Vaccine hesitancy toward childhood vaccination in the Madinah Region, Saudi Arabia: A cross-sectional study

Study design and setting

This was a cross-sectional survey conducted among the parents recruited from selected healthcare facilities in the Madinah region of Saudi Arabia between December 2024 and March 2025, with supplementary in online participation. The study was conducted in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cross-sectional studies [Supplementary Material].

Study population and eligibility criteria

The study targeted parents (mothers or fathers) of children aged 1–10 years residing in Madinah. For the purposes of this study, “residency” was defined as living in the Madinah region for at least six months before the study, regardless of nationality, including both Saudi citizens and non-Saudi residents. Inclusion criteria were (1) parent with at least one child aged 1–10 years, (2) residency in the Madinah region as defined above, (3) attendance at one of the selected healthcare facilities during the study period or receipt of the online questionnaire, (4) ability to read and understand Arabic, and (5) provision of verbal informed consent. Exclusion criteria included (1) non-residents or temporary visitors staying <6 months, (2) parents who declined participation, (3) children with documented medical contraindications to vaccination, and (4) incomplete questionnaire responses.

Sampling strategy and sample size

Participants were recruited using a convenience sampling technique from five healthcare facilities representing different geographic clusters of Madinah: Al-Salam Primary Health Care (PHC) Center (Central Cluster), Al-Badrani PHC (North Cluster), Al-Hijra PHC (South Cluster), Aldaitha PHC (East Cluster), and the Pediatric Outpatient Clinics at Maternity and Children’s Hospital. Within each facility, parents attending for routine vaccination, pre-school entry examination, or any other clinic visit were approached consecutively by trained research assistants during clinic hours (8:00 am to 2:00 pm, Sunday through Thursday) and invited to participate. All parents meeting the eligibility criteria who attended during the data collection period were approached until the target sample size was achieved.

The required sample size was calculated using the standard formula for single proportions with a 95% confidence level (CI), 5% margin of error, and an expected prevalence of vaccine hesitancy of 30% based on previous studies in similar settings. This calculation yielded a required sample size of 323 participants. A total of 322 valid responses were collected and analyzed.

Data collection tool

The study used the validated Arabic version of the parent attitudes about childhood vaccines (PACV) questionnaire. The PACV was originally developed by Opel et al.,^[21] and has been validated in multiple languages including Arabic.^[22,23]The questionnaire consists of 15 items across three domains: (1) Vaccination behavior (5 items): History of vaccine refusal, delay, or compliance; (2) Beliefs about vaccine safety and efficacy (6 items): Perceptions of side effects, effectiveness, and trust in science; and (3) Trust in healthcare providers (4 items): Confidence in pediatricians and the healthcare system regarding vaccines.

Each item was scored as 0 (non-hesitant response), 1 (intermediate response), or 2 (hesitant response). The total raw score was converted to a 0–100 scale using the formula (raw score/30) × 100. Parents with scores ≥50 were classified as vaccine hesitant, while those with scores <50 were considered non-hesitant, following standard PACV scoring guidelines. A supplementary sociodemographic questionnaire captured information on parental age, gender, education level, monthly household income, employment status, number of children, previous adverse events following vaccination, and sources of vaccine information including social media exposure.

Data collection procedures

Data were collected through both on-site paper-based questionnaires and a secure online platform. Parents attending the selected healthcare facilities were approached by trained research assistants who explained the study purpose and obtained informed verbal consent. Participants could choose to complete the questionnaire immediately using a tablet device or paper form at the facility or to access it later through a secure online link sent to their mobile phones. The online version was hosted on a password-protected survey platform with data encryption. All questionnaires, whether completed on paper or electronically, were reviewed for completeness before acceptance. Paper forms were entered into the electronic database by two independent data entry clerks, with discrepancies resolved through reference to the original forms. A comparison of responses between paper-based and electronic formats revealed no significant differences in PACV scores (p = 0.612) or hesitancy prevalence (p = 0.584), indicating no substantial mode effects.

Statistical analysis

Data was analyzed using IBM Statistical Package for the Social Sciences Statistics version 28.0 (IBM Corp., Armonk, NY, USA). Descriptive statistics including frequencies, percentages, means, and 95% confidence intervals were used to summarize participant characteristics and vaccine hesitancy prevalence. The normality of continuous variables was assessed using the Shapiro–Wilk test and visual inspection of histograms.

Bivariate associations between vaccine hesitancy (dichotomized as hesitant vs. non-hesitant based on PACV score ≥50) and independent variables were examined using Pearson Chi-square tests or Fisher’s exact test where appropriate. Variable demonstrating associations with p < 0.20 in bivariate analysis were selected as candidates for inclusion in the multivariate model. Multivariate logistic regression analysis was performed to identify independent predictors of vaccine hesitancy, with adjusted odds ratios (aOR) and 95% CI reported. Statistical significance was set at p < 0.05 for the final model.

