Psychosocial predictors of cervical cancer screening intention among female teachers in Karaj, Iran: An integrated behavioral model

Abstract

Objective

To examine psychosocial factors associated with cervical cancer screening intention among employed women in Iran using an integrated framework based on the Theory of Planned Behavior and the Health Belief Model.

Methods

This cross-sectional study included 200 female teachers in Karaj, Iran, in 2023. A validated questionnaire assessed knowledge, psychosocial constructs, and screening intention. Path analysis of observed composite scores tested the hypothesized relationships.

Results

The model showed good fit and explained 47.70% of intention. Perceived behavioral control (β = 0.26, P < 0.01), perceived benefits (β = 0.24, P < 0.01), subjective norms (β = 0.22, P < 0.01), and attitude (β = 0.17, P = 0.01) were positively associated with intention, whereas perceived barriers were negatively associated (β = −0.15, P = 0.01). Knowledge was positively associated with attitude, subjective norms, and perceived behavioral control. Attitude and perceived behavioral control were positively associated with benefits, while attitude was negatively associated with barriers.

Conclusions

Screening intention was associated with multiple psychosocial perceptions. Interventions should strengthen perceived control, communicate screening benefits, foster supportive norms, and address barriers.

Highlights

• •Perceived control was the strongest predictor of screening intention.
• •The integrated model explained 47.70% of screening intention.
• •Benefits increased screening intention, while barriers reduced it.
• •Knowledge was associated with attitudes, norms, and perceived control.
• •Findings support culturally tailored, multicomponent interventions.

1. Introduction

Cervical cancer remains a major cause of morbidity and mortality among women, particularly in low- and middle-income countries, where access to preventive services is limited (Arbyn et al., 2020; Canfell et al., 2020). Although screening and treatment of precancerous lesions can largely prevent the disease, approximately 660,000 new cases and 350,000 deaths occurred worldwide in 2022 (W.H. Organization, 2023). Iran has a lower incidence than the global average, but participation in cervical cancer screening remains suboptimal (Akbari et al., 2023; Momenimovahed and Salehiniya, 2018).

Screening decisions are influenced by knowledge, fear, embarrassment, stigma, perceived susceptibility, and access to services (Petersen et al., 2022). In culturally sensitive settings, social approval and support may also shape preventive behavior (Manning, 2009). Identifying these determinants is therefore important for designing contextually appropriate interventions.

1.1. Conceptual model development and hypothesized pathways

The conceptual framework integrated constructs from the Theory of Planned Behavior and the Health Belief Model. The Theory of Planned Behavior proposes that intention, the most proximal antecedent of behavior, is shaped by attitude, subjective norms, and perceived behavioral control (Ajzen, 1991; Fishbein and Ajzen, 2011). These constructs represent evaluation of screening, perceived approval from influential others, and perceived capability and control, respectively. Their associations with cancer-screening intention and participation may differ across sociocultural settings (Dhakal et al., 2023; Dsouza et al., 2022; Khani Jeihooni et al., 2021); therefore, each was specified as a direct predictor of screening intention.

Knowledge was treated as a background variable. Within the Theory of Planned Behavior, background factors are expected to influence intention mainly by shaping behavioral, normative, and control beliefs (Ajzen, 1991; Fishbein and Ajzen, 2011). Knowledge of cervical cancer, screening benefits, eligibility, and procedures may foster favorable attitudes, perceived social support, and confidence in obtaining screening (Hoque et al., 2014; Kirubarajan et al., 2021; Vallone et al., 2022). Accordingly, knowledge was specified to predict attitude, subjective norms, and perceived behavioral control.

The Health Belief Model complements the Theory of Planned Behavior by distinguishing the perceived benefits of preventive action from barriers to its performance (Champion and Skinner, 2008). Screening benefits include early detection and prevention of disease progression, whereas barriers may include embarrassment, fear, privacy concerns, cost, limited time or access, and insufficient family support. Both constructs, particularly perceived barriers, are associated with cervical screening participation (Bukowska-Durawa and Luszczynska, 2014). Consistent with the Health Belief Model, benefits and barriers were specified as positive and negative predictors of intention, respectively.

