This page is currently being updated from the earlier version of my website. Sorry that it is not yet fully available.
Someone on the MedStats email discussion group asked about how to analyze adverse event data. He noted that adverse event data is not one of the primary or secondary outcome measures, and wondered if it would be appropriate to provide statistical analysis of this data.
Adverse events (and safety data in general) represent a special type of analysis that does not fit in well with the listing of primary/secondary outcomes. The main reason for this is the number of possible adverse event categories is very broad and it is not always possible to anticipate in advance what type of adverse events are of greatest interest.
You have to analyze adverse events, of course. To ignore them is just bad research. But there is no clear consensus on how to analyze adverse event reports or safety data in general. Here are some of my random thoughts.
- Some safety issues can be anticipated in advance because of knowledge about other drugs in the same class, information gleaned from animal experiments and early phase clinical trials, and suspicions based on plausible biological mechanism. These should be specified in the protocol and should be considered primary outcomes alongside the efficacy outcomes.
- You should not use a Bonferroni adjustment even though the safety data will add multiple outcomes on top of the efficacy measure(s). In particular, a global null hypothesis that includes both efficacy and safety measures is nonsensical.
- Any adverse events which were not anticipated during protocol development but which occur at a greater than expected rate, either with respect to the control group or to a historical background rate, should be reported. An analysis of these events should be largely qualitative in nature.
- A differential drop out rate between treatment arms is suspicious and may indicate that a particular drug or treatment is more difficult to tolerate.
- How the adverse events color your overall recommendation is a complex process. The severity and frequency of the adverse events need to be balanced against the value of a cure and the improvement in the probability of observing a cure with the new treatment. You should discuss the plausibility of the adverse events in light of the size of the effect, previous research, biological mechanisms, etc.
All of the above are just my opinions, of course, and I don’t want to pretend that they are drawn from any consensus opinion among researchers.
You can find an earlier version of this page on my original website.
Also see this page.