Model diagnostics were performed to assess goodness-offit using the Hosmer–Lemeshow test (p > 0.05 indicating acceptable fit). Multicollinearity among predictor variables was evaluated using variance inflation factors (VIF), with VIF<5 considered acceptable. The final model’s discriminatory ability was assessed using the area under the receiver operating characteristic curve (AUC-ROC).

Ethical considerations

The study protocol received approval from the Institutional Review Board of Madinah Health Affairs, which is the officially designated research ethics committee for health research conducted within the Madinah region under the Saudi Ministry of Health (approval number: IRB log No. 24-126, dated December 1, 2024). The study was conducted in accordance with the Declaration of Helsinki and all applicable national and institutional regulations for research involving human participants. Participation was entirely voluntary, and informed verbal consent was obtained from all participants after explanation of the study purpose, procedures, potential risks, and benefits. No personal identifiers were recorded in the dataset, and all data were stored securely with access restricted to authorized research team members only. Participants were informed of their right to withdraw from the study at any time without consequences.

Participant characteristics

A total of 356 parents were approached, of whom 322 completed the questionnaire, yielding a response rate of 90.4%. The participant flow is shown in Figure 1. The mean age of participating parents was 34.8 ± 6.7 years (range: 22–52 years). The majority were mothers (n = 198, 61.5%), married (n = 314, 97.5%), and held at least a bachelor’s degree (n = 187, 58.1%). Detailed sociodemographic characteristics stratified by vaccine hesitancy status are presented in Table 1.

Close

Figure 1:

Participant flow diagram. Strengthening the reporting of observational studies in epidemiology flow diagram showing participant recruitment, exclusions, and final analysis sample.

Export to PPT

Prevalence of vaccine hesitancy

Of the 322 parents, 89 (27.6%, 95% CI: 23.0–32.7%) scored ≥50 on the PACV scale and were classified as vaccine hesitant [Figure 2]. The mean PACV score for the entire sample was 38.4 ± 18.2 (median: 36.7, range: 0–86.7). Among hesitant parents, the mean score was 62.1 ± 10.4, while non-hesitant parents scored 28.9 ± 11.8 (p < 0.001). The distribution of PACV scores is illustrated in Figure 3.

Close

Figure 2:

Prevalence of vaccine hesitancy. Donut chart showing the proportion of vaccine-hesitant (parent attitudes about childhood vaccines [PACV] score ≥50) and non-hesitant (PACV score <50) parents with 95% confidence intervals (CI).

Export to PPT

Close

Figure 3:

Distribution of parent attitudes about childhood vaccines (PACV) scores. Histogram showing the frequency distribution of PACV scores across the study population, with the hesitancy threshold (PACV score ≥50) clearly marked.

Export to PPT

Factors associated with vaccine hesitancy

In bivariate analysis, several factors showed significant associations with vaccine hesitancy. Parents with lower educational attainment (high school or below) were significantly more likely to be hesitant compared to those with university education (38.5% vs. 20.9%, p = 0.001). Unemployed parents demonstrated higher hesitancy rates than employed parents (35.7% vs. 22.1%, p = 0.009). Parents who reported exposure to anti-vaccine information on social media platforms had substantially elevated hesitancy rates compared to those without such exposure (44.3% vs. 17.8%, p < 0.001).

In addition, parents with four or more children showed higher hesitancy (36.1% vs. 23.4%, p = 0.022), as did those reporting previous adverse events following immunization in their children (42.4% vs. 22.0%, p < 0.001). Neither parental age, gender, monthly household income, nor child’s age group showed significant associations with vaccine hesitancy in bivariate analysis.

Multivariate analysis

Variables with p < 0.20 in bivariate analysis were entered into the multivariate logistic regression model. The final model demonstrated good fit (Hosmer-Lemeshow χ² = 8.34, df = 8, p = 0.400) and acceptable discrimination (AUC-ROC = 0.78, 95% CI: 0.73–0.83). All VIF values were below 2.0, indicating no problematic multicollinearity.