The integrated model additionally examined whether planned behavior constructs were associated with intention through benefit and barrier appraisals. Screening cognitions may be interconnected and responsive to cognitive and contextual conditions rather than fixed, parallel determinants (Taflinger and Sattler, 2024). Thus, the model did not reverse conventional Health Belief Model relationships: benefits and barriers remained direct predictors of intention, while pathways from planned behavior constructs to these beliefs represented an extension of the two frameworks.

A favorable attitude, supportive norms, and greater perceived control were expected to increase perceived benefits and reduce the salience of barriers. These associations may be relevant in collectivist settings, where family and social expectations influence reproductive health decisions (Manning, 2009; Dsouza et al., 2022). Perceived behavioral control and barriers were treated as distinct: the former reflects global capability and control, whereas the latter represents specific anticipated obstacles.

As shown in Fig. 1, knowledge was hypothesized to predict the three planned behavior constructs; these constructs to predict intention, benefits, and barriers; and benefits and barriers to predict intention positively and negatively, respectively. Both direct and indirect pathways were examined. Because the design was cross-sectional, all paths represent theoretically informed predictive associations, not temporal or causal effects.

Fig. 1.

Proposed integrated conceptual model of cervical cancer screening intention among female teachers in Karaj, Iran, 2023.

Evidence for an integrated model linking planned behavior constructs to cervical cancer screening intention both directly and through benefits and barriers remains limited, particularly in low- and middle-income and Middle Eastern settings (Dsouza et al., 2022; Kirubarajan et al., 2021). Evaluating these relationships simultaneously may improve understanding of screening decisions and identify culturally relevant intervention targets.

This study therefore examined direct and indirect pathways linking knowledge, planned behavior constructs, perceived benefits, perceived barriers, and cervical cancer screening intention among employed women in Iran using path analysis of observed composite scores.

2. Methods

2.1. Study design and population

This cross-sectional study was conducted in 2023 among married female teachers aged 18 years or older in public schools in District 2 of Karaj, Alborz Province, Iran. Women who were pregnant, had undergone hysterectomy, had a history of cancer or another condition affecting eligibility, or were unwilling to participate were excluded.

Sample size was determined according to model complexity and recommendations for path analysis (Hair et al., 2014; Kline, 2023; Wolf et al., 2013). The model contained seven observed composites and 31 freely estimated parameters; 200 women provided 6.45 observations per parameter, exceeding the commonly recommended minimum of five. The recruitment target was increased by 20% to 240 to allow for ineligibility, nonparticipation, and incomplete questionnaires.

From the Department of Education sampling frame, 240 teachers were randomly selected using computer-generated numbers. Eighteen were ineligible because of pregnancy (n = 6), hysterectomy (n = 7), cancer history (n = 3), or other medical conditions (n = 2). Of 222 eligible women, 22 declined or did not complete the questionnaire; 200 were analyzed, yielding a 90.10% response rate (Fig. 2). Questionnaires were distributed during staff meetings after consent and collected anonymously upon completion.

Fig. 2.

Random sampling, eligibility assessment, recruitment, and analysis of female teachers in Karaj, Iran, 2023.

2.2. Measures

Data were collected using a structured questionnaire developed from the two behavioral theories and cervical cancer screening literature. The first section assessed age, education, marital characteristics, menopause, previous Pap testing, and age at first pregnancy.

The second section assessed attitude (four items), subjective norms (four items), perceived behavioral control (four items), perceived barriers (three items), perceived benefits (four items), and screening intention (three items) on seven-point scales from 1 (strongly disagree) to 7 (strongly agree). Higher scores indicated higher levels of each construct. Knowledge was assessed with six items answered “yes” (2), “no” (1), or “I do not know” (0); total scores ranged from 0 to 12, with higher scores indicating greater knowledge.

Experts in health education, health promotion, obstetrics, and gynecology evaluated face and content validity. All retained items met acceptable Content Validity Ratio and Content Validity Index criteria (Vakili and Jahangiri, 2018). All constructs had Cronbach's alpha coefficients above 0.70. In 30 women retested after two weeks, intraclass correlation coefficients were also above 0.70.

2.3. Statistical analysis

Descriptive statistics summarized the women and study constructs. Composite scores were calculated according to scoring instructions; higher perceived-barrier scores represented greater barriers. Pearson correlations assessed bivariate relationships. Skewness and kurtosis were within the prespecified limits of ±3 and ±10 (Kim, 2013).

