Summary

Objective: To find out the attitude and knowledge of evidence-based practice (EBP) among doctors in Hong Kong.  

Design: Questionnaire survey of attitudes, perceived barriers and knowledge of EBP.

Subjects: Three groups of doctors: (i) a random selection from all registered medical practitioners (N=276/3,000), (ii) members of the Hong Kong College of Family Physicians (N=422/1,488), and (iii) participants in EBP training courses (N=159).

Main outcome measures: Attitude and barriers towards EBP, mean knowledge score in searching.

Results: Most respondents agreed that EBP was useful in clinical practice (73%), few agreed that they would prefer to search the literature to find answers to clinical questions (46%). Most ranked "poverty of local evidence" as an important barrier (44%). Though few thought that "lack of knowledge and skills" was important (20%), the mean EBP knowledge score was below the pass-mark of 50% (43%, CI 41-46%). Having internet access and training were predictive of better knowledge, while a more positive attitude was associated with regular searching. Conclusion: Respondents showed a positive attitude towards EBP, but knowledge of searching was barely satisfactory, especially for doctors in private practice. The low response rate precludes generalization but the results probably represent the most favourable scenario. Providing training, improving internet access, encouraging regular searching and alerting to pre-appraised evidence are likely to improve attitude and knowledge of EBP.

Keywords: Evidence-based practice, questionnaire survey, Hong Kong doctors

摘要

目的：研究香港醫生對實證醫學的態度和知識。

設計：就對實證醫學的態度、已知障礙和知識做問卷調查。

對象：三組醫生：（一）從所有註冊醫生隨機選擇(N=276/3,000)，（二）香港家庭醫學學院會員(N=422/1,4 88)，（三）實證醫學培訓班學員(N=159)。

主要測量內容：對實證醫學的態度，施行的障礙，尋找實證相關所需知識的平均分數。

結果：雖然73%受訪者認為，實證醫學對臨床實踐有用，但只有46%受訪者會選擇從文獻中找出解決臨床問題的答案。“缺乏本地證據”是主要的實踐障礙(44%)。儘管只有20%受訪者認為“缺乏知識和技能”是重要的實踐障礙，若以50%為合格，受訪者尋找實證所需知識的平均分數並不及格，僅為(43% CI 41-46%)。能使用互聯網和曾接受實證醫學的培訓的受訪者，在尋找實證所需知識方面上有較佳表現，此外經常搜索與對實證醫學的積極態度相關。

結論：受訪者對實證醫學的態度積極，但是受訪者特別是私人執業醫生尋找實證所需的知識卻差強人意。由於問卷的回應率偏低，結果的概化性受到限制，但此結果可能是可測試的最佳表現。提供培訓，加強互聯網運用，鼓勵經常性的搜索，提供預評證據，有助改善醫生對實證醫學的態度和增進尋找實證所需的知識。

主要詞彙：實證醫學，問卷調查，香港醫生

---

Introduction

In order to assure quality care, the practice of medicine should be evidence-based.^1,2 Evidence-based practice (EBP) was structurally introduced into the medical undergraduate curricula in 1999 for the University of Hong Kong and in 2000 for the Chinese University of Hong Kong. While EBP training has been widely available to doctors practicing in hospital settings and the public sector,^3,4 few formal training courses are available for primary care doctors working in the community. Barriers in the adoption of EBP have been reported among undergraduates,⁵ but studies of attitude and knowledge of EBP on practicing doctors are scarce.⁶ It is not known whether doctors in Hong Kong support EBP, and whether they have the required knowledge and skills. Our objective is to find out the attitude and knowledge of EBP among doctors in Hong Kong.

Design and subjects

The study was a postal questionnaire survey. Subjects consisted of: (i) randomly selected doctors from all registered medical practitioners residing in Hong Kong, (ii) all members of the Hong Kong College of Family Physicians (HKCFP) and (iii) the 159 participants in EBP training courses. Ethical approval for our study was provided by the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster.

Method

In June 2005, we sent postal questionnaires to 3,000 doctors randomly selected by the computer from 10,000 doctors registered with the Hong Kong Medical Council in 2004.⁷ The three-paged structured questionnaire consisted of four parts: 18 attitudinal questions, 5 questions on barriers, 4 knowledge tests and demographic data. The questions on attitude and barriers were modified from previous studies^8-13 and the knowledge test was developed by an international expert in EBP. A lucky draw was offered to respondents and a reminder was sent to non-respondents one month later in July 2005.

