data mining 2

I currently use a health records program that has several input screens to collect various data on a patient, on every encounter. Some of the examination data can be selected and reapplied on the next examination, if it has not changed. But...much of the data cannot. Often, a patient is returning for the same complaint with the same history, and examination, and I must re-input all of the data. After each encounter, I preserve the record as a PDF. This contains the data that I input under specific headings. I would like to be able to collect the data from the PDF and re-insert it into the next encounter, but on a field by field basis. This would require a program that acquires the data from the PDF and then goes into the program, screen by screen, inserting the appropriate input into the appropriate filed and advancing to the next screen until complete. I attached a demo. There are about 10 additional fields that I did not illustrate for the sake of time and brevity