Spline-Based Self-Controlled Case Series Method

• Activity Number: 183
• Type: Contributed
• Date/Time: Monday, August 10, 2015 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Epidemiology
• Author(s): Yonas Ghebremichael Weldeselassie* and Paddy Farrington and Heather J. Whitaker
• Companies: The Open University and The Open University and The Open University
• Keywords: Epidemiology ; Integral of I-splines ; M-splines ; Self-controlled case series ; Smooth risk functions

Abstract

The self-controlled case series method is one of the study designs, such as cohort and case control, used to investigate potential associations between the timing of vaccine or other drug exposures and adverse events. It requires data only on cases and automatically controls all fixed confounding variables that could modify the true association between exposure and adverse event. Time-varying confounders such as age must be allowed for explicitly. The original SCCS and recent extensions used step functions to represent exposure and/or age effects. The use of step functions requires pre-specifying exposure risk periods or age groups a priori. However,Poor choice of group boundaries may lead to bias in the association between exposure and adverse event. We propose a spline-based SCCS method in which both age and exposure effects are represented by cubic M-splines. To avoid a numerical integration of the product of these two spline functions in the likelihood function of the SCCS, we defined the first, second and third integrals of I-splines based on the integrals of M-splines. Simulation studies showed that the new method performs well. It is applied to a data on paediatric vaccines.