Grant: Project Grant

Persons who inject drugs (PWID) are at high risk of infective endocarditis, an infection of the heart's inner lining that can lead to abnormal growths and damage the heart valves. A major consideration of the risk-benefit of surgery compared with medical management in PWID is reinfection of the heart valve from drug use. Infective endocarditis treatment guidelines are based studies of older patients without substance use disorder and with more health problems than PWID; many studies are methodologically flawed. Without data on outcomes of PWID after infective endocarditis, treatment in PWID will continue to be informed by non-PWID patients and PWID mortality will remain high. This problem is substantial: there has been a >160% increase in infective endocarditis in PWID in Canada with the ongoing drug crisis. To address these knowledge gaps, we will use existing health record data (www.ICES.on.ca) for Ontario residents from 2012-2024 to assemble the largest population-based cohort of PWID with endocarditis (N~4,000). To control for addiction and other factors that can affect death, we will 1) restrict the population to PWID with infective endocarditis, 2) use propensity scores to estimate and adjust for the patient's probability of surgery compared with medical treatment based on a wide range of measured patient characteristics, and 3) use novel approaches to incorporate hospital chart abstracted data on cardiac imaging, bacterial blood test results, and in hospital drug use for a subset of 900 patients to further balance patient characteristics by endocarditis treatment. Our methods address study error that typically arises from the absence of randomization to treatment. These results will provide needed evidence on treatment and outcomes to optimize care in this growing population.