Semi-automatic generation of medical tele-expert opinion for primary care physician

A computer-based system has been developed for the generation of medical expert opinions on the insulin-resistance syndrome, based on clinical data obtained from primary care physicians.

An expert opinion for each patient was generated by using a decision tree for entering individual text modules and by adding optional free text. The expert opinions were returned by e-mail, telefax or by ordinary mail.

1389 primary care physician sent anonymous data sets and requested expert opinions for a total of 3768 patients. Through the set up of a rule-based system an automation of the generation of the expert opinions could be achieved and the generation time dropped from initially 40 minutes to less than 5 minutes at the end.

By using predefined text modules and a rule based system, a large number of medical expert opinions can be generated with relatively few additional resources.