You have a legal obligation to report such suspicions to the board. If you work in a bigger hospital, you can accomplish the same by reporting it to your peer review committee. They are considered local 'agents' for the board, for purposes of well, peer review.
If your boss doesn't take you seriously, that DOES NOT relieve you of your obligation to report. She is merely the FIRST link in a CHAIN OF COMMAND.
Also, you might report it to the ethics committee.
With the advent of pyxis, etc., there is no reason for hospitals not to be on top of this. It's merely a matter of two statistical reviews: 1. How many total narcs a nurse gives in relationship to other nurses on similar units. 2. The percentage of total narcs an individual's group of pts is allowed to have compared to how many are actually being signed out, and a comparison of THAT relationship to other nurses and THEIR group of pts on similar units.
Even the most conscientious pain control nurse should fall within the standard deviations for sign outs.
Any nurse that falls significantly outside the standard deviation (say more then 1.5 standard deviations) should be subject to a drug test EVERYTIME for EVERY shift they fall outside the standard deviation. On the other side, a nurse that signs out signficantly LESS then the standard deviation should be counselled on effective pain control management. With their push for adequate pain control, I'm surprised JCAHO doesn't already mandate this.
That might not solve drug diversion for uses other then personal, but it WOULD significantly reduce the number of 'impaired' nurses due to diversion.
I don't think hospitals do this because 1. they subscribe to the 'any warm body' philosophy, and 2. they don't want to know about such liabilities (See no evil . . .)
But, I think this WILL ultimately become a standard. And, when it does, those hospitals will almost certainly backdate their analysis for the entire employment history of nurses (really, it's just a matter of pushing a few more buttons). You ARE being monitored in this as we speak, even if it's nothing more then a passive collection of records for later analysis at this point.
And let me say this, one of the things I do BEFORE I leave any shift is to check my diebold reports on my pts to see to it that EVERY narc signed out on my pts were either signed out by me, or at my request. If not, that is an incident report. The TWO times I've filed such an incident report: 1 was a diverter that was fired that night (but, he had diverter narcs on 9 pts not his that night), one signed out a xanax on my patient accidentally as his pt was right next to mine on the list AND had the same med ordered, and he had charted he had given it earlier in his shift and there was no record it was signed out on his pt: a simple and correctable accounting error.