Summary

Objective: To investigate if the knowledge, attitude and practice (KAP) on cervical cancer screening of private general practitioners (GPs) can be modified by education.

Design: A controlled cross-over trial of continuing medical education (CME).

Subjects: 116 volunteer GPs in private practice who responded to a CME invitation were assigned to study (pap smear workshop, 60) and control (women's health workshop, 56) groups. Sixty-one were randomised while 55 were allocated by date of availability, before knowing which topic occurred on which dates.

Main outcome measures: KAP results from self-administrated questionnaires before and 4 months after the workshop.

Results: Ninety-four (81% of 116) attended the workshops (46 intervention and 48 controls). Among the 78 doctors (83% of 94) who completed both pre- and post-test questionnaires, while attitude was unchanged, knowledge (in 3 out of 8 items) and behaviour (self-reported Pap smear performing frequency and opportunistic advice for screening) were significantly improved only in the study group.

Conclusion: Private GPs acquired more knowledge and reported positive changes in behaviour after an interactive CME workshop. More research is needed in how to maximise the screening capabilities of private GPs and thereby increase the screening rate for cervical cancer in Hong Kong.

Keywords: Cervical cancer screening; general practice; knowledge, attitude, practice; Hong Kong; continuing medical education, postgraduate

摘要

目的：研究私人開業全科醫生（GP）有關子宮頸癌篩查的知識、態度和技能（KAP）是否能因教育而改變。

設計：醫學繼續教育（CME）的對照性交叉試驗。

對象：116名對研討邀請做出回應的私人開業全科醫生。他們被分為研究組（宮頸塗片，60人）和對照組（婦女健康研討會，56人）。在了解那一天將有那個 題目之前，61人隨機地、55人按照能參加的日期分組。

測量內容：研討會前和會後四個月自我問卷的KAP結果。

結果：共有94人（116人中的81%）參加了研討會（46個在研究組，48個在對照組）。其中78位（94人的83%）完成了測驗前、後問卷。他們的態度沒有變化；知識（8項中的3項）和行為（自我報告的宮頸塗片實施頻率和機會性篩查建議）僅在干預組有明顯的改善。結論：在一次互動式的CME研討會後，私人全科醫生獲得了更多的知識，並報告了在行為上正面的變化。如何最大程度地提高私人全科醫生篩查的能力，從而提高香港的子宮頸癌篩查率，還需要做更多的研究。

主要詞彙：子宮頸癌篩查；全科醫療；知識、態度、技能；香港；醫學繼續教育、畢業後

Introduction

In Hong Kong, cancer of the cervix uteri was the fourth commonest cancer and the seventh cause of death for women in 2000.¹ Over the past 15 years, there has been a slow overall decrease in the incidence and mortality rates in Hong Kong.² This may be attributed to the low coverage of cervical cancer screening test - the Papanicolaou (Pap) smear. A survey done by the Family Planning Association of Hong Kong in 1997 found that the coverage of cervical cancer screening was 52%³ which is low, when compared to 80%-90% for some developed countries like England and Finland where population-based screening programme has been implemented.^4,5

Most screening services in Hong Kong (73%) are provided by government or semi-government organisations including Women's Health Centres of the Department of Health, well-women clinics in different hospitals, and clinics of the Family Planning Association, which together have capacity to provide only 16% of screening services required by the eligible women per year.^6-8 The Harvard report showed that in Hong Kong, one person makes an average of 9.1 visits to a doctor per year, and 79% of the 2048 female responders stated that they preferred private to Government or public clinics for outpatient care.⁹ The commonest reason given by women for obtaining a cervical cancer screening test is at their doctor's recommendation.¹⁰ In New Zealand and Australia, general practitioners (GPs) are the most effective health professionals in persuading women to have Pap smears.^11,12 Hence, Hong Kong GPs in private practice are in a favourable position to provide Pap smear to patients.

