[Development of a clinical decision support system using clinical laboratory test data].

OBJECTIVES

An alert system in a hospital information system (HIS) would be helpful for physicians to prescribe medication appropriately. In this study we developed an alert system that extracts target problems from the raw data and verifies whether contraindicated medication is prescribed.

METHODS

This system scrutinizes data handled in the HIS. It picks up data needed to ascertain problems and data on medication entered. If patient data indicate impaired renal function, or the dosage is over prescribed according to a patient's renal function, the system sends an alert message. The system also determines whether a patient has liver diseases, renal diseases or diabetes mellitus based on clinical laboratory and prescription data, and checks whether a patient is prescribed medication that is contraindicated to his/her problem. When a physician selects patients from an inpatient or an outpatient list, the system presents alerts concerning these patients.

RESULTS

We counted the number of alerts for inpatients. The frequency of alerts for overdose due to patients' renal function and contraindication for patients with liver disease were more frequent than alerts for other reasons. Of patients with renal dysfunction who were prescribed contraindicated medication, 24% had their medication discontinued before the alert system was put into operation. In contrast, that rate significantly increased to 54% after the alert system was initiated.

CONCLUSION

We developed an alert system that extracts target problems and verifies whether inappropriate medication is prescribed. The alerts generated by this system were effectively contributed to discontinuing contraindicated medication.