The response rate of the two mailings was very low (9.5%, 276 responses), as compared to similar local surveys.^14-16 Hence, we shortened the questionnaire to one two-paged format which consisted of 4 attitudinal questions and 1 knowledge test sent in a second sampling to members of the HKCFP in November 2005, after excluding 75 who had already responded to the earlier survey.

Pre-training data collected from 159 participants of EBP training courses (100 from two 2-hour seminars, 59 from two 6-hour workshops) around the period of November 2004 to Feb 2006 were also included in the survey.

Analysis

Frequency tables and descriptive statistics on the attitude and knowledge were done overall by pooling the 3 samples: i.e. all Hong Kong doctors (HKD), members of the HKCFP (CFP) and training course participants (TCP). Student's t tests and chi-square tests were used to test the differences between the subgroups. Multiple linear regressions (forward stepwise at p<0.05) were carried out to identify predictors of a positive attitude or high knowledge scores. All analyses were carried out using Statistical Package for Social Science version 14.0 for Windows.¹⁷

Results

Study sample

Of the 3,000 questionnaires sent, 101 were returned (the address was incorrect, the doctor had retired or died). We received 276 (9.5%) responses to the remaining 2899 questionnaires.

Out of 1488 members of HKCFP, 75 had already responded to the general survey above. From the remaining members, we received 422 (30%) responses to the 1413 questionnaires sent.

Pre-training questionnaires were collected from all 159 participants in four EBP training courses.

In total, we received 857 responses from all 3 groups. Some of the responses had incomplete replies leading to lower response rates for some questions.

Respondents

Table 1 presents the demographic details of the respondents. Most (70%) were men and the mean age was 38 years. They graduated from medical school a mean of 14 years ago (range, 1 - 55 years). Most (79%) graduated in Hong Kong. About three-quarters (74%) were general practitioners (GPs). More than half (64%) worked in group practices. A similar proportion worked in the private sector (46%) and the Hospital Authority (HA). 7% worked in the Department of Health (DH). 226 were fellows of the Hong Kong Academy of Medicine (HKAM), and were certified specialists in family medicine (8%), internal medicine (4%), surgery (4%), etc. Most (72%) had obtained postgraduate qualifications, including fellowship of the HKCFP (35%), various diplomas and master, MD or PhD degrees.

Table 1: Demographic data of all respondents and the subgroups surveyed

All respondents N=857(%)     HKD N=276 (%)     CFP N=422(%)     TCP N=159(%)   Age 21-40       538(64) *   157(58) *   263(63)     118(79) *     41-55   213(25) *   88(32) *   101(24)     24(16) *     Sex Male   591(70) *   201(74)     301(71)     89(60) * Local graduates       652(79)     219(81)     311(76)     122(83)   Field   GP/FM   610(74) *   121(45) *   369(90) *   120(83) * Type   Solo   291(36) *   81(34)     174(41)     36(25) *     Group   515(64) *   159(66)     246(59)     110(75) * Sector   Private   384(46) *   111(41) *   222(53) *   51(35) *     HA   361(43) *   134(49) *   148(35) *   79(54) *     DH   60(7)     22(8)     28(7)     10(7)   Postgraduate   FHKAM   226(29) *   127(49) *   82(21) *   17(13) *     qualification Others   335(43) *   929(35) *   192(49) *   51(40) *     Nil   220(28) *   41(16) *   118(30) *   61(47) * Attended training       474(55) *   135(49) *   290(69) *   49(31) * Has a computer       744(89)     244(88)     370(88)     130(82)   Internet access       576(71)     208(75)     290(69)     119(75)                                 HKD=Respondents from all registered doctors in Hong Kong CFP=Respondents from Hong Kong College of Family Physicians TCP=Training Course Participants GP/FM=General practice /Family Medicine HA=Hospital Authority DH=Department of Health FHKAM=Fellow of Hong Kong Academy of Medicine *Differences between subgroups and all respondents significant at p<0.05 (chi-square test)

Fifty-five percent of respondents had attended training on EBP and most had a computer at work (89%). Seventy-one percent stated they could access the internet for medical databases. Medline was the most frequently cited (18%), followed by PubMed (12%), e-KG (9%) and university libraries (6%). Websites like e-KG established by HA or links from university libraries provide access to multiple databases, for example, Ovid. Few doctors directly cited the Cochrane Library (4%), pre-appraised journals (0.7%) or meta-searching engines (0.5%).

