Investigating Methods to Reduce the Burden of Cervical Cancer in Highly Impacted Countries

Abstract

Cervical cancer, which primarily develops in middle-aged women, is a large public health issue in most low- and middle-income countries (LMICs), causing significant morbidity and mortality among women during peak working and child-rearing ages. In contrast, high-income countries, which tend to have comprehensive cytologic (e.g. Pap smear) screening programs, have low rates of cervical cancer, demonstrating the effectiveness of secondary prevention on incidence and mortality of this disease. Unfortunately, as of yet, LMICs have not effectively implemented screening programs; thus, it is critical to investigate non-traditional forms of cervical cancer interventions in low-resource settings. It is estimated that 90% of cervical cancers are caused by the human papillomavirus (HPV), so novel interventions focus on HPV detection and prevention. This dissertation investigates the impact of testing for and vaccinating against HPV among women in two LMICs: Guatemala and Thailand. First, we evaluate the acceptability of self-collection HPV testing as a primary form of cervical cancer screening among Buddhist and Muslim women in southern Thailand. We survey women to assess knowledge of cervical cancer and screening and offer self-collected HPV tests. We find high acceptability across both cultural groups, particularly among Muslim women, who report lower rates of prior cervical cancer screening. Second, we analyze barriers to cervical cancer screening among women in Guatemala, a country with about 50% indigenous population and over 25 spoken languages and dialects, using nationally representative cross-sectional data from 2014-2015. Specifically, we explore the association between reported health barriers (e.g. needing permission, cost, distance, reluctance to travel alone, language barriers) and prior cervical cancer screening, using multivariate negative binomial regressions to estimate prevalence ratios. We find that up-to-date screening remains low, particularly among rural and indigenous women, and that all reported health barriers, particularly discordant languages between a woman and the healthcare personnel in her community, are inversely associated with past screening. Third, we investigate the impact of exposure to self-collection HPV testing on participation in cervical cancer screening programs. We replicate the study from Aim 1 in a cohort of Guatemalan women, finding similar acceptability of HPV testing, and follow the cohort for two years to identify further participation in screening programs. We find that women who participate in an HPV testing campaign are more likely to receive additional screening. Discrepancies in the data indicate that many women do not accurately report screening history, suggesting that reported screening rates could be misleading. Fourth, we develop computational models of HPV transmission and cervical cancer development among a simulated cohort of Guatemalan women. This is the first model developed for the country using local data. The model suggests that HPV vaccination and effective screening schedules similar to those currently in practice in the United States would produce considerable reductions in cervical cancer incidence. However, even a one-time screen around age 40 or vaccination of 10-14-year-old females would cause a significant reduction in cervical cancer cases and deaths. This model can be adjusted to imitate the population of other LMICs and can be used by policymakers to evaluate different strategies when deciding which cervical cancer intervention strategies might best fit their country-specific context. Overall, this dissertation provides much needed insight into the utility of culturally sensitive novel cervical cancer interventions in highly impacted regions of the world.