Population pharmacokinetic studies have suggested that high polymyxin B (PMB) doses (≥30,000 IU/kg/day) can improve bacterial kill in carbapenem-resistant Gram-negative bacteria (CR-GNB). We aim to describe the efficacy and nephrotoxicity of patients with CR-GNB infections prescribed high-dose PMB. A single-centre cohort study was conducted from 2013 to 2016 on septic patients with CR-GNB infection and prescribed high-dose PMB (~30,000 IU/kg/day) for ≥72 h. Study outcomes included 30-day mortality and acute kidney injury (AKI) development. Factors associated with AKI were identified using multivariable regression. Forty-three patients with 58 CR-GNB received high-dose PMB; 57/58 (98.3%) CR-GNB were susceptible to PMB. The median daily dose and duration of high-dose PMB were 32,051 IU/kg/day (IQR, 29,340–34,884 IU/kg/day) and 14 days (IQR, 7–28 days), respectively. Thirty-day mortality was observed in 7 (16.3%) patients. AKI was observed in 25 (58.1%) patients with a median onset of 8 days (IQR, 6–13 days). Higher daily PMB dose (aOR,1.01; 95% CI, 1.00–1.02) and higher number of concurrent nephrotoxins (aOR, 2.14; 95% CI; 1.03–4.45) were independently associated with AKI. We observed that a sizable proportion developed AKI in CR-GNB patients described high-dose PMB; hence, the potential benefits must be weighed against increased AKI risk. Concurrent nephrotoxins should be avoided to reduce nephrotoxicity.

Background: Ertapenem (preferred choice for ESBL-producing organisms) use exhibited an increasing trend from 2006 to 2008. As extensive use of ertapenem might induce the mutation of resistant bacteria strains to ertapenem, we aimed to assess the appropriateness and impact of ertapenem-use, on ESBL production, the trends of gram-negative bacterial resistance and on the utilization of other antibiotics in our institution. Methods: Inpatients who received a dose of ertapenem during 1 January 2006 to 31 December 2008, were reviewed. Pertinent patient clinical data was extracted from the pharmacy databases and assessed for appropriateness based on dose and indication. Relevant data from Network for Antimicrobial Resistance Surveillance (Singapore) (NARSS) was extracted, to cross-correlate with ertapenem via time series to assess its impact on hospital epidemiology, trends of gram-negative resistance and consumption of other antibiotics from 2006 to mid-2010. Results: 906 cases were reviewed. Ertapenem therapy was appropriate in 72.4% (93.7% success rate). CNS adverse events were noted in 3.2%. Readmission rate (30-day) due to re-infection (same pathogen) was 5.5%. Fifty cases had cultures growing Pseudomonas aeruginosa within 30 days of ertapenem initiation, with 25 cases growing carbapenem-resistant Pseudomonas aeruginosa. Ertapenem use increased from 0.45 DDD/100 patient days in 2006 to 1.2 DDD/100 patient days in mid-2010. Overall, the increasing trend of ertapenem consumption correlated with 1) increasing incidence-densities of ciprofloxacin-resistant/cephalosporin-resistant E. coli at zero time lag; 2) increasing incidence-densities of ertapenem-resistant Escherichia. coli and Klebsiella spp. at zero time lag; 3) increasing incidence-density of carbapenem-resistant Pseudomonas aeruginosa, at zero time lag. Increasing ertapenem consumption was significantly correlated with decreasing consumption of cefepime (R² = 0.37344) 3 months later. It was significantly correlated with a decrease in imipenem consumption (R² = 0.31081), with no time lag but was correlated with subsequent increasing consumption of meropenem (R² = 0.4092) 6 months later. Conclusion: Ertapenem use was appropriate. Increasing Ertapenem consumption did not result in a decreasing trend of ESBL producing enterobacteriaceae and could result in the selection for multi-drug resistant bacteria.

The use of newer azoles as prophylaxis in hematological patients undergoing stem cell transplantation or immunosuppressive chemotherapy has been shown to decrease the risk of developing invasive fungal disease (IFD). However, the cost-effectiveness of such a strategy is dependent on the local epidemiology of IFD. We conducted an audit of hematological patients with IFD in our institution in order to derive the prevalence and types of IFD that occur locally.

Colonization of patients occurs before development into invasive candidiasis. There is a need to determine the incidences of Candida colonization and infection in SICU patients, and evaluate the usefulness of beta-D-glucan (BDG) assay in diagnosing invasive candidiasis when patients are colonized. Clinical data and fungal surveillance cultures in 28 patients were recorded from November 2010, and January to February 2011. Susceptibilities of Candida isolates to fluconazole, voriconazole, amphotericin B, micafungin, caspofungin and anidulafungin were tested via Etest. The utilities of BDG, Candida score and colonization index for candidiasis diagnosis were compared via ROC. 30 BDG assays were performed in 28 patients. Four assay cases had concurrent colonization and infection; 23 had concurrent colonization and no infection; three had no concurrent colonization and infection. Of 136 surveillance swabs, 52 (38.24 %) were positive for Candida spp, with C. albicans being the commonest. Azole resistance was detected in C. albicans (7 %). C. glabrata and C. tropicalis were, respectively, 100 and 7 % SDD to fluconazole. All 3 tests showed high sensitivity of 75–100 % but poor specificity ranging 15.38–38.46 %. BDG performed the best (AUC of 0.89). Despite that positive BDG is common in surgical patients with Candida spp colonization, BDG performed the best when compared to CI and CS.

Traditional in vitro time-kill studies (TKSs) require viable plating, which is tedious and time-consuming. We used ATP bioluminescence, with the removal of extracellular ATP (EC-ATP), as a surrogate for viable plating in TKSs against carbapenem-resistant Gram-negative bacteria (CR-GNB). Twenty-four-hour TKSs were conducted using eight clinical CR-GNB (two Escherichia coli, two Klebsiella spp., two Acinetobacter baumannii, two Pseudomonas aeruginosa) with multiple single and two-antibiotic combinations. ATP bioluminescence and viable counts were determined at each timepoint (0, 2, 4, 8, 24 h), with and without apyrase treatment. Correlation between ATP bioluminescence and viable counts was determined for apyrase-treated and non-apyrase-treated samples. Receiver operator characteristic curves were plotted to determine the optimal luminescence threshold to discriminate between inhibitory/non-inhibitory and bactericidal/non-bactericidal combinations, compared to viable counts. After treatment of bacteria with 2 U/mL apyrase for 15 min at 37 °C, correlation to viable counts was significantly higher compared to untreated samples (p < 0.01). Predictive accuracies of ATP bioluminescence were also significantly higher for apyrase-treated samples in distinguishing inhibitory (p < 0.01) and bactericidal (p = 0.03) combinations against CR-GNB compared to untreated samples, when all species were collectively analyzed. We found that ATP bioluminescence can potentially replace viable plating in TKS. Our assay also has applications in in vitro and in vivo infection models.