procedures. Each participant was asked to complete a questionnaire while in the clinic. Demographic data from all participants were obtained. This information included age, gender, monthly family income (1 =
Table 1 Characteristics of subjects recruited to survey of Hong Kong patients n (total = 503)
(%)
Gender Male Female
165 336
(32.93) (67.07)
Age (years) 18–19 20–29 30–39 40–49 50+
37.21 30 215 46 67 140
[18.33]a (6.02) (43.17) (9.24) (13.45) (28.11)
Education level Post-secondary or below College or above
223 280
(44.33) (55.67)
Family income (HK$/month) 5000 or below 5001–10,000 10,001–20,000 20,001–30,000 30,001–40,000 40,001+
65 77 151 107 39 52
(13.24) (15.68) (30.75) (21.79) (7.94) (10.69)
History of hospitalisation Yes No
219 284
(43.54) (56.46)
£1 = HK$ 12.3. a Mean [S.D.].
the clinic (1 = <30 min, 3 = >60 min), history of hospitalisation (yes or no), illness status including discomfort (1 = not at all, 5 = very much), severity (0 = no, 5 = very serious), and type (1 = chronic, 3 = acute) of present illness. Subjects’ attitudes towards TCM and WM were measured using the Chinese-Western Medical Beliefs (CWMB) Scale developed by Liang.8 The CWMB questionnaire contains 22 items measured with a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Four subscale scores related to TCM can be obtained from the questionnaire (12 items), restorative effects (3 items), side effects (3 items), therapeutic effects (3 items) and priority (3 items). These reflect an overall belief in the superiority of TCM. Two subscale scores related to WM can also be obtained from the questionnaire (10 items), efficiency of the treatment and therapeutic effects (3 items), and better quality and facilities (7 items). These reflect an overall belief in the superiority of WM. Two additional questions with a 5-point Likert scale were used. The first measured participants’ trust of TCM and WM physicians (1 = not at all, 5 = very much). The second measured their preference pattern (1 = lowest, 5 = highest) on four factors: consultation fees, attitude of health care professional, efficacy of service and waiting time during consultation.
STATISTICAL METHODS Factor analysis was used to identify number of factors of the CWMB Scale and Cronbach’s alpha
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
coefficients were used to examine the internal reliability of this scale. A cluster analysis was used to group subjects based on five main variables: (1) demographic data, (2) current illness status, (3) CWMB scores, (4) trust scores, and (5) preferences ranking. The K-means method, nearest centroid sorting, was used for this study.9,11 This is a conservative method suggested by Anderberg9 and SPSS10 if sample sizes are more than 200. The squared standardised Euclidian distance was chosen as the measure of similarity.7 The analysis used standard scores (z) based on 467 out of 503 subjects for the five main variables described previously. Differences in means and proportions between clusters on the above factors were tested for statistical significance by ANOVA and χ2 -test, respectively. In case of a significant F-statistic at P < 0.05, post hoc comparisons were performed with the Scheffé test, one of the most conservative with respect to Type I errors. Differences between cluster groups were delineated both descriptively and statistically.
RESULTS Cluster analysis Cluster analysis yielded three clusters, consisting of 23.8% (n = 111), 62.3% (n = 291) and 13.9% (n = 65) of the subjects, respectively (Table 2). The three subgroups of subjects (clusters) were formed on the basis of the similarity of the subjects’ responses to the questionnaire. The three subject profiles were compared with regard to five main variables. The variables included: (1) demographic data (i.e. age, gender, education level, family income, length of diagnosis, travelling time to the clinic, history of hospitalisation); (2) current illness status (i.e. discomfort, severity and type); (3) CWMB scores; (4) trust scores; and (5) preferences ranking order on four factors (i.e. consultation fees, efficacy of service, waiting time, attitude of health care professionals). Profiles are depicted in Figure 1a–f.
Demographic data Significant differences were noted for all demographic variables (Table 2). The subjects in cluster 1 were older (mean = 55.5 years) than subjects in the other two clusters (cluster 2: mean = 26.0; cluster 3: mean = 48.2). Further, the subjects in cluster 1 showed less favourable scores on education level (mean = 2.3) and family income (mean = 2.4), compared with subjects in the other two clusters. The subjects in cluster 1 had a longer length of diagnosis (mean = 2.9) than subjects in cluster 2 (mean = 1.1, P < 0.05) and cluster 3 (mean = 1.6, P < 0.05). More subjects in clusters 1 (60.4%) had a history of hospitalisation compared to subjects in cluster 2 (34.7%, P < 0.001). Moreover, the proportion of male subjects in cluster 1 (21.6%) was
105
smaller than in the other two subject profiles clusters (cluster 2: 37.8%; cluster 3: 38.5%) (Fig. 1a).
