Using administrative data to predict health outcomes: a machine learning approach

Adverse events due to benzodiazepine use and falls are important clinical outcomes in older adults. As well, high-cost users of the health care system, although small in number, are a substantial burden with regards to health costs. Continued use of BZRAs in older adults is concerning from a public health standpoint; 1 in 3 older adults experience a fall in the community; 5% of health care users consume about 60% of hospital and home care spending. Being able to identify these individuals would add to the current efforts to reduce health care burden.

Using Machine Learning to Predict 30-Day Risk of Hospitalization, Emergency Visit or Death Among Albertans Who Received Opioid Prescriptions

When utilizing and implementing ML for prediction using administrative health data, two key issues are ML algorithm evaluation and generalizability21. Current approaches evaluate model performance by quantifying how closely the prediction made by the model matches known health outcomes. Evaluation metrics include sensitivity, specificity, and positive predictive value, as well as measures such as the area under the receiver operating characteristic (ROC) curve, the area under the precision-recall curve, and calibration.

Predicting risk of unplanned hospital readmission within 30-days of discharge using machine learning approaches

Unplanned hospital readmissions are a preventable and costly outcome in the health care system. There are limited tools to estimate risk of readmission. The machine learning process offers an opportunity to develop a risk predictor to identify those at high risk of readmission upon discharge. OKAKI has an opportunity to diversify the commercial products it can offer to health care administrators.

AI in Ophthalmology triage automation

There are currently 18 retina specialists in the province of Alberta, approximately half in Calgary and half in Edmonton. Retinal diseases are common. For example, approximately 6.5 percent of people age 40 and older have some degree of macular degeneration. Diabetes retinopathy affects approximately 500,000 Canadians. Many retinal conditions are treatable when detected early, however retinal specialists are concentrated in large urban centres. There are 700+ optometrists across the province today capable of taking and transmitting retinal images for consultation with a retinal specialist.

Benzodiazepine and Opioids Use in Alberta

Opioids have received much attention in the media, public and government because of the risks associated with them, including fatalities. Concurrent use of BZRAs (benzodiazepines used for treating anxiety and insomnia) and opioids is of particular concern because this is a recognized risk factor for fatal opioid overdoses. Despite this warning, concurrent use is still occurring. An outcome study using Alberta data on concurrent use has not been published in the literature.