TriHealth Cancer Institute’s collaboration with the Tempus AI TIME program impact on clinical trial operations and enrollment.

Abstract

1553 Background: TriHealth Cancer Institute has the largest personalized medicine program serving Cincinnati. TriHealth’s EMR integration with Tempus AI TIME allows for patient screening using an AI powered software TApp that matches to a broad clinical trial portfolio. Each patient is followed using LINK portal and trials can be activated rapidly via TIME’s operational model. The goal of this collaboration is to increase TriHealth’s patients' access for clinical trials. Methods: The TApp is software utilizing subject data, trial eligibility criteria, and NLP models for AI matching to TIME studies. Unique TApp searches occur whenever there is new patient data or changes to eligibility criteria. All TApp matches are reviewed by Tempus AI nurses followed by confirmatory screening by TriHealth’s staff. Trials were activated using the TIME program’s standardized operational methods including a pre-negotiated rate card, trials agreement, and a central IRB. Data included number of patients screened, unique TApp searches, trial activations, consents, operational metrics, and financial impact. Results: From January 2022 - March 2023, TApp performed 8,653,397 unique searches for 135 TIME trials on TriHealth’s population of 18,823 patients. During this period, 985 trial eligibility criteria were modified and 200,776 patient clinical updates occurred. There were 2,032 potential matches (1,944 interventional, 88 observational) for 45 different trials. Tempus AI nurses spent 389 hours reviewing all 1,944 interventional matches and of these sent 284 to TriHealth for evaluation. For TIME studies, 28 patients consented (18 interventional, 10 observational) and 7 trials were activated (4 interventional, 3 observational). Total trial billable revenue for TIME studies was approximately $241,500 ($34,500 average per trial). Regarding non-TIME trials, 3 patients consented on 3 interventional studies. Activation time for non-TIME trials compared to TIME trials was 186 days versus 37 days. Conclusions: TriHealth’s partnership with TIME evaluated 18,800+ patients for 135 trials. Over 8.6M unique TApp searches resulted in 2,000+ patients screened by nurses. Operationally, TIME trials resulted in a 2.3X increase in the number of trial activations, an 80% reduction in days to activation, and 9.3 fold increase in enrollment compared to non-TIME studies. The use of AI enabled patient screening, combined with a structured trial activation process, improved patient enrollment at a community health system. [Table: see text]