Path analysis of observed composite scores using maximum likelihood estimation tested the hypothesized relationships. No latent measurement model or confirmatory factor analysis was performed. Residuals of attitude, subjective norms, and perceived behavioral control were allowed to covary on theoretical grounds. Model fit was assessed using the chi-square statistic divided by degrees of freedom, Comparative Fit Index, Tucker–Lewis Index, Incremental Fit Index, Normed Fit Index, Relative Fit Index, and Root Mean Square Error of Approximation. Values below 3 for the chi-square statistic divided by degrees of freedom, at least 0.90 for incremental indices, and below 0.08 for the Root Mean Square Error of Approximation indicated acceptable fit (Kline, 2023; Hu and Bentler, 1999). Standardized coefficients, P-values, and explained variance were reported. An alternative model adding a direct knowledge-to-intention path was also tested. Tests were two-sided with P < 0.05. All descriptive, correlational, and path analyses were conducted in R software version 4.5.1(R Foundation for Statistical Computing, Vienna, Austria).

The Ethics Committee of Iran University of Medical Sciences approved the protocol and verbal consent procedure (IR.IUMS.REC.1401.634). Procedures followed the Declaration of Helsinki and the university's requirements for protecting the safety, privacy, confidentiality, and anonymity of human participants. Women received information about the study, voluntariness, confidentiality, and withdrawal rights before consenting. Verbal rather than written consent preserved anonymity in view of the sensitive reproductive-health content, and no identifying information was collected. The corresponding author confirms that the approval is on file at the authors' institution.

3. Results

3.1. Participant characteristics and correlations

The analytical sample comprised 200 teachers with a mean age of 40.71 ± 7.44 years. Mean marriage duration was 15.56 ± 8.47 years and mean age at first pregnancy was 23.13 ± 10.33 years. Overall, 79.50% had previously undergone Pap testing and 12.50% were menopausal.

Table 1 presents descriptive statistics and correlations. Mean scores were 25.18 ± 3.22 for benefits, 23.56 ± 3.88 for attitude, 21.25 ± 3.21 for perceived control, 16.10 ± 4.61 for intention, and 7.13 ± 3.20 for knowledge. Intention correlated positively with knowledge (r = 0.29), attitude (r = 0.50), subjective norms (r = 0.47), perceived control (r = 0.54), and benefits (r = 0.52; all P < 0.01), and negatively with barriers (r = −0.19, P = 0.01). Perceived control and barriers were weakly and nonsignificantly correlated (r = 0.03, P = 0.66).

Table 1.

Descriptive statistics and Pearson correlations among female teachers in Karaj, Iran, 2023 (N = 200).

Variable	Mean ± SD	1	2	3	4	5	6	7
1. Knowledge	7.13 ± 3.20	1.00
2. Attitude	23.56 ± 3.88	0.28*	1.00
3. Subjective norms	19.12 ± 5.76	0.23*	0.36*	1.00
4. Perceived behavioral control	21.25 ± 3.21	0.28*	0.45*	0.46*	1.00
5. Perceived benefits	25.18 ± 3.22	0.28*	0.47*	0.32*	0.45*	1.00
6. Perceived barriers	13.16 ± 4.23	−0.07	−0.19*	0.04	0.03	−0.12	1.00
7. Behavioral intention	16.10 ± 4.61	0.29*	0.50*	0.47*	0.54*	0.52*	−0.19*	1.00

Note. SD = standard deviation. *P < 0.01 (two-sided Pearson correlation test).

3.2. Path model evaluation and hypothesis testing

Path analysis examined the hypothesized relationships. Residuals of the three theoretically related planned behavior constructs were allowed to covary.

The model fitted the data well: χ²(4) = 5.96, P = 0.20; χ²/df = 1.49; Comparative Fit Index = 0.99; Tucker–Lewis Index = 0.97; Incremental Fit Index = 0.99; Normed Fit Index = 0.98; Relative Fit Index = 0.91; and Root Mean Square Error of Approximation = 0.05 (Table 2).

Table 2.

Fit indices for the integrated behavioral path model among female teachers in Karaj, Iran, 2023 (N = 200).

