Operational machine learning and mechanistic modelling for supporting patient flow at GOSH

Patient flow describes the movement of patients through a hospital as they transition along their care pathway. Good patient flow is central to patient safety, hospital efficiency and patient experience, and is therefore of great importance in the NHS.

Patient flow modelling can be used to understand patient flow and help improve efficiency in areas such as demand and capacity planning, appointment scheduling and patient pathway optimisation. Great Ormond Street Hospital (GOSH) has recently developed a cloud-based data and analytics platform, the digital research environment (DRE), and an associated digital research and informatics unit, DRIVE. The DRE collects electronic patient records data from the hospital for research, presenting the opportunity to model patient flow at GOSH to inform future flow planning and strategy within the trust.

Using patient-level data, this project aims to identify trust-wide patient flow pathways using network analysis and clustering, apply statistical and machine learning methods to predict markers of patient flow such as length of stay, build department-specific patient flow models using mechanistic modelling and deep learning techniques, and develop a predictive model of patient flow across the hospital.