Errors associated with applying decision support by suggesting default doses for aminoglycosides

Abstract

Background: Medication errors, and the resultant adverse drug events (ADEs), are one of the main preventable causes of morbidity and mortality. Computerised physician order entry (CPOE) is reported to reduce the frequency of these errors. However, CPOE systems themselves may be associated with errors. The aim of this study was to investigate the effects of a CPOE system that displays an initial default dose for gentamycin and tobramycin administration on the frequency of medication errors and potential ADEs in patients with renal insufficiency. Methods: Gentamycin and tobramycin prescriptions from the CPOE records of a Dutch tertiary adult intensive care unit were retrospectively compared with doses recommended by a locally developed guideline. The default dose for gentamycin and tobramycin in the CPOE system is 240 mg/day. A dose prescribing error was defined as an administered dose that exceeded the recommended dose by >10%. Results: Three hundred and ninty two prescriptions, relating to 253 patients (of whom 184 had renal insufficiency), were analysed. There was a high frequency (58%, 227 of 392) of prescriptions that used the CPOE system's default dose of 240 mg/day. The dose was wrong in 73% (165) of these orders. Default orders for patients with renal insufficiency amounted to 52% (132 of 259). A total of 86% (113 of 132) of these resulted in potential ADEs compared with 53% (66 of 124) for the rest of orders (p < 0.0001). Discussion: A markedly high frequency of prescriptions followed the default dose value and, in patients with renal insufficiency, there was a high frequency of doses exceeding the guideline recommendation (+10%), amounting to potential ADEs. Conclusion: Initial CPOE dose values for prescribing gentamycin and tobramycin, which are based on a fixed default value, form a source of potential ADEs for patients with renal insufficiency. © 2006 Adis Data Information BV. All rights reserved.