Fit index	Recommended criterion	Final model
Chi-square statistic	–	5.96
Degrees of freedom	–	4
P-value for chi-square statistic	>0.05	0.20
Chi-square statistic/degrees of freedom	<3.00	1.49
Comparative Fit Index	≥0.90	0.99
Tucker–Lewis Index	≥0.90	0.97
Incremental Fit Index	≥0.90	0.99
Normed Fit Index	≥0.90	0.98
Relative Fit Index	≥0.90	0.91
Root Mean Square Error of Approximation	<0.08	0.05

Note. The P-value was obtained from the chi-square test of exact model fit.

Knowledge predicted attitude (β = 0.28, P < 0.01), subjective norms (β = 0.23, P < 0.01), and perceived control (β = 0.28, P < 0.01), explaining 7.80%, 5.10%, and 7.80% of their variance, respectively.

Attitude predicted greater benefits (β = 0.31) and fewer barriers (β = −0.27), and perceived control predicted greater benefits (β = 0.27; all P < 0.01). Subjective norms did not predict benefits (β = 0.09, P = 0.20) or barriers (β = 0.08, P = 0.29), and perceived control did not predict barriers (β = 0.12, P = 0.16). The model explained 29.50% of benefits and 5.80% of barriers.

All direct predictors of intention were significant: perceived control (β = 0.26, P < 0.01), benefits (β = 0.24, P < 0.01), subjective norms (β = 0.22, P < 0.01), and attitude (β = 0.17, P = 0.01) were positive predictors, whereas barriers were negative (β = −0.15, P = 0.01). The model explained 47.70% of intention. Table 3 and Fig. 3 report all significant and nonsignificant paths.

Table 3.

Path coefficients for the integrated behavioral model among female teachers in Karaj, Iran, 2023 (N = 200).

Hypothesized path	B	SE	Critical ratio	β	P-value	Result
Knowledge → Attitude	0.34	0.08	4.12	0.28	<0.01	Significant
Knowledge → Subjective norms	0.41	0.13	3.26	0.23	<0.01	Significant
Knowledge → Perceived behavioral control	0.28	0.07	4.11	0.28	<0.01	Significant
Attitude → Perceived benefits	0.26	0.06	4.57	0.31	<0.01	Significant
Subjective norms → Perceived benefits	0.05	0.04	1.29	0.09	0.20	Not significant
Perceived behavioral control → Perceived benefits	0.27	0.07	3.77	0.27	<0.01	Significant
Attitude → Perceived barriers	−0.30	0.09	−3.45	−0.27	<0.01	Significant
Subjective norms → Perceived barriers	0.06	0.06	1.05	0.08	0.29	Not significant
Perceived behavioral control → Perceived barriers	0.15	0.11	1.40	0.12	0.16	Not significant
Attitude → Behavioral intention	0.20	0.08	2.61	0.17	0.01	Significant
Subjective norms → Behavioral intention	0.18	0.05	3.72	0.22	<0.01	Significant
Perceived behavioral control → Behavioral intention	0.37	0.09	4.03	0.26	<0.01	Significant
Perceived benefits → Behavioral intention	0.35	0.09	3.97	0.24	<0.01	Significant
Perceived barriers → Behavioral intention	−0.16	0.06	−2.80	−0.15	0.01	Significant

Note. B = unstandardized regression coefficient; SE = standard error; β = standardized regression coefficient. P-values were obtained from two-sided tests of critical ratios under maximum likelihood estimation. All tested pathways are reported.

Fig. 3.

Standardized path coefficients and explained variance for cervical cancer screening intention among female teachers in Karaj, Iran, 2023 (N = 200).

Adding a direct knowledge-to-intention path yielded a nonsignificant coefficient (β = 0.05, P = 0.41) and negligible change in explained variance (47.70% to 47.80%); the parsimonious model was retained.

4. Discussion

The integrated behavioral model fitted the data well and explained 47.70% of cervical cancer screening intention. Perceived control, benefits, subjective norms, and attitude were positively associated with intention, whereas barriers were negatively associated. Knowledge was related to the planned behavior constructs but had no additional direct association with intention.

Perceived control was the strongest direct predictor (β = 0.26), consistent with the Theory of Planned Behavior and meta-analytic evidence that control is important when health behavior requires time, resources, service access, or cooperation (Ajzen, 1991; Hagger et al., 2018). Women may value screening yet hesitate if arranging appointments, managing responsibilities, or coping with the procedure seems difficult. Flexible appointments, clear referrals, reminders, affordable care, and female providers may therefore strengthen control.

