Mailed Outreach for Colorectal Cancer Screening in Community Health Centers

Mailed FIT-DNA outreach reached 27.9% screening participation at 90 days versus 22.6% with mailed FIT, an adjusted 4.7-point difference among patients due for colorectal cancer screening in the CARES pragmatic cluster randomized trial. The study was conducted in Boston and Los Angeles community health centers, with a 95% CI of 0.8 to 8.6 and P=.02. The comparison reflects two mailed outreach workflows tested in routine safety-net primary care.

Key Takeaways

• FIT-DNA outreach was associated with modestly higher screening participation than mailed FIT, and the difference persisted through 180 days.
• Boston had higher overall participation than Los Angeles, while the between-group advantage was concentrated in Los Angeles and several subgroup patterns favored FIT-DNA.
• Among patients with abnormal stool tests, 36 of 100 completed colonoscopy within 180 days, and navigation attempts still left many patients unreached.

The CARES pragmatic cluster randomized clinical trial was conducted in 8 community health centers in greater Boston and Los Angeles, with enrollment between June 7, 2023, and October 24, 2023. Eligible participants were English- or Spanish-speaking primary care patients aged 45 to 75 years who were due for colorectal cancer screening. The randomized sites enrolled 5127 participants, including 2435 in FIT and 2692 in FIT-DNA, while a parallel nonrandomized FIT-DNA site in Rapid City, South Dakota, was reported separately. The primary outcome was screening participation by any modality within 90 days, and secondary outcomes included 180-day participation and time to screening.

The FIT arm used a primer text, a mailed kit with graphic instructions, and reminder texts at days 14 and 28. FIT-DNA outreach paired mailed letters with manufacturer-assisted telephone calls, text messages, and occasional emails. In that arm, outreach included letters for 99.5%, calls for 97.4%, texts for 94.0%, and emails for 4.8%, and median outreach attempts totaled 10. At 180 days, participation was 31.7% with FIT-DNA and 26.7% with FIT, for an adjusted 4.5-point difference, and time to screening was shorter. Higher participation occurred alongside a more intensive manufacturer-supported outreach workflow.

Regional findings showed higher overall screening participation in Boston than Los Angeles, and subgroup analyses generally favored FIT-DNA in older, Hispanic, Spanish-speaking, Medicaid-insured, and uninsured participants. Between-group completion was similar in Boston, while the FIT-DNA advantage was concentrated in Los Angeles. Among screened participants in the randomized sites, 100 of 1435 had an abnormal stool test result. That included 42 of 623 in FIT and 58 of 812 in FIT-DNA, and 36 of 100 completed colonoscopy within 180 days. Colonoscopy completion was similar by outreach group but higher in Boston than Los Angeles, and navigation reached 59 patients, with 37 never reached and 4 refusing, underscoring incomplete diagnostic follow-up.

The trial did not assess cost or cost-effectiveness, and it did not include a usual-care control group. Different FIT brands were used in Boston and Los Angeles, which may have contributed to differing abnormal-result rates. The investigators also said the analysis could not separate the contribution of follow-up intensity from the effect of test type. Overall, randomized community health centers saw higher screening participation with FIT-DNA outreach, while colonoscopy after abnormal stool tests remained limited.