Followup: I tried all kinds of combinations in case there was a data error, on the theory that if I omitted data that contained invalid values. Changing to different CQM Alternates did not help it compile. However, adding children immunizations before the CQM Alternates suddenly made it ok, and all numbers were counted, no matter what date range. So now I have the report, just one more set of data (children immunizations) than I need. I don't understand how adding a data set makes it ok, if there was an invalid entry in the data set.
