Refining Risk Prediction Algorithms for Post-Discharge Death Through Edge Case Analysis

In many low-resource communities, children who are treated for serious infections still face a high risk of dying after they leave the hospital. Most of these deaths happen within weeks or months, often due to poverty and poor access to follow-up care. To address this, a team from BC Children’s Hospital and the University of British Columbia developed ‘Smart Discharges’. Smart Discharges helps health workers identify children at the highest risk of dying after discharge and create personalized care plans for them.

In Uganda, Smart Discharges has been introduced in 19 hospitals. The system uses health data collected at hospital admission to predict risk, but researchers believe it could be improved. This project will explore “edge cases,” or children who died even though they were predicted to be low-risk, using artificial intelligence and machine learning to find other risk factors.

These findings may also help improve healthcare in rural and remote areas of British Columbia, where children often face worse health outcomes and long, difficult journeys to access specialized care. The project could lead to better care planning and use of resources. It also helps attract passionate health professionals to the province through meaningful global health experiences.

Faculty Supervisor:

Matthew Wiens

Student:

Partner:

BC Children's Hospital

Discipline:

Computer science

Sector:

Health and Related Sciences & Technology

University:

The University of British Columbia

Program: