Abstract
Purpose:
Cancer awareness can help reduce the associated morbidity and mortality. There is no scale for the assessment of general cancer awareness in Turkiye. This study aimed to evaluate the cultural relevance and psychometric properties of the cancer awareness measure (CAM) for use in adolescents in the Turkish population.
Methods:
This methodological study included 400 students attending seven high schools in western Turkiye, recruited by convenience sampling. The CAM was translated into Turkish and assessed for content validity by expert panel opinion and content validity index (CVI). The Turkish CAM was administered online, and its reliability was assessed using Cronbach’s alpha coefficients, item-total and item-subscale correlation coefficients, and test-retest analysis.
Results:
The Turkish CAM had excellent content validity (scale CVI = 0.99). Cronbach’s alpha for the entire scale was 0.72, and item-subscale correlation coefficients were 0.33–.75 (p < 0.001). Test-retest reliability was good, with an intraclass correlation coefficient of 0.83 for the entire scale (p < 0.001). The students’ mean score on the Turkish CAM was 45.60 ± 7.23 (on a scale of 11–77), indicating moderate cancer awareness.
Conclusion:
The Turkish CAM is a valid and reliable measure that can be used to assess cancer awareness in adolescents in Turkiye. Interventional studies should be performed to measure and increase cancer awareness and promote cancer prevention among young people.
Introduction
In many countries, knowledge about the causes of cancer is an important tool in cancer prevention.1,2 Numerous studies conducted in Turkiye have shown that people lack accurate and sufficient information about cancer.3–5 Identifying risk factors and raising awareness are essential in the fight against cancer. 6 A major contributor to the high mortality rates of cancer is late diagnosis and treatment resulting from low awareness. 7 Therefore, people need to gain awareness, even from a young age, and make the necessary lifestyle modifications to prevent cancer and ensure early diagnosis. 8 Although awareness alone may not be enough to motivate change, it is unrealistic to expect a change in behavior if people are not aware of the risk factors. 6
Primary prevention of cancer should focus on modifiable risk factors. 9 The four main behavioral risk factors associated with one-third of cancer deaths are tobacco and alcohol use, low fruit and vegetable intake, high body mass index, and lack of physical activity. It is estimated that many cancers can be prevented by changing these behaviors, and approximately half of all cancers can be prevented through lifestyle modification.10,11
Adolescents, defined as people aged 10–19 years, constitute approximately 16% of the world population and 15.1% of the population in Turkiye.12,13 Adolescents can encounter many health-related risk factors during this period of body, personality, and social changes. 14 The negative effects of risky lifestyles on health are an important problem. An inverse relationship has been observed between actual risk behavior and the desire to engage in health-promoting behaviors. Therefore, students with a higher risk profile have lower intentions to engage in health-promoting behaviors.15,16
It should be kept in mind that cancer-related risk factors may emerge at a young age, and therefore, health promotion should begin early. 16 Unfortunately, it has been reported that adolescents’ health promotion behaviors are not at the desired level.17,18
Most school-age children are poorly informed and have low awareness about cancer risk factors.18,19 Only a few studies worldwide have provided data on the cancer awareness of adolescents.18–21 Similarly, there has been little research examining how informed and aware adolescents in our country are of cancer. In the literature, Turkish authors have only studied awareness of breast cancer, 22 oral cancer, 23 skin cancer, 24 and knowledge levels and attitudes toward cancer in general 14 among high school student populations.
Health education interventions can be designed to increase cancer awareness in the community. However, reliable and validated tools are needed to evaluate the effectiveness of these interventions. Different questionnaires have been developed to assess awareness of specific cancers, 25 and there are scales to measure general cancer awareness in the adult population.26,27 As a result, researchers use various assessment tools, reducing the comparability of existing studies. The cancer awareness measure (CAM) was developed in England to assess cancer awareness in society. 28 However, in our review of the national and international literature, we were unable to find any studies adapting the CAM to different cultures.
Materials and Methods
Purpose
The present study aimed to fill the gap in the literature and facilitate research on cancer awareness in Turkish society by translating the CAM and evaluating the cultural appropriateness and psychometric properties of the Turkish version of the CAM (Tur-CAM) in adolescents.
