Safety and clinical outcomes of outpatient parenteral antibiotic therapy for infective endocarditis in Christchurch, New Zealand: A retrospective cohort study.

OBJECTIVES: We examined the safety and clinical outcomes of outpatient parenteral antibiotic therapy (OPAT) for patients with infective endocarditis (IE) in Christchurch, New Zealand. METHODS: Demographic and clinical data were collected from all adult patients treated for IE over 5 years. Outcomes were stratified by receipt of at least partial OPAT vs entirely hospital-based parenteral therapy. RESULTS: There were 172 episodes of IE between 2014 and 2018. OPAT was administered in 115 cases (67%) for a median of 27 days after a median of 12 days of inpatient treatment. In the OPAT cohort, viridans group streptococci were the commonest causative pathogens (35%) followed by Staphylococcus aureus (25%) and Enterococcus faecalis (11%). There were six (5%) antibiotic-related adverse events and 26 (23%) readmissions in the OPAT treatment group. Mortality in OPAT patients was 6% (7/115) at 6 months and 10% (11/114) at 1 year and for patients receiving wholly inpatient parenteral therapy was 56% (31/56) and 58% (33/56), respectively. Three patients (3%) in the OPAT group had a relapse of IE during the 1-year follow-up period. CONCLUSION: OPAT can be used safely in patients with IE, even in selected cases with complicated or difficult-to-treat infections.

Frequency of pharmacological target attainment with flucloxacillin and cefazolin in invasive methicillin-susceptible Staphylococcus aureus infection: a prospective cohort study in hospitalized patients.

BACKGROUND: The proportion of patients with invasive methicillin-susceptible Staphylococcus aureus (MSSA) infection who achieve target concentrations of flucloxacillin or cefazolin with standard dosing regimens is uncertain. This study measured drug concentrations in a prospective cohort of patients with invasive S. aureus infections to determine the frequency of target concentration attainment, and risk factors for failure to achieve target concentrations. PATIENTS AND METHODS: Unbound flucloxacillin and cefazolin plasma concentrations were measured at the midpoint between intravenous doses. Adequate and optimal targets were defined as an unbound plasma concentration of ≥1 and ≥2 times the minimum inhibitory concentration (MIC) (flucloxacillin 0.5 mg/L, cefazolin 2 mg/L), respectively (50%fT≥1MIC, 50%fT≥2MIC). RESULTS: There were 50 patients in each of the flucloxacillin and cefazolin groups. Eighty-five (85%) patients met the target of 50%fT≥2MIC and 95 (95%) patients met the target of 50%fT≥1MIC. The median unbound flucloxacillin concentration was 2.6 mg/L [interquartile range (IQR) 1.0-8.1]. The median unbound cefazolin concentration was 15.4 mg/L (IQR 8.8-28.2). A higher proportion of patients in the flucloxacillin group failed to achieve the optimal target compared with the cefazolin group [13 (26%) vs 2 (4%); P=0.002]. Younger age and higher creatinine clearance were associated with lower plasma concentrations. CONCLUSIONS: Standard dosing of flucloxacillin and cefazolin in the treatment of invasive MSSA infections may not achieve target plasma concentrations for a subgroup of patients. Measuring drug concentrations identifies this subgroup and facilitates dose individualization.

Population pharmacokinetics of free flucloxacillin in patients treated with oral flucloxacillin plus probenecid.

Oral flucloxacillin may be coadministered with probenecid to reduce flucloxacillin clearance and increase attainment of pharmacokinetic-pharmacodynamic (PK/PD) targets. The aims of this study were to develop a population PK model of free flucloxacillin when administered orally with probenecid, and to identify optimal dosing regimens for this combination. METHODS: We performed a prospective observational study of adults (45 participants) treated with oral flucloxacillin 1000 mg and probenecid 500 mg 8-hourly for proven or probable staphylococcal infections. Steady-state mid-dose-interval flucloxacillin measurements (45 concentrations) were combined with existing data from a crossover study of healthy participants receiving flucloxacillin with and without probenecid (11 participants, 363 concentrations). We developed a population pharmacokinetic model of free flucloxacillin concentrations within Monolix, and used Monte Carlo simulation to explore optimal dosing regimens to attain PK/PD targets proposed in the literature (free drug time above minimum inhibitory concentration). RESULTS: Flucloxacillin disposition was best described by a 1-compartment model with a lag time and first-order absorption. Free flucloxacillin clearance depended on probenecid, allometrically-scaled fat free mass (FFM) and estimated glomerular filtration rate (eGFR). Predicted PK/PD target attainment was suboptimal with standard dosing regimens with flucloxacillin alone, but substantially improved in the presence of probenecid. CONCLUSION: The simulation results reported can be used to identify dose regimens that optimise flucloxacillin exposure according to eGFR and FFM. Patients with higher FFM and eGFR may require the addition of probenecid and 6-hourly dosing to achieve PK/PD targets. The regimen was well-tolerated, suggesting a potential for further evaluation in controlled clinical trials to establish efficacy.

