Writing our book (Kopeikin et al. 2011), we suggested the following definition of event:
Let us consider a physical process taking place within a sufficiently small volume
of space and a sufficiently short period of time. One assumes that it is possible to
reduce the actual physical process to its limiting content by shrinking the volume
and the time interval to zero. This brings to life an idealised notion of event,
an often used abstraction of a physical phenomenon taking place at one point
in space and at one instant of time. Examples of the events are a photon’s
emission or absorption, a collision of two elementary particles, a meteorite’s burst
in planetary atmosphere, etc. The event is mathematically identified with a
point, and the physical world is considered as consisting of a continuous set of
points modeling its underlying mathematical structure. The set of events is further
equipped with additional mathematical paraphernalia which convert it into a well explored mathematical object called manifold.
I regret that we did not mention what some call "The Third Postulate of Special Relativity", the principle of invariance of coincidences. Tacitly implied since Einstein's times, this principle was spelled out explicitly by David Mermin in his book "Space and Time in Special Relativity":
When one observer says two events coincide in space and time, so will all other observers.
The principle defines an important aspect of the notion of event.
For more on this, see this discussion.
Now, you are asking: "What can be said (for example) of the particle before it collides with the other particles?" In my opinion, this question is answered by the fact that a particle is always interacting with virtual particles. In this sense, an event is happening to the particle at each point of its trajectory.