Grant: CIHR Fellowship

Inflammatory arthritis (IA) affects approximately 3 of every 100 Canadians. This condition results from immune dysfunction, leading to damage to organs and joints. Consequently, patients may experience disability, loss of work, pain, fatigue, and even premature death. To prevent such damage and potential early death, all IA patients require disease-modifying antirheumatic drugs (DMARDs). However, these medications can have adverse effects on the liver. A key indicator of this is a change in liver enzymes, which may signal potential liver damage. Regularly monitoring these enzymes in the blood is crucial to initiate timely treatments, prevent severe problems, and reduce both health resource usage and healthcare costs. Despite its importance, there is limited research on the frequency of missed liver enzyme checks. Regardless of whether monitoring is conducted, having a predictive tool to assess the future risk of liver damage would be invaluable to both clinicians and patients. Our study has two main goals. 1) Understanding Monitoring Gaps: We want to see how often liver checks are missed. We suspect that many IA patients will have some missed checks, but males with lower income, more severe diseases, or living far from their doctors might miss more. 2) Predicting Liver Problems: We're comparing two methods to predict changes in liver enzymes. One uses traditional statistical techniques, and the other uses advanced computer techniques called machine learning. We believe the machine learning method might be better because it can handle complex information more effectively. We will use a massive health database that contains information from over 2.5 million patients. We hope our findings will make it safer for those taking DMARDs. Our research will 1) identify any shortcomings in current liver health monitoring practices for DMARD users; and 2) determine if computerized tools can assist doctors in providing personalized advice about potential risks to IA patients.