Study setting and sample
The study consisted of 6248 students studying in seven public high schools in the western part of Turkey during the 2021 academic year. The sample was recruited by convenience sampling and data were collected by online survey (Google Forms). Inclusion criteria were being a student in one of the target schools, volunteering to participate in the study, signing the informed consent form, and having internet access. There were no exclusion criteria. The minimum necessary sample size for this study was determined based on the recommendation to include five or more samples per scale item in methodological studies.29,30 As the CAM consists of 47 items, the minimum sample size was determined to be 235. However, to maximize the representativeness of the sample, the study was conducted with all students who agreed to participate. Therefore, the sample comprised a total of 400 students who met the inclusion criteria. In terms of distribution, 40 students were recruited from three of the seven schools (10% of the sample each), 60 students from two of the schools (15% each), and 80 students from two of the schools (20% each).
Data collection tools
The online survey used for data collection consisted of basic demographic questions and the Tur-CAM. The demographic part of the form asked about the student’s age, gender, class, and their perceived awareness of cancer symptoms and cancer prevention.
The CAM is a 47-item scale developed by Cancer Research UK to assess cancer awareness levels and identify risk factors for poor cancer awareness.28,31 The items are grouped into the following subscales: warning signs (10 items: 1 open-ended question and 9 recognition items); seeking help (10 multiple-choice items); barriers to seeking help (11 items: 4 emotional, 3 practical, 3 service-related, and 1 other barrier); risk factors (12 items: 1 open-ended question, 11 recognition items); cancer and age (1 multiple-choice item); most common cancers (6 open-ended items asking the top three cancers for men and women); and Ministry of Health (MOH) screening programs (6 items: cervical, breast, and bowel screening programs and the age at which screening is offered for each).
Scoring is based on the number of correct responses in the subscales of warning signs (9 recognition items, score 0–9 points), risk factors (11 recognition items rated on a 4-point Likert-type scale, score 11–55 points), cancer and age (1 item, score 0–1 point), most common cancers (6 items, score 0–6 points), and MOH screening programs (6 items, score 0–6 points). Thus, the total score ranges from 11 to 77, with higher scores reflecting higher cancer awareness.
The researchers who developed the CAM reported a Cronbach’s alpha coefficient of 0.77–.90 for the entire scale, good test-retest reliability, and item-total correlations above 0.70. 28 In the assessment of item difficulty, most items were answered correctly by > 20% and < 80% of respondents. 28
Translation and content validity of the CAM
Linguistic validity
For language validity, the measure was translated from English to Turkish by a linguist and a subject-matter expert, both of whom are proficient in both languages and familiar with both cultures. These translations were combined by the researcher into a single Turkish draft version of the CAM by one of the researchers. This was then back-translated into English by a linguist and a subject-matter expert who had not seen the English version before and had a good command of both languages and cultures. The conceptual and language equivalence of the Tur-CAM were assessed.
Content validity
The Tur-CAM was submitted to a panel of seven nurse specialists for their expert opinion. The nurse specialists were members of the nursing faculty and were knowledgeable and experienced in the area of cancer. The experts were given the original and Tur-CAM and asked to score the Turkish items as very appropriate (1), appropriate (2), minor revision needed (3), or major revision needed (4).
Pilot study
After ensuring content validity, the intelligibility of the items was assessed by pilot testing the Tur-CAM with a small group of age-matched adolescents who attended a different school and shared the same characteristics as the sample group. 32 Twenty or more participants are recommended in a qualitative pilot study for item analysis. 32 Therefore, 20 adolescents were recruited for the pilot test. No negative feedback was received from the pilot group, and therefore no changes were made to the Tur-CAM.
Data collection and reliability analyses
Data were collected from the study sample using an online survey hosted on Google Forms. After obtaining written permission from the Ministry of National Education and the university ethics committee, a link to the survey (including informed consent form) was sent to the WhatsApp accounts of the participants’ families by the vice principals of the schools. It took the students 5–7 minutes to complete the survey.
To determine the invariance of the Tur-CAM over time, test-retest reliability was assessed by administering the scale to 20 students again after an interval of 3 weeks. The results were compared using Wilcoxon’s signed rank test and intraclass correlation coefficient (ICC). Internal consistency reliability was evaluated by calculating Cronbach’s alpha coefficients and analyzing item-total and item-subscale score correlations.
