© The authors, 2025

Introduction

Oral cancer ranks 16th among the most common malignant neoplasms in the world [1], and squamous cell carcinoma (SCC) accounts for 91% of these cases. Nearly 50% of individuals diagnosed with this condition die within five years [2], and those who survive experience a negative impact on their quality of life, as their functions, aesthetic appearance, and psychological well-being are compromised [3]. It is necessary to develop strategies for the prevention, identification, and reduction of individual and collective risks, early detection of the disease, and the implementation of health education plans to increase cure and survival rates [4,5].

For the development of any health strategy, factors related to the individual must be considered including their beliefs, [6]. as the perceptions or convictions that individuals have about the likelihood of an event occurring and the consequences of that event. In health, the Health Belief Model (HBM) addresses this need by including a cognitive/perceptual component. The HBM seeks to explain behaviours such as individual responses to symptoms, adherence to medical treatments, self-examination practices, and individuals' participation in promotion and prevention programs [7,8]. The main value of the model lies in its applicability to large population groups, serving as a quantitative instrument, and after validation, it allows for method replication in other populations, associating beliefs with the cultural environment [8,9].

The model focuses on four dimensions: (1) Perceived Threat: assesses how individuals perceive their vulnerability to disease, as well as their beliefs about the severity of contracting it, the consequences of not treating it, and the social impact of suffering from it. (2) Evaluation of Behaviours-Barriers: assesses how individuals perceive the obstacles or barriers they face when attempting to change behaviour. (3) Evaluation of Behaviours-Benefits: assesses the importance an individual places on committing to a behaviour that reduces the disease risk. (4) Cues to Action: identifies the types of stimuli necessary to trigger health-related decision-making, whether internal (symptoms) or external (recommendations from others or information from the media). Some versions include the dimension of self-efficacy as a key component in the self-management of healthy behaviours [8].

One strategy to understand health beliefs is to apply instruments, as it is a low-cost technique that allows for reaching more participants and facilitates data analysis [10]. Additionally, if the instrument is validated, it provides better insight into the behaviour of people facing or experiencing an event, enabling more accurate intervention planning. In this regard, the number of studies and validated instruments related to the HBM in oral cancer is limited. However, it is effective in characterizing a population and designing educational strategies that positively impact early detection and reduce morbidity and mortality from this cause in our setting [10].

Oral health beliefs have been explored in some studies; however, beliefs regarding oral cancer have been less explored, possibly due to a professional-focused approach or the lack of instruments for measuring them [11–13]. Additionally, studies addressing beliefs about oral cancer tend to focus on healthcare professionals, where moderate knowledge and screening practices are reported [12]. However, from the patient's perspective, the topic is underexplored. Previous studies in Asia with immigrants identified a general lack of knowledge about the risk of oral cancer (43–76%), and more than 50% of people with one or more risk behaviours for oral cancer, such as smoking [13]. Nevertheless, the subject is rarely studied in Colombia and Latin America.

For all these reasons, it was considered necessary to develop, validate, and apply an instrument on beliefs about oral cancer that could be used and implemented to guide future educational, promotional, and preventive activities for this disease in Colombia.

Methodology

Design and population

A cross-sectional study was conducted and an instrument on beliefs about oral cancer was validated. The general population treated in the clinics of the School of Dentistry at the University of Antioquia participated, using non-probabilistic sampling according to the clinic's schedule.

Design phase and content validity

Seven experts and two regular patients from the clinics participated in the evaluation process, with the aim of achieving a Content Validity Index (CVI) of 0.58 or higher [14–16]. The experts were dentists with a postgraduate degree and with at least two years of clinical experience.

Application phase and validity of structure and internal consistency

A sample of no fewer than five participants per item was calculated to reach adequate validity levels [16]. The original version of the instrument consisted of 62 items, which were reduced to 48 items after the first round of development for application to the population. Based on this number of items, a sample of 240 individuals was required. The final sample consisted of 318 individuals.

Instruments

The Survey of Beliefs about Oral Cancer (SBOC) is a 48-item, self-administered scale from the first phase of this study. Each item was rated on a six-point Likert scale, where 1 = strongly disagree, 2 = moderately disagree, 3 = somewhat disagree, 4 = somewhat agree, 5 = moderately agree, and 6 = strongly agree. Scores were calculated by dimension and for the total scale, corresponding to the sum of the responses to each item.

