This presentation is part of: Oral Concurrent Session A - Clinical Strategies and Guidelines
MEDICATION ERRORS IN THE INTENSIVE CARE UNIT: COMPARISON BETWEEN COMPUTERIZED VERSUS PAPER-BASED PHYSICIAN ORDER ENTRY
Kirsten Colpaert, MD1, Barbara CLaus, PharmD2, Annemie Somers, PharmD2, Hugo Robays, PharmD, PhD2, Koen Vandewoude, MD1, Sandra Oeyen, MD1, and Johan Decruyenaere, MD, PhD1. (1) University Hospital Ghent, Intensive Care Dpt, Ghent, Belgium, (2) Ghent University Hospital , Belgium, Pharmacy Dpt, Ghent, Belgium
Purpose: Medication errors (ME) in the Intensive Care Unit (ICU) are frequent and lead to attributable patient morbidity and mortality, increased length of ICU stay and substantial extra costs. We investigated if the introduction of a computerized ICU system (Deio Clinisoft, General Electric) reduces the incidence and severity of ME. Methods: A one-month (March 2004) prospective trial was conducted comparing ME in a computerized unit (CU) versus a paper-based unit (PBU) in a university hospital. Every medication entry was registered and evaluated for ME, Medication Prescription Errors (MPE) and Rule Violations (RV) by an experienced and unit-independent clinical pharmacist. An independent panel evaluated the severity class of every ME according the NCC MERP guidelines. Results: 160 patient-days resulting in 2662 medication prescriptions were evaluated. There were no differences in the CU-patients compared with the PBU-patients regarding age, admission reason, severity of illness (APACHE, SAPS, SOFA score), renal failure and number of administered medications. The incidence of ME, MPE, RV was significantly lower in the CU compared with the PBU (5,1% vs. 24,7%, 0,9% vs. 15,0% and 0,17% vs. 7,90%, respectively; all P<0.001). The most significant reduction were seen in dosing errors (1,6% vs. 7,4%), dose errors in renal failure (0,4% vs. 4,5%) and wrong name errors (0,42% vs. 8,87%)(all P<0.001). The most frequent drug classes involved were cardiovascular medication and antibiotics in both groups, but more errors in sedative, analgesic and gastro-intestinal medication prescription were seen in the PBU. The severity of ME was mainly class A (36% vs. 52%) and C (54% vs. 44%) according to the NCC MERP classification, but in the PBU more class D errors occurred (1,4% vs. 2,5%, P=0.025). No fatal errors were seen. Allergy status was recorded in 69% of CU patients vs. only 2% in PBU. Patient weight was recorded in 60% of CU patients vs. 18% of PBU patients. Conclusions: The ICU computerization, including the medication order entry, resulted in a significant decrease in the occurrence and severity of medication errors in the ICU. The main reasons are the more adequate recording of allergy status and patient weight and the drug dose recommendation by the ICU computerized system.
