|
KMID : 0869620070240010035
|
|
Journal of Korean Society of Hospital Pharmacists
2007 Volume.24 No. 1 p.35 ~ p.53
|
|
|
Implementation and Evaluation of the Computerized Physician Order Entry Systems using Knowledge base
|
|
Kim Ka-Eun
Lee Jae-Hyun
Chung Seon-Young
In Yong-Won
Sohn Kie-Ho
|
|
|
Abstract
|
|
|
|
Medication errors and preventable adverse drug events are common, and nearly half of medication errors occur during medication ordering. Within the Computerized physician order entry (CPOE) systems, clinical decision support systems (CDSS) such as automated order checks and more commonly known as drug safety alerts, may offer possible benefits from patients care. Samsung Medical Center (SMC) developed the CPOE systems which is linked with the CDSS to prevent potential medication errors (Duplicate drug class, Drug-drug interactions) based on drug-code coincidence. However, our systems were newly developed and operated to improve integrated management with aim in April, 2006. Thereafter, inappropriate prescriptions were detected as individual drug-drug or drug-class. After new system applying, the inappropriate prescriptions which are detected by alerting systems for 1 month (August, 2006) were 5,043 cases. The detected duplicate prescriptions, 4,649 (211.3/day) cases, which was 6.7% of total ambulatory prescriptions. This showed increase of detection rate in 39.9% compared with last year (August, 2005). In addition, the change of the detection criteria caused additional increase of detection rate in 43.1%. However, the correction rate of duplication errors was only 11.3% (524 cases). The duplicate drug classes ranked in the order of Antiplatelet agents (501 cases, 10.8%), NSAIDs (291 cases, 6.3%), Calcium Channel Blockers (175 cases, 3.8%), and Anticonvulsants (152 cases, 3.3%). Drug safety alerts were overridden by 15.7% of 394 (17.9/day) cases which was detected as drug interaction. Among the combinations of drug-drug interactions, the combination of ketorolac tromethamine and aceclofenac was observed most frequently (97 cases, 24.6%). Although adding alerts into order entry systems seem compelling, our research suggests that more warnings are ignored or overridden rather than followed. The causes of high override rates are not entirely clear because we did not examine the causes of override in SMC. However, drug alerts can slow the ordering process and may lack relevance for a given clinical situation. As well, physicians may ignore or undervalue alerts as the results of "alert fatigue" generated by falsepositive alerts. This study was designed to introduce and evaluate medication safety alerts as CDSS which is linked with the CPOE systems. This presented the possibility that it will be able to apply the knowledge base which is constructed a drug information provision to the integrated CPOE systems. Our finding suggests that other types of safety alerts, such as those related to prescribing in the pregnancy, drug-disease interactions, drug-allergy interactions or drug-laboratories, are also valuable.
|
|
|
KEYWORD
|
|
|
Computerized physician order entry systems, Clinical decision support systems, Drug alerts, Drug interaction, Duplicate drug class
|
|
|
FullTexts / Linksout information
|
|
|
|
|
|
Listed journal information
|
|
|
|