A threshold hazard model for estimating serious infection risk following anti-tumor necrosis factor therapy in rheumatoid arthritis patients

摘要

Over recent years novel biologic agents have been developed for the treatment of rheumatoid arthritis. The most common type of biologic agent in use in the United Kingdom is the anti-tumor necrosis factor inhibitor class. To fully appreciate the potential risks of anti-tumor necrosis factor therapy in patients, knowledge about the baseline hazard (risk pattern) and the characteristics of patients associated with serious infection is important. We propose a nonproportional hazard model for estimating the infection risk, by including the drug exposure history information into the baseline hazard. We found that the infection risk reaches a peak within 1 month after drug exposure starts and then declines steadily for nearly 2 years before stabilizing out.