As an experimentalist, I will first address this last summary of your question:
How do you know the particle is still there?
Let us define the terms of the question:
a) In particle physics we know a particle was there by the tracks it leaves in a bubble chamber.
b)By the signals it sets off as it passes and ionizes
The measurements have shown us that we deal with very small dimensions in all quantities, mass, size etc.
We also have found that particles follow quantum dynamics and the solutions of the appropriate equations of motion.
Generally: can I trap one particle and "know" it is there? I have not done it, but it is being done billions of times a second at the accelerators. If I went to the trouble to design an experiment that has trapped a single proton in a magnetic configuration, I would know it was there from the radiation it would emmit as it oscillated in the magnetic trap.
Usually though, because of the very small values accompanying the existence of a particle one deals with a flux of them at a time.
Now coherence. Coherence is the term describing the quantum mechanical solution of the equations of more than one particle, and refer to the phase differences between those particles : i.e. coherence means that those phase differences remain constant . Described as quantum mechanical waves, the particles are "in step". If you only have one particle, as in my gedanken experiment above, the quantum mechanical solution is known and phases can only be defined with respect to the field. As long as energy is supplied to my proton this description will hold.
The "know the particle" phrase should become "know the particles" phrase.
Coherence is observed macroscopically:
in laser light
in superconducting magnets, over kilometers of wire length.
All these require zillions of particles and no question should arise if they are there or not. The answer by Peter Morgan addresses the question of stability of such systems.
Now I suspect you are asking the question from statements of coherence and the density matrix formulation. This has to do with the quantum mechanical statistical behaviour
of many particles, so again, your one particle question does not compute.
You should maybe clarify in your head what you really want to learn about coherence. Maybe the density matrix formalism confuses you?