Obviously, it's hard for many to believe oil filters filter better as they are used. Go to bobistheoilguy and spend some time reviewing actual UOA's (used oil analysis) and pay attention to the Insoluables values. They do NOT go up on the second nor third OCI if a filter is left in for more than one OCI.
Be sure to also read the theory of filtration information. ALL Filters (that aren't disentegrating) filter better and better as they become loaded.
Think about this: when a fluid is run through a filter, more fluid flows through the larger pathways than the smaller pathways. Big particles roll off the areas of small pathways until they become lodged in one of the more open areas where the flow is pushing them to go. It's a bit like a log being shoved down a stream...it will keep rolling around until it gets hung up somewhere...at which point it begins to catch other trash coming down the stream. At the same time, water is forced around the log-jam to find other ways downstream. The more big areas that are plugged by big particles, the more the flow is forced to go to other, smaller pathway areas. That's better filtration.
Unless the particles are broken up into smaller particles which can then pass through the pathway that initially caught them, these large particles plugging the larger holes are why filters become more efficient as they are used.
It's one of the arguements the stainless steel screen filter manufacturers use to say they are "better". Metal Screen filters have more uniform opening sizes than filter or mesh. However, mesh's irregularly shaped openings may snag more particles, especially if the particle is not a sphere.
I won't go on about it...if interested, the information is out there. If not interested, change the filter every time...although science can prove using a new filter every time isn't as effective as leaving them in longer. Assuming, of course, an appropriate oil and OCI is being followed for the engine system.
PS: For those of you that study UOA's, you'll periodically notice comments related to increased rate of wear metals produced per mile, commonly seen after an oil change. Many are now questioning if this increased rate of wear is due to lowered filter efficiency provided by the virgin filter, or, if it is due to anti-wear additives in the oil that needs some heat-cycling to become maximally effective, or due to increased detergent action of new oil, or a combination of these factors. Tribology is interesting stuff for those that are truly curious about lubrication science, and aren't relying upon Arnold Palmer (or some mechanic, the vast majority that are not be informed about modern tribological science) sitting on a tractor talking about how well Pennzoil has worked on the farm for so long.