Most ppl who use oil analysis probably do so on a consistent basis. Pointless, random sampling won't really let you spot trends...

The lab (Blackstone) has a database (rather large!) of all the samples it has analyzed, and the AVERAGE data for YOUR type of engine is included with, and compared against your current sample data. Thus, you DO have a "baseline" to compare to. So, when you see "xxx" ppm of copper in your sample, and it is much higher than the average, you can be somewhat sure that there is some abnormal bearing wear occuring. (aren't most engine bearings basically the same composition? copper, lead, tin, etc ?)

When you submit your sample, you are instructed to indicate the "service miles" of the sample. So, 1000 vs 4000 miles doesn't really matter. It's all factored into the results.

Bottom line is that a random (say yearly) sample won't be of as much value as one every oil change, but can give you data that is relevant to the general condition of your engine. I think you would have to be pretty "retentive" to want to sample at every change.

I'm getting ready to send a sample in soon, the first oil change after changing out my head. I'm mainly interested in anti-freeze intrusion, and they can detect very small amounts. As you probably know, anti-freeze will trash your bearings in a hurry!