Abstract 3451: Association of multi-level factors with cervical cancer screening in healthcare settings

Abstract Background- Few studies have examined the independent contribution of patient, clinician and system factors associated with completing cervical cancer screening. Methods- Using a large retrospective cohort created from electronic health record data, we examined the association of patient- and provider-level characteristics with cervical cancer screening among unknown- and average-risk, screen eligible women who entered the cohort at a primary care visit in 1 of 3 U.S. healthcare systems (Kaiser Permanente Washington, Parkland Health, Mass General Brigham) between 2010-2019. The outcome was time, in days, from cohort entry to screening. We fit piecewise exponential (PWE) survival models, nesting patients hierarchically within provider seen and the facility at the time of their analytic cohort entry visit. We excluded patients whose visit was with providers at their non-principal facility (principal defined as the facility that they saw most patients in 2010). The analysis was stratified by healthcare setting. Results- Median time from cohort entry until screening varied across settings (202-1063 days) and across providers and clinics (between cluster variance= 0.11-0.77, 0.06-1.18, respectively). Many patient-level characteristics (age, cohort entry year, race/ethnicity, number of comorbidities, smoking status, primary care visit within 12 months before cohort entry, and insurance), some of the provider-level characteristics (provider specialty and sex) and social determinants of health measures (Local Isolation Score Hispanic) were associated with time to screening. In the adjusted PWE models, for all three settings: a) Younger patients (21-49 years old) (vs older patients (50-65 years)); b) Hispanic and Non-Hispanic Asian/Pacific Islander (vs Non-Hispanic White); and c) patients seeing ob-gyn providers (vs family medicine providers) had shorter time to screening. Patients with a) higher (vs lower) comorbidity burden, and b) Medicare insurance (vs Commercial) were less likely to receive screened. Conclusion- By identifying characteristics at multiple levels that are associated with the outcome, our results help us identify targets for interventions to improve and increase timely cervical cancer screening in healthcare settings. Citation Format: Rutu Rathod, Ellen Hu, Sandi Pruitt, Jasmin A. Tiro, Jessica Chubak, Jennifer Haas, Steven J. Atlas, Gina Kruse, Amy E. Hughes. Association of multi-level factors with cervical cancer screening in healthcare settings [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 3451.