Health Articles

 

Untitled

Continuous Infusion of Beta-Lactam Antibioticsin Severe Sepsis: A Multicenter Double-Blind,Randomized Controlled Trial Joel M. Dulhunty,1 Jason A. Roberts,1 Joshua S. Davis,2 Steven A. R. Webb,3 Rinaldo Bellomo,4 Charles Gomersall,5Charudatt Shirwadkar,6 Glenn M. Eastwood,4 John Myburgh,7 David L. Paterson,8 and Jeffrey Lipman1 1Department of Intensive Care Medicine, Royal Brisbane and Women's Hospital, and Burns, Trauma and Critical Care Research Centre, University of Queensland, Brisbane, 2Menzies School of Health Research, Charles Darwin University and Royal Darwin Hospital, 3Royal Perth Hospital, and Schoolof Medicine and Pharmacology, University of Western Australia, Perth, 4Department of Intensive Care, Austin Hospital, Melbourne, Australia; 5Princeof Wales Hospital and Chinese University of Hong Kong, Hong Kong; 6Blacktown Hospital, 7Critical Care and Trauma Division, George Institute forGlobal Health, Sydney, and 8Infectious Diseases Unit, Royal Brisbane and Women's Hospital, and University of Queensland Centre for ClinicalResearch, Brisbane, Australia (See the Editorial Commentary by Drusano and Lodise, on pages 245–7, and the Invited Article by Falagas et al, onpages 272–82.) Background. Beta-lactam antibiotics are a commonly used treatment for severe sepsis, with intermittent bolus dosing standard therapy, despite a strong theoretical rationale for continuous administration. The aim of this trialwas to determine the clinical and pharmacokinetic differences between continuous and intermittent dosing inpatients with severe sepsis.
Methods. This was a prospective, double-blind, randomized controlled trial of continuous infusion versus inter- mittent bolus dosing of piperacillin-tazobactam, meropenem, and ticarcillin-clavulanate conducted in 5 intensive care units across Australia and Hong Kong. The primary pharmacokinetic outcome on treatment analysis was plasmaantibiotic concentration above the minimum inhibitory concentration (MIC) on days 3 and 4. The assessed clinicaloutcomes were clinical response 7–14 days after study drug cessation, ICU-free days at day 28 and hospital survival.
Results. Sixty patients were enrolled with 30 patients each allocated to the intervention and control groups.
Plasma antibiotic concentrations exceeded the MIC in 82% of patients (18 of 22) in the continuous arm versus 29%(6 of 21) in the intermittent arm (P = .001). Clinical cure was higher in the continuous group (70% vs 43%; P = .037),but ICU-free days (19.5 vs 17 days; P = .14) did not significantly differ between groups. Survival to hospital dischargewas 90% in the continuous group versus 80% in the intermittent group (P = .47).
Conclusions. Continuous administration of beta-lactam antibiotics achieved higher plasma antibiotic concentra- tions than intermittent administration with improvement in clinical cure. This study provides a strong rationale forfurther multicenter trials with sufficient power to identify differences in patient-centered endpoints.
Keywords. pharmacokinetics; clinical outcome; meropenem; piperacillin-tazobactam; ticarcillin-clavulanate.
Severe sepsis is a major cause of mortality worldwide. In (ICU) admissions are associated with severe sepsis (over Australia and New Zealand, 11.8% of intensive care unit 17 000 episodes per annum) with in-hospital mortalityof 37.5% and a mortality burden 4 times the Australianannual road toll [, ]. This burden is evident globally [– Received 4 March 2012; accepted 6 August 2012; electronically published 16 ]. Early administration of antibiotics active against October 2012.
Correspondence: Joel Dulhunty, MBBS, MTH, PhD, Dept of Intensive Care Med- the infecting organism is a cornerstone of effective man- icine, Royal Brisbane and Women's Hospital, Herston, QLD 4029, Australia agement In a recent point prevalence study of ICU antibiotic usage in Australia and New Zealand, 3 of the 4 Clinical Infectious Diseases The Author 2012. Published by Oxford University Press on behalf of the Infectious most commonly used antibiotics in treatment were Diseases Society of America. All rights reserved. For Permissions, please e-mail: beta-lactams, with ticarcillin-clavulanate, meropenem, DOI: 10.1093/cid/cis856 and piperacillin-tazobactam accounting for 56% of all 236 • CID 2013:56 (15 January) • Dulhunty et al antibiotics used . Given that subtherapeutic dosing is associat- planned commencement or commencement within the previ- ed with poorer clinical outcomes and increased incidence of drug ous 24 hours of ticarcillin-clavulanate, piperacillin-tazobactam resistance ], optimal dosing of beta-lactam antibiotics has or meropenem; and (3) an expected or actual ICU stay greater the potential to improve the outcome for critically ill patients than 48 hours. Patients were excluded if they were <18 years with severe sepsis.