Data analysis
The data were analyzed using SPSS version 27. Continuous data were analyzed for normal distribution using graphical methods and skewness (between −0.36 and 0.78) and kurtosis (between −0.49 and 1.70) values. As they were normally distributed, the students’ descriptive characteristics and Tur-CAM scores were expressed using mean and standard deviation for continuous variables and number and percentage for categorical variables. In addition, the validity and reliability analyses described above were performed. The statistical significance level was accepted as p < 0.05.
Ethical considerations
Permission was obtained from Kirstie Osborne via email to conduct the validity and reliability of the CAM in Turkiye. The study was also approved by the Dokuz Eylül University Ethics Committee (meeting no: 2021/01-1, protocol no: 5880-GOA). In addition, necessary permissions were obtained from the participating schools via the İzmir Provincial Directorate of National Education. A parental consent form was sent to the students and their families, and those who provided consent were invited to participate in the study. The participants provided online consent again at the start of the survey.
Results
Sociodemographic features
The students in the study had a mean age of 15.89 ± 0.92 years (range: 15–18 years) and 75.5% (n = 302) were female. Most of the students were in the 9th (33.5%) and 10th (37.5%) grades (Table 1).
Basic Demographic Characteristics of Students (n = 400)
Knowledge Level of Students on Cancer Symptoms and Prevention (n = 400)
Test-Retest Analysis Results of CAM (n = 20)
Bold face values are significant with p < 0.05. >
Z*, Wilcoxon-signed Rank Test. CAM, cancer awareness measure.
When asked to rate their knowledge about cancer symptoms and cancer prevention on a scale of 1 to 10, the proportion of students who rated their knowledge as 1–5 was 67.8% for symptoms and 65% for prevention (Table 2).
Psychometric properties of the Tur-CAM
Validity
The content validity of the Tur-CAM was assessed using the Davis method. Content validity index (CVI) values were determined based on the experts’ scores. The CVI of the items ranged from 0.86 to 1, and the CVI of the whole scale was found to be 0.99. Therefore, the CVI values for both items and scale were above the acceptability threshold of 0.80.30,33
Reliability
The Tur-CAM showed good consistency over time, with no significant differences between the mean scores of the two measurements made 3 weeks apart (p > 0.05). ICCs between the test and retest scores were 0.83 for the entire CAM (p < 0.001) and ranged from 0.77 to 0.93 for the subscales (p ≤ 0.001) (Table 3).
In our analysis of the internal consistency reliability of the Tur-CAM and its subscales, the Cronbach’s alpha coefficient was found to be 0.72 for the whole scale. Cronbach’s alpha coefficients for the Tur-CAM subscales were 0.74 for warning signs, 0.62 for risk factors, 0.62 for the most common cancers, and 0.72 for the MOH screening programs and age at first invitation.
The results of item-total and item-subscale score correlation analysis of the 33 items with numerical scores are presented in Table 4. One item showed a very weak and nonsignificant item-total correlation (question 7: age at which cancer risk is highest) (r = 0.06, p > 0.05), while two other items had weak but highly significant correlation (question 8 b: lung cancer is the second most common cancer in women [r = 0.14, p < 0.001] and question 8c: intestinal cancer is the third most common cancer in women [r = 0.13, p < 0.001]). The item-total correlation coefficients of the other 30 scored items ranged from 0.20 and 0.53 and showed high statistical significance (p < 0.001).
Item Total and Item Subdimension Total Score Correlations of CAM (n = 400)
Item-subdimension analysis could not be performed because there is only one item.
rs*, Spearman correlation analysis; r**, Pearson correlation analysis. CAM, cancer awareness measure.
Item-subscale score correlation coefficients were 0.46–.63 for warning signs (p < 0.001), 0.33–.56 for risk factors (p < 0.001), 0.37–.70 for the most common cancers (p < 0.001), and 0.48–.75 for MOH screening programs and their starting ages (p < 0.001). These results indicated that all items were reliable. As the cancer and age subscale consisted of only one question, item-subscale correlation could not be analyzed.
