Having a look at AE reporting

Adverse event reporting seems easy enough: document a diagnosis (ideal, but not always the case...obviously), get a start date/time, get an end date/time, and ask your friendly investigator to assess it for causality. He or she may not ask questions about the adverse event and simply make an assessment as to whether or not it may be related to study product. Let's stop right there. What is the problem with this?

Most adverse event logs only list the information that is required by the case report form, and very little else. This is a problem because unless you are going to hold the investigator's hand while he is assessing the adverse events, there is very little that is presented to him which might allow him to make an informed clinical judgement about the causality of a particular adverse event. Granted this may not seem like a big deal at the site level, but if this situation is occurring at most sites, then the whole process is flawed to the point where inaccurate safety information is being gathered.

I believe there is a simple resolution to this issue that would not take up the valuable time of the coordinator while allowing the investigator to assess adverse events accurately. An adverse event log could be very easily amended to include the following information along with is required by the CRF:

-last dose of study medication (if applicable)

-whether or not the AE was listed in the investigator brochure/informed consent (simple yes/no) as a known risk

-whether or not the AE is of "special interest" to the sponsor

-any other study specific information that may of use: is there a titration period? is the AE intermittent? persistent? any other events that may have caused it?

This sort of information made readily available to investigators would greatly increase the accuracy of adverse event reporting across the board. I could think of one hypothetical example to drive home the point looking at adverse events in a vaccine trial:

Subject is vaccinated on day 0. On day 20 the subject reports itching at the injection site after being bitten by a mosquito. The bite becomes inflamed and the subject reports itching and pain in the injection site (say, left shoulder) to the study coordinator who documents the two adverse events on a common AE log and leaves it for the investigator to assess. Without knowing the last vaccination date or any other historical information, injection site pain/itching would seem like a dead ringer to be absolutely related to the vaccine. Thus the investigator chooses that option and moves onward, leaving the sponsor with inaccurate data that could have been rectified by simply providing more information to the investigator.

I realize that many adverse event logs are convenient because they allow for collection of multiple events in limited space, but I believe it's time to re-examine the way that AEs are evaluated by providing the investigators with information that is completely necessary for an appropriate clinical judgement of causality.