Home

ISLAMIC MEDICAL EDUCATION RESOURCES01

9711-KNOWLEDGE AND ATTITUDE ABOUT CANCER PREVENTION: A PERSPECTIVE OF A MULTI- NATIONAL ISLAMIC UNI

Draft of paper presented at the 14th Asia Pacific International Cancer Conference  Hong Kong 17-23 November 1997 by  Prof Dr Omar Hasan Kasule, Sr.;  MB ChB, MPH, DrPH (Harvard) Deputy Dean for Research and Post-graduate Affairs Faculty of Medicine, International Islamic University, Malaysia  PO Box 70 Jalan Sultan Petaling Jaya Selangor Malaysia 46700 Phone 60 3 755 3433 Fax 60 3 757 7970

ABSTRACT

OBJECTIVES: This questionnaire survey elicited information on the following: (a) knowledge of cancer sites/organs, symptoms & signs, methods of early detection, methods of treatment, risk factors, prevention measures, (b) opinions, perceptions, attitudes, and health seeking behaviours in relation to cancer (c) family history of cancer (d) previous exposure to cancer risk factors (e) sources of information about cancer. Data analysis sought to identify knowledge and attitude gaps that should be emphasised in cancer education programs using the medium of the Islamic religion that is common among the study sample.

BACKGROUND: There is a lot of socio-cultural similarity among Muslims of different races and countries because they follow a strict and a common religious law (shariah). This justifies a common international approach to cancer prevention using religious authority to induce behavioral with regard to cancer. The International Islamic University in Malaysia is an ideal opportunity as a multi-national epidemiological field laboratory because it has students and faculty from more than 90 nationalities.

METHODS: Self-coding questionnaires were sent to a stratified sample of 629 academic staff and students at the International Islamic University Malaysia. Univariate and bivariate analyses were carried out to describe sample characteritics and to identify variables associated with knowledge of and attitudes about cancer. Multivariate analysis (both Mantel-Haenszel and Logistic regression) was carried out to determine independent determinants of knowledge and attitudes about cancer. Nine representative variables were used in the regression analyses as response covariates

RESULTS: The response rate was 282/629 ( %).  The proportion of correct knowledge of  and attitudes to …. Items on the questionnaires ranged from a minimum of 33.5% for knowing that immunotherapy is a method of cancer treatment to a maximum of  98.9% for knowing that the breast is a primary cancer site. There was high correlation and agreement of responses to questionnaire items from the same group. Gender was found to be an independent determinant of:    . Age was found to be an independent determinant of     . Region of birth was found to be an independent determinant of…..   CONCLUSIONS & FOLLOW-UP: The results of the study reveal a diversity in knowledge and attitudes that requires customising cancer education programs according to the target audience and the material being taught.

KEY WORDS: CANCER, KNOWLEDGE, ATTITUDES, PREVENTION

 

1.0 INTRODUCTION

Research on knowledge and attitudes to cancer and its treatment is undertaken to identify new preventive approaches that involve behavioral changes and can be a basis for new program initiatives. Review of recently published reports on MEDLINE revealed no comprehensive studies on knowledge and attitudes in the Muslim world (      ).  The only study identified was a limited one done on smoking among medical students in Tunisia (   ).

The Muslim world is one large expanse of land extending from the Atlantic seaboard in Morocco through North Africa and Central Africa, the Middle-east, Central Asia, South Asia and ends in the Indonesian and Philippine archipelagos. Muslims also live as minorities in Europe, the Americas, The Far East and the Pacific, and Sub-Saharan Africa. There are about 1.0 billion people who despite their racial diversity, have a common Islamic culture which influences their lifestyles as well as opinions, perceptions, and attitudes to diseases including cancer. The common culture is based on the shared Islamic creed, law, the Arabic language, a shared history of 14 centuries of interactions that is reinforced by the annual pilgrimage to Makka at which 2-3 million faithful from all over the world congregate.

This study based at the International Islamic University in Malaysia on the outskirts of Kuala Lumpur. It is a 15,000-strong international community of students, faculty, and their families representing more than 90 nationalities from all over the world. This is a unique opportunity for the epidemiologist who thrives on studying the impact of racial, geographical, and socio-social diversity as they relate to disease and its prevention. The community is not only diverse but it also has many commonalities based on Islam that offer possibilities for an international approach to intervene against cancer.

 

2.0 METHODS

 

2.1 SAMPLE SELECTION AND QUESTIONNAIRE ADMINISTRATION

A self-coding questionnaire on basic socio-demographic information, knowledge and attitudes on cancer, and health-protective behaviours was prepared and was pre-tested. The study sample was selected from academic faculty and students.

An explanatory letter with a questionnaire was sent to all 191 academic staff in the Faculty of Islamic Revealed Knowledge and Human Sciences of the International Islamic University in Petaling Jaya Malaysia. The questionnaire was also sent to 50 staff and 450 students in the 10 residential colleges (hostels). Each hostel houses an average of 800 students. All five (5) administrative staff in each college were included. Five (5) out of the 10 student leaders called mushrif and 20 out of 50 student leaders called nuqaba were selected at random from each of the 10 colleges. In addition 20 international (ie non-Malaysian) students were selected at random from each college care being taken to select at least 2 from each of the following ten (10) regions: (1) Western and Eastern Europe including Bosnia, Albania, Russia and Belorussia (2) Australia, New Zealand and the Pacific (3) The Americas including USA, Canada, the Caribbean, and Latin America  (4) Middle-east including Arab countries, Turkey, Palestine, and Iran (5) North Africa (6) Sub-Saharan Africa (7) Central Asia including all Muslim Republics of the former Soviet Union (8) South Asia including Afghanistan, India, Bangladesh, Pakistan, Sri Lanka, Maldives (9) South Asian nations excluding Malaysia (10) East Asia, North Asia, and the Far-East including China, Japan, Korea.

Trained research Assistants followed up with academic staff to make sure they returned the completed questionnaires. College staff followed up with the students to make sure questionnaires were returned on time.

