Extent of Follow-Up on Abnormal Cancer Screening in Multiple California Public Hospital Systems: A Retrospective Review

Abstract

Background: Inequitable follow-up of abnormal cancer screening tests may contribute to racial/ethnic disparities in colon and breast cancer outcomes. However, few multi-site studies have examined follow-up of abnormal cancer screening tests and it is unknown if racial/ethnic disparities exist.

Objective: This report describes patterns of performance on follow-up of abnormal colon and breast cancer screening tests and explores the extent to which racial/ethnic disparities exist in public hospital systems.

Design: We conducted a retrospective cohort study using data from five California public hospital systems. We used multivariable robust Poisson regression analyses to examine whether patient-level factors or site predicted receipt of follow-up test.

Main measures: Using data from five public hospital systems between July 2015 and June 2017, we assessed follow-up of two screening results: (1) colonoscopy after positive fecal immunochemical tests (FIT) and (2) tissue biopsy within 21 days after a BIRADS 4/5 mammogram.

Key results: Of 4132 abnormal FITs, 1736 (42%) received a follow-up colonoscopy. Older age, Medicaid insurance, lack of insurance, English language, and site were negatively associated with follow-up colonoscopy, while Hispanic ethnicity and Asian race were positively associated with follow-up colonoscopy. Of 1702 BIRADS 4/5 mammograms, 1082 (64%) received a timely biopsy; only site was associated with timely follow-up biopsy.

Conclusion: Despite the vulnerabilities of public-hospital-system patients, follow-up of abnormal cancer screening tests occurs at rates similar to that of patients in other healthcare settings, with colon cancer screening test follow-up occurring at lower rates than follow-up of breast cancer screening tests. Site-level factors have larger, more consistent impact on follow-up rates than patient sociodemographic traits. Resources are needed to identify health system-level factors, such as test follow-up processes or data infrastructure, that improve abnormal cancer screening test follow-up so that effective health system-level interventions can be evaluated and disseminated.

Keywords: breast cancer; cancer disparities; cancer screening; colon cancer; safety-net system.

Fig. 1

a Rates of follow-up colonoscopy after abnormal FIT. b Rates of timely follow-up biopsy (21 days) after a highly suspicious mammogram (BIRADS 4/5).

Fig. 2

a Risk of failing to receive colonoscopy after abnormal FIT. Reference categories are age (50–54 yo), gender (male), race/ethnicity (White), preferred language (English), insurance (private), and site (A). This figure presents results of the adjusted model that controlled for age, gender, race/ethnicity, preferred language, insurance, and site. b Risk of failing to receive a timely biopsy (within 21 days) after a suspicious/highly suspicious mammogram (BIRADS 4/5). Reference categories are age (50–54 yo), race/ethnicity (White), preferred language (English), insurance (private), and site (A). This figure presents results of the adjusted model that controlled for age, race/ethnicity, preferred language, insurance, and site.