46. Sequential Statistical Methods for Prospective Postmarketing Safety Surveillance

  1. Brian L. Strom MD, MPH2,3,4,5,
  2. Stephen E. Kimmel MD, MSCE4,5 and
  3. Sean Hennessy PHARMD, PHD4,5
  1. Martin Kulldorff

Published Online: 3 JAN 2012

DOI: 10.1002/9781119959946.ch46

Pharmacoepidemiology, Fifth Edition

Pharmacoepidemiology, Fifth Edition

How to Cite

Kulldorff, M. (2012) Sequential Statistical Methods for Prospective Postmarketing Safety Surveillance, in Pharmacoepidemiology, Fifth Edition (eds B. L. Strom, S. E. Kimmel and S. Hennessy), Wiley-Blackwell, Oxford, UK. doi: 10.1002/9781119959946.ch46

Editor Information

  1. 2

    George S. Pepper Professor of Public Health and Preventive Medicine, Philadelphia, PA, USA

  2. 3

    Department of Biostatistics and Epidemiology, Philadelphia, PA, USA

  3. 4

    Center for Clinical Epidemiology and Biostatistics, Philadelphia, PA, USA

  4. 5

    Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA

Author Information

  1. Department of Population Medicine, Harvard Medical School and Harvard, Pilgrim Health Care Institute, Boston, MA, USA

Publication History

  1. Published Online: 3 JAN 2012
  2. Published Print: 17 FEB 2012

ISBN Information

Print ISBN: 9780470654750

Online ISBN: 9781119959946



  • sequential analysis;
  • sequential probability ratio test;
  • group sequential method;
  • pharmacovigilance;
  • drug safety surveillance;
  • vaccine safety


Near real-time postmarketing drug and vaccine safety surveillance systems are increasingly being developed and used to quickly detect potential safety problems. Many of these systems use automated weekly or monthly data updates from electronic medical health records or health insurance claims, which contain information about both drug/ vaccine exposure and potential adverse event outcomes from a well-defined population. When statistical analyses are repeatedly conducted on the same slightly expanded data, sequential statistical methods are needed to adjust for the multiple testing inherent in the many looks at the data. In this chapter, we describe such methods and their use for drug and vaccine safety surveillance.