ABSTRACT

Meta-analysis is the application of statistical techniques to quantitatively combine the results of independent studies, deemed similar and appropriate to be combined, to allow inferences to be made to the population of interest. It has been used in regulatory decision making, especially in safety evaluations, quite extensively. This framework can be motivated for several reasons:

• To provide a more precise estimate of a measure of risk (and its uncertainty) associated with the use of a regulated product than a single trial can provide

• To rule out a certain threshold risk value for a potential class effect (e.g., cardiovascular risk in type 2 diabetes mellitus products)

• To search for adverse events not previously encountered

CONTENTS

13.1 Introduction ................................................................................................ 237 13.2 Design Issues .............................................................................................. 239 13.3 Outcome Ascertainment ........................................................................... 240 13.4 Statistical Methods .................................................................................... 241 13.5 Reporting .................................................................................................... 243 13.6 Example 1: Antidepressants and Suicidal Events ................................. 244 13.7 Example 2: Rosiglitazone and Cardiovascular Outcomes ................... 246 13.8 Conclusion .................................................................................................. 249 References .............................................................................................................250