ML-CDSS

Participants believed the ML-CDSS could improve blood result triaging, reduce unnecessary hospital visits, cut treatment delays, reduce drug and resource waste, and lower overall NHS costs. The ML-CDSS was perceived as potentially mitigating laboratory delays by reducing the total number of blood tests required. The ML-CDSS was not in active clinical use at any participating site at the time of the study, so all findings are based on hypothetical discussion rather than observed practice. The ML-CDSS was developed through a collaboration between UCL, Durham University, and Evergreen Life, with the ML model built by the universities and the web platform by Evergreen Life. The ML-CDSS analyses blood chemistry data specifically from treatment cycles three and four to generate risk predictions. The ML-CDSS was designed for use with chemotherapy regimens that treat breast, colorectal, and lymphoma cancers. The two intended clinical outputs of the ML-CDSS are reducing blood test frequency for lower-risk patients and enabling enhanced monitoring for higher-risk patients.