Using Machine Learning to Predict 30-Day Risk of Overdose Hospitalization, Emergency Visit or Death Among Albertans Who Receive a Dispensation of Prescription Opioids
Prescribed opioids are a key factor in driving the current opioid crisis in Alberta (AB), a province of Canada, and nationwide. Over 400,000 patients a year currently receive at least one opioid prescription in AB, out of a population of 4.5 million. Currently, no risk calculators are available that quantify the risk of hospitalization or death following a dispensation of prescribed opioids and prescribers must rely on guidelines to assess risk. OKAKI is a public health informatics social enterprise that manages Alberta’s prescription monitoring program. OKAKI and its partners undertook a study to use machine learning (ML) with the province’s large health administrative datasets to develop and test a model to predict the 30-day risk of hospitalization, emergency department visit (ED) or death, at the time of an opioid dispensation. A model that performed better than current approaches could potentially be used as a clinical decision aid or to support province-wide monitoring and interventions. The presentation will summarize the study context, methods and results, as well as key challenges in implementing this type of prediction model in clinical or public health practice.
Dr. Samanani is a physician executive with more than 20 years of clinical, public health, and health system analytics experience. In 2008, he founded OKAKI (Blackfoot meaning "be wise"), a public health informatics social enterprise, headquartered in Alberta, Canada. For over a decade, OKAKI has leveraged IT and data to help address challenges in indigenous health, environmental health, emergency and disaster response, immunization, chronic disease and prescription monitoring. In 2019, OKAKI and its partners launched a precision public health initiative, with an initial focus on predicting and reducing opioid overdoses.