 

2.2 DATA MANAGEMENT AND ANALYSIS

All questionnaires were collected together in the study office. Validation and consistency checks were carried out. Data entry was by trained part-time data entry clerks supervised by an experienced systems analyst. Random checks were made to make sure there were no data entry errors. SAS was used for data management and data analysis. All covariates were dichotomised for multivariate analyses as explained below. The response covariates (knowledge and attitudes) were recoded to be dichotomy (yes=1 no=0 or agree=1 do not agree=0). Responses such as ‘not sure’ were included in ‘no’ or ‘do not agree’ categories. The explanatory covariates were dichotomised as follows: gender (female=1 male=0), age (20-40=1, 41-80=0), and Region of birth (ROB) as follows ROB1 (1=Europe, (Europe, Australia, New Zealand, and the Americas=1, others=0), ROB2 (Middle-east and North Africa=1, others =0), ROB3 (sub-saharan Agrica=1, others=0), ROB4 (central Asia=1, others=0), ROB5 (south Asia=1, others=0), and ROB6 (South-east Asia and the Far east=1, others=0).

Exploratory one-way analysis was carried out to provide an overall picture. Two-way cross-tabulations were then made. Gender, region of birth, and age were identified as potential determinants or confounders and were used in the 2-way cross-tabulations and bivariate analyses for identification of covariates significantly associated with knowledge of and attitudes to various aspects of cancer using the chi-square statistic; the category ‘not sure’ was included in the computation the reported conservative p-values.

Multivariate analysis (stratified Mantel-Haenzsel) was carried out to identify independent correlates of cancer knowledge and attitudes while controlling for confounding covariables. The MH chi-square for association as well as the MH odds ratio with 95% confidence bounds were computed for the relation between knowledge and attitude covariates on one hand and explanatory covariate of gender, age, and region of birth while stratifying for the relevant potential confounders.

Multivariate analysis (logistic regression): The following 9 representative variables were selected as response variables in logistic regression models:  knowledge of the cervix as a primary cancer site, knowledge of pain as a cancer symptom, knowledge of pap smear as a method of early cancer detection, knowledge of chemotherapy as a method of cancer treatment, knowledge of tobacco cigarettes as a cancer risk factor, knowledge that stopping smoking prevents cancer, the perception that cancer is a serious disease, the attitude to early detection of cancer in order to cure it, and seeking to have an annual physical examination as health protection. Gender, age, and region of birth were used as explanatory variables in the models. The selection of the representative variables was based on the findings of univariate and bivariate analysis as well as the high correlation of the variables with other variables in their group. The respective groups with the total number of variables in each group are: primary cancer sites 14, symptoms and signs 19, methods of early detection of cancer 9, methods of cancer treatment 4, cancer risk factors 15, cancer prevention measures 13, opinions and perceptions 12, attitudes 6, and health-seeking behaviours 6.

 

3.0 RESULTS

 

3.1 DESCRIPTION OF THE SAMPLE AND THE RESPONDENTS

 

3.1.1 RESPONSE RATES

A total of 629 questionnaires were distributed, 191 to academic staff (   male and … female), 50 to non-academic administrative staff (25 male and 25 female), and 187 to students (225 male and 225 female). Two hundred and eighty two (282) questionnaires were returned, 58 from academic staff (43 male and 14 female), 31 from non-academic administrative staff (11 male and 20 female), and 186 from students (59 male and 127 female). The university position of one questionnaire could not be classified. The overall response rate was 282/629 (  %). The response from females was proportionately higher than males for all three categories of respondents: academic staff 14/  (%) vs   53/…(%), administrative staff 20/25 (   %) vs 11/25 (   %), and students 127/225 (   %) vs 59/225 (   %).

 

3.1.2 AGE DISTRIBUTION

Virtually all students respondents were aged below 40; 184/185 (99.5%) of all student respondents were aged below 40. On the other hand 42/73.6%) of the academic staff were aged above 40 years. The proportion of respondents aged 20-40 as a total of all respondents from each region of birth varied: Sub-Saharan Africa 26/28(93.9%), Central Asia 9/10(90.0%), South-east Asia and the Far east 140/156(89.7%), Europe, Australia, and the Americas 9/12(75.0%), South Asia 17/23(73.9%), and the middle-east 10/29(34.5%). The lower proportion of young respondents from the middle-east is explained by the fact that a large number of the older academic staff are from that region.

 

GENDER DISTRIBUTION

There were more female respondents than male 162/279(%) vs 117/279(%). Female respondents were younger than male respondents. The overall proportion of those aged 20-40 was: females 155/161 (96.3%) vs males 76/116 (65.52%).

 

MARITAL STATUS

 The marital status distribution was: married 89/271 (32.8%), Widowed 1/271 (0.4%),  Divorced 0/271 (0%), and Never Married 181 (66.8%). Nobody reported divorce.

 

3.1.5 HIGHEST LEVEL OF EDUCATION ATTAINED

The study sample being a University community was highly educated holding qualifications as follows: doctorate 45/279 (16.1%), Masters 31/279 (11.1%), Batchelor 108/279 (38.7%), and undergraduates with only a school certificate 95/279 (34.1%). Under-graduates who hold only a school certificate were relatively under-represented in the sample; they form over 80% of the university community.

 

3.1.6 ETHNIC DISTRIBUTION

The respondents came from 21 ethnic groups from all over the world; the major ones being Malay 158/   (66.4%), Black African 19/   (8.0%), Albanian 18/   (7.6%), and Indo-Pakistani 7/  (2.9%).

 

REGION OF BIRTH

The respondents were from the following regions of birth: Region 1 which comprises of Europe, Australia, New Zealand, and the Americas 12/259 (4.6%), Region 2 which comprises the Middle-east and North Africa 31/259 (12.0%), Region 3 which are the countries of Sub-Saharan Africa 28/259 (10.8%), Region 4 which comprises the Central Asian Muslim Republics of the ex-Soviet Union 9/259 (3.4%), South Asia comprising India, Pakistan, Bangladesh, and Sri Lanka 23/259 (8.9%), and South-East Asian and the Far-east 156/259 (60.2%). South-east Asia, including the host country Malaysia, had the highest proportion of female respondents which mirrors the situation in the university that admits more females than males overall and markedly more males from other regions of the world than females. Academic staff from South-east Asia and the far east were 24/57 (42.1%), from the Arab world (middle-east and North Africa)  21/57 (36.4%), and a few numbers from each of the other regions.

 

EXPOSURE TO CANCER RISK FACTORS

 

3.2.1 FAMILY HISTORY OF CANCER:

(a) Univariate analysis:

Relatively few respondents recalled a history of cancer in a close relative; .the highest proportions being reported for history of cancer in the grandfather 8/131 (6.1%), and grandmother 12/134 (9.0%), and cousin 13/132 (10.0%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: females had a higher proportion of recalling cancer in relatives than males as follows: cancer in grandmother 7/64 (10.9%) vs. 1/67 (1.5%) p=0.05; cancer in grandfather 10/66 (15.2%) vs 2/68 (2.9%) p=0.03. There were no marked gender differences in reports of cancer from other relatives.

By region of birth: It was not possible to undertake a reliable analysis of family cancer history by region of birth because of the small numbers in several cells.

By age: A family history of cancer was more likely to be reported by the younger respondents aged below 40 years as compared to those aged above 40; the respective proportions for history of cancer in the grandfather were 7/8(87.5%)  vs 1/8(12.5%), for cancer in the grandmother 12/12(100%) vs  0/12(0%), and for cancer in the cousin  8/12 (66.8%)  vs  4/12(33.3%).

 

3.2.2 HISTORY OF CIGARETTE SMOKING:

(a) Univariate analysis:

The distribution of smoking status was: never smoked 221/262 (83.4%), current smoker 15/262(5.7%), stopped last year 5/262 (1.9%), stopped several years ago 24/262 (9.1%). Of those who reported being current smokers 12/16 (75.0%) smoked less than 20 cigarettes a day. Among those who reported being current smokers or to have smoked at some time in the past and have now stopped, 19/35 (54.3%) reported having started smoking in their teen years (age 10-20).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: Males reported being current smokers more than females 11/111 (19.8%) vs 4/151 (0.7%)  p=0.001. The numbers were too small for a reliable analysis number of cigarettes smoked and age at start of smoking by gender.

By region of birth: The distribution of those who never smoked revealed a marked varying regional pattern: ROB1 7/12 (58.3%), ROB2 23/30 (76.7%), ROB3 23/25 (92.0%), ROB4 10/11 (90.9%), ROB5 18/23 (78.3%), and ROB6 121/143 (84.6%). It is also worth noting that all 5 current smokers from ROB1 started in their teens whereas only 10/17 (58.8%) of those from ROB6 started in their teens.

By age: A higher proportion of the older respondents aged above 40 were current smokers compared to the younger ones aged below 40 years 4/39(10.3%) vs. 10/216(4.6%). A higher proportion of the younger current smokers started smoking in their needs compared to the older current smokers 12/17(70.6%) vs 7/16(  %). 

 

3.2.3 DIETARY HABITS:

(a) Univariate analysis:

Consumption of the following food items more than 7 times a week was: green vegetables 68/269 (25.3%),  fruits 43/268 (16.0%), and meats 71/268 (26.5%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: Males reported a higher proportion of consuming more meat more than 7 times a week 39/115 (42.9%) vs 32/153 (20.9%) p=0.02. There was no marked gender difference in the consumption of green vegetables and fruits.

By region of Birth: The proportion consuming green vegetables more than 7 times a week was highest in ROB4 5/11 (45.5%). The proportion consuming fruits more than 7 times a week was highest in ROB2 12/31 (38.7%). ROB1 had the highest proportion of meat consumption more than 7 times a week 5/12 (41.7%); the proportions for other regions were: ROB2 10/31 (32.3%), ROB3 10/28(35.7%), ROB4 3/11(27.3%), ROB5 4/22(18.2%), and ROB6 (34/148 (22.97%).

By age: There was marked age variation of weekly food frequency more than 7 times. Green vegetables: for age 20-40  47/221 (12.3%0, for age 41-60 20/41 (48.8%) , and for age 61-80 2/6(33.3%). The respective proportions for fruits were 23/220 (10.5%), 17/41 (41.5%), and 4/6 (66.7%). There was no marked age variation in meat consumption.

 

 

KNOWLEDGE ABOUT CANCER

 

3.3.1 SOURCES OF INFORMATION:

(a) Univariate analysis:

The respondents were well informed about cancer and got information from a variety of sources. The most important sources of information about cancer in order of popularity were: newspapers 232/235 (98.7%), magazines 221/224 (98.7%), radio 172/176 (97.7%), pamphlets 165/169 (97.6%), national TV 184/189 (97.4%), books 161/166 (97.0%), physicians 133/138 (96.4%),  friends 133/142 (93.7%),  school teachers 95/104 (91.4%), outdoor advertisement 87/98 (88.8%), foreign TV including cable and satellite 74/84 (88.1%), family members 92/105 (87.6%), nurses at health centres 62/71 (87.3%), nurses in hospitals 59/69 (85.5%), Other health staff 51/61 (83.6%), volunteer non-health staff 51/61 (83.6%), religious leaders 50/63 (79.4%), social clubs 54/79.4%), neighbours 53/68 (77.9%), political and government officials 37/49 (75.5%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: There were marked gender differences in the source of information about cancer. Females reported a higher proportion of obtaining cancer information from nurses in hospitals 47/52 (90.4%) vs 12/17 (70.6%)  p= 0.04, volunteer non-health staff 42/46 (91.3%) vs 9/15 (60.0%) p=0.01,  national TV 123/124 (99.2%) vs 61/65 (93.9%) p=0.05, radio 112/112 (100%) vs 60/64 (93.8%) p=0.02, newspapers 141/141 (100%) vs 91/94 (96.8%) p=0.06, magazines 140/140 (100%) vs 81/84 (96.4%) p=0.05, books 104/105 (99.1%) vs 57/61 (93.4%) p=0.06, pamphlets 122/122 (100%) vs 43/47 (91.5%) p=0.001. There were no marked gender differences for the rest of cancer information sources.

By region of birth: There were marked differences by region of birth in the source of cancer information with ROB3 (Sub-Saharan Africa) consistently showing a lower proportion than any other region: hospital nurse, national TV, radio, newspapers, magazines, books, pamphlets, friend, school teacher, outdoor advertisement; it was not possible to test for statistical significance due to small numbers in some cells.

By age: There was not marked variation by age in the sources of cancer information.

 

 

3.3.2 PRIMARY CANCER SITES:

(a) Univariate analysis:

Knowledge of primary cancer sites arranged in descending order from the highest was as follows: breast 263/266 (98.9%), lung 247/252 (98.0%), skin 205/216 (94.9%), brain 212/227 (93.4%), mouth 200/216 (92.6%), cervix 170/187 (90.9%), bone 187/207 (90.3%), stomach 172/197 (87.3%), uterus 153/186 (82.3%), kidney 128/157 (81.5%), blood 168/207 (81.2%), intestine 94/146 (64.4%), eye 89/143 (62.2%), and bladder 72/120 (60.0%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: A higher proportion of females knew the following primary cancer sites: breast 161/161 (100%) vs 102/105 (97.1%) p=0.06, cervix 125/132 (94.7%) vs 45/55 (81.8%)  p=0.01, brain 137/143 (95.8%) vs 75/84 (89.3%) p=0.05.  A higher proportion of males knew blood as a primary cancer site 70/79 (88.6%) vs 98/128 (76.6%)  p=0.04. There were no other marked gender differences for the rest of  the cancer sites.

By region of birth: There was marked variation in the proportion of those who knew of the following primary cancer sites by region of birth: cervix, mouth, bone, bladder, and brain. For these sites and others for which the variation was not marked, there was a consistent observation that the Sub-Saharan African region had the lowest proportion of those who knew the primary cancer sites.

By age: A higher proportion of older respondents aged 41-80 knew the following primary sites of cancer than younger respondents aged below 40: blood 35/36(    %) vs 132/170(77.6%) p=0.02 and intestine 18/20(    %) vs 77/127(60.6%) p=0.03. There was no other marked difference by age with regard to the rest of cancer sites investigated

 

(a) Multivariate analysis:stratified MH analysis

Gender effect: There was a significant association between gender and knowledge of the following primary cancer sites after controlling for the effect of age: breast (MH chi=4.1, p=0.04), cervix (MH chi=11.1, p=0.001). The association between gender  and knowledge of breast  and cervical cancer was not significant when controlled for region of birth (MH chi=2.57, p=0.11).

Region of birth effect:: Significant relations were found only between knowledge of the cervix as a primary cancer site and regions ROB3 (after controlling for gender and age) and ROB5 (after controlling for age). There was a significant relation between knowing the mouth as a primary cancer site and region ROB3 after controlling for age.

Age effect: There was no significant relation between knowledge of primary cancer sites and age after controlling for gender and region of birth

 

3.3.3 CANCER SYMPTOMS AND SIGNS:

(a) Univariate analysis:

Knowledge of cancer symptoms and signs in descending order was as follows: severe pain in the body 143/213 (67.1%), chronic cough 129/207 (62.3%), swelling or thickening of a part of the body 128/207 (61.8%), bloody cough 132/214 (61.7%), abnormal secretion or bleeding 139/230 (60.4%), decreased appetite 117/210 (55.7%), severe chest pain 103/204 (50.5%), non-healing ulcer 117/228 (51.3%), bone pains or fractures 101/203 (49.8%), pain on passing motion 96/201 (47.8%), pain on passing urine 95/200 (47.5%), difficulty in swallowing 96/203 (47.3%), chronic diarrhoea 74/192 (38.5%), high fever 77/223 (34.5%), frequently passing urine at night 58/191 (30.4%), insomnia 64/214 (29.9%), difficulty in passing motion, constipation 55/191 (28.8%),  increased appetite 21/208 (10.1%),  insanity 17/190 (9.0%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

Gender effect: Females reported a higher proportion of knowing cancer symptoms and signs as follows: non-healing ulcer 87/141 (61.7%) vs 30/87 (34.5%) p= 0.001, abnormal bleeding or secretions 97/143 (67.8%) vs 42/87 (48.3%)  p=0.01, decreased appetite 85/136 (62.5%) vs 32/74 (43.2%) p=0.03, swelling or thickening of the body 89/131 (67.9%) vs 39/76 (51.3% p=0.05, severe pain in body 103/135 (76.3%) vs 40/78 (51.3%)  p=0.001, chronic cough 94/132 (71.2%) vs 35/75 (46.7%) p=0.001, bloody cough 94/138 (68.1%) vs 38/76 (50.0%)  p=0.03,  severe chest pain 75/130 (57.7%) vs 28/74 (37.8%)  p=0.02, difficulty in passing motion for more than 2 weeks 43/125 34.4%) vs 12/66 (18.2%) p=0.05, frequent bone fractures and pains 80/132 (60.61%) vs 21/71 (35.0%).

Region of birth effect:: Variation in knowledge of cancer symptoms and signs by region of birth was marked for the following: non-healing  ulcer,  abnormal secretions or bleeding, decreased appetite, pain while passing motion, severe pain in the body, frequently passing urine at night, frequent bone fractures and pains. ROB2 (middle-eastern and North Africa) and ROB3 9sub-saharan African) regions consistently showed the lowest proportions of those who knew the above mentioned symptoms. The numbers in several cells were too small for testing statistical significance.

Age effect: The younger respondents aged below 40 had a higher proportion of those who mentioned the following as cancer symptoms compared to those aged above 40 years: severe body pain 127/182(69.8%) vs 15/30(50.0%) p=0.05; chronic cough 116/178(65.2%) vs 12/28 (   %) p=0.04; severe chest pain 98/177 (55.4%) vs 5/26(  %), and frequent bone pains or fractures 95/174(54.0%) vs 5/28(   %)  p=0.004.   

 

(c) Multivariate analysis:MH stratified analysis

Gender effect: There was a significant association between knowledge of the following cancer symptoms and signs and gender after controlling for age: abnormal secretions and bleeding (MH chi=5.44, p=0.02), swelling of body (MH chi=7.78, p=0.01),  severe body pain (MH chi=9.7, p=0.002), chronic cough (MH-chi=7.7, p=0.01), bloody cough (MG chi=5.2, p=0.2). The relationship was not significant when controlled for region of birth with the exception of severe body pain (MH chi =7.4, p=0.01).

Region of birth effect:  There were significant relations between knowledge of the following symptoms and the indicated regions of birth: severe body pain with ROB1 & ROB5 after controlling for age and for gender. There was a significant relation between knowledge of severe body pain as a symptom and ROB3 after controlling for age. 

Age effect: There was a significant relation between knowledge of severe body pain as a symptom of cancer and age after controlling for region of birth (MH chi=4.9, p=03) but not for after controlling for gender (MH chi=0.26, p=0.61).

