Computer aided prescribing

Editor—Ferner's article is a rational appraisal of computer aided prescribing.¹ Such prescribing has great potential but if the software is designed without knowledge of how doctors prescribe it won't be used effectively. The old “garbage in, garbage out” rule still applies.

I receive computer generated letters from pharmacy benefit managers or consultant companies almost daily, telling me what I am doing wrong on the basis of their incorrect data. Last month I was scolded for not giving one of my diabetic patients ramipril (she has been taking it for three years).

I spent my residency years using an electronic record system that flagged almost everything conceivable. For example, prescribing any drug to a patient who had ever been prescribed a wheelchair or a cane produced a warning of a potential unknown interaction between “wheelchair” and “penicillin.” Perhaps each user could be allowed to set a certain level for flagging of potential problems, producing “just the right” number of warnings for an individual prescriber. Last month I was scolded by another company's computer for having placed a patient who is taking glyburide on a short tapering course of prednisone—did I know it could affect glycaemic control?

With electronic friends like these, who needs enemies?