Attitude towards EBP

It can be seen from Table 2 that most respondents agreed that EBP was useful in clinical practice (73%), and thought they knew how to read a paper (62%). Fewer agreed they would prefer to search the literature to find answers to clinical questions (46%), or felt confident about applying the best current evidence (39%).

Table 2: Attitude of respondents towards evidence-based practice

All respondents N=857 (%)   Mean attitudinal score for subgroups #     Disagree or Strongly Disagree   Neutral   Agree or Strongly agree   HKD N=276   CFP N=422   TCP N=159 Q1 I believe evidence-based medicine is useful in my clinical practice. 14 (1)   213 (26)   612 (73)   5.1 *   4.8 *   5.0 Q2 I prefer to search the literature to find answers to my clinical questions. 33 (4)   420 (50)   385 (46)   4.4   4.4   4.3 Q3 I know how to read a medical paper. 38 (5)   275 (33   )522 (62)   2.2 *   2.3 ^   3.0 *^ Q4 I feel confident about applying the best current evidence in daily patient care. 53( 6)   460 (55)   321 (39)   4.4 *   4.2 ^   3.8 *^ # Mean attitudinal score by Likert scale: 6=strongly agree, 1=strongly disagree HKD=Respondents from all registered doctors in Hong Kong CFP=Respondents from Hong Kong College of Family Physicians TCP=Training Course * ^ Differences in mean attitudinal score significant at p<0.05 (student's 2-sample t-test)

Those who were doing more than one search per week were more likely to have a positive attitude towards EBP, prefer to search and feel confident about applying the evidence, but they were less likely to think that they know how to read a medical paper. (Table 3).

Table 3: Factors that could affect attitude towards evidence-based practice

Do more than 1 search per week   Have a computer in practice   Attended training   FHKAM Fellow   Being a GP   Male   Age < 40   CFP   HKD Q1   0.129*   -   -   -0.072*   -   -   -   -0.114*   - Q2   0.165*   -   -   -0.114*   -   -   -   -   - Q3   -0.146*   0.068*   -0.068*   0.098*   -0.092*   0.107*   0.077*   -0.260*   -0.240* Q4   0.140*   -   0.125*   -   0.136*   -0.120*   -0.143*   -   - All standardized beta coefficients are significant at p<0.05* (multiple linear regression) DH=Department of Health FHKAM=Fellow of Hong Kong Academy of Medicine GP=General Practitioner CFP=Respondents from Hong Kong College of Family Physicians CFP=Respondents from Hong Kong College of Family Physicians

Barriers to EBP

As shown in Table 4, "poverty of local evidence" was ranked by most respondents as "a very important" or "the most important" barrier (44%), while "lack of knowledge/skills" was similarly ranked by only 20%. Respondents working in HA or who had attended training were more likely to find time constraint being a barrier. Respondents working in DH were more likely to find it difficult to access evidence, especially local evidence, at the point of care (Table 5).

Table 4: Barriers perceived by respondents towards evidence-based practice

All respondents N=857 (%)   Mean barrier score for Subgroups #       Least to quite important   Neutral   Very to most important   HKD N=276   CFP N=422   TCP N=159 Q1 Time Constraint   151 (20)   282 (38)   309 (42)   3.2   3.3   3.4 Q2 Difficulty in accessing evidence at the point of care   170 (23)   271 (37)   297 (40)   3.2   3.2   3.2 Q3 Poverty of local evidence   143 (19)   275 (37)   323 (44)   3.2   3.4   3.3 Q4 Lack of knowledge and skills   324 (44)   270 (36)   145 (20)   2.6*   2.6^   3.0*^ Q5 Limited relevance of research to daily practice   276 (37)   285 (39)   175 (24)   2.7   2.8   2.8                             # Mean barrier score by Likert scale : 5=most important, 1=least important HKD=Respondents from all registered doctors in Hong Kong CFP=Respondents from Hong Kong College of Family Physicians TCP=Training Course Participants * ^ Differences in perceived barriers significant at p<0.05 (student's t-test)

Table 5: Factors that could affect perceived barriers towards evidence-based practice

