The provision of guidelines may not be sufficient in ensuring appropriate prescribing and monitoring of vancomycin when barriers relating to the prescribing environment exist. Developing interventions targeted toward these barriers, such as having dedicated phlebotomists for vancomycin blood sampling, fostering better handover processes, and educating staff on poorly understood aspects of guidelines, is likely to improve the uptake of guideline recommendations for vancomycin and other medications requiring therapeutic drug monitoring.

Despite the availability of international consensus guidelines, vancomycin dosing and therapeutic drug monitoring (TDM) remain suboptimal. This study aimed to assess concordance of vancomycin dosing and TDM with institutional guidelines and to identify factors taken into consideration by clinicians when prescribing vancomycin. Methods: A retrospective audit of 163 patients receiving vancomycin therapy (≥48 hours) was undertaken. Data collected included patient characteristics, dosing history and plasma vancomycin and creatinine concentrations. Concordance of dosing and TDM with institutional guidelines was evaluated. Semi-structured interviews, including simulated prescribing scenarios, were undertaken with prescribers (n = 17) and transcripts analysed. Results: Plasma vancomycin concentrations (n = 1043) were collected during 179 courses of therapy. Only 24% of courses commenced with a loading dose with 72% lower than recommended. The initial maintenance dose was concordant in 42% of courses with 34% lower than recommended. Only 14% of TDM samples were trough vancomycin concentrations. Dose was not adjusted for 60% (21/35) of subtherapeutic and 43% (18/42) of supratherapeutic trough vancomycin concentrations, respectively. Interview participants reported that patient characteristics

Background Limited research has focused on understanding if and how evidence of health information technology (HIT) effectiveness drives the selection and implementation of technologies in practice. This study aimed to explore the views of senior hospital staff on the role evidence plays in the selection and implementation of HIT, with a particular focus on clinical decision support (CDS) alerts in electronic medication management systems. Methods A qualitative descriptive design was used. Twenty senior hospital staff from six Australian hospitals in New South Wales and Queensland took part in a semi-structured interview. Interviews were audio-recorded, transcribed, and a general inductive content analysis approach used to identify themes. Results Participants acknowledged the importance of an evidence-base, but reported that selection of CDS alerts, and HIT more broadly, was rarely underpinned by evidence that technologies improve patient care. Instead, investments in technologies were guided by the expectation that benefits will be achieved, bolstered by vendor assurances, and a perception that implementation of HIT is unavoidable. Postponing implementation of a technology until an evidence-base is available was not always feasible. Although some technologies were seen as not requiring an evidence-base, stakeholders viewed evidence as extremely valuable for informing decisions about selection of CDS alerts. Conclusion : Decisions to implement CDS, and technology more broadly, are rarely evidence-based. In the absence of evidence, evaluation or monitoring of technologies post-implementation is critical, particularly to identify new errors or risks associated with HIT implementation and use. Increased transparency from vendors, with technology evaluation outcomes made directly available to healthcare organizations, may result in less reliance on logic, intuition, and vendor assertions, and more evidence-based selection of HIT.

Aims and Method To compare prescribing practice in a community mental health service with evidence-based guidelines and identify factors related to sub-optimal prescribing. All current patients (n=640) were assessed regarding six key aspects of prescribing (polypharmacy, high-dose treatment, use of thioridazine/maintenance benzodiazepine/maintenance hypnotic or routine anticholinergic treatment). The relationship of quality of prescribing practice to demographic, illness and service variables was examined by regression analysis. Results Five-hundred and five (79%) patients were receiving psychotropic medication. Of these, 232 (46%) had evidence of sub-optimal prescribing practice. Mean prescribing practice quality score was 0.75 ± 0.99. Maintenance benzodiazepine/ hypnotic (31%) and anticholinergic (30%) use were particularly common. Prescribing practice quality score was higher in those receiving depot antipsychotic treatment (P < 0.01) and in older patients (P < 0.01). Scores were significantly lower in patients whose principal medical contacts were with a consultant rather than a junior doctor (P < 0.001). Clinical Implications Prescribing practices in real-world settings frequently deviate from evidence-based guidelines. The quality of prescribing is related to patient, illness and service variables. In particular, greater contact with consultant staff is linked to better practices. Patients receiving depot antipsychotics are especially liable to less judicious prescribing practice.