Automated patient identification for clinical trials using oncology pathways.

605 Background: Current manual processes for identifying candidates for clinical trials are inefficient and contribute to an unacceptably low participation rate in adult cancer patients. Our team has built and scaled the PACT (Pathways for Accrual to Clinical Trial) program, which uses Clinical Pathways navigations to identify potential candidates for clinical trials in near-real-time. Methods: Providers are asked to navigate the Pathway decision platform each time they write a new treatment order. Each navigation provides structured data about patient and disease characteristics, relevant biomarkers, and the selected treatment. Using our PACT Tableau Dashboard, principal investigators (PIs) and clinical research coordinators (CRCs) can access a current list of potentially eligible patients. The dashboard currently supports 10 non-treatment trials and uses row-level security to ensure that the investigators can only access data relevant to their respective trials. The analysis includes accrual rate data from the opening of each trial to May 2025, as well as the results of our pre- and post-data delivery survey that measures the impact of our PACT program on its users. Poisson analysis was used to calculate accrual rate confidence intervals, and chi-square analysis was used to assess differences between pre- and post-PACT rates. Results: Across the 10 trials, the total accrual rate increased from 0.64 accruals/trial-month (95% CI .53 - .77) before PACT involvement to 1.54 accruals/trial-month (95% CI 1.32-1.78) with PACT involvement, more than doubling enrollment (p<0.0001). Out of the 8 investigators surveyed (2 trials predated the survey creation), 100% reported improved workflow effectiveness and 75% reported improved workflow efficiency. 100% also agreed that the characteristics of the patients identified through Pathways fit the eligibility criteria as expected. Conclusions: Our PACT program has more than doubled the accrual rate for the trials it supports, while providing a more effective and efficient accrual workflow for clinical trial staff. We are working to scale this automated patient identification program across a broader swath of trials. Trial Accruals, Pre-PACT Accrual Rate Pre-PACT (Accruals/Trial-Month) Accruals, Post-PACT Accrual Rate Post-PACT (Accruals/Trial-Month) p-value 21-519 16 in 14 mo 1.14 76 in 29 mo 2.62 p=0.0019 22-028 0 in 7 mo 0 34 in 29 mo 1.17 p=0.0042 21-741 29 in 26 mo 1.12 13 in 9 mo 1.44 p=0.4373 23-109 4 in 14 mo .29 4 in 6 mo .67 p=0.2170 21-299 21 in 39 mo .54 6 in 6 mo 1 p=0.1742 20-222 29 in 45 mo .64 4 in 7 mo .57 p=0.8215 23-341 1 in 8 mo .13 3 in 8 mo .38 p=0.3173 24-300 0 in 2 mo 0 22 patients in 8 mo 2.75 p=0.0190 21-559 16 in 29 mo .55 3 in 8 mo .38 p=0.5369 23-668 8 in 9 mo .89 10 in 4 mo 2.5 p=0.0227 Total 124 in 193 mo .64 175 in 114 mo 1.54 p<0.0001