Does the availability of therapeutic drug monitoring, computerised dose recommendation and prescribing decision support services promote compliance with national gentamicin prescribing guidelines?

作者：Diasinos N; Baysari M^*; Kumar S; Day R O

来源：Internal Medicine Journal, 2015, 45(1): 55-62.

DOI：10.1111/imj.12627

• gentamicin
• therapeutic drug monitoring
• computerised dose prediction
• decision support
• interviews

BackgroundGentamicin is an aminoglycoside antibiotic that is highly effective in treating Gram-negative infections, but inappropriate use leads to toxicity. In 2010, the Australian Therapeutic Guidelines (Antibiotic) were revised to recommend the use of computerised methods to individualise dosing of gentamicin and optimise therapy, rather than traditional nomogram approaches.
AimTo determine whether gentamicin prescribing was compliant with the Australian Therapeutic Guidelines, version 14 (2010) in a setting where computerised dose recommendation resources and computerised decision support were available, and to determine why the resources were effective or ineffective in achieving compliance to guidelines.
MethodsDuring phase 1, a retrospective audit of gentamicin prescribing from 1 January 2012 to 31 December 2012 (n=826) at a 320-bed teaching hospital in Sydney was undertaken. In phase 2, 12 doctors from specialties with high-volume prescribing of gentamicin were interviewed.
ResultsIntravenous gentamicin was used in 545 cases, 81% of which were for short-term therapy (48h). Doctors feared inducing toxicity in patients, but limited the dose rather than altering the dosing interval according to renal function. Of the continued' dosing cases, 55% went unmonitored and the computerised dose recommendation service was rarely used. Doctors were unaware of its availability despite electronic alerts accompanying prescriptions of gentamicin.
ConclusionsIn comparison with the national guidelines, there was significant under-dosing and monitoring practices were haphazard. Computerised electronic alerts were ineffective in informing users. To improve prescribing practices, we recommend exploring alternative computerised decision support approaches (e.g. pre-written orders) and more pervasive and persuasive implementation strategies.

• 出版日期2015-1