I have a method for the analysis of impurities related to 1 of 3 active ingredients in a drug product. One of the other active ingredients adheres to the column. Unless I clean the column on 100 % Methanol after any sample injections for 1-2 hours, I will get carryover. I have been able to demonstrate that the carryover is not due to the LC. On occaison we are seeing a significant impurity peak which we have determined is a degradant of the active ingredient that is adhering to the column.
I am using a 5-µm 4.6x150-mm Zorbax Eclipse XDB-C18. The mobile phase is methanol: pH 3.0 25mM sodium phosphate, monobasic,(20:80). Would a guard column help this situation?

A guard column would only help temporarily. You need to do some sample cleanup, probably by SPE.

Think about why this active ingredient sticks! Most likely, it is more hydrophobic, and a simple wash with a higher % organic will solve your problem. I personally would do it after every single run, but you may get away with doing it less frequently.
I can not see why you would need to wash the column for 1-2 hours in methanol. It may be that the problem compound is retained for other reasons than hydrophobic interaction, and washing with 50/50 methanol-buffer would do a better jop than washing with methanol. If indeed the interaction is hydrophobic, you could wash the column with acetonitrile or THF after each run.