 

3.3.4 METHODS OF EARLY CANCER DETECTION:

(a) Univariate analysis:

The proportions of those who knew early cancer detection methods were as follows: pap smear 130/142 (91.6%), mammography 97/111 (87.4%), breast self-examination 170/180 (94.4%),  x-ray 208/219 (95.0%), blood examination 149/162 (92.0%), physician regular check-up 160/172 (93.0%), examination of urine for blood 116/128 (90.6%), examination of stool for blood 82/94 (87.2%), and use of the CAT scan 94/105 (89.5%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age:

By gender: Females had a markedly higher proportion of knowledge of all methods of early cancer detection: pap smear 103/106 (97.2%) vs 27/36 (75.0%)  p=0.001, mammography 73/77(94.8%) vs 24/34(70.6%) p=0.002, breast self-examination 133/134 (99.3%) vs 37/46 (80.4%)  p=0.001, x-ray 132/134 (98.5%) vs 76/85 (89.4%) p= 0.01, blood examination 94/96(97.9%) vs 55/66(83.3%)  p=0.002, physician regular check-up 111/112(99.1%) vs 49/60(81.7%)  p=0.001, examination of urine for blood 79/80(98.8%) vs 37/48(77.1%)  p=0.001, examination of stool for blood 58/59(98.3%) vs 24/35(68.6%)  p=0.001, CAT scan 69/70(98.6%) vs 25/35 (71.4) P=0.001.

By region of birth: Regional differences in knowledge of early cancer detection methods was marked for all the methods investigated but testing for statistical significance was not reliable due to small numbers in some cells. South Asia had the lowest proportion of those who knew that pap smear and x-ray were methods of early cancer detection. The proportions of those who knew that CAT scan was a method of early cancer detection was the same in both sub-Saharan Africa and South Asia and was the lower than all other regions. For the remaining detection methods, the middle east, North Africa, and sub-Saharan African regions consistently had the lowest proportion of people who knew that these were methods of early cancer detection.

By age: Younger respondents aged below 40 had a higher proportion of those who knew the following methods of early cancer detection compared to older respondents aged above 40: pap smear 114/118(96.1%) vs 16/24(   %), p=0.001, breast self-examination 158/160(98.8%) vs 12/19 (  %) p=0.001,  x-ray examination 180/183 (98.4%) vs 28/35(  %) p=0.001, blood examination 128/133(96.2%) vs 21/28(   %) p=0.001, annual physician check 147/151(97.4%) vs 13/20(  %)  p=0.001, urine examination 110/114(96.4%) vs 7/14(  %) p=0.001, examination of stool for blood 76/80(95.0%) vs 7/14(   %) p=0.001, CAT scan 86/89(96.6%) vs 9/16(  %)  p=0.001.

 

(c) Multivariate analysis:MH stratified analysis

Gender effect: There was a significant relation between gender and report of knowledge of mammography as a method of early cancer detection (MH chi=4.1, p=0.04) after controlling for region of birth. The relation was not significant after controlling for age. There was a significant relation between knowledge of breast self-examination and gender after controlling for age (MH chi=3.45, p=0.06) and region of  birth (MH chi=5.1, p=0.02).  Similarly there was a significant relation between gender and the knowledge of knowledge of annual physician examination as a method of early cancer detection after controlling for age (MH chi=5.3, p=0.02) and region of birth (MH chi = 5.3, p=0.02).

Region of birth effect:

Age effect: There was a significant association between age and knowledge of pap smear as a method of early cancer detection after controlling for gender (MH chi=4.5, p=0.03) but not after controlling for region of birth (MH chi=1.8, p=0.19). A similar relation was found for mammography the respective statistics being for gender MH=5.83 p=0.02 and for region of birth MH chi=3.49,  p=0.06. There was a significant relation between knowledge of breast self-examination as a method of early cancer detection and age after controlling for gender (MH chi=12.3, p=0.001) and after controlling for region of birth (MH chi=10.3, p=0.001). A similar relationship was found for x-ray examination (for gender MH chi=9.16, p=0.002 for ROB MH chi=4.78, p=0.02), and for annual physician examination (for age MH chi=8.545, p=0.003, for ROB MH chi=4.28, p=0.04).

 

3.3.5 METHODS OF CANCER TREATMENT:

(a) Univariate analysis:

Knowledge of cancer treatment methods was as follows: chemotherapy 123/210 (58.6%), radiotherapy 169/225 (75.1%), surgery 230/250 (92.0%), and immunotherapy 65/192 (33.9%)..

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

Gender effect: There were no marked gender differences in knowledge of the various cancer treatment methods.

Region of birth effect: There was no marked variation by region of birth of the proportion of people who knew the methods of cancer treatment that were investigated.

Age effect: Older respondents aged above 40 had a higher proportion of persons who knew chemotherapy as a method of cancer treatment than young ones aged below 40 29/37(  %) vs 94/173 (54.3%) p=0.01. There was no marked age difference for other methods of treatment.

 

(c) Multivariate analysis:MH stratified analysis

Gender effect: There was a significant relation between gender and knowledge of chemotherapy as a method of cancer treatment (MH chi = 6.3, p=0.02) after controlling for age but not after controlling for region of birth (MH chi=0.95, p=0.33).

Region of birth effect:

Age effect: There was a significant association between age and report of knowledge of chemotharapy as a method of cancer treatment after controlling for gender (MH chi=11.7, p=0.001) and region of birth (MH chi=8.02, p=0.01). There was a significant relation between knowledge of radiotherapy as a method of cancer treatment and age after controlling for gender (MH chi=4.35, p=0.04) but not region of birth (MH chi=0.13, p=0.71).

 

 

3.3.6 CANCER RISK FACTORS:

(a) Univariate analysis:

The proportions of those who knew the specific risk factors for cancer were: tobacco cigarettes 249/252 (98.8%), tobacco chewing 182/190 (95.8%), high lipid diet 106/118 (89.8%),  diet low in  vegetables & fruits 135/144 (93.8%), diet low in fiber 103/116 (88.8%), viral infection 97/109 (89.0%), bacterial infection 137/149 (92.0%), early and frequent sexual exposure 136/148 (91.9%), frequent pregnancy & child-bearing 120/135 (88.9%), alcohol 164/174 (94.3%), occupational exposure 111/127 (87.4%), environmental exposure 143/156 (91.7%), irradiation 108/122 (88.5%), contaminated food 121/134 (90.3%), chemicals and drugs 150/160 (93.8%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

Gender effect: Females had a higher proportion of persons knowing the following to be cancer risk factors: chewing tobacco 119/120(99.2%) vs 63/70(90.0%)  p=0.01, high-lipid diet 75/77(97.4%) vs 31/41(75.6%)  p=0.001, diet low in fruits and vegetables 99/99(100%) vs 36/45(80.0%)  p=0.001, low-fiber diet 75/75(100%) vs 28/41(68.3%) p=0.001, viral infection 71/72(98.6%) vs 26/37(70.2%) p=0.001, bacterial infection 103/103(100%) vs 34/46(73.9%) p=0.001, early and frequent sexual exposure 107/108(99.1%) vs 29/40(72.5%) p=0.001, frequent pregnancy and child-bearing 91/95(95.8%) vs 29/40(72.5%) p=0.001, alcohol 115/115(100%) vs p=0.001, occupational exposure 85/87(97.7%) vs 26/40(65.0%) p=0.001, environmental exposure 105/107(98.1%) vs 38/49(77.6%) p=0.001, irradiation 78/82(95.1%) vs 30/40(75.0%) p=0.004, contaminated food 91/93(97.9%) vs 30/41(73.2%) p=0.001 and chemicals or drugs 99/99(100%) vs 51/61(83.6%).