Illness status data The subjects in cluster 1 showed less favourable scores on chronicity (mean = 1.6) and less favourable scores on discomfort (mean = 3.3) and severity (mean = 3.0), compared with subjects in the others two clusters (Fig. 1b).
The Chinese-Western Medical Beliefs Scales Two factors were generated accounting for 32.1% of the variance (Table 3). Factor 1 accounted for 16.2% of the variance (factor loading ranged from 0.40 to 0.63) and reflected concerns to subjects’ beliefs towards TCM. Factor 2 contributed 15.9% of the variance (factor loading ranged from 0.48 to 0.68) and focused on subjects’ beliefs towards WM. Overall, the Cronbach’s alpha for 22 items, was 0.75. The internal reliability of the subscales for TCM and WM were 0.77 and 0.78, respectively. Reliability scores for the CWMB instruments were found to be moderate on average. On the CWMB question, significant differences among the three clusters were noted, and subjects in cluster 3 evidencing higher both on the TCM superior scores (mean = 4.1) and the WM superior scores (mean = 4.2) compared with subjects in the other two clusters (Table 2). For each TCM sub-scores, subjects in cluster 3 exhibited higher scores than subjects in other two clusters in all sub-scores (Fig. 1c). For each WM sub-scores, subjects in cluster 3 exhibited higher scores than subjects in other two clusters in all sub-scores (Fig. 1d).
Trust data In Table 2, the trust scores of TCM physicians, subjects in cluster 3 exhibited more favourable scores (mean = 3.8) than subjects in clusters 1 (mean = 3.4) and 2 (mean = 3.3). For the trust scores of WM physicians, subjects in cluster 1 exhibited lowest scores (mean = 3.5) than subjects in cluster 2 (mean = 3.9) and cluster 3 (mean = 3.9) (Fig. 1e).
Preferences ranking data Comparisons were made between the subjects in the three clusters of their preferences order on four factors: consultation fees, attitude of health care professionals, efficacy of service and waiting time (Table 2). For all subjects, efficacy of service ranked as the top priority, followed by consultation fees, then waiting time and attitudes of health care professionals ranked lowest. There were significant differences in the scores of subjects’ preference data between the three clusters for consultation fees (P = 0.002),
106
Table 2
Comparison between subjects in the three clusters
Demographic Age Education levela Family incomeb Traveling time to clinicc Length of diagnosisd Gender (number of males in percent) History of hospitalisation (number of yes in percent) Illness status Discomfort of present illnessf Severity of present illnessg Type of present illnessh Chinese-Western Medical Beliefs Scalei TCM restorative effect TCM fewer side effect TCM therapeutic effect TCM prioritised Sub-total of TCM superior scores WM quality and facilities WM efficiency of the treatment and therapeutic effects Sub-total of WM superior scores Trustj TCM physician WM physician Preferences ranking Consultation fees Attitude of the health care professional Efficacy of the service Waiting time
Cluster 1 (n = 111)
Cluster 2 (n = 291)
Cluster 3 (n = 65)
ANOVA significance
Mean
(S.D.)
Mean
(S.D.)
Mean
(S.D.)
55.48 2.26 2.35 2.15 2.86
(14.76) (0.96) (1.17) (0.62) (1.86)
25.98 4.73 3.82 2.43 1.43
(9.88) (0.73) (1.31) (0.66) (1.14)
48.20 2.77 2.54 2.31 2.17
(14.92) (1.18) (1.30) (0.64) (1.62)
P P P P P P P
21.62 60.36
37.80 34.71
38.46 53.85
< 0.001 < 0.001 < 0.001 < 0.001 < 0.001 = 0.007e < 0.001e
Scheffe´ test
Total (n = 467) Mean
(S.D.)
