The first key element to a good UOA program is how you draw your sample. The analysis is only as good as the sample and a externally contaminated sample will show erroneous results. The best way is to either use a oil sampling pump or a oil sampling port mounted in the system being tested. If you use a sampling pump make sure you use a new piece of nylon tubing each time. You can draw a sample while you are draining but you stand the chance of getting contaminates in the oil.
Next you need to set up a sampling schedule. If you don't plan on extending your drain intervals this is rather simple, draw the sample at the time of service. If you plan on extending drain intervals you need to do a little more planning. Draw your first sample at what would be your normally scheduled service (I'll get to filter's later). Now let's say your are sampling a OTR truck engine oil who's regular service intervals are 10,000 miles. Take your second analysis at 15,000 miles, then 20,000 miles adding half of the regular service interval to each additional analysis. At some point no matter how high of a quality lubricant you are using it is going to need changing. Fuel dilution, soot loading, moisture, contamination, additive package depletion, etc... At some point the analysis lab is going to flag one of the samples marginal in some category, probably contamination or additive depletion. When you get this result change the oil. Let's say you got all the way out to 80,000 miles. At this point the oil has reached it's usable life but we want a safety margin. Now that you have your initial analysis cycle done you know at about 80,000 miles the oil will be at the limit. On the second oil change take your first sample at around 20,000 miles, next one at 40,000 and then at 60,000 miles draw a sample and change the oil. This gives you a safety margin so you are not always running the oil to it's limit. You now have your extended drain interval schedule.
As time goes on you will get a base line of what the normal wear metal readings are. Let's say that the lab would normally flag a analysis at 200 ppm (parts per million) but your equipment usually runs at about 20 to 30 ppm per 20,000 mile sampling interval. If you were to get a analysis back that showed 180 ppm in this metal the lab wouldn't flag it because it has not reached it's acceptable limit yet but you know it would regularly be at only 20 or 40 ppm. Draw your next sample early to monitor and see if it is increasing at a increased rate. This is a indication of something abnormal going on and it's time for some diagnosis to track down the culprit. It will stick out like a sore thumb on the analysis results in the graph section as there will be a spike in the what would normally be a gradual sloped line. This is known as trend analysis since a piece of equipment has a "trend" of how quickly it wears out the lubricant and contaminates it. If your analysis is provided by your lubricants supplier they will probably also get a copy of all your results. In my case all of the customers I service, I get a copy of their results. If I notice something that seems out of place on the analysis I contact the customer and inform them. Not all lube reps do this and I'm probably a rare bird but I think that my customers deserve the extra attention. I work with the customer to try to track down the problem weather it's external contamination, a mechanical problem or just particle scarring.
Filter maintenance is probably the most difficult to establish a change schedule for since there is a wide degree of filter quality. If in doubt just change the filter at the regular service interval. Check to see if your filter supplier offers a extended drain interval filter, many do. Extended drain filters usually have a lot more filter media in them and of much higher quality. Many have built in bypass filtration units. Some even offer extended drain filters that contain a additive booster's. I'm not a big fan of the additive booster filters since every oil has a different additive package and different levels. The filter has no way of knowing how much of what is needed.
If in doubt about anything on a analysis or how to implement one contact the analysis lab or your lube rep if you have one. It's better to play it safe than sorry.