*sighs*
Beyond the obvious you-don't-need-to-worry-because-the-pagefile-usage-is-a-drop-in-the-proverbial-bucket, page file usage is hard to measure because not everything that gets paged out is paged out to the pagefile.
Say you run a program. It has a 100MB EXE file. When that program loads, that EXE file is loaded and mapped into memory. If a region of that EXE file is seldom used and the system decides to page it out, the obvious thing to do would be to simply evict it from physical memory because that data is already on the disk. It still counts as committed memory belonging to that process because any attempt to access that mapped region will page it back in from the disk.
There are other complications, too, like how Windows has memory de-duplication where two identical pages will initially share the same physical memory until one of them is written to and changed.
This, I suspect, is why such numbers are almost always bogus because most programs just naively measure the amount committed and the amount in in RAM and assume the difference is in the pagefile.
I really shouldn't be encouraging or enabling berryracer's usual obsessiveness, but you could try AIDA64--in my experience, their pagefile figures tend to be right more often than not.