By region of birth: There was marked regional difference in the knowledge that the following were cancer risk factors: diet low in vegetables and fruits, high lipid diet,  low fiber diet, viral infection, bacterial infection, early and frequent sexual exposure, alcohol, frequent pregnancy and child-birth, occupational exposures, environmental exposures, irradiation, contaminated food, chemicals and drugs. The middle-east and north Africa region had consistently the lowest proportion of persons who knew that the above were risk factors for cancer. The next lowest region was consistently sub-Saharan Africa. Testing for statistical significance was unreliable due to small numbers in some cells.

By age: Virtually all younger and older respondents knew tobacco as a cancer risk factor; the respective proportions being  208/208 (100%) vs 41/42 (  %)  for tobacco cigarettes and 153/157 (97.5%) vs 29/33 (   %) for tobacco chewing. There was a marked difference in knowledge of diet as cancer risk factor between the younger respondents aged below 40 and older respondents aged above 40 years, the respective proportions being 98/104(94.2%) vs 8/14 (   %)  for diets high in lipid and 125/128(97.7%) vs 11/17(   %) for diets low in green vegetables and fruits, and 95/101(94.0%) vs 9/16(  %) for diets low in fiber. Younger respondents had a higher proportion of knowledge of infective risk factors the respective proportions being 89/95(93.7%) vs 8/14(   %) for viral infections and 129/135(95.6%) vs 8/14(  %) for bacterial infections. The respective proportions for other risk factors were: early and frequent sexual exposure 130/136(95.6%) vs 7/13( %),  early and frequent pregnancy 112/121 (92.6%) vs 9/15 (   %),  alcohol 151/155(97.4%0 vs 13/19 (  %), occupational exposure 99/109 (90.3%) vs 12/18 (  %), and environmental exposure 128/135 (94.8%) vs 15/21 (71.4%), irradiation 94/102(92.2%) vs 14/21 (  %), contaminated food 114/121(94.2%) vs 7/13(%), chemicals and drugs 133/137(97.1%) vs 17/23(%).

 

(c) Multivariate analysis:MH stratified analysis

Gender effect: There was a significant relation between gender and knowledge of the following cancer risk factors after controlling for age and region of birth (rob): chewing tobacco (for age MH chi=5.3 p=0.02 and for ROB MH chi=4.0, p=0.05), diet low in green vegetables and fruits (for age MH chi=9.3, p=0.002, for ROB MH chi=6.7, p=0.01), alcohol (for age MH chi=10.8, p=0.001, ROB MH chi=5.2, p=0.02), chemicals and drugs (for age MH chi=9.7, p=0.002, for ROB MH chi=4.8,  p=0.03).

Region of birth effect:

Age effect: There was a significant association between age and knowledge of tobacco cigarettes as a cancer risk factor after controlling for gender (MH chi=5.1, p=0.02) but not region of birth (MH chi=3.2, p=0.7). There was a significant relation between knowledge of the following as risk factors for cancer after controlling for gender and region of birth: diet low in green vegetables and fruits (for gender NH chi=6.4, p=0.01, for ROB MH chi=9.69, p=0.002), alcohol (for gender MH chi=6.48, p=0.011, for ROB MH chi=6.4, p=0.011), chemicals and drugs (for gender MH chi=4.57 p=0.03, for ROB MH chi=2.7, p=0.10).

 

 

3.3.7 CANCER PREVENTION MEASURES:

(a) Univariate Analysis:

The proportions of those who knew various measures for preventing cancer were: stop smoking 247/254 (97.2%), low calorie diet 164/74.6%), vegetarian diet 151/231 (65.4%), regular exercise 210/240 (87.5%), avoiding alcohol 224/245 (91.4%), include green vegetables and meat in the diet 197/234 (84.2%), increase fiber in the diet 169/223 (75.8%), wash hands before meals 179/226 (79.2%), and avoid sex with multiple partners 192/229 (83.8%). The proportion who could recognise wrong measures were lower than expected: frequently eating preserved food 104/212 (49.0%), increase intake of meat 135/215 (62.8%), practise traditional medicine 46/212 (21.7%),

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

By gender: Females had a higher proportion of knowledge of the following cancer prevention measures: low calorie diet 122/146(83.6%) vs 42/74(56.7%) p=0.001, taking a vegetarian diet 96/148(64.9%) vs 55/83(66.3%) p=0.03, regular exercise 134/150(89.3%) vs 76/90(84.4%) p=0.03, increasing fiber in the diet 126/150(84.0%) vs 43/73(58.9%) p=0.001, hepatitis B immunisation 88/124(71.0%) vs 30/55(54.6%) p=0.02. There were no marked gender differences with regard to other preventive measures.

By region of birth: There were marked regional differences in the knowledge of preventive measures mentioned above with the exception of  taking vegetarian diet and avoiding sex with multiple partners.

Age: There were no marked variations by age in knowledge of stopping smoking as a cancer preventive measure. There was a marked age difference in knowledge of the following cancer preventive measures with younger respondents aged below 40 having a higher proportion of those who knew than those aged above 40 years: low calorie diet 148/188(78.7%) vs 16/31(%), vegetarian diet 132/197(67.0%) vs 19/31(%), regular exercise 181/204(88.7%) vs 28/35(%), increase in diet fiber 146/187(78.1%) vs 23/35( %), and hepatitis B immunisation 108/157(68.8%) vs 9/22(%).   

 

(c) Multivariate analysis MH stratified analysis:

Gender effect: There was a significant relation between gender and knowledge that increased fiber in the diet was a preventive factor in cancer after controlling for age (MH chi=13.1, p=0.001) and region of birth (MH chi=8.0, p=01).

