Hello:
By "artifact of injection", I'm assuming you mean one of two things : the small dip in the baseline that sometimes occurs on injection, or the so-called "solvent peak" that occurs if the sample is not dissolved in mobile phase, but another solvent instead. Both these peaks can't be used for system suitability calculations, since (a) they may not be reproducible and (b) they may be distorted in shape.
What you need is a reproducible clean peak of an unretained solute for validatable calculations.
If validation is an issue, then you would definitely need to use an unretained solute for K' calculation. For a C18 column, uracil is commonly used.
The best thing to do would be to check the test chromatogram and certificate of analysis that should accompany every column on purchase. These documents would tell you all need to know about the column and its validation protocols.
If you don't have these for your column, the original vendor would provide them, if you tell him the column's serial number.
Rgds