Determinants of vancomycin nephrotoxicity when administered to outpatients as a continuous 24-hour infusion.

Vancomycin continuous infusion (VCI) is used to treat serious Gram-positive infections in outpatients. This study was conducted to retrospectively investigate the rate of nephrotoxicity and associated risk factors in out-patients on VCI between May 2013 and November 2018. Vancomycin concentration was monitored twice-weekly to ensure adequate concentrations while avoiding high concentrations linked to nephrotoxicity (a rise in serum creatinine of ≥50% or 44 µmol/L from baseline). The likelihood of developing nephrotoxicity was evaluated using multivariable logistic regression. The 223 patients treated had a mean (standard deviation) age of 61 (16.7) years, baseline serum creatinine of 83.9 (21.2) µmol/L and estimated glomerular filtration rate (eGFR) of 80.6 (20.1) mL/min/1.73m2. Most patients (66%) were treated for bone and joint infections. Eight patients (3.6%) developed nephrotoxicity. In the most parsimonious model, nephrotoxicity was independently associated with an increased median (interquartile range) weighted-average serum vancomycin concentration (28.0 [24.3-32.6] vs. 22.4 [20.2-24.5] mg/L; odds ratio [OR] 1.25; 95% confidence interval [95% CI] 1.09-1.46; P<0.002) and Charlson co-morbidity index (OR 1.62; 95% CI 1.07-2.47; P=0.02). Post-hoc analysis identified 26 patients with a lower nephrotoxicity threshold (rise in serum creatinine of ≥30% or 27 µmol/L). Independent predictors of nephrotoxicity in this group were an increased weighted-average vancomycin concentration, diabetes, con-gestive heart failure and exposure to non-loop diuretics. The nephrotoxicity rate during VCI in this study was lower than previously reported (3.6% vs 15.0-17.0%).  Reducing the weighted-average serum vancomycin concentration may reduce nephrotoxicity while maintaining efficacy.

Legionnaires' disease caused by Legionella longbeachae: Clinical features and outcomes of 107 cases from an endemic area.

BACKGROUND AND OBJECTIVE: Legionella longbeachae is a predominant cause of Legionnaires' disease in some parts of the world, particularly in Australasia. Clinical reports of L. longbeachae infection are limited to case reports or small case series, and culture-confirmed cases. METHODS: We reviewed the clinical characteristics and outcomes of L. longbeachae pneumonia in a large case series from Christchurch, New Zealand during a 4-year period when both PCR and cultures were used as routine diagnostic tools for Legionnaires' disease. Cases of Legionella pneumophila pneumonia were reviewed for comparison. RESULTS: A total of 107 cases of L. longbeachae infection were identified by PCR and/or culture. The median age was 65 years (range 25-90 years), 63% were male, and most became unwell during spring or summer. Presenting clinical features were similar to those reported for community-acquired pneumonia, with headache, myalgia and diarrhoea being common. Elevated C-reactive protein, hyponatraemia and abnormal liver function tests were also common. History of productive cough, involvement of both lungs, and high bacterial load were independently associated with culture of Legionella from lower respiratory samples. One quarter required intensive care unit admission, and 5% died. Among patients given antimicrobial therapy before admission, those given agents without anti-Legionella activity were more likely to be admitted to the intensive care unit. Limited comparisons were made with the 19 L. pneumophila cases over the same time period. CONCLUSION: Characteristics of L. longbeachae pneumonia are broadly similar to those reported for community-acquired pneumonia from a variety of other populations, except for the spring/summer seasonality.

Corynebacterium accolens-associated pelvic osteomyelitis.

Corynebacterium accolens is a rare human pathogen. We encountered a case of C. accolens isolated from a thigh collection in a man with osteomyelitis of the adjacent pubic symphysis.

Use of ethanol locks to prevent recurrent central line sepsis.

Catheter-related sepsis (CRS) is a common complication of long-term parenteral nutrition. Conventional antibiotic therapy is often effective in the short-term but, because of poor activity against intraluminal microbial biofilms, may not prevent relapse. Ethanol is an effective antiseptic. We describe a case of a patient with recurrent CRS successfully treated with 70% ethanol locks.