Working in HA   Working in DH   FHKAM   Attended training   Do more than 1 search per week   Male   Age <40 Q1   0.106*   -   -   0.091*   -   -   - Q2   -   0.079*   -   -0.085*   -0.107*   -   - Q3   -   0.070*   -   -   -   -   - Q4   -   -   0.149*   -0.143*   -0.149*   0.095*   -0.118* Q5   -   -   0.078*   -   -   -   -0.104* All standardized beta coefficients are significant at p<0.05* (multiple linear regression) HA=Hospital Authority DH=Department of Health FHKAM=Fellow of Hong Kong Academy of Medicine

Knowledge

The distribution of the EBP knowledge scores is shown in Table 6. 76 and 77% of doctors understood the terms "AND" and "OR", respectively, about half knew that the Cochrane Library contains the largest database of RCTs (60%) and that PubMed is a Medline interface available free via the internet (52%). Fewer knew that Medline contains the largest database of non-randomized studies (31%), or understood MeSH terms (33%), wildcard* (26%) or explode (19%). The mean total score for the 8 matching items was 43% (95% CI 41-46%).

Table 6: Knowledge test by matching items (% correct)

Statement (Matching item)   All respondents   HKD   CFP   TCP Requires that an article contains BOTH words (AND)   76   72   77   80 Keywords coded by the National Library of Medicine (MeSH terms)   33   37   33   23 Is used to find words with the same stem (Wildcard*)   26   22   29   24 Contains the largest database of randomized controlled trials (Cochrane Library)   60   48   66   66 Requires that an article contains EITHER word (OR)   77   73   80   78 A Medline interface available free via the internet (PubMed)   52   52   50   56 To use all subheadings of a MeSH term you would use (Explode)   19   21   20   16 Contains the largest database of non-randomized studies (Medline)   31   19   38   30   Mean Knowledge Score (%)   43 (CI 41-46)   43*   49*^   42^ HKD=Respondents from all registered doctors in Hong Kong CFP=Respondents from Hong Kong College of Family Physicians TCP=Training Course Participants * ^ Differences in mean knowledge score significant at p<0.05 (student's 2-sample t-test)

A high knowledge score was more likely to be found among doctors who searched regularly, had attended training, or aged less than 40. Doctors who were in private practice, fellows of the HKAM, and respondents from all registered doctors tended to score lower (Table 7).

Table 7: Factors that could affect mean knowledge score

Higher Score   Lower Score Do more than 1 search per week   0.900*   In private practice   -0.146* Attended training   0.171*   FHKAM   -0.115* Age <40   0.147*   HKD   -0.104* All standardized beta coefficients are significant at p<0.05* (multiple linear regression) FHKAM = Fellow of Hong Kong Academy of Medicine HKD=Respondents from all registered doctors in Hong Kong

Discussion

Methodological issues and limitations

Our study aimed at gaining an insight into the attitude and knowledge of EBP among doctors in Hong Kong, but generalisability was limited by the low response rates. We compared our response rates with three other local studies: 72% in a survey on 1016 HKCFP members on upper respiratory tract infection,¹⁴ 60% in a study involving 600 private primary care doctors on cervical screening,¹⁵ and a population-based survey of 4850 doctors on clinic computerization yielded a low response rate of 18.5%¹⁶ which still nearly doubled ours. Limited resources did not permit us to use strategies like personal telephone calls or giving out monetary incentives.¹⁶ Personal communications with some non-respondents revealed that they found the topic of EBP unfamiliar and irrelevant, the knowledge test too difficult and the questionnaire too long. Reasons for the poor response and possible methods for improvement need to be further explored by qualitative studies.

Our data were pooled from several sampling frames and the differences between the subgroups are shown in Table 1. Without attempting to generalize the results, our findings did reflect the situation of 857 doctors in Hong Kong. Comparing the characteristics of 5498 doctors surveyed in the Health Manpower Survey in 2005,¹⁸ our respondents were similar to all medical practitioners residing in Hong Kong in sex ratio (male 70%) and in the practice sector (46% private and 43% HA) but, our respondents were younger (median age less than 40 vs 42) and more were GPs (79% vs 34%). Doctors spending most of their working time in specialist practice were under-represented (26% vs 58%). We do not know if the HKCFP sample was representative of all members of HKCFP because of lack of comparison data. Given the likely bias towards more knowledgeable respondents, our findings represent the most favourable scenario in the knowledge of (and arguably the attitude towards) EBP among doctors in Hong Kong.

