Cervical Cancer Precursors Among Low-Income Women Living in Pima County, Arizona: Prevalence and Risk Factors

Abstract

BACKGROUND: Compared to non-Hispanic Whites, Hispanics are more likely to be diagnosed with cervical cancer (CC) in the United States (9.9 per 100,000 among Hispanics vs. 7.3 among Non-Hispanic Whites) and more likely to die of the disease. In the United States (US), between 250,000 to 1 million women are diagnosed with a precursor to CC every year. While most cases (70%) of mild dysplasia, low-grade squamous intraepithelial lesions (LSIL) regress, moderate and severe dysplasia, high-grade squamous intraepithelial lesions (HSIL) are less likely to resolve. Without treatment, 30% to 50% of cervical dysplasia may progress to ICC. Using a sample population of low-income women residing in Pima County, AZ, the aims of this dissertation are to compare the prevalence of cervical dysplasia among low-income women Hispanics living in a border region (Pima County, AZ) with the national prevalence among Hispanics; assess whether Hispanics born in Mexico have a higher prevalence of cervical dysplasia compared to Hispanics born in the US; and to compare the prevalence of cervical dysplasia during a time frame that coincided with the implementation of a promotora led outreach and education program, REACHcervix, relative to time frames preceding and following program implementation. METHODS: Secondary data analysis of Pap test results funded through Pima County’s Well Women HealthCheck (WWHP) program between 2003-2016 was conducted. Logistic regression was used to assess prevalence and risk factors associa2ted with cervical dysplasia (LSIL + HSIL/ICC). Multinomial logistic regression was used to explore the relationship between the predictor variables (Hispanic ethnicity, Mexican nativity & intervention time period) and three different categories of cervical cytology (negative, LSIL & HSIL/ICC). Additional variables included age, income, number of visits, household size, and previous Pap history. RESULTS: In the study that compared cervical cancer precursors between Hispanics and non-Hispanic Whites (NHWs), ethnicity was modestly associated with higher prevalence of LSIL (odds ratio (OR)=1.39, 95%CI=1.01-1.91)), but this association was not statistically significant after adjusting for confounders. However, in the final models, lower income was independently associated with LSIL (adjusted odds ratio [aOR]=1.55, 95%CI=1.30-1.44), while smoking (aOR=2.88, 95%CI=1.21-6.84) and no history of Pap test within five years (aOR=3.54, 95%CI=1.61-6.99) were independently associated with HSIL. In the study that stratified Hispanic women by nativity (US born vs Mexican born), women born in Mexico were more likely to be diagnosed with an abnormal Pap smear (OR=2.31, 95%CI=1.25-4.27) and LSIL (OR=2.32, 95%CI=1.18-4.57) compared to Hispanic women born in the US. There was no significant difference in Pap test history between the two groups. In the third study that compared the prevalence of cervical cancer precursors among Hispanic women throughout three time periods to assess a potential effect of an intervention (before intervention, intervention after intervention), the risk of being diagnosed with LSIL was higher during the pre-intervention period (OR=1.79, p-value<.001) and the post-intervention time period (OR=1.83, p-value<.01). When compared to the time period before the implementation of the prevention program (REACHcervix), there was a significant increase in the percentage of Hispanics who received Pap smears (74.2% vs 80.4% respectively, p-value<.001), and the percentage of Hispanics did no decrease significantly in the years following the program (79.1%, p-value=0.30). There was also a significant increase in the percentage of Hispanic immigrants screened during the intervention time period compared to the period prior to program implementation (73.9% vs 68.8% respectively, p-value<.001); however, the percentage of Hispanic immigrant participation dropped significantly after the program concluded (70.2%, p-value<0.01). Conclusion: After adjusting for confounding in a sample of low-income women with comparable Pap screening rates, ethnicity was not associated with greater prevalence of abnormal pap smears. However, other variables were independently associated with LSIL and HSIL. The higher proportion of LSIL cases among lower income individuals compared to those with higher incomes, and the higher proportion of HSIL cases observed among those screened least regularly stresses the importance of programs like WWHP, while research focusing on risk and protective factors unique to Mexican born women may help reduce the prevalence of invasive cervical cancer and precancerous lesions among this population. These findings suggest that country of birth informs on cervical cancer risk and should be factored-in for this at-risk group. More frequent pap testing may be advised in Mexican born women; however, this approach to early detection should be tested in randomized, controlled trials. Despite the introduction of the HPV vaccine, continued Pap screening is needed to prevent cervical cancer. Further research is needed to assess the effects of community-based health worker-led interventions on cervical dysplasia prevalence among high-risk women. Keywords: low-grade squamous intraepithelial lesions; high-grade squamous intraepithelial lesions; Well Women HealthCheck Program; National Breast and Cervical Cancer Early Detection Program, COVID-19