The SBOC was built on the theoretical foundation of the Health Belief Model, which is based on the construct of beliefs, attitudes, and practices, with four dimensions: 1. Perceived threat: consists of 18 items. 2. Evaluation of behavioral barriers: consists of 9 items. 3. Evaluation of Behaviors-Benefits: consists of 7 items, and 4. Cues for Action: consists of 14 items.

Procedure and analysis

Design phase and content validity

In the initial phase, the first version of the survey with 62 items was developed based on the literature. Subsequently, the appearance and content of the survey was evaluated by six experts and two patients, resulting in a revised version with 49 items. After a second round with seven experts (four from the first round and three new participants), two items were added, three items were removed, and a final version of 48 items was approved for administration to patients. The process used the Delphi method, supported by an instruction manual and an Excel template which was used by the experts to assess the sufficiency, clarity, coherence, and relevance of each item. Items were rated on a scale of 1 to 4, with responses categorised as “essential” (ratings of 3 and 4) or “non-essential” (ratings of 1 and 2).

For the essential/non-essential options, the content validity index (CVI) proposed by Lawshe [14] was calculated as the average of the acceptable items. The minimum acceptable CVI was 0.58 [11,12]. In addition, Fleiss' [14] kappa was used to assess inter-expert agreement, with kappa values considered fair (0.40 to 0.60), good (0.60 to 0.75), and excellent (greater than 0.75).

Application phase and validation of structure and internal consistency

In order to assess the structural validity of the survey, a confirmatory factor analysis (CFA) was conducted using a structural equation modelling approach. The Weighted least squares (WLS) method and oblique geomin rotation were used. Model fit was assessed using the Comparative Fit Index (CFI) and the Tucker-Lewis Index (TLI), with values close to 0.90 indicating good fit, and the Root Mean Square Error of Approximation (RMSEA), with acceptable values ranging between 0.04 and 0.08 [17]. The internal consistency of the survey and its dimensions was assessed using Cronbach's alpha, with values between 0.7 and 0.9 considered acceptable [14,15,16].

Additionally, differences in SBOC scores by socio-economic variables were examined. Due to the non-normal distribution of the scores and their dimensions, non-parametric tests (Mann–Whitney U and Kruskal–Wallis) were used. A significance level of 0.05 and a confidence level of 95.0% were accepted for the tests. Analyses were performed using STATA version 14.

Results

Of the 318 participants in the study, 214 were women (67.3%) and 85 were men (26.7%), ranging in age from 18 to 83 years, with a mean age of 42 years (SD 15.1). Of the participants, 37.4% had completed high school and 21.1% had a technical education. The majority of participants belonged to the middle (46.9%) and lower socioeconomic classes (45.9%) (Tab. I).

Face validity and expert judgment

Two rounds of Delphi were conducted. In the first round, 62 items were validated, resulting in 49 items. In the second round, the 49-item scale was evaluated, resulting in the elimination of 3 items: “Healthy food is more expensive”, “I have more important things to think about than mouth cancer”, and “I never talk about mouth cancer with my friends or family”. Two new items were added, resulting in a final version of 48 items to be administered to the population. Coherence and clarity indicators showed adequate CVI and kappa values; relevance indicators showed moderate results, with kappa values close to 0.6 (Tab. II).

Structural validity and internal consistency

The 48-item version, resulting from the content validation process with experts, was subjected to Confirmatory Factor Analysis (CFA) to assess whether the theoretical four-dimensional model fit the empirical data. During the CFA process, the model was reduced from 48 to 18 items. This final version of 18 items with three dimensions (perceived threat, benefits, and cues to action) showed the best fit to the theoretical model (Fig. 1). The CFI and TLI indicators were approximately 0.92, and the RMSEA was 0.051 (95% CI: 0.038–0.063) with an SRMR of 0.063.

The SBOC, with 18 items validated by experts and confirmed through CFA, showed an internal consistency of 0.82 (95% CI: 0.78–0.85) for the entire scale. Among its dimensions, the highest Cronbach's alpha was found in the perceived hreat dimension (0.75, 95% CI: 0.71–0.80). For the Benefits dimension, it was 0.65 (95% CI: 0.59–0.71), and for Cues to Action, it was 0.85 (95% CI: 0.82–0.88). The final list of items can be found in Supplement 1.

