Kaleido registry: A multi-center registry platform with site staff and automation to enable accelerated clinical research and drug development at scale in community practices.

Abstract

11020 Background: The number of candidate drugs and clinical trials in oncology is rapidly increasing, necessitating innovative approaches to streamline patient identification, enrollment, and trial execution. However, data collection for patient eligibility assessments remains highly manual, placing significant burden on clinical teams. Technology solutions leveraging Electronic Health Records (EHR) data are emerging, but missing or inaccurate data often impede high-fidelity and low-latency prescreening. In 2022, the Kaleido registry was initiated to address these challenges by implementing on-site staff, abstraction prior to each visit, and draft notes for medical oncologists in the registry. Methods: The Kaleido registry enrolls patients with oncological/hematological diagnoses at four community practice sites. Real-time data abstraction is performed in all patients prior to each visit in an expanding set of indications (NSCLC, CRC, prostate cancer, and myeloproliferative neoplasms) to provide systematic and unbiased trial pre-screening. An automated prescreening tool has been developed and validated in 394 pts to identify potential trial candidates. Standardized note drafts free up staff time. A questionnaire addressing social determinants of health (SDOH) and self-reported performance status was introduced to further reduce missing data in the EHR. Results: To date, over 10,000 patients (61% female, 39% male) have been enrolled. 6% of approached patients declined to consent. The most common diagnoses include breast, prostate, lung, and colorectal cancer. The automated prescreening tool was validated with 90% sensitivity and 97% specificity. The patient questionnaire has a completeness rate of 99%. Real-time abstraction was demonstrated to be feasible, with completeness above 90% for key variables. Among patients with metastatic NSCLC, the genomic testing rates exceeded 80% supporting the sites’ suitability for clinical trials. Conclusions: With high consent rates, data completeness, and a validated automation tool, the Kaleido registry offers a scalable model for systematic, unbiased patient recruitment and comprehensive data collection across oncology/hematology. This approach has the potential to accelerate clinical research and drug development in oncology and hematology, ultimately enhancing trial efficiency and improving patient outcomes.