Benefits were the second strongest predictor (β = 0.24), supporting the Health Belief Model proposition that perceived advantages promote readiness for preventive action (Janz and Becker, 1984; Al-Ani et al., 2024). Communication should explain that screening can identify precancerous changes before symptoms or invasive disease develop. Subjective norms also predicted intention (β = 0.22), indicating that encouragement from family, peers, and professionals may reduce uncertainty and provide practical or emotional support (Ajzen, 1991; Hagger et al., 2018). Such engagement should preserve women's autonomy and confidentiality.

Attitude predicted intention and was associated with greater benefits and fewer barriers. This suggests that overall evaluations of screening are connected to how women weigh its advantages and difficulties. However, cross-sectional paths cannot establish direction; favorable attitudes and benefit–barrier appraisals may influence one another (Jones et al., 2015).

Barriers independently reduced intention, consistent with previous evidence (Al-Ani et al., 2024). Embarrassment, fear, privacy concerns, cost, time, and limited access should be addressed through confidential and respectful care, clear procedural information, female providers, shorter waiting times, and services outside working hours.

Knowledge predicted attitude, norms, and perceived control, but its direct path to intention was nonsignificant. Thus, knowledge may operate mainly through more proximal psychosocial beliefs, as proposed by the Theory of Planned Behavior (Ajzen, 1991). Although education based on the Health Belief Model can improve screening beliefs and performance in Iran (Simbar et al., 2020), information alone is unlikely to be sufficient; it should be combined with strategies that strengthen control, benefits, supportive norms, and barrier reduction.

Planned behavior constructs were selectively, rather than uniformly, associated with health beliefs. Attitude and perceived control predicted benefits, but norms did not; only attitude predicted barriers. Importantly, perceived control and barriers were empirically distinct (r = 0.03), and their path was nonsignificant. The former represents global capability and control, whereas the latter captures specific anticipated obstacles; retaining both therefore adds complementary information.

These findings favor multicomponent interventions over information-only approaches. Education should communicate benefits and correct misconceptions; trusted professionals should provide supportive recommendations; and services should increase women's control through flexible scheduling, confidential counseling, female providers, simple appointments, and affordable or workplace-linked referrals (W.H. Organization, 2020).

Caution is required because intention, not screening uptake, was measured. Although intention is a proximal determinant of behavior, it does not invariably lead to action (Ajzen, 1991; Sheeran, 2002). Access, cost, time, and family responsibilities may prevent screening despite favorable intentions. Longitudinal and intervention studies should therefore assess verified screening behavior.

Limitations include the cross-sectional design, which precludes causal or temporal inference; reliance on self-report; and assessment of intention rather than behavior. The women were teachers from one province and may have higher education and health literacy than the general population, limiting generalizability. Although response was high (90.10%), information on nonparticipants was unavailable, so nonresponse bias cannot be excluded. Future longitudinal studies should test whether intention translates into screening uptake and validate the model in socioeconomically, educationally, occupationally, and geographically diverse populations.

5. Conclusions

The integrated model explained 47.70% of cervical cancer screening intention. Perceived control, benefits, subjective norms, and attitude were positively associated with intention, whereas barriers were negatively associated; knowledge was linked to proximal planned behavior constructs but not directly to intention.

Interventions should combine education with supportive norms, greater control, clear benefits, and barrier reduction. Because this cross-sectional study assessed intention rather than behavior, longitudinal studies in diverse populations are needed to determine whether these factors predict screening uptake.

Declaration of generative artificial intelligence and artificial intelligence-assisted technologies in the writing process

During preparation of this manuscript, the authors used artificial intelligence-assisted tools solely for language editing and improvement of readability. The authors reviewed and edited the content and take full responsibility for the final version.

CRediT authorship contribution statement

Minoo Aryanpour: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation. Jamileh Abolghasemi: Writing – review & editing, Writing – original draft, Visualization, Validation, Formal analysis. Seyedeh Melika Kharghani Moghadam: Writing – review & editing, Writing – original draft, Investigation, Conceptualization.

Ethical approval

The Ethics Committee of Iran University of Medical Sciences approved the study and verbal informed consent procedure (IR.IUMS.REC.1401.634). All women voluntarily provided verbal informed consent.

Funding

This project was supported by Iran University of Medical Sciences as a master's thesis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The datasets generated and analyzed during this study are available from the corresponding author upon reasonable request.