Reaching underscreened women: correlates of cervical cancer underscreening, reactions to mailed HPV self-sampling kits, and cost-effectiveness of HPV self-sampling programs

Colin Malone | 2020

Advisor: Rachel L. Winer

Research Area(s): Cancer Epidemiology, Clinical Epidemiology, Epidemiologic Methods


Underscreening is a major risk factor for the development of cervical cancer. This dissertation examines factors associated with clinic-based Pap screening and human papillomavirus (HPV) self-sampling in a U.S. healthcare system and summarizes the existing evidence on the cost-effectiveness of HPV self-screening in order to support health system implementation of new strategies to increase cervical cancer screening. The first chapter is an analysis of electronic health record data from Kaiser Permanente Washington (KPWA) to identify correlates of underscreening among 48,711 women aged 30-64 years, with different levels of interaction with the health care system, namely primary care visits and/or use of an online health portal. Recent healthcare interactions were associated with lower odds of underscreening. Compared to screening-adherent women, underscreening was associated with older age, higher body mass index (BMI), current smoking, and non-adherence to breast and colorectal cancer screening guidelines. These screening disparities persisted even among women who interacted with the healthcare system. The second chapter identifies attitudes, experiences, and reactions to mailed HPV self-sampling kits among women who were enrolled in a pragmatic randomized trial at KPWA. Using a web-based survey, women who did and did not use and return a kit were queried about potential psychosocial correlates of HPV self-sampling (including knowledge, barriers to Pap screening, trust in physicians, and perceived risk), their experiences with using a kit (or reasons for non-return), and their reactions to the kit (trust in HPV self-sampling, screening preferences, and future intentions). Most kit returners had positive reactions to using a kit and would use them again, and many non-returners reported openness to using a kit in the future. Many kit returners and non-returners, however, lacked trust in HPV self-sampling as a cervical cancer prevention method. The third chapter is a systematic literature review on the cost-effectiveness of HPV self-sampling to prevent cervical cancer. Sixteen eligible studies were identified, eleven conducted in high-income countries and five set in low/middle-income countries. Fourteen of sixteen studies reported that HPV self-sampling was cost-effective under certain conditions. Overall, studies found that the most important factor for cost-effective HPV self-sampling implementation was the observed increase in screening uptake, with HPV self-sampling kit cost, sensitivity for detecting cervical precancers, and the screening status of self-sampling users also having an impact. There are gaps in the literature on the effects of vaccination and new HPV self-sampling triage strategies on cost-effectiveness in high-income settings, as well as a need for additional studies based on real-world implementation of HPV self-sampling, especially in low/middle-income countries. New technologies such as HPV-self-sampling offer health systems opportunities to increase screening uptake among underscreened women. The results of this study suggest that existing health care interactions (such as those between patient and provider) may be an important avenue for building women’s trust and confidence in HPV self-sampling. Many women are open to using HPV self-sampling; to optimize the effectiveness and cost-effectiveness of HPV self-sampling programs, health systems should focus on leveraging existing points of patient interaction and continuing to research ways to reach those women who persistently do not engage with the health care system.