Timely Ordering of Pharmacogenetic Testing

This study aims to evaluate the effectiveness of a ML model in predicting patients at high risk of requiring a "targeted medication" within the next three months. A machine learning model will predict, the morning following admission to any inpatient service, whether there will be receipt of a targeted medication within the next three months. The research team will be notified regarding eligible patients each morning, and the research team or pharmacogenomics team will approach the patient's primary care team as applicable. By leveraging ML, this study seeks to enhance the identification of patients who would benefit from such medications in a timely and resource-efficient manner.

The study team identified specific medications as indications for pharmacogenetic testing based on prevalence and level of evidence for modifying prescribing practices. These pre-selected medications are referred to as "targeted medications" and are as follows: azathioprine, brivaracetam, clobazam, clopidogrel, flecainide, phenytoin, tacrolimus, voriconazole and warfarin. Only systemically administered (oral, subcutaneous, intramuscular or intravenous) medications or prescriptions (e.g. not topical, intrathecal or intravitreal) are included. Phenytoin was only considered if given orally (to exclude emergency administration without a plan for ongoing treatment).

Pharmacogenetic testing will be offered to participants and conducted as addressed in an associated pharmacogenetic testing protocol (REB# 1000053445 PI: Iris Cohn).

Inclusion criteria

• Inpatient at The Hospital for Sick Children
• Between 6 months to 18 years old

Exclusion criteria

• Prior pharmacogenetic testing and/or prior receipt of a targeted medication
• Current Intensive Care Unit (ICU) admission
• Expected hospital discharge is prior to midnight on the day of admission