Integrating epic aura and shared decision-making to improve colorectal cancer screening participation at the University of Washington Medicine.

[OBJECTIVE] Colorectal cancer (CRC) remains a leading cause of cancer-related mortality in the United States. Multi-target stool DNA (mt-sDNA) testing is a noninvasive, highly sensitive screening option, but its effectiveness is limited by workflow inefficiencies and suboptimal patient adherence. University of Washington Medicine (UWM) implemented Epic Aura, an electronic health record (EHR) interface that automates test ordering, patient outreach, and results communication, alongside shared decision-making (SDM) interventions. This study evaluated the impact of these implementations on mt-sDNA utilization and adherence.

[METHODS] This retrospective cohort study used de-identified laboratory data from Exact Sciences for UWM patients aged ≥45 years. The pre-implementation period spanned December 5, 2022-December 4, 2023, and the post-implementation period December 5, 2023-December 4, 2024. The primary outcome was adherence, defined as return of a valid mt-sDNA kit within 365 days of shipment. Descriptive analyses and multivariable logistic regression compared adherence across periods and identified associated factors.

[RESULTS] A total of 125 pre-implementation and 1,160 post-implementation mt-sDNA shipments were analyzed. Utilization increased nearly 10-fold following implementation. Adherence improved from 64.0% pre-implementation to 79.1% post-implementation ( < .001), with faster kit return post-implementation (59.0% vs 48.8% within 30 days). Post-implementation was independently associated with higher adherence (adjusted odds ratio [aOR] = 1.98; 95% CI = 1.30-3.02; = .002). Higher adherence was also associated with insurance type, higher household income, and digital outreach.

[CONCLUSIONS] EHR-integrated workflows combined with SDM were associated with marked improvements in mt-sDNA utilization, timeliness, and adherence, supporting scalable strategies to enhance CRC screening in real-world settings.