In general, most private GPs in Hong Kong learned gynaecology only to a minimal extent, and many have had no clinical gynaecology experience after graduation. Our previous study in 2001 found that only about half (53%) of the questioned private primary care doctors performed Pap smear test.¹³ There is a need to devise interventions to enhance GP's knowledge and improve their practice on Pap smear test. However, changing doctors' behaviour through educational interventions is difficult.¹⁴ We sought to measure whether private GPs in Hong Kong will change as a result of continuing medical education (CME).

Aims

We aim to find out whether doctors in private general practice will change their KAP after attending a behaviour-focused, brief but intensive, education workshop on cervical screening.

Methods

Design

We recognised that only interested doctors would volunteer for this programme, therefore we used a wait-list control technique to provide two groups. The control group was offered an initial workshop on women's health issues, and a second workshop on Pap smears about four months later. The initial Pap smears education group was offered the women's health workshop at the end. This crossover design allowed us to administer measurement instruments under the same conditions each time.

Subjects

An introductory letter explaining the purpose of this free workshop together with an application form, in both English and Chinese, was sent to 1,400 CME course members of the Hong Kong Medical Association (HKMA). We enrolled private GPs who were not currently taking cervical smears but willing to consider doing smears, and those undertaking very few smears but would like to increase their number.

The application form included information on personal particular, enrolment requirements, question on preference in teaching language and dates preferred. Three continuing medical education points were awarded by HKMA and The Hong Kong College of Family Physicians to participating doctors for attending each workshop.

Randomisation

The doctors who were able to attend both the study workshop (Pap smears) and the control workshop (General women's health) dates were randomly allocated in pairs by a computer-generated method. Other doctors who were not available on both dates were assigned to the study or the control workshop according to the date they were available, so that 30 doctors were in each workshop. At the time of application, neither we nor the doctors knew which workshop would occur on which date.

Study intervention

The workshops lasted for about 3.5 hours on a Sunday afternoon. Each study and control workshop was conducted twice, one in English and the other in Cantonese. Topics covered included updates on information about cervical cancer screening, video shows, role play in counselling women on Pap smear and skills practice on pelvic models.

Instrument

Changes in knowledge, attitude and practice (KAP) of doctors were evaluated by a self-administered questionnaire at the beginning of the workshop and four months later. By incorporating the ideas and modifying the questions from similar local and overseas studies^13,15-17 the questionnaire included questions on who needs Pap smears, attitude towards Pap smears, how often and under what circumstances the doctors would provide Pap smear to patients and some demographic data. The questionnaire was content validated by experts and professionals in cervical cancer screening and was pilot-tested on 10 GPs.

Main outcome measures

KAP results from the self-administrated questionnaires before and 4 months after the workshop were analysed.

Statistics

SPSS 10.1.0 (Statistical Package for Social Science) was used. Chi-square or Fisher's exact test was used to compare independent categorical data and Mann-Whitney U was used to compare independent continuous data between study and control groups. McNemar's test for categorical variables and Wilcoxon Signed Ranks test for continuous variables were used to ascertain significant pre- to post-test changes by individuals. Ordinal data were considered continuous in analysis and a two-tailed p value of less than 0.05 was taken as significant.

Results

A. Enrolment

Two hundred and twenty-one application forms were received (response rate 15.8%). Only 116 doctors met the selection criteria and were enrolled in this study. Among them, 61 (53%) could be randomised, and 55 were assigned by availability. A total of 60 doctors were assigned to the study group.

B. Participants

Only 94 (81%) of the enrolled doctors attended the workshops: 46 (77%) doctors attended the study workshop and 48 (86%) doctors attended the control workshop. There were no significant differences in mean age, mean years graduated and mean years in private general practice between attendees and those who were absent. However, significantly more male doctors (20/85 males, 2/31 females, c²=4.8, P=0.029) and Hong Kong graduates (15/52 HK graduates, 7/62 mainland and other) c²=6.4, P=0.042) were absent.

