High-Touch vs Low-Touch Strategy for Implementing a CRC Screening Digital Health Intervention A Randomized Clinical Trial

IMPORTANCE Patient, clinician, and system-level barriers contribute to low colorectal cancer (CRC) screening rates. OBJECTIVE To determine whether a more intensive implementation strategy results in greater use of a mobile app targeting multilevel barriers to CRC screening. DESIGN, SETTING, AND PARTICIPANTS This hybrid type 3 effectiveness-implementation pragmatic trial of a tablet app for CRC screening was conducted from March 2021 to March 2023. Eighteen community-based primary care practices were randomized to either a low-touch (n = 8) or a high-touch implementation strategy (n = 10). Analysis was completed in October 2024. INTERVENTIONS All clinics received on-site training and technical support. High-touch clinics received at-elbow support during launch, identification of a clinic champion, audit and feedback, and performance coaching. MAIN OUTCOMES AND MEASURES The primary outcome included reach-the percentage of eligible patients who completed the CRC app in the month 6 after implementation-and effectiveness-the percentage of screening-eligible individuals after implementation who completed a CRC test within 16 weeks of their index visit, compared with a preimplementation cohort. RESULTS Overall, 50 176 patients 18 years and older were seen (33 435 at high-touch and 16 741 at low-touch practices) in the first 6 months after implementation. The mean (SD) age of patients was 55.6 (17.7) years; most patients were female (59%) and had commercial insurance (51%). Overall, 2.0% of participants were American Indian or Alaska Native, 11.2% were Black or African American, 0.1% were Native Hawaiian, 81.1% were White, 4.5%were other race, and 0.3%were unknown. CRC app use started at 5.8% in the high-touch group and 5.3%in the low-touch group in the first month (P = .55), and decreased to 0.9%and 1.0%, respectively, in the sixth month (P = .93). Compared with low-touch clinics, high-touch clinics had similar use of the CheckIn app (odds ratio [OR], 2.8 [95%CI, 0.9-9.0]) and CRC app (OR, 1.7 [95%CI, 1.0-3.0]). After adjusting for month, there was no observed difference in CRC screening after the CRC app was implemented compared with 8 months before implementation (OR, 0.9 [95%CI, 0.7-1.2]). CONCLUSIONS AND RELEVANCE This study found that use of a patient-facing health app for CRC was low, regardless of implementation intensity. CRC screening completion rates remained unchanged. Challenges of staff turnover, postpandemic fatigue, multiple handoffs in the workflow, and competing time demands may be difficult to overcome with implementation strategies.