Syndromic Surveillance for Post-vaccination Adverse Events Using SPRT: A Retrospective Analysis

• Activity Number: 392
• Type: Invited
• Date/Time: Thursday, August 12, 2004 : 8:30 AM to 10:20 AM
• Sponsor: Section on Government Statistics
• Author(s): Margarette S. Kolczak*+ and Robert Davis and Edwin Lewis and Michael J. Goodman and Richard Platt and Robert T. Chen and David Shay
• Companies: Centers for Disease Control and Prevention and University of Washington and Kaiser Permanente of Northern California and HealthPartners Research Foundation and Harvard Medical School and Harvard Pilgrim Healthcare and Centers for Disease Control and Prevention and Centers for Disease Control and Prevention
• Keywords: sequential probability ratio test ; risk-adjustment ; vaccination

Abstract

Methods are needed to quickly detect adverse events following the introduction of new vaccines. Here we propose a method that uses sequential probability ratio test (SPRT) charts. To evaluate this method, we use retrospective data from the Vaccine Safety Datalink (VSD) to look for expected changes in adverse events following the introduction of the rotavirus and diphtheria-tetanus-acellular pertussis (DTaP) vaccines. To construct a SPRT chart, a test must be formulated with specific values for the probability of a post-vaccination event under a null hypothesis (p0) and under an alternative hypothesis (p1). To test these hypotheses, we created weekly cohorts of vaccinated children. The test compares the hypothesized probability of an adverse event in the surveillance period which follows the introduction of a new vaccine to the probability obtained using a comparison vaccination in the baseline period before the introduction of the new vaccine. To control for potential confounders such as HMO, age, calendar time, season, and gender, we extended the SPRT by using the risk-adjustment methods described by Steiner et. al.