Region of birth effect::

Age effect: There was no significant relation between age and knowledge of cancer prevention measures after controlling for gender and region of birth

 

3.4 OPINIONS, PERCEPTIONS, ATTITUDES & BEHAVIOURS WITH REGARD TO CANCER

 

3.4.1 OPINIONS AND PERCEPTIONS:

(a) Univariate analysis:

The following are proportions of those who agreed with the following correct propositions that reflect opinions and perceptions about cancer: cancer can be cured if detected early 241/263 (91.6%), cancer is a serious problem 252/266 (94.7%), some cancers are not death-threatening 139/247 (56.3%), some cancers can be detected early 224/258 (86.8%), . The proportion of those who disagreed with wrong propositions were: cancer is a punishment for sins and can not be cured 185/242 (76.5%), only smokers will get cancer 204/244 (83.6%), there are no facilities for early detection of cancer 185/244 (75.8%), cancers can not be cured even if detected early 180/235 (76.6%), all cancers can be detected early before any symptoms and signs 116/242 (47.9%), cancer is a very rare disease that affects only those who are unlucky 193/247 (78.1%), cancer is contagious 107/223 (48.0%). The responses to the item on cancer being hereditary reflected good understanding of the involvement of both hereditary and non-hereditary factors in cancer etiology; 83/233 (35.6%) agreed with the proposition, 77/233 (33.1%) disagreed, and 73/233 (31.3%) were not sure.

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

By gender: A higher proportion of females had correct opinions/perceptions about the following aspects of cancer: cancer is a serious problem 155/159(97.5%) vs 97/107(90.7%) p=0.04. Females also had a higher proportion of those who recognised wrong assertions and disagreed with them: only smokers get cancer 137/155(88.4%) vs 67/89(75.3%) p=0.01, there are no facilities for early detection of cancer 124/154(80.5%) vs 61/90(67.8%) p= 0.03, cancer is a rare disease that affects only those who are unlucky (135/155(87.1%) vs 58/92(63.0%) p=0.001. There were no other marked gender differences.

By region of birth: The middle-east, North African, and Sub-Saharan African regions had consistently the lowest proportion of correct perceptions on 2 items: cancer is a serious problem, and no facilities for early detection of cancer. Only sub-Saharan Africa region had the lowest proportion of correct perceptions on the item ‘only smokers get cancer’.

By age: There were marked age differences in the proportion of those who had correct opinions on cancer. The proportions in the younger respondents compared to the older respondents were as follows: cancer is a serious problem 214/221(96.8%) vs 39/44( %), some cancers can be detected early 189/212(89.2%) vs 34/45( %), cancers can not be cured even if detected early 155/195(79.5%) vs 25/39(%).  

 

(c ) Multivariate analysis:

Gender effect: There was a significant relation between the perception that cancer is rare after controlling for age (MH chi=15.4, p=0.001) and region of birth (MH chi=8.1, p=004).

Age effect:: There was a significant association between age and the perception that cancer is a serious problem after controlling for region of birth (MH chi=8.2, p=0.004). There was a significant  relation between age and the perceptions that some cancers can be detected early after controlling for gender: (MH chi=4.1, p=0.04).

 

3.4.2 ATTITUDES:

(a) Univariate analysis:

The proportions of those who agreed with the following positive attitudes were: I prefer to detect cancer early in order to save life 252/259 (97.3%), all individuals are at risk of getting cancer 196/258 (76.0%), I do not like sitting next to someone who is smoking 231/258 (89.5%), . The proportion of those who disagreed with the following negative attitudes were: I prefer traditional treatment if I have breast cancer 109/240 (45.4%), I will not see a doctor even if I have early signs of cancer 223/248 89.9%), I am not concerned about what I eat 212/251 (84.5%).

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

By gender: females had a higher proportion of those who agreed with the attitudinal assertion that all individuals are at risk of getting cancer 128/159(80.1%) vs 68/99(68.7%)  p=0.02,. There were no other marked gender differences.

By region of birth: The central Asian region had the lowest proportion of correct perceptions regarding the assertion ‘I am not concerned about what I eat’. There were no remarkable differences among the regions on other attitude items.

By age: The proportion of those who agree with the attitudinal assertion that all individuals are at risk of getting cancer was higher among the younger respondents aged below 40 compared to the older ones aged above 40, 166/215(77.2%) vs 29/41(%).

 

(c) Multivariate analysis:MH stratified analysis

Gender effect: There was a significant relationship between gender and the attitude to detect cancer early in order to save life after controlling for age (MH chi=8.1, p=0.004) and region of birth (MH chi=3.9, p=0.05).

Age effect: There was no significant association after controlling for gender and region of birth.

 

3.3.8 HEALTH PROTECTIVE BEHAVIORS:

(a) Univariate analysis:

The proportions who had undertaken health-protective behaviours at any one time in the past were: hepatitis B immunisation 117/242 (48.35%), breast examination by a doctor 52/221 (23.5%), mammography examination 16/203 (7.9%), breast self-examination 115/214 (53.7%), pap smear 15/174 (7.9%). The proportion reporting good dietary habits represented by eating the following foods more than 10 times a week were: vegetables 37/258 (14.3%), fruits 24/257 (9.3%), fish 14/257 (5.5%). Only 9/256 (3.5%) reported eating meat more than 10 times a week.

 

(b) Bivariate analysis: 2-way cross tabulation by gender, region of birth, and age

By gender: It is interesting to note that 11 males reported breast self examination, 1 did it once a week, 2 once every 6 months, and 8 once a year. 50/154 (32.5%) of the females had never undertaken breast-self examination. 39/154(25.3%) did it once week, 42/154(27.3%) did it once every 6 months, and 23/154(14.9%) did it once a year. Males had an annual physical examination more often than females 50/92(54.4%) vs 38/153(24.8%)  p=0.001. Males reported eating meat more than females. There was no significant gender difference in the frequency of green vegetables, fruits, and fish consumption.

Region of birth:

 

By age: There were marked age differences in the proportion of those who had correct attitudes on cancer, the respective proportions for the younger respondents compared to the older ones were: immunisation against hepatitis B 105/201(52.2%) vs 12/40(%), breast self-examination 109/190(%) vs 4/23(%).

 

(c) Multivariate analysis:

Gender effect: There was a significant relation between gender and report of breast self examination after controlling for age (MH chi=33.5, p=0.001) and region of birth (mh=23.2, P=0.001).. A similar relation was found for annual physical examination, the respective statistics being for age MH=14.8, p=0.001 and for region of birth MH=10.5, p=0.001.