After controlling for potential confounders, five factors remained independently associated with vaccine hesitancy [Table 2]. Parents with a high school education or below had nearly three times the odds of being vaccine-hesitant compared to those with a university education (aOR: 2.87, 95% CI: 1.52–5.41, p = 0.001). Unemployment was associated with more than double the odds of hesitancy (aOR: 2.14, 95% CI: 1.18–3.89, p = 0.012). The strongest predictor was exposure to anti-vaccine information on social media, which increased the odds of hesitancy nearly fourfold (aOR: 3.76, 95% CI: 2.08–6.79, p < 0.001). Parents with four or more children had approximately double the odds of being hesitant compared to those with fewer children (aOR: 1.98, 95% CI: 1.05–3.73, p = 0.034). Finally, parents who reported previous adverse events following immunization in their children had more than double the odds of vaccine hesitancy (aOR: 2.52, 95% CI: 1.31–4.85, p = 0.006).

Sources of vaccine information

Analysis of vaccine information sources revealed significant differences between hesitant and non-hesitant parents [Table 3]. Healthcare providers were the most commonly cited source overall (78.3%) but were significantly more trusted by non-hesitant parents (85.4%) compared to hesitant parents (59.6%, p < 0.001). Social media platforms were used by 62.1% of hesitant parents compared to 37.8% of non-hesitant parents (p < 0.001). Television and newspapers were less commonly used, with no significant differences between groups. Notably, family and friends were more frequently consulted by hesitant parents (41.6%) than non-hesitant parents (28.3%, p = 0.019).

PACV domain analysis

Analysis of responses across the three PACV domains revealed distinct patterns [Figure 4]. The behavior domain showed the highest proportion of hesitant responses (32.4%), followed by the safety and efficacy domain (28.9%), and trust in healthcare providers domain (22.7%). Specific vaccine concerns among hesitant parents are detailed in Table 4. Within the safety domain, concerns about vaccine side effects were most frequently cited (45.3% expressing moderate to high concern), followed by doubts about vaccine necessity given low disease prevalence (38.2%), and questions about the number of vaccines given to children (36.0%).

Close

Figure 4:

Vaccine hesitancy by parent attitudes about childhood vaccines (PACV) domain. Bar chart displaying the percentage of parents expressing hesitant responses across the three domains of the PACV questionnaire: Vaccination behavior, beliefs about safety and efficacy, and trust in healthcare providers.

Export to PPT

Limitations

Several study limitations warrant consideration. The convenience sampling approach and focus on healthcare facility attendees may introduce selection bias. Specifically, parents who are actively disengaged from health services or who completely avoid the healthcare system may be systematically excluded from our study, potentially leading to an underestimation of the true hesitancy prevalence in the general population. Parents who attend healthcare facilities, even if hesitant, demonstrate at least some level of engagement with the health system that non-attendees do not share. Future community-based studies using probability sampling methods would provide more generalizable estimates.

The cross-sectional design precludes the determination of temporal relationships or causal inference. While we identified associations between various factors and vaccine hesitancy, we cannot establish causality or determine whether exposure to anti-vaccine information preceded or followed the development of hesitant attitudes. Social desirability bias may have influenced responses, although the anonymous nature of data collection likely minimized this effect. The study was conducted during a specific four-month period, and seasonal or temporal variations in vaccine attitudes were not captured.

While our study defined residency as living in the Madinah region for at least 6 months regardless of nationality, the dataset did not systematically capture whether participants were Saudi citizens or non-Saudi residents. This represents a limitation, as non-Saudi residents, including those born and raised locally, constitute an important part of the community and may have different vaccine attitudes shaped by their countries of origin or cultural backgrounds. Future research should specifically examine vaccine hesitancy among non-Saudi residents to provide a more comprehensive understanding of community-level vaccination attitudes and to inform culturally tailored interventions for diverse population subgroups.

Finally, while the PACV has been validated in Arabic, cultural nuances in how questions are interpreted may vary across different regions of Saudi Arabia. Despite these limitations, this study provides valuable insights into vaccine hesitancy in an understudied yet epidemiologically important region of Saudi Arabia.

Recommendations and future directions

The findings suggest several evidence-based recommendations for the Madinah region. First, targeted health education campaigns should prioritize populations with lower educational attainment, using clear, accessible language and culturally sensitive messaging. Second, healthcare providers require training in effective communication techniques for addressing vaccine concerns, particularly regarding safety and adverse events. Third, collaboration with social media platforms and religious leaders to promote accurate vaccine information and counter misinformation is essential. Fourth, systematic monitoring of vaccine hesitancy trends would enable early detection of declining confidence and timely intervention.

Future research should examine vaccine hesitancy trends longitudinally to assess the effectiveness of interventions and identify emerging concerns. Qualitative studies exploring the nuanced beliefs and experiences underlying hesitancy would provide deeper insights for intervention development. In addition, studies examining the role of religious leaders and cultural norms in shaping vaccine attitudes in the Madinah region, given its unique status as a holy city, would be particularly valuable.