The adverse events observed in a clinical trial are conventionally presented as a set of multiple tables. An individual table contains the counts of adverse events of a certain type (e.g. treatment-related events or events with a certain severity grade) observed within a prespecified set of subsystems (preferred terms) within multiple body systems for different treatment groups. The inspection of the tables by the DMC or FDA reviewer is expected to detect a small subset of body systems/preferred terms where the differences in the proportions of adverse events across the treatment groups are the most striking.

The detection is somewhat complicated by the large number of multiple comparisons to be made. We present a simple visualization display that helps the reviewers to focus their initial attention on the potentially relevant items in the adverse events table. Our display may be viewed as a variation of the suggestions of Amit, Heiberger and Lane (2008) , Harrell (2010) and Mehrotra and Adewale (2012).