After determining the factorial structure, scores were calculated for each dimension and the total score of the survey. All scores were between 1 and 3 for all dimensions. The highest average score was observed in the Cues to Action dimension, with a mean of 2.02 (SD 0.48), while the lowest score was for Benefits (median = 1.09, SD 0.26). Differences in the SBOC scores were found by gender, age and marital status. A higher median score was observed for women in the perceived threat dimension (1.86, p-value = 0.040) and in common law union (2.00, p-value =0.016). Statistically significant differences were found, with higher medians (above 2.0) in participants under 40 years of age for the Perceived threat and Cues to Action dimensions (p-value < 0.05) No statistically significant differences were found for the other variables analysed (Tab. III).

Table IV shows the proportions of participants according to the response options. In dimension 1 (perceived threat), 75.6% of participants agreed with the statement that they could get cancer; another significant proportion (43.5%) also agreed that a sore mouth could be a sign of cancer. In Dimension 2, more than 80% of participants agreed with each of the five statements. In the Cues to Action dimension, a significant proportion of participants indicated that they did not know how to respond to certain concerns.

Fig. 1 SBOC model with three dimensions and 18 Items. Note: Chi2: 210.873, p-value = 0.000. RMSEA: 0.051, 95% CI: 0.038–0.063. AIC: 6781.21. CFI: 0.93. TLI: 0.916. CD: 0.99. SRMR: 0.063.

Discussion

Based on the Health Belief Model, this study presents the structural and content validation of an 18-item questionnaire with adequate content validity (CVI > 0.58), good and excellent inter-rated agreement among the participating experts (kappa 0.6), and acceptable and good internal consistency values (Cronbach's alpha 0.82). It also showed a good factorial structure (RMSEA: 0.051; CFI and TLI 0.92). In this sense, the study aims to propose an instrument to assess beliefs related to oral cancer, applying the theoretical foundation of the HBM [7,8] taking into account dimensions such as perceived threat, behaviours-benefits, and cues to action, since the evaluation of behaviours-barriers dimension did not show an adequate CFA.

Using the validated survey, results showed that the majority of the study population believed they could develop oral cancer and had a clear understanding of the benefits. Women, common law union and those under the age of 40 had the clearest beliefs about the threats. The latter group also provided the impetus and direction for educational campaigns, although there was still a lack of clarity about how to reduce the risks.

The study was facilitated by the fact that most of the study population was a captive audience from the school clinics. However, a limitation is that this population, as clinic attendees, may receive more information and education than the general population, and the results cannot be extrapolated to the Colombian population.

HBM has been widely used in the literature for various diseases, including oral health [18]. However, there is a paucity of studies on oral cancer. Topics such as public knowledge, awareness of oral cancer, risk factors and manifestations have been studied [19], as well as attitudes and practices in different populations. These studies conclude that knowledge is generally low, especially in lower socioeconomic groups [20]. Self-examination habits are lacking, risk factors are not well understood, and certain populations are strongly influenced by religious beliefs [20]. However, few studies have used the HBM. Those that have used it have demonstrated the effectiveness of educational interventions in changing beliefs about various diseases and habits, such as smoking and oral hygiene, and their influence on oral cancer [21].

According to socio-demographic data, this study had a higher participation of women than men, which is consistent with other studies showing that women are more likely to seek health services and participate in survey-based studies.

Although there is limited literature on the use of HBM in oral cancer, an important finding of this study is that a high percentage of the population (75.8%) believe that they could develop oral cancer, similar to other studies in countries such as Malaysia (83%) [22]. This is important because previous studies have shown that the perception of vulnerability and seriousness of the disease motivates preventive behaviours. The perception of disease and the usefulness of preventive measures influence the implementation of preventive behaviours.

In this study, we did not ask about beliefs about known risk factors such as tobacco and alcohol, as several knowledge-based surveys already shown that the population is well informed about these. Instead, we focused on less well-known risk factors. For example, the belief that factors such as sun exposure, oral sex, and poor diet could cause oral cancer was low. To the best of our knowledge, this is the first study using the HBM that examined these types of risk factors.

Another engagement result is that there is an association between perceived threat, gender, common law union and age under 40, suggesting that young women are more aware than men of the possibility of developing oral cancer, early signs of oral cancer, and certain risk factors. Despite the fact that the relationship between marital status and oral cancer beliefs has been investigated in other studies, not such relationship has been identified [23]. This finding is novel and should be explored further.

This underscores the importance of creating educational campaigns to increase accurate beliefs about oral cancer risk factors other than tobacco and alcohol, particularly in the under-40 population, where there is an increasing number of individuals developing oral cancer without traditional risk factors. A significant percentage of this group still believes that oral cancer mainly affects older adults.