of age, had an allergy to one or more of the study medications, Beta-lactam antibiotics are administered almost exclusively were receiving palliative or supportive treatment only, were re- by intermittent bolus dosing However, there are strong ceiving continuous renal replacement therapy, did not have pharmacodynamic data suggesting that this mode of adminis- central venous catheter access with at least 3 lumens (a dedi- tration may be less effective than administration by continuous cated lumen was required for study drug administration), or infusion. Bacterial killing for beta-lactam antibiotics is related to had received the study drug for >24 hours.
the duration of time that bacteria are exposed to a concentrationof antibiotic that exceeds the minimum inhibitory concentra- tion (MIC), that is, T>MIC Administration of beta-lactam Patients were randomized to receive either (1) active infusion antibiotics by infusion produces higher blood and interstitial and placebo boluses (intervention arm) or (2) placebo infusion fluid concentrations with greater time above the MIC compared and active boluses (control arm). The 24-hour dose was clini- with intermittent dosing, particularly for bacteria with high cian-chosen and unaffected by randomization. Ticarcillin- MIC values, which are common in the ICU clavulanate and piperacillin-tazobactam (or placebo) infusions Although continuous infusion has been shown to be superior were changed every 24 hours, while meropenem (or placebo) to intermittent administration in animal and ex vivo models, 2 infusions were changed 8 hourly, as determined by antibiotic meta-analyses of the human trials to date have not demonstrat- stability at room temperature ]. Labeling was used to ed differences in clinical cure or survival [, These human conceal the syringe contents for bolus administration. Infusion trials, however, have been primarily conducted in noncritically contents were concealed by dilution of medication in 100– ill patients and were underpowered, even when pooled, limiting 250 mL infusion bags. Both methods of administration were their applicability to patients with severe sepsis. In addition, 13 used with the active treatment contained in only one adminis- of the 14 studies included in a recent meta-analysis used non- tration route. Clinical staff, data collectors, and patients were equivalent dosing in the treatment arms limiting direct compar- blinded to allocation status.
isons between the 2 delivery methods ]. The aim of this trial was to determine the clinical and pharmacokinetic differences Antibiotic Plasma Levels between continuous and intermittent dosing in critically ill pa- A maximum of 3 blood samples per patient were taken imme- tients with severe sepsis to establish feasibility to proceed with a diately prior to the active (or placebo) bolus dose during a 48- larger multicenter trial.
hour window period on days 3 and 4 to determine plasmatrough levels. Blood samples were centrifuged at 3000 rpm for 10 minutes and the plasma stored at −80°C until batchedanalysis at a central laboratory; samples were stored at −20°C Study Design and Setting for <30 hours at one site until storage at −80°C. Antibiotic This prospective, multicenter, double-blind, concealed, ran- concentration was determined by validated high performance domized controlled trial was conducted at Royal Brisbane and liquid chromatography which included within-batch cali- Women's Hospital, Austin Hospital, Blacktown Hospital, and brators and quality controls Samples were prepared by Royal Darwin Hospital, Australia, and Prince of Wales Hospi- protein precipitation with a dichloromethane wash, and the tal, Hong Kong. Recruitment occurred between April 2010 extracts separated on a C18 stationary phase and monitored and November 2011. Institutional ethics approval for the by ultraviolet. Accuracy and precision of the assays were vali- study was obtained at each site. Consent was obtained from dated at high, medium, and low concentrations of the calibra- the patient or from a substitute decision maker prior to study tion range. All results met the bioanalysis acceptance criteria enrollment. The study was registered with the Australian New of the US Food and Drug Administration []. Free Zealand Clinical Trials Registry (ACTRN12610000238077).
(unbound) drug concentrations were determined using pub-lished protein binding values (2% for meropenem, 21% for pi- Selection Criteria peracillin, and 45% for ticarcillin) [ Patients were eligible if they met all of the following inclusioncriteria: (1) severe sepsis in the previous 48 hours, defined as Outcomes and Measurements confirmed or suspected infection with new organ dysfunction The primary pharmacokinetic endpoint was plasma antibiotic based on diagnostic criteria published elsewhere [(2) concentration above MIC, scored as a dichotomous variable.
