Detection of safety signals in randomised controlled trials using groupings

Research output: Contribution to conferencePoster

34 Downloads (Pure)


Many different types of adverse event are routinely recorded during clinical trials. The statistical analysis of this data may need to take into account:
1.potential multiple comparison issues;
2.low power - effect sizes of adverse events in clinical trials are generally small.
The use of methods which use possible groupings of adverse events (e.g. by System Organ Class) in their statistical analyses may result in an increase in the power to detect adverse event incidence while maintaining control over the Type-I error rate.
Original languageEnglish
Number of pages1
Publication statusPublished - 5 Sep 2017
EventRoyal Statistical Society Conference 2017 - Glasgow
Duration: 4 Sep 20177 Sep 2017


ConferenceRoyal Statistical Society Conference 2017
Abbreviated titleRSS2017


  • safety
  • clinical trials
  • adverse events
  • body system
  • system organ class
  • Bayesian hierarchy
  • false discovery
  • rate
  • interim analysis

Cite this