On the other hand, regardless of socio-demographic and economic variables, most respondents believe they can benefit from reducing their risk of developing oral cancer by being informed, making lifestyle changes, seeking timely medical advice, and investing in their health. This underscores the importance of promoting health promotion at both the community and individual levels during consultations, as there is a willingness to change. Studies such as Khwankong S. and colleagues [20] show different results, where a strong religious influence (Muslim) influenced beliefs about oral cancer.

It is important to note that, although a large proportion of participants stated that they did not know how to respond to beliefs in the “cues for action” dimension, 65% believe that “social media and the Internet are good tools for learning about oral cancer”. Some studies confirm that these tools play a crucial role in how young people and adults receive information and education today, making digital communication a key strategy for educating the general population and those in clinics about oral cancer and its risk factors [24,25].

To the best of our knowledge and based on our literature review, this is the first study in Colombia to use the HBM to provide information on beliefs about oral cancer in a population attending a dental school clinic and the general public. It shows that women and those under 40 have more deeply rooted beliefs about risk factors other than tobacco and alcohol, and highlights social media as an important educational tool. It also shows that beliefs about the benefits of healthy behaviors are clear for most respondents, regardless of socioeconomic factors such as education level or social status.

There is a clear need to conduct health promotion campaigns to change public beliefs about oral cancer risk factors and to use digital media, such as social networking sites, with short and effective messages to educate the population.

Dental schools, public and private health services, and health professionals should play an important role in educating the public about “new” risks factors and oral cancer, through social media, especially targeting younger populations under the age of 40. Therefore, having validated instruments on this topic is crucial for both clinical and public health practice [26,27], as it allows application to the general population and provides results that facilitate the design of educational programs for prevention, risk reduction, and early detection, ultimately positively impacting communities and reducing the problems associated with late diagnosis of oral cancer.

The future goal of this work is that with the validated survey and its widespread use, educational strategies can be designed to improve community beliefs as reported in the literature [23]. After validation, the educational strategies will be implemented and the health behaviors of the target groups will be analyzed, demonstrating that education is effective in promoting the desired behavior [23–28]. Therefore, the development and implementation of educational programs based on the health promotion and disease prevention model is encouraged, as shown by Jeihooni et al. [23], who used a validated survey based on the HBM and found that six months after an educational intervention on oral cancer and its prevention, the construct scores increased when the survey was administered again.

Conclusion

A three-dimensional questionnaire based on the Health Belief Model was validated, showing that people are aware that they could develop oral cancer. However, further multimodal risk factor education is needed, targeting different populations and focusing on risk factors, their control, and oral self-examination.

Acknowledgments

The author wish to acknowledge all the people who agreed to be part of the survey.

Funding

This study was funded by the Universidad de Antioquia, School of Dentistry.

Conflicts of interest

The authors state that there is no conflict of interest with people or organizations.

Data availability statement

Data supporting the findings of the study titled 'Beliefs in Oral Cancer in an Adult Population in Medellín, Colombia: A Cross-Sectional Study', are available from the corresponding author upon reasonable request. Due to ethical considerations and confidentiality agreements approved by the institutional ethics commiNee, the dataset is not publicly available. However, de-identified data may be shared with qualified researchers for academic purposes, provided they comply with data protection regulations and obtain institutional approval.

Author contribution statement

Conceptualization, design and writing: A Hurtado, GJ Alvarez. Methodology, analysis, and writing: LS Giraldo. Bibliographic research, data collection, analysis and writing: M Villada, MF Mejía, MA Ramos, NR Serna, MA Díaz, L. All authors approved the final version of the manuscript.

Ethics approval

This study was reviewed and approved by the Bioethics Committee of the School of Dentistry of the University of Antioquia, as documented in Act Number 10 of September 13, 2019. The ethical conditions established in Resolution 8430 of 1993 were followed.

Informed consent

Written informed consent was obtained, authorizing participation in the project and the use of the data and results for academic dissemination.

References

All Tables

All Figures

	Fig. 1 SBOC model with three dimensions and 18 Items. Note: Chi2: 210.873, p-value = 0.000. RMSEA: 0.051, 95% CI: 0.038–0.063. AIC: 6781.21. CFI: 0.93. TLI: 0.916. CD: 0.99. SRMR: 0.063.
In the text