Discussion
This study was conducted to investigate the cultural relevance and psychometric properties of the CAM in Turkish adolescents. The results show that the Tur-CAM has high content validity according to expert panel opinion. Confirmatory and exploratory factor analyses could not be performed because not all CAM items were scored on a Likert-type scale. These analyses were also not included in the validity and reliability study of the original measure. 28
In the literature, it is stated that expert opinion should be obtained from at least 3 and at most 10 people to evaluate the content validity of an assessment tool. 32 In the present study, seven experts were consulted, and CVI was used as a measure of concept validity. Therefore, the CVI of 0.99 (p < 0.001) we obtained for the entire scale indicates excellent agreement among the experts who evaluated the content validity of the Tur-CAM.30,33
In test-retest reliability analysis, comparison of the students’ mean scores obtained 3 weeks apart showed no significant differences (p > 0.05). An ICC value of at least 0.70 is sought between test and retest results. ICC values of 0.70−.84 are interpreted as moderate, 0.85–.94 as high, and 0.95-1 as excellent reliability. 34 In this study, the ICC values for the total scale and subscales ranged from 0.77 to 0.93, indicating that the Tur-CAM has good reliability across time. Similarly, in the original development and validation study, the CAM was administered again to 94 people at an interval of 2 weeks, and high Pearson’s correlation values (r = 0.72–.86) were found for all subscales except the most common cancers (r = 0.33; p < 0.001). 28
The Cronbach’s alpha coefficient is a measure of internal consistency reliability. In our study, the Cronbach’s alpha of the entire Tur-CAM was 0.72, indicating adequate reliability and similar to the value of 0.77 reported in the original development study by Stubbings et al. 28
In reliability analysis, item-total score correlation coefficients are used to indicate that the items can measure the behavior at the desired level. 32 The lower limit for the item-total score correlation value is 0.20, items with values of 0.30–0.40 are acceptable, and those with values higher than 0.40 are interpreted as very discriminating and reliable. 34 The item-total score analysis is an analysis that reflects both titles in validity and reliability studies and tests the construct validity of the measures. 34 The item-total correlation coefficients in our study ranged between 0.06 and 0.53 and were statistically significant for all items except one. In the original study, all item-total correlation coefficients were greater than 0.20. 28 The item-subscale correlations in our study were significant for all items in all subscales except for the cancer and age subscale, which could not be analyzed because it contains a single item.
Although the question about the relationship between cancer and age and the questions about the second and third most common types of cancer in women had low item-total correlation coefficients, we decided not to remove these items because of their theoretical importance, the positive expert opinion about these items, their high item-subscale correlations, and the fact that removing them did not significantly change the Cronbach’s alpha of the measure (α = 0.72 before and after removing these 3 items). The total measure score and reliability coefficients of the other 30 items in the measure were found to be sufficient.
Strengths and limitations of study
A strength of the present study was the size and diversity of the sample, which included students from seven different high schools and representing all four grade levels. However, the study had some limitations. First, the scale structure was not conducive to exploratory and confirmatory factor analysis. However, we believe the large sample size of this study may reduce the impact of this limitation. Second, the lack of instruments measuring similar constructs in the Turkish literature prevented us from assessing the convergent validity of the scale. Third, there are no studies evaluating the psychometric properties of CAM in other languages, so we could not compare our results with those obtained in different cultures.
Implications for practice
Enabling measurement tools with proven validity and reliability in one culture to be used to evaluate the results of interventional studies conducted in the relevant age groups of another culture is an efficient and cost-effective approach. It allows the collection of data and evaluation of changes over time and facilitates the generation of interventional evidence.
Conclusions
To the best of our knowledge, this is the first time the validity and reliability of the CAM have been evaluated in another culture. Tur-CAM was found to have linguistic, conceptual, and substantive validity consistent with the original scale. Our results demonstrate that the Tur-CAM is a valid and reliable tool that can be used to assess cancer awareness among Turkish adolescents.
This study could help to develop better-tailored and thus superior interventions to prevent cancer in adolescents. Interventional studies should be conducted and their outcomes evaluated with the Tur-CAM to measure adolescent cancer awareness in Turkiye. We believe that the results of this descriptive research will inspire numerous interventional studies.
Footnotes
Acknowledgments
The authors thank all the participating students for their role in this study.
Authors’ Contributions
Both authors participated in study design, literature review, data analysis, interpretation of the results, writing, and critical review of the article.
Author Disclosure Statement
The authors report no actual or potential conflicts of interest.
Funding Information
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