36.08 3.87 3.29 2.35 1.87
(17.79) (1.41) (1.44) (0.66) (1.53)
A, C
3.07 2.81 2.08
(0.95) (0.97) (0.98)
A, B, C B, C B, C A, B, C A, B, C B, C B, C B, C
3.66 3.78 3.40 2.89 3.45 3.68 3.77 3.71
(0.52) (0.58) (0.69) (0.79) (0.45) (0.59) (0.59) (0.49)
A, A, A, A A, A, A,
B, C B, C C B, C B C
3.30 2.97 1.56
(0.90) (1.13) (0.86)
3.04 2.80 2.33
(0.94) (0.86) (0.93)
2.86 2.62 1.86
(1.00) (1.13) (0.98)
P = 0.007 P = 0.060 P < 0.001
3.59 3.95 3.19 3.04 3.48 3.54 3.66 3.60
(0.52) (0.46) (0.68) (0.74) (0.35) (0.55) (0.55) (0.43)
3.56 3.59 3.35 2.66 3.31 3.63 3.67 3.65
(0.47) (0.54) (0.63) (0.67) (0.37) (0.57) (0.56) (0.46)
4.20 4.33 3.96 3.69 4.07 4.14 4.25 4.18
(0.43) (0.53) (0.67) (0.84) (0.41) (0.49) (0.57) (0.45)
P P P P P P P P
3.41 3.50
(0.91) (0.80)
3.30 3.90
(0.76) (0.65)
3.82 3.86
(0.93) (0.93)
P < 0.001 P < 0.001
B, C A, B
3.40 3.80
(0.84) (0.75)
3.50 2.59 4.50 2.95
(0.99) (0.87) (0.85) (1.00)
3.22 2.92 4.88 2.82
(0.88) (0.89) (0.40) (0.87)
3.57 2.86 4.72 2.85
(0.85) (0.95) (0.63) (0.87)
P P P P
= 0.002 = 0.004 < 0.001 = 0.443
A, C A A, B
3.33 2.84 4.76 2.85
(0.91) (0.90) (0.59) (0.90)
< < < < < < < <
0.001 0.001 0.001 0.001 0.001 0.001 0.001 0.001
A, B
A: significant difference of means between subjects in cluster 1 and cluster 2 (P < 0.05); B: significant difference of means between subjects in cluster 1 and cluster 3 (P < 0.05); C: significant difference of means between subjects in cluster 2 and cluster 3 (P < 0.05). a Scores are 1 (never been to school), 2 (primary), 3 (secondary), 4 (post-secondary/vocational), 5 (college/university). b 1 (
Complementary Therapies in Medicine
Variables
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
Fig. 1 Cluster analysis of subjects’ demographic data and attitudes.
107
108
Complementary Therapies in Medicine Table 3
Factor analysis of Chinese-Western belief itemsa
Items
Factor loading
Factor 1b : Chinese medicine superior beliefsc Compared with WM, the therapeutic effect of TCM for chronic diseases is greater Compared with WM, the therapeutic effect of TCM for medical diseases is greater TCM treats the causative agent, while WM only relieves symptomatic distress Compared with WM, TCM have fewer side effects Compared with WM, TCM cause less gastrointestinal side effects Compared with WM, TCM are slower and less potent in their action WM is used only when all other treatments in TCM have failed While relying mainly on TCM, when necessary WM may be used as an adjunct TCM should be prioritised over WM when they are equally effective TCM can cure diseases and also promote health TCM can have a restorative effect which can promote health over time TCM can improve one’s constitution
0.44 0.46 0.56 0.59 0.63 0.61 0.53 0.60 0.60 0.40 0.46 0.45
Factor 2d : Western medicine superior beliefse Compared with TCM, WM can diagnose illness more accurately Compared with TCM, WM has better facilities available Compared with TCM, the overall quality of WM is better Compared with TCM, the process of education and training in WM is more rigorous Compared with TCM, WM has a more scientific approach The labelling of content of WM is more accurate than TCM Compared with TCM, WM is specialised towards particular diseases Compared with TCM, WM acts more quickly Compared with TCM, the therapeutic effect of WM for acute disease is greater Compared with TCM, the therapeutic effect of WM is more apparent
0.62 0.56 0.59 0.65 0.68 0.54 0.48 0.49 0.50 0.64
Reliability for all 22 items, α = 0.75. Variance percentage: 16.2. c Chinese medicine belief items, α = 0.77. d Variance percentage: 15.9. e Western medicine belief items, α = 0.78. a
b
attitude of health care professionals (P = 0.004) and efficacy of service (P < 0.001), but no significant differences in scores for waiting time (P = 0.44) (Fig. 1f ).
