Pattern Detection in Clinical Laboratory Data for Patient Identification and Quality Assurance

BACKGROUND Therapeutic drug monitoring (TDM) is a useful tool for preventing Adverse Drug Events related to patient variability or inappropriate dosing. Typical physicians find proper interpretation of monitoring data difficult, however.' Patient follow-up by TDM experts is an effective method of improving TDM utilization, but the cost of having such a service track all patient visits across an entire medical center is prohibitive. Therefore, we have developed a rule-based expert system called LabScanner that analyzes clinical laboratory drug concentration data from a laboratory information system (LIS) to identify a population of patients enriched in those who could benefit from expert follow-up.