Using machine learning to predict and attribute mortality to non-adherence to medication among patients with asthma and/or COPD

This project has two objectives. At the individual level, we propose to develop, using artificial intelligence, a predictive tool useful in the doctor-patient relationship. It will indicate the five-year risk of death for asthma and COPD patients based solely on individual electronic information already recorded by the provincial health insurance organism: sociodemographic factors and the history of adherence to each prescribed medication. At the population level, we will estimate the proportion of deaths among asthma and COPD patients that could be prevented by better medication adherence. This estimate will be particularly useful for public policy managers when allocating public resources to preventive measures.

Faculty Supervisor:

Lucie Blais

Student:

Partner:

AstraZeneca Canada Inc

Discipline:

Life Sciences

Sector:

Manufacturing; Professional, scientific and technical services

University:

Université de Montréal

Program: