First, Catch Your Signal! I. Ralph Edwards and Marie Lindquist Uppsala Monitoring Centre, Uppsala, Sweden 1. The Trouble with Definitions Over the years there have been many attempts to define what we mean by a signal. We have come to the view that attempts to provide a sin- gle, concise definition cannot succeed because a signal in pharmacovigilance is not an entity. There are several signal endpoints possible that depend on a multiplicity of factors, defying a deconstructural approach. It is a little like defin- ing ‘love’, which is very subjective and, even al- lowing for that, it seems that there are different kinds of love. No definition allows us to under- stand the range of love and its attributes! This does not mean that the definitions are wrong, only that they do not give a full picture. One might start by saying that a signal in pharmacovigilance is any new information about a drug or its use that is interesting/important, and which may have some impact on the way the drug will be used. This simple concept is incorporated in all definitions, which are then expanded in a variety of ways. But the difficulty with this starting point is that a signal is part of a process that starts with an observation that arouses in- terest or concern. This is a value judgement by the observer, leading him or her to more thought, investigation and activity around the observation to the point where the observer feels they have a hypothesis to share with others. That point seems intuitively right to designate a ‘signal’ in the sense of information worth passing on to others, for them to know about and to consider further ac- tion. But all this is vague and subjective: what is ‘interesting’ or ‘important’? What kinds of ‘thought’ and activity’ are required? How does one decide whether one has sufficient informa- tion and argument to warrant telling others? There is another problem arising from the wide use of ‘signal’ and ‘alert’. Some would say that a signal is the point at which one individual starts to do more investigation, and an alert is the point at which one tells others about the result (who the others are is not clear, but seems to be usually used for the wider health professional audience). Some people use ‘signal’ and ‘alert’ in the opposite sense, but we will not complicate matters by mentioning ‘alert’ further! We propose that a signal can only be under- stood by answering the following questions: What do I think is an important observation af- fecting drug use and patient safety? How do I find these observations and consider them? How do I know when I have found enough for a signal? What should be done with the signal? These questions can be considered both generally and in relation to an individual case. 2. What Do I Think Is an Important Observation Affecting Drug Use and Patient Safety? A useful guide is to consider whether an ob- servation might improve the use of a drug for some patients, and how the new information should be developed to make it optimal. It is important to take into account how dif- ferent stakeholders might need to use the in- formation. One challenge is that there are many potential useful bits of information and little time to consider them all, but the above consideration is surely helpful in making priorities. The context of the signal reviewer’s situation and experience are critical and, if we are con- cerned with case reports, this also applies to EDITORIAL Drug Saf 2010; 33 (4): 257-260 0114-5916/10/0004-0257/$49.95/0 ª 2010 Adis Data Information BV. All rights reserved.