Rheumatoid arthritis (RA) is a common autoimmune condition affecting 1% of the population and is associated with substantial morbidity and mortality. Treatment advances have improved disease control, but infections remain a major complication of immunosuppressive treatment. Glucocorticoids are used as long-term therapy in 30-60% of patients with RA and are potentially a major contributor to infection risk in RA. It is well known that higher glucocorticoid doses carry very high infection risk, but risk with more commonly used low-dose therapy remains unknown. Defining infection risk with low-dose glucocorticoids has been identified as a critical area for research, needed to inform treatment strategies to improve outcomes in patients with RA. The objective of this study is to determine whether low-dose glucocorticoids are associated with an increase in the risk of 1) serious infections requiring hospitalization or 2) common infections that do not lead to hospitalization. We will examine the risk of serious infections requiring hospitalization by applying advanced pharmacoepidemiologic methods to two large claims databases and a longitudinal registry that includes disease activity measures. By conducting a propensity-adjusted analysis and a novel instrumental variable analysis based on the variability in the prescribing of glucocorticoids, we will provide the most comprehensive evaluation of infection risk with glucocorticoids to date. A different approach is needed to study non- hospitalized infections. These infections are more than 20 times more common than hospitalized infections but cannot be accurately identified in existing data sources. We will conduct a prospective study with frequent assessments to measure these infections, recruiting patients from a large online patient registry. An additional key goal of this work will be to determine the impact of common infections on patients. Dr. George's career goal is to become an independent investigator studying treatment safety and effectiveness to improve outcomes in patients with rheumatic disease. He and his mentors have created a comprehensive career development plan that takes advantage of the unique skills of his mentorship team and the resources available at Penn. With formal coursework and hands-on experience guided by his mentorship team, he will build on his current training in epidemiology and experience using administrative claims data. He will gain skills in advanced pharmacoepidemiology including instrumental variable analysis, and he will develop new expertise in prospective patient-oriented research, a critical skill needed to develop future interventional studies to complement observational data. Through this work, and with the guidance of his mentorship team, he will gain the skills and experience needed to launch a successful career as an independent physician scientist and become a leader in rheumatoid arthritis research.
Infections are a major cause of morbidity and mortality in patients with rheumatoid arthritis, and understanding the contribution of immunosuppressing medications to infection risk is a key step to improving patient outcomes. A key unanswered question is the risk of infection with low-dose glucocorticoids, one of the most common medications used in the treatment of rheumatoid arthritis. The purpose of the proposed study is to determine the risk of both serious (hospitalized) infections and common (non-hospitalized) infections with low- dose glucocorticoids.
