Effects of religion, politics and distance to providers on HPV vaccine attitudes and intentions of parents in rural Utah

Abstract

Purpose: Human papillomavirus (HPV) causes an estimated 300,000 high grade cervical dysplasias and 36,000 preventable cancers each year in the United States alone. Despite having a safe, effective and long lasting vaccine since 2006, the rate of uptake has been suboptimal, particularly in rural areas. In 2020, Utah ranked near last in teenage HPV vaccination rates with rural areas trailing urban areas by about 18 percent. In this study, we identified factors that affect the intent of rural Utah parents to vaccinate their children against HPV.

Methods: A survey was distributed electronically to Utah residents in rural areas. Recruitment was carried out through targeted advertising, community organizations, and professional survey panels. The survey was open from Nov. 15, 2022 to April 15, 2023. A total of 410 respondents were used for analysis. Survey results were analyzed using exploratory factor analysis, confirmatory factor analysis, and structural equation modeling.

Findings: Distance to care was shown to negatively influence direct intent to vaccinate, while trust in government, general vaccine attitudes, and HPV knowledge positively influence intent to vaccinate. It was found that religious practice decreased vaccine hesitancy while cautious sexual attitudes, distance to care, and general negative vaccine attitudes increased vaccine hesitancy. Conservative political identity and high income were both shown to decrease vaccine hesitancy as covariates.

Fig 1. Pre-analysis data cleaning.

Responses that were deemed low quality were removed (e.g., duplicate email addresses, a response time that was shorter than one standard deviation below the mean time, etc.) Cleaning reduced our sample size from 1,348 to 410.

Fig 2. Survey reach.

We obtained responses from every targeted county. Counties with higher response rates (e.g., Washington, Wasatch, Tooele, etc.) were found to be correlated with counties that have a higher census population.

Fig 3. Model A: Logistic structural model.

Latent variables concerning logistical factors and covariates were combined into a structural model. SEM was used to determine which factors impact parental intent to vaccinate their children against HPV as well as HPV vaccination hesitancy. Greater distance from care predicts greater HPV vaccine hesitancy and decreased intent to vaccinate against HPV. Greater HPV knowledge predicted greater intent to vaccinate against HPV. There was a slight correlation between political ideology with more conservative leanings predicting lower vaccine hesitancy. Additionally, higher income slightly predicted decreased HPV vaccine hesitancy. All other latent variables and covariates were not significantly predictive of intent to vaccinate in the overall structural mode. Values are beta coefficients. Value in red shows negative correlation. Model A showed robust fit with the following fit statistics: RMSEA 0.064, CFI 0.908, TLI 0.879, SRMR 0.060.

Fig 4. Model B: Belief structural model.

Latent variables concerning belief factors and covariates were combined into a structural model. Increased governmental trust and positive general vaccine attitudes increases intent to vaccinate. Greater cautious sexual attitudes as well as positive attitudes about vaccines in general predicted increased vaccine hesitancy. Similar to Model 1, greater conservative ideology as well as greater income predicted decreased vaccine hesitancy. Values are beta coefficients. Values in red show negative correlations. Model B showed robust fit with the following fit statistics: RMSEA .057, CFI .907, TLI .889, SRMR .057.

Fig 5. Split structural models for belief.

Trust in government and positive general vaccine attitudes predicted greater intent to vaccinate. No other latent variables or covariates significantly affected intent to vaccinate. Cautious sexual attitudes and positive general vaccine attitudes both predicted heightened vaccine hesitancy while religious practice predicted reduced hesitancy. More conservative political views and higher income correlated with decreased hesitancy. No other latent variables or covariates had a significant effect on vaccine hesitancy. Values are beta coefficients. Values in red show negative correlations. The fit statistics for Model B-1 are: RMSEA .057, CFI .914, TLI .896, SRMR .053. The fit statistics for B-2 are: RMSEA .060, CFI .911, TLI .894, SRMR .056.

Fig 6. Response distributions for the questions that represent the latent variable ‘cautious sexual attitudes’.

All questions were answered on a 5-point scale. In the figure, 1 represents the response that reflects the most sexually cautious attitude and 5 represents the response that reflects the least sexually cautious attitude. The vast majority of respondents (at least 87%) chose either 1 or 2 for each question, representing a population that has highly cautious sexual attitudes. Our sample population included little diversity in terms of cautious sexual attitudes and this likely affected the viability of our latent variable ‘cautious sexual attitudes’.