Interpretation of findings

Most respondents agreed that EBP was useful in clinical practice, but few agreed that they preferred to search, or felt confident in applying the evidence. The general positive attitude was in line with the results from GPs in the UK,⁹ Australia,¹⁰ Saudi Arabia¹² and Malaysia.¹³

Lack of local data was the most commonly cited barrier to EBP, which might explain why doctors in Hong Kong did not prefer to search the literature or had low confidence in applying the evidence. This calls for more research and better dissemination of data for our local population. Lack of knowledge and skills was not ranked as an important barrier, but a knowledge gap was revealed in the low score in searching (below the "pass-mark" of 50%). Training was predictive of a higher knowledge score. Respondents working in HA and DH are likely to have more training opportunities;^3,4 yet, their scores were no more than 60%. Perceived barriers reported by respondents: lack of time for doctors working in HA and suboptimal internet access for doctors working in DH (broadband not readily available, personal communication) are modifiable factors that could be tackled. Though training opportunities are supposed to be ample in these two sectors,^3,4 knowledge and skills learnt from one-shot training is difficult to become retained later without continual support and practice.

It is disappointing although not unexpected that private doctors who are providing more than 70% of primary care¹⁹ had a significantly lower knowledge score. Reforms in the Hong Kong's health system has been widely discussed.²⁰ Among other measures to rectify the compartmentalized health care delivery, HA has initiated a public-private interface programme.²¹ In the 2007-08 policy address, it has been announced that HK$150 million per year would be spent to subsidize the purchase of primary medical care services by the private sector for all Hong Kong citizens aged 70 or above from 2008 to 2010. Introducing more services through public-private partnership and purchasing a full range of medical services by the private sector have also been mentioned.²² Doctors in private practice need to face up to the challenge and be better equipped with the most up-to-date medical evidence in order to provide the best practices for their patients. On a personal level, knowledge and skills in EBP is a tool for life-long self-directed learning that can prevent rust-out.²³

There is some evidence from our study that training was associated with better EBP knowledge. The Diploma in Family Medicine of the HKCFP for practising doctors has included EBP in the curriculum since 2003. The low proportion of diploma students from the private sector (16 to 33% for years 2003 to 2006) reflects an "inverse care" phenomenon. The incentives and barriers in continuing education for doctors working in private practice deserve further studies.²⁴

Regular searching was associated with a positive attitude and predicted a better knowledge of EBP. The majority of respondents had a computer in their practice, but a local survey showed that the computer was seldom used for meeting information needs in clinical decisions.¹⁶ Efforts in promoting clinical computer use should emphasize its potential in supporting evidence-based care. The proportion of pre-appraised evidence (the Cochrane Library, Evidence-based Medicine) cited by respondents as accessible databases was disappointingly low, probably because these are not freely available. Providing free/ready access to and promoting the use of these methodologically filtered summaries are some of the strategies in ensuring evidence-based care for busy clinicians.^25-26 It is also a strategy that the government should consider in the course of shifting health care from the hospitals to primary care.

Conclusion

Our study aimed to find out the attitude and knowledge of EBP among doctors in Hong Kong. Low response rate precluded generalization. Respondents showed a positive attitude towards EBP, but their knowledge was barely satisfactory, particularly for doctors in the private sector. Providing training, improving internet access, encouraging regular searching and alerting to pre-appraised evidence are likely to improve the attitude and knowledge of EBP.

Two areas for further research are identified: 1) whether the low survey response reflects a grave deficiency in the knowledge and skills on EBP among doctors in Hong Kong, and if so, what are the strategies for rectification; 2) incentives and barriers of learning about EBP, especially for doctors in private practice.

Acknowledgement

This study was funded by the Research Fellowship of The Hong Kong College of Family Physicians 2005. Ethics approval was granted by the HKU/HAHKW IRB 001F3.

We wish to thank all doctors who contributed by answering the questionnaires, Mr Nam Tat in randomisation of the survey list, Ms Teresa Lee and college secretariat in posting questionnaires to members of HKCFP, Ms Felice Liu for posting and data entry and Ms Ida Chan for statistical advice.

Key messages

1. In order to assure quality care the practice of medicine should be evidence-based.
2. A survey of Hong Kong doctors showed that 73% agreed that evidence-based practice was useful in clinical practice.
3. "Poverty of local evidence" was ranked as the most important barrier.
4. The mean knowledge score in searching for evidence from the internet was below the pass-mark of 50% in the survey.
5. Providing training, improving internet access, encouraging regular searching and alerting to pre-appraised evidence are likely to improve attitude and knowledge of evidence-based practice.

---

---

Reference