Who listens and who doesn't? Factors associated with adherence to antibiotic stewardship intervention in a Singaporean tertiary hospital.

INTRODUCTION: Antibiotic stewardship programmes (ASPs) can improve patient outcomes by prospective audit and feedback with interventions. However, adherence to ASP interventions is not mandatory. Identifying factors associated with improved adherence may help to enhance ASP recommendations and activities. METHODS: A retrospective cohort study was conducted, comprising all ASP interventions performed as part of the prospective audit and feedback strategy in our institution (an acute tertiary-care hospital in Singapore) from January 2016 to July 2018. Adherence to ASP intervention was ascertained based on documented compliance with the recommended interventions within 48h. Factors associated with adherence to ASP interventions, such as patient demographics, clinical condition, type of infection, and characteristics of ASP interventions were identified using the χ2 test for categorical variables. On multivariate analysis, factors independently associated with adherence to ASP intervention were identified using logistic regression. RESULTS: Adherence to ASP intervention was 81.9% (5758/7028). On univariate and multivariate analysis, interventions coupled with direct communication via phone call (adjusted odds ratio [aOR] 1.61, 95% CI 1.23-2.08) were associated with higher odds of adherence, whereas admission to a surgical unit, intervention involving carbapenem use, and recommendation to de-escalate or discontinue antibiotics were associated with lower odds of adherence to ASP interventions. CONCLUSION: Although adherence rates to ASP interventions were relatively high, interventions made to the surgical unit and recommendations related to carbapenem use were not so well received. Interventions communicated verbally via phone call were well received, highlighting the need for a close working relationship between ASP teams and hospital physicians.

The utility of voriconazole therapeutic drug monitoring in a multi-racial cohort in Southeast Asia.

OBJECTIVES: Voriconazole serum concentration, which is affected by several factors, is associated with treatment response and toxicity. There is paucity of data on voriconazole therapeutic drug monitoring (TDM) among Southeast Asians, who exhibit a higher prevalence of CYP2C19-poor metabolisers compared with Caucasians and East Asians. Hence, there are concerns for higher risk of voriconazole accumulation and toxicity. We aim to determine the utility of voriconazole TDM through establishing: (1) proportion of patients achieving therapeutic troughs without dose adjustments; (2) characterisation of patients with sub-therapeutic, therapeutic and supra-therapeutic levels; (3) appropriate dose titrations/dose required for therapeutic troughs; (4) correlation between troughs and adverse events, treatment response/fungal breakthrough. PATIENTS AND METHODS: A single-centre retrospective analysis of data from adults (≥21 years old) with ≥1 voriconazole trough measured at Singapore General Hospital from 2015 to 2017 was performed. RESULTS: Thirty-two patients (45.7%) among 70 patients achieved therapeutic troughs (defined as 2.0-5.5 mg/L) without dose adjustments. Eleven patients (15.7%) experienced hepatotoxicity (troughs 0.5 to >7.5 mg/L). Neurotoxicity occurred in three patients (4.3%) (troughs ≥6.7 mg/L) and all patients had symptom resolution upon dose reduction. Treatment failure of invasive fungal infection appeared less in patients with therapeutic troughs compared with sub-therapeutic troughs (11.4% vs. 14.2%). Two patients experienced treatment failure despite supra-therapeutic voriconazole troughs. CONCLUSIONS: TDM should be implemented due to significant unpredictability in dose exposure. TDM can reduce unnecessary switches to alternatives due to intolerability and rule in the possibility of resistant organisms in the event of treatment failure despite therapeutic troughs, alerting clinicians to switch to alternatives promptly.