COMPARATIVE-STUDY ON THE EFFECT OF THE AP PLICATION OF AN EXPERT-SYSTEM ON DRUG PRESCRIPTION HABITS

Citation
C. Codina et al., COMPARATIVE-STUDY ON THE EFFECT OF THE AP PLICATION OF AN EXPERT-SYSTEM ON DRUG PRESCRIPTION HABITS, Medicina Clinica, 109(14), 1997, pp. 538-541
Citations number
8
Categorie Soggetti
Medicine, General & Internal
Journal title
ISSN journal
00257753
Volume
109
Issue
14
Year of publication
1997
Pages
538 - 541
Database
ISI
SICI code
0025-7753(1997)109:14<538:COTEOT>2.0.ZU;2-O
Abstract
wBACKGROUND: The aim of this study was to evaluate the impact of a exp ert drug system implementation on the drug prescription habits and on drug cost in an university hospital. MATERIAL AND METHODS: During a 3 months period, the drug prescriptions to patients admitted to the Inte rnal medicine, lung, gastroenterology and hepatology units, have been evaluated through the expert drug system (Medisource (R)). This expert drug system functions in according to patients characteristics such a s age, weight, height, sex, renal function and liver function. It reco mmends the correct dose, detects interactions and adverse effects and makes suggestions in pregnancy and lactation. It also offers alternati ve drugs with their cost. During the study period physicians were unaw are of the investigation being performed. RESULTS: 836 patients (63.9 +/- 16.5 years) with an average hospital stay length of 11.6 +/- 6.7 d ays were studied. The most common diagnoses were: lung obstructive chr onic disease, cirrhosis, gastrointestinal hemorrhage and cancer. The t otal amount of drug prescribed was 6,308. The expert system detected 4 58 overdosages and 33 underdosages, mainly in antibiotics and antiulce r drugs, and 1,722 interactions. The drug costs reduction that could b e obtained following the expert system recommendations was 4.5% in ant ibiotic drugs and 23% in antiulcer drugs. CONCLUSIONS: The frequency o f drug overdosage and underdosage in patients admitted in an universit y hospital is relatively high. The expert systems available for drug d ecisions could solve this problem.