Cohort 4

Rachael Lear

In my 12 months as a Topol Digital Fellow I will develop natural language processing pipelines for turning free-text documentation into insights around why patients are falling in hospital.

My Topol fellowship problem / project:

Problem Statement

Falls are the most frequently reported serious incident in hospitals, associated with physical and psychological harm and increased costs. Understanding why falls have occurred currently relies on manual review by clinical leads. This process is time-consuming meaning the focus is on investigation rather than improving care.

Opportunity

There is a need for innovative tools that help clinical teams to understand their falls data while reducing their workload burden and releasing time for improvement activity. One approach is to develop the infrastructure and interface for a new clinical dashboard that automatically displays near-real-time insights into why patients are falling, based on documentation from electronic patient records and incident reporting systems.

Project goals

  • Develop natural language processing pipelines for turning free-text documentation into insights around why patients are falling in hospital.
  • Co-design a prototype dashboard for displaying near-real-time data on inpatient falls for clinical teams.
  • Evaluate usability of the dashboard and its impact on safety improvement activity

I currently work as a Research Fellow in Digital Health. My work involves designing, implementing, and evaluating digital innovations that address the needs of patients and the health service. I am a registered nurse and previously worked in a busy vascular unit caring for patients with conditions affecting their arteries and veins. During that time, I experienced how digital health technologies can bring substantial benefits providing that the diverse needs of patients and professionals are understood and addressed. These experiences made me want take on more of a leadership role in digital health, to ensure that technologies are being used meaningfully with proper evaluation of their impacts.