One of the reason could be a change in temperature, you need to start all runs at the same temperature. If there is any difference in temperature of the runs then you will observe drift in retention time. So you need to check your temperature program and be sure it is the same you have been using all along. Another possible reason might be the column might be contaminated, in that case you need to solvent rinse the column. At a lab I used to work to we used to bake the column, that is condition it at 250 degrees C. Another reason often overlooked is the septum. If the septum is worn out and leaking then that is the main reason. I suspect that be what you check first, then if septum is fine try solvent rinsing then finally baking. Good luck.