Characteristics of the doctors

Ninety doctors completed the pre-test questionnaire. Results of KAP scores are based on the analyses of the data received from the 78 doctors who completed both pre- and post-test questionnaires (Table 1). No significant differences in demographic data were found between those doctors who only completed the pre-test questionnaire and those who returned both questionnaires. The characteristics of doctors are shown in Table 2.

Table 2: Characteristics of participants

Characteristics   Total N=78   Study group N=37   Control group N=41 Gender*               Male   54 (69)   25 (68)   29 (71)   Female   24 (31)   12 (32)   12 (29)                 Age (Mean SD years)   55 11   55 11   55 12                 Medical education               Hong Kong   31 (40)   14 (38)   17 (42)   China   38 (49)   19 (51)   19 (46)   Others   9 (11)   4 (11)   5 (12)                 Years graduated (Mean SD years)   29 11   29 11   30 11   Mode (Min - Max)   33 (3-46)   32 (8-45)   34 (3-46)                 Years in private general practice (Mean SD years)   15 9   15 10   15 8   Mode (Min - Max)   15 (1-39)   13 (1-39)   15 (1.5-32)                 Nature of Practice#               Solo   68 (88)   35 (95)   33 (83)   Small group (<5 doctors)   9 (12)   2 (5)   7 (17)                 Office district#               Hong Kong Island   14 (18)   9 (25)   5 (12)   Kowloon   29 (38)   13 (36)   16 (39)   New Territories   34 (44)   14 (39)   20 (49)                 Patient characteristics             Number treated per day#               Mean SD   33 24   32 22   34 25   Median (Min - Max)   26 (6-100)   28 (6-100)   25 (8-100)                 Estimated proportion of female patients#               (Mean SD)%   53 14   53 14   53 15   Median (Min - Max)   50 (20-90)   50 (30-80)   50 (20-90)                 * Randomised Vs Non-randomised (80% Vs 55% Male, c2=5.8, df=1, P=0.016) # A few missing data

C. Outcomes

Knowledge

Eight items were used to assess knowledge of Pap smears. There was little difference between the intervention and control groups initially, demonstrating the comparability of the groups. In the post-test, there was a small difference in most of the answers, slightly favouring the intervention group. The learning effect was greatest for those giving three years as the correct interval for smears (Tables 3a-c).

Table 3c: Responses to the knowledge questions of "Who needs Pap smear test?"

Study (n=34-37#)     Control (n=39-41#)         Correct   Incorrect     Correct   Incorrect         No   Yes Don't know     No   Yes Don't know a. Women older than 25 years of age and never had sexual intercourse Pre   30 (88)   4 (12) -     30 (77)   8 (21) 1 (3) Post   29 (85)   3 (9) 2 (6)     31 (79)   6 (15) 2 (5)         Yes   No Don't know     Yes   No Don't know b. Women 65 years and never had any Pap smear Pre   34 (92)   3 (8) -     36 (88)   4 (10) 1 (2) Post*   36 (97)   1 (3) -     34 (83)   6 (15) 1 (2)         Yes   No Don't know     Yes   No Don't know c. Married women who have never been pregnant Pre   34 (94)   1 (3) 1 (3)     38 (96)   1 (2) 1 (2) Post   36 (100)   - -     39 (98)   - 1 (2)         Yes   No Don't know     Yes   No Don't know d. Women with only one sexual partner and previous negative Pap results Pre   31 (86)   4 (11) 1 (3)     34 (85)   5 (13) 1 (2) Post**   35 (97)   1 (3) -     35 (88)   5 (12) -         No   Yes Don't know     No   Yes Don't know e. Women 70 years with regular negative Pap results Pre   19 (56)   15 (44) -     30 (75)   8 (20) 2 (5) Post   25 (73)   8 (24) 1 (3)     33 (83)   6 (15) 1 (2)         Yes   No Don't know     Yes   No Don't know f. Women who had hysterectomy for cervical malignancy Pre   15 (44)   18 (53) 1 (3)     14 (35)   26 (65) - Post   13 (38)   20 (59) 1 (3)     14 (35)   24 (60) 2 (5)                             # Numbers who answered each question varied slightly * Correct Vs Incorrect: Randomised Post study Vs Post control (Fisher's Exact Test, P=0.022) ** Correct Vs Incorrect: Randomised Post study Vs Post control (Fisher's Exact Test, P=0.049)

