We have always had an annual 'diff test' in my lab. At the beginning of the year I select abnormal diffs as they come along, make and stain several slides of the blood or body fluid specimen and then when I have 25 or so slides present them as sequentially numbered test slides. The staff then performs a diff on each one as a test. Grading can be done by comparison to the original patient report with the assumption that it is correct or by having your medical director perform the diff on each one and using their result as the answer key. i had done this for years and CAP was satisfied with this method of competency testing. Graded tests along with documented review of any 'missed' slides were presented to the inspector for competency testing. This year I did something different. Instead I used CAP Kodachrome slides from surveys of previous years and had a multilple choice answer for each slide. My staff seemed to perfer this format as it was easier to take the test and it is much easier to grade. In the future with the anticipation that CAP will be going to a digital photo format for survey microscopic pix I plan to present the test in a PowerPoint format with the case history and possible answers displayed right on the slide.
For a physicians office lab you might simply have the same diffs done by all staff at the different locations and try to achieve reasonably close results as a test, similar to what I do. Of course your medical director can always be used to provide the answer key for the slides and the techs should get similar results within certain percentages. There is your 'split sample' and 'education' all rolled into one.