Using Text Messages and Fotonovelas to Increase Return of Home-Mailed Colorectal Cancer Screening Tests: Mixed Methods Evaluation (Preprint)

Abstract

BACKGROUND Colorectal cancer (CRC) is currently the second-leading cause of cancer death in the US; yet it is mostly preventable with appropriate screening and often treatable when found at early stages. Many patients within an urban Federally Qualified Health Center (FQHC) clinic were found to be out-of-date with screening guidelines and therefore not benefitting from screening. Texting campaigns and motivational fotonovelas comics have both showed promise to encourage prosocial behaviors in various contexts. OBJECTIVE This project sought to use tailored bidirectional texting with fotonovela comics to boost participation in a campaign to mail Fecal Immunochemical Test (FIT) colorectal cancer screening kits to patients. METHODS The clinic mailed FIT screening kits to 11,000 unscreened patients in July 2021. Consistent with usual care, all patients received two text messages and a patient navigator call within the first month of mailing. Three months after mailing the kits, 5,241 patients who did not return their FIT kit, were aged 50 to 75, and spoke either English or Spanish were randomized to either intervention (4-week tailored two-way texting campaign with a fotonovela comic and remailing kits if requested) or control (no further intervention). RESULTS 39.5% of intervention patients engaged with bidirectional texting, and they were on average younger (aged 50-60) and more likely to have commercial insurance than those who did not engage with bidirectional texting. Of the 1,026 who engaged bi-directionally, 31% clicked the fotonovela (12.2% of the intervention population). Patients in the 50-60 age group were more likely to click the fotonovela than those 61-75 (95% CI for difference of means 2.6% to 7.6%, P<.001). Patients who clicked on the fotonovela link generally reported enjoying it, with 54.2% of patients who rated the fotonovela responding that they loved it (n=32 of 59) and 35.6% saying they liked it (n=21 of 59). The intervention group was significantly more likely to get screened (18.8%, N=2,597) than controls (11.6%, N=2,644; P<.001), and this pattern held, regardless of demographic subgroup (gender, age, screening history, preferred language, and payer type). Data from qualitative interviews (N=16) indicated that generally the text messages, navigator calls, and fotonovelas were well received and not unduly invasive. Respondents noted several important barriers to CRC screening and had several suggestions for reducing barriers and increasing screening. CONCLUSIONS Texting with automated conversational responses to all eligible patients appears valuable in increasing CRC screening. Intervention participants had significantly greater rates of FIT return, regardless of demographic factors. A significant number of patients remained unscreened, underscoring the need for continued education and multi-level interventions to reduce barriers.