Attitude

One question about the importance of assessing Pap smear history and seven questions about roles for private doctors in Pap smear testing assessed the attitude of participating doctors. The patterns of response given by both study and control groups were similar. The vast majority (85-90%) of the doctors attending this programme showed a positive attitude towards Pap smear test: They disagreed that women who need Pap smear should go to government or public clinic rather than private general clinic; disagreed that male doctors are not suitable to perform Pap smear; agreed that Pap smear is an effective preventive measure for cervical cancer, and that doctors should take more responsibility than their patients for initiating Pap smear discussion. Only minor and insignificant changes in the responses to the attitude questions from pre-test to post-test were found in both groups (Tables 4a-b). Almost half of both groups still felt that counselling women to have Pap smear is difficult. About one-third of the study group were still worried that most of their patients would not take Pap smear even if they explained the need for one in the post-test.

Self-reported behaviour

a. Self-reported frequency of performing Pap smears

While the groups were similar in pretest results, in the post-test more doctors of the study group performed Pap test more often when compared to the control group (Mann-Whitney U=457.5, P=0.002), though few reported doing smears more than twice weekly. A significant decrease in the frequency of Pap test was reported post-test when compared to its pre-test (Wilcoxon Signed Ranks Test, Z=-3.2, P=0.001) by the control group.

b. Discussing Pap smear tests with patients

In the pre-test, the doctors in the study group answered that they discussed Pap smear test with 41 29% (mean SD, median=30%) of female patients in a usual working day, while the proportion in the control group was 34 24% (mean SD, median=30%). Improvement was found in the study group, with the median percentage increased from 30% in the pre-test to 50% in the post-test (Wilcoxon Signed Ranks Test, Z=-2.3, P=0.002), and the improvement in the mean was significantly higher than that of the control group (51 26% Vs 34 27%, Mann-Whitney U=1269.5, P=0.007).

c. Opportunistic screening

After the education workshop, doctors in the study group reported that they increased their offers of opportunistic screening to their female patients. Out of the example situations given (Table 5), the largest improvements were reported for "when female patients came for treatment of respiratory infection or vague backache" (Tables 5c-d). For both cases, more doctors in the study group reported they would suggest Pap smears under these situations, and the proportion increased more in the post-test when compared to the pre-test (Wilcoxon Signed Ranks Test, P=0.046 and P=0.011 respectively) or to the control group post-test (Mann-Whitney U=471.5, P=0.01 and 510.5, P=0.02 respectively). However, no change was reported for women attending for treatment related to gynaecology or abdominal conditions.

Summary of major findings

Discussion

Our study trial shows that general practitioners who have been in private practice for an average of 15 years and have not previously undertaken much cervical smears can change. The participants were volunteers, and already had some interest in the topic, as shown by their positive attitudes at pre-test. The workshop therefore did not make measurable changes in attitudes. There were some improvements in knowledge. Most importantly, through its focus on the manual skill in taking smears and the communication skill in persuading women to have a smear, behaviour changes seemed to have occurred. The self reported changes in behaviour to increase opportunistic screening of women presenting with respiratory infection and vague backache matched the examples used in the role-play section, demonstrating that they have learnt from the workshop, or at least agreed with our suggestions to do so.

While only some of the doctors were absolutely randomised, the others were allocated to the workshop date that was acceptable to them, but they had made this choice in ignorance of which group would be held on which date: even before the researchers made the decision. Therefore there was no possibility of selection bias on this account. Further, detailed analysis of only the randomised groups showed almost identical results.