DISCUSSION This study aimed to explore some major health beliefs of a group of Hong Kong Chinese. This research identified three clusters of attitudes among Hong Kong Chinese on the basis of five aspects related to health care (i.e. demographics variables, illness status, the CWMB Scale, trust scale and preferences ranking). People in cluster 1 tend to be older (mean 55 years), are predominantly female, and not particularly well educated or wealthy. They have chronic conditions which they perceive as more serious with greater discomfort. They are more likely to have spent time in hospital. Their belief in Western medicine physicians is lower than other groups. Their chronic illness may have made them somewhat sceptical. They represent about a quarter of the sample. People in cluster 2 are clearly different: they are more likely to be younger (mean 26 years), and are likely to be male. They have had longer education, they earn more, and they have more short-lived conditions on the whole. This group have the greatest faith in Western physicians and the least belief in TCM. They represent more than half of this sample.
The outstanding feature of patients in the smallest group, cluster 3 is their trust and belief in medicine— both Western medicine and TCM. In other respects they are rather similar in sociodemographic status and aspects of their illness to cluster 1, though they are rather younger (mean 48 years) and more likely to be male. This study shows the CWMB scores of some Hong Kong Chinese, on TCM (range: 3.31–4.07) and WM (range: 3.60–4.18) range from moderate to great, while their trust scores on TCM (range: 3.30–3.82) and WM (range: 3.50–3.90) physicians were moderate. These might partly explain why Chinese choose TCM or WM for a specific type of illness that they are suffering from.5–8,12,13 By understanding patients’ views on WM and TCM, physicians may begin discussions about TCM or WM and consider how they should be advising their patients. It may also improve the patient–doctor relationship and patient care if physicians can show an understanding of patient attitudes and beliefs while being proactive in discussing whether their patients are using TCM or WM. Some caution must be exercised in generalising the results of the present study. One potential limitation is that data for this study were obtained from a self-report questionnaire and open to a possible response bias.14 The results are limited due to the use of a rather small sample collected from only two clinics. The sample contains an overrepresentation of subjects between the ages of 20–29 (43.17%)
Attitudes of Hong Kong Chinese to traditional Chinese medicine and Western medicine
and 50+ (28.11%). No adolescents (below 18) were included in the sample. Future research should evaluate process and outcome of TCM and WM focusing on longitudinal designs and larger samples of patients.
REFERENCES 1. Wang DIC, Leung JC, Wu IC, Gao N. Biotechnology and Hong Kong. In: Berger S, Lester R, eds. Made in Hong Kong. Hong Kong: Oxford University Press, 1997. p. 249. 2. CMCHK http://www.cmchk.org.hk. The Chinese Medicine Council of Hong Kong, 2001. 3. Tung CH. Building Hong Kong for a new era. Policy address for Hong Kong Government Special Administrative Region, the PRC, 8 October 1997. 4. Tsang JC, Lo YL. Biotechnology development in Hong Kong: a renewed scientific interest in traditional Chinese medicine. Biochem Educ 1998; 26: 301–303. 5. Lau BWK. When and how patients seek medical help? An exploration in Hong Kong. Hong Kong Pract 1996; 18: 109–115.
109
6. Lau JTF, Yu A. The choice between Chinese medicine and Western medicine practitioners by Hong Kong adolescents. Am J Chin Med 2000; 28(1): 131–139. 7. Lam TP. Strengths and weaknesses of traditional Chinese medicine and Western medicine in the eyes of some Hong Kong Chinese. J Epidemiol Commun Health 2001; 55: 762–765. 8. Liang C. The development and examination of the Chinese-Western Medical Beliefs Scale. Nurs Res (in Chinese version) 1999; 7: 445–458. 9. Anderberg MR. Cluster analysis for applications. New York: Academic Press, 1973. 10. SPSS Base 7.5 Applications guide, SPSS Inc., 1997. 11. Morris R, Blashfield R, Satz P. Neuropsychology and cluster analysis: potentials and problems. J Clin Neuropsychol 1981; 3: 79–99. 12. Wong TW, Wong SL, Donnan SPB. Traditional Chinese medicine and Western medicine in Hong Kong: a comparison of the consultation processes and side effects. J Hong Kong Med Assoc 1993; 45(4): 278–284. 13. Wong TW, Wong SL, Donnan SPB. A study of traditional Chinese medicine practitioners in Hong Kong. J Hong Kong Med Assoc 1993; 45(4): 285–290. 14. Polit DF, Hungler BP. Nursing research: Principles and methods, 5th edn. Philadelphia: Lippincott, 1995.