I assume you are referring to simple testing during initial commissioning, and then later on during PM etc.
But let me guess ... the medics are complaining that they are "inaccurate", and you need a quick way to prove otherwise. Hint:- nine times out of ten, the problem will be "operator error".
With items like these I believe it is best to check them out in batches (that is, a good number of each type in one session). That way you can look for (and confirm) consistency across the batch. These thermometers must have some sort of "self-test" or "Cal" - so that would be the first thing to check. Any that fail, or come outside the "expected norm", should be set aside.
There's no way that you're going to be able to test them for absolute accuracy - unless you have a "black body" (see Annex A1 at the link below). But, following the "test by comparison" approach I mention, you may be able to find an ad-hoc
heat source somewhere; a handy brick wall, for example. I was also wondering about the side of a water bath ... you may have to experiment a bit with all this!
It's probably not a good idea to try and test them individually, either - such as when coming across one out on the wards, for example. As I say, you shall need a "good one" (at least) to compare it with.
If you have a lot of these, it may be worth keeping a "known good one" of each type back (hidden) in your workshop; and then bring (or have sent) any reported faulty ones back there for checking (comparing) under some semblance of "controlled conditions".
Meanwhile, what are you planning to do with those that fail? Apart from battery condition, battery connections, sticking buttons, cracked cases and so forth - what else can you do with them (apart from binning them, that is)?
Note that the link you have given states:- "Accuracy level: According to ASTM E 1965-98:2009". An earlier version of E 1965 is available (.pdf) here
. There is a lot of technical background there; see also Section 6 : Test Methods. There is a newer version of E 1965 available (2009) but you would probably need to pay for that.
For reference, here is another data sheet