In a trial of three organisation approach to cervical cancer screening in the Netherlands, the total Pap attendance and coverage rates were highest in the family practice-based approach, compared to the community-based approach or the combination approach.¹⁸ Regular care in a local family medicine clinic, where the practice organisation promotes prevention, was associated with higher cervical cancer screening uptake in women.¹⁹ More women in Hong Kong may benefit from regular screening if the family practice-based approach is also promoted here. The total number of smears done at an established Well Women's Clinic over the past ten years was 61,101.²⁰ If we can train and encourage 100 private GPs to each do 2 smears a day, or 250 GPs to each do 5 smears a week, or 600 GPs to do 2 smears a week, this would amount to 60,000 smears a year, covering 180,000 women at three-yearly intervals.

Primary care doctors in Hong Kong should be provided with more opportunities and incentives to enhance their skills in how to bring up the subject of Pap smears with eligible women in a culturally acceptable manner, and to perform Pap smear test at regular intervals in an ongoing basis for patients who are under their care. Pap smear campaigns should communicate to the public that Pap smear services are also available in private general practitioners' clinics, in addition to the public well women's clinics. The doctors in the study represent an older group, having graduated for an average of 29 years. It is reasonable to expect more recently trained family medicine graduates to be able to provide more cervical screening for their patients when they enter private practice. Further research in assessing any future increase in Pap smear services provided by private GPs, and women's acceptance of having Pap smear test at private GP's clinic should be conducted. We can only decrease the incidence and mortality of cancer of the cervix in Hong Kong by maximising all efforts in Pap smear screening.

Limitations

Only 78 (83%) of the 94 participants' completed both questionnaires, though no significant differences in demographic data were found between those included and those excluded for analysis. The doctors were all volunteers. Necessarily they were interested in the topic, so making it difficult to achieve changes in knowledge and attitudes, though they were selected as performing few or no smears, giving the greatest opportunity for changing behaviour. The improvements were self-reports. Whether behaviour change actually occurred or not needs to be verified. This has been done and will be reported in another paper.

Conclusion

Private general practitioners acquired more knowledge and reported positive changes in behaviour, in terms of advising their patients and performing Pap smears for them, after an interactive CME workshop. More research is needed in how to maximise the screening capabilities of private general practitioners and thereby increase the screening rate for cervical cancer in Hong Kong.

Acknowledgement

This study is funded by the Health Care and Promotion Fund (HSRC Grant #217007). We are grateful to the Hong Kong Medical Association, the Hong Kong College of Family Physicians, the Association of Licentiates of the Hong Kong Medical Council, the Department of Obstetrics and Gynaecology of The Chinese University of Hong Kong, and the Clinical Education Center at the Prince of Wales Hospital for cooperation and collaboration. We thank all the doctors in the study, the dietitian Ms A Lam for helping to conduct the control workshop, Professor J L Tang for advice on the design of the study, and the research nurses Ms L Lau, Ms F Cheng and research assistants Ms W Yeung and H Tse. The study was performed while the principal authors were still on staff at the Department of Community and Family Medicine of the Chinese University of Hong Kong.

Key messages

1. Family physicians are well suited to be the most effective health professional to persuade women to have Pap smears and provide long-term regular cervical screening service.
2. Our previous study showed that only 50% of private primary care doctors in Hong Kong perform Pap smear test.
3. Our present study showed that an interactive CME workshop with emphasis on skills training can make a difference to private general practitioners' knowledge and behaviour.
4. Both opportunistic advice and frequency in performing Pap smears increased in the study group.
5. A family practice-based approach is recommended in addition to existing programmes to increase local cervical cancer screening uptake.
6. If we can train and encourage 100 private GPs to each do 2 smears a day, this would amount to 60,000 smears a year, covering 180,000 women at three-yearly intervals.