A decision support system for microbiology quality control.

Abstract

Manual review of antibiotic sensitivity testing results is an essential component of a microbiology laboratory's quality control process. Such review is tedious and prone to human error, however. An expert system is described that remembers which susceptibility patterns are considered typical or atypical by expert reviewers, then uses these to prescreen future isolates. It uses a similarity function to allow matching against this library when two patterns are close, but not identical. Use of this system allows more efficient and reliable review of the laboratory's antibiotic sensitivity testing results.