Age effect: There were significant relation between age and the following health-seeking behaviours after controlling for region of birth: breast examination by a physician (MH chi=4.8, p=0.03),, breast self-examination (MH chi=6.44, p=0.011), getting a pap smear (MH chi=4.02, p=0.05), and an annual physical examination (MH chi=5.03, p=0.03). 

 

4.0 DISCUSSION

 

5.0 CONCLUSIONS AND FOLLOW-UP

The paper proposes incorporation of cancer education information in the weekly Friday sermon that is attended by a high proportion of Muslims varying 50 -90 % in different countries. The paper explains how religious leaders can be recruited as major players in cancer education. It uses the results of the questionnaire survey to propose an outline of a cancer education program suitable for the Muslim World. The program incorporates concepts from the Qur’an, the primary Islamic religious  scripture. This program will be piloted in Malaysia and perhaps one other Muslim country in 1998; the results will be reported at the 15th Asia- Pacific Cancer Conference in 1998.

 

REFERENCES

1. Andrykowski-MA; Munn-RK; Studts-JL: Interest in learning of personal genetic risk for cancer: a general population survey. Prev-Med. 1996 Sep-Oct; 25(5): 527-36

 

2. Bonelli-L; Branca-M; Ferreri-M; Barizzone-D; Rossi-E; Cedri-S; Bruzzi-P; Santi-L: Attitude of women towards early cancer detection and estimation of the compliance to a screening program for cervix and breast cancer. Cancer-Detect-Prev. 1996; 20(4): 342-52

 

3. Fakhfakh-R; Hsairi-M; Ben-Romdhane-H; Achour-N; Ben-Ammar-R; Zouari-B; Nacef-T: [Smoking among medical students in Tunisia: trends in behavior and attitudes]: Le tabagisme des etudiants en medecine en Tunisie: tendances des comportements et des attitudes. Sante. 1996 Jan-Feb; 6(1): 37-42

 

4. Fusco-Karmann-C; Gangeri-L; Tamburini-M; Tinini-G: Italian consensus on a curriculum for volunteer training in oncology. J-Pain-Symptom-Manage. 1996 Jul; 12(1): 39-46

 

5. Ghali-WA; Freund-KM; Boss-RD; Ryan-CA; Moskowitz-MA: Menopausal hormone therapy: physician awareness of patient attitudes. Am-J-Med. 1997 Jul; 103(1): 3-10

 

6. Harris-MA; Byles-JE: A survey of screening compliance among first degree relatives of people with colon cancer in New South Wales. J-Med-Screen. 1997; 4(1): 29-34

 

7. Hodge-FS; Fredericks-L; Rodriguez-B: American Indian women's talking circle. A cervical cancer screening and prevention project. Cancer. 1996 Oct 1; 78(7 Suppl): 1592-7

 

8. Hubbell-FA; Chavez-LR; Mishra-SI; Valdez-RB: Beliefs about sexual behavior and other predictors of Papanicolaou smear screening among Latinas and Anglo women. Arch-Intern-Med. 1996 Nov 11; 156(20): 2353-8

 

9. Hubbell-FA; Chavez-LR; Mishra-SI; Valdez-RB: Differing beliefs about breast cancer among Latinas and Anglo women. West-J-Med. 1996 May; 164(5): 405-9

Jacob-TC; Penn-NE; Brown-M: Breast self-examination: knowledge, attitudes, and performance among black women. J-Natl-Med-Assoc. 1989 Jul; 81(7): 769-76

 

10. Lacey-LP; Phillips-CW; Ansell-D; Whitman-S; Ebie-N; Chen-E: An urban community-based cancer prevention screening and health education intervention in Chicago. Public-Health-Rep. 1989 Nov-Dec; 104(6): 536-41

 

11. Lantz-PM; Weigers-ME; House-JS: Education and income differentials in breast and cervical cancer screening. Policy implications for rural women. Med-Care. 1997 Mar; 35(3): 219-36

 

12. Mahoney-MC; Michalek-AM; Wiggins-CL; Tenney-M; Bad-Wound-D; Burhansstipanov-L: Native American Cancer Conference III. Cognitive correlates and impressions of attendees. Cancer. 1996 Oct 1; 78(7 Suppl): 1533-7

 

13. Monneuse-MO; Bellisle-F; Koppert-G: Eating habits, food and health related attitudes and beliefs reported by French students. Eur-J-Clin-Nutr. 1997 Jan; 51(1): 46-53

 

14. Morrison-G: Sun exposure and skin cancer development: nurses' attitudes. Nurs-Stand. 1996 May 29; 10(36): 39-42

 

15. Myers-RE; Wolf-TA; McKee-L; McGrory-G; Burgh-DY; Nelson-G; Nelson-GA: Factors associated with intention to undergo annual prostate cancer screening among African American men in Philadelphia. Cancer. 1996 Aug 1; 78(3): 471-9

 

16. Myhre-SL; Li-VC; Guan-JH; Wang-ZJ: Cancer knowledge and perceptions among Chinese factory workers: implications for cancer control and prevention. Cancer-Detect-Prev. 1996; 20(3): 223-33

 

17. Patterson-RE; Kristal-AR; White-E: Do beliefs, knowledge, and perceived norms about diet and cancer predict dietary change? Am-J-Public-Health. 1996 Oct; 86(10): 1394-400

 

18. Portnoy-B; Christenson-GM: Cancer knowledge and related practices: results from the National Adolescent Student Health Survey. J-Sch-Health. 1989 May; 59(5): 218-24

 

19. Souvignier-ST; Mayer-JA; Eckhardt-L: Educating the public about skin cancer prevention: a role for pharmacists.: J-Clin-Pharm-Ther. 1996 Dec; 21(6): 399-406

 

20. Vennin-P; Giard-S; Julian-Reynier-C; Sailly-F; Peyrat-JP; Fournier-C; Eisinger-F; Sobol-H [Attitudes towards screening and prevention of breast and ovarian cancers with hereditary predisposition. Survey by female gynecologists in the north of France]: Attitudes envers le depistage et la prevention des cancers mammaires et ovariens avec predisposition hereditaire. Enquete aupres des femmes gynecologues de la region Nord de la France.: Bull-Cancer. 1996 Sep; 83(9): 697-702

Professor Omar Hasan Kasule November 1997