I think that once the original findings have been evaluated / confirmed and no cause can be found (against a backdrop of a history of great stability) I would get approval from QA/RA to add an interval at first opportunity to see if there is a trend beginning or something else caused the prior aberrant results.