For clinical trial coordinators, busy emergency departments usually act as good locations to identify people who might be good study candidates. According to Yizhao Ni, PhD, Division of Biomedical Informatics and the study’s lead investigator, the ACTES system is specifically designed to simplify a clinical trial recruiting process that usually proves to be inefficient and does not invariably identify a sufficient number of qualified candidates.
Because of the large volume of data documented in EHRs, the recruiting processes used now to find relevant information are very labor-intensive within the short time frame needed. By leveraging natural language processing and machine learning technologies, ACTES was able to quickly analyze different types of data and automatically determine patients' suitability for clinical trials.
Yizhao Ni, Study Lead Investigator and PhD, Division of Biomedical Informatics, Cincinnati Children’s Hospital Medical Center
How it Works
Natural language processing in the ACTES system enables computers to infer and decode human language as the system examines a great deal of linguistic data. Through machine learning, computerized systems can automatically learn and progress from experience without particularly being programmed. This approach makes it feasible for computer programs to extract information, process data, and create knowledge autonomously.
Structured data like clinical assessments and patient demographics from EHRs is extracted by the automated system. In addition, the system detects unstructured data from clinical notes, such as the patients’ treatments, symptoms¸ clinical conditions, and so on. The information, thus obtained, is subsequently matched with eligibility criteria to establish the suitability of a subject for a certain clinical trial.
According to the researchers, the machine learning component of the system enables it to learn from historical enrollments to enhance upcoming recommendations. Carefully designed AI algorithms handled most of the analyses. These include essentially formulas or procedures that can be used by computers to solve issues by executing a set sequence of definite actions.
Advanced to Live Clinical Setting
In 2015, the Journal of the American Medical Informatics Association reported the successful pilot testing of the system. In the present study, the solution was prospectively tested and in real time in a busy emergency department setting. Here, patients were recruited by clinical research coordinators for six varied pediatric clinical trials that involved a range of diseases.
Applying the novel technology in a live clinical setting involved considerable association between information service technicians, application developers, data scientists, clinical staff, and the end-users.
Thanks to the institution’s collaborative environment, we successfully incorporated different groups of experts in designing the integration process of this AI solution.
Yizhao Ni, Study Lead Investigator and PhD, Division of Biomedical Informatics, Cincinnati Children’s Hospital Medical Center
The small number of clinical trials employed in the study, all from one clinical department, was listed as limitations by the researchers. The team also emphasized some persistent problems involving the precision of the system at inferring data. In upcoming studies, these problems will be resolved via ongoing improvements to the technologies as well as by testing the system in a boarder range of clinical departments, added the investigators.
The study was partly funded by the National Institutes of Health (1R01LM012230, 1U01HG008666, 5U18DP006134), Agency for Healthcare Research and Quality (1R21HS024983), and from Cincinnati Children’s Hospital Medical Center. The production of ACTES is licensed via the medical center’s office for technology commercialization, Innovation Ventures.
Source: https://www.cincinnatichildrens.org/
Original Paper: Ni Y, Bermudez M, Kennebeck S, Liddy-Hicks S, Dexheimer J
A Real-Time Automated Patient Screening System for Clinical Trials Eligibility in an Emergency Department: Design and Evaluation
JMIR Med Inform 2019;7(3):e14185