Improving Colorectal Cancer Screening in a Regional Safety-Net Health System over a 10-Year Period: Lessons for Population Health.

Abstract

Despite national policy efforts to increase colorectal cancer (CRC) screening, rates in vulnerable populations remain suboptimal. Many types of interventions have been employed, but their impact on improving population-level rates of CRC screening over time is uncertain. Assess the impact of 10 years of different in-reach and outreach strategies to improve CRC screening and identify factors associated with being screen up-to-date (SUTD). Observational cohort study. Patients aged 50-74 years from 12 community-based primary care clinics in an integrated, regional safety-net health system. Multiple system-level interventions were implemented over time (visit-based electronic health record [EHR] reminders, quality measurement, annual preventive service letters, and mailed fecal immunohistochemical stool tests [FIT]). CRC SUTD rates by calendar year among those with a primary care (PC) visit in the prior 1 and 3 years and their multivariable correlates. The sample included 31,786-40,405 patients/year. In 2011, mean age was 58.9, 63.9% were women, 37.0% were Hispanic, 39.3% Black, 16.8% White, and 6.6% Asian/Other, and 60.5% were uninsured/Medicaid. Three-quarters of patients had ≥ 1 PC visit in the prior year. Lower-intensity interventions (EHR reminders, quality measurement, annual prevention letters) had limited impact on SUTD rates (2-3% rise). Implementing system-wide mailed FIT increased rates from 51.2 to 61.9% among those with a PC visit in the past year (40.5 to 46.8% with a PC visit ≤ 3 years). Stopping mailed FIT due to COVID wiped out these gains. Higher screening rates were associated with the following: older age; female; more comorbidities, PC clinic visits, and prior FITs; and better insurance coverage. Hispanics had the highest SUTD rates followed by Asians, Blacks, and Whites (p < 0.05). Implementation of a system-wide mailed FIT program had the greatest impact on SUTD rates. Lower-intensity interventions (EHR reminders, quality measurement, and patient letters) had limited effects.