Continuous Infusion of Beta-Lactams • CID 2013:56 (15 January) • 237 MIC breakpoints for Pseudomonas aeruginosa (16 mg/L for Adequacy of blinding was assessed by clinician survey. A piperacillin and ticarcillin, and 2 mg/L for meropenem) were nurse on day 1 or 2 and a medical officer at a later date during used and scored as positive if all measured free plasma antibi- study enrollment were asked whether they thought the patient otic concentrations exceeded the breakpoint was receiving continuous or intermittent treatment and the Secondary endpoints included clinical response rated by degree of certainty in this decision using a 5-point scale [ blinded clinicians at a test of cure date 7–14 days after studydrug cessation (Table ) Time to clinical resolution was Statistical Analysis defined as the number of days from randomization to the first An on-treatment analysis of all patients with plasma antibiotic identified date of clinical resolution; this was set at 28 days for samples taken on days 3 and 4 was performed for the primary patients who did not achieve clinical cure within a 28-day pharmacokinetic endpoint (n = 22 and 21 for the intervention period. Vital status at ICU and hospital discharge and ICU- and control group, respectively). Free plasma antibiotic con- free days at day 28 were also evaluated. "ICU-free days" was centration differences were analyzed by Mann-Whitney U test defined as the number of days alive and free of ICU admission and expressed as box (median and interquartile range [IQR]) in the first 28 days postrandomization. Daily sequential organ and whiskers (10–90 percentile). An intention-to-treat analysis failure assessment (SOFA) scores were recorded [The of all randomized patients was performed for clinical end- focus of infection, concomitant antibiotic use, and duration of points (n = 30 in each group). The primary outcome was eval- therapy were recorded. Adverse events during treatment were uated by Fisher exact test. Secondary outcomes were analyzed evaluated as, almost certainly, probably, possibly, or unlikely by Student t test or Mann-Whitney U test depending on caused by study medications.
whether inspection of a normal Q-Q plot confirmed or reject-ed the normality assumption, respectively. A Kaplan-Meier curve, with follow-up until hospital discharge, was plotted to A sample of 60 patients was calculated to achieve a power of show survival trend; a log-rank test was used to compare treat- 80% to detect a 15% absolute difference in the primary ment groups. Mean ± standard deviation are reported for nor- outcome at a significance level of 5%, with a target of 8–16 mally distributed variables and median [IQR] for nonnormal participants per site.
variables. A 2-sided P value <.05 was considered statisticallysignificant. Statistical analysis was conducted using IBM SPSSStatistics 19 (IBM Corporation, Armonk, New York). James Randomization and Masking and Bang blinding indices were computed using Stata Randomization was stratified by institution with 1:1 allocation software (StataCorp LP, College Station, Texas). Box and to treatment arm. Following study enrollment, an unblinded whisker plots were generated in GraphPad Prism 5 (GraphPad research nurse or pharmacist responsible for preparation of Software, Inc, La Jolla, California).
the blinded medications determined allocation status byopening a sequentially numbered sealed envelope.
Clinician-Rated Outcome De Recruitment and Baseline Characteristics Sixty patients were enrolled; 16 at Royal Brisbane and Women's Clinical response Hospital, 14 at Austin Hospital, 12 at Blacktown Hospital, 10 atRoyal Darwin Hospital, and 8 at Prince of Wales Hospital.
Resolution—disappearance of all signs and symptoms Forty-four patients (73%) completed 4 or more days of ran- related to the infection domized treatment, with equal distribution between treatment Improvement—a marked or moderate reduction in theseverity and/or number of signs and symptoms of infection arms (Figure Four patients were discharged from the ICU Failure—insufficient lessening of the signs and symptoms of within 48 hours of randomization, and 2 patients died during infection to qualify as improvement, including death orindeterminate (no evaluation possible, for any reason) this period. The 24-hour antibiotic dose for the interventionand control groups was comparable: 13.5 [13.5–13.5] g versus 13.5 [11.3–13.5] g for piperacillin-tazobactam, 3.0 [3.0–3.8] g Resolution—as above versus 3.0 [3.0–3.0] g for meropenem, and 12.4–13.5 g (2 partici- All other findings (ie, sum of 2 and 3 above) pants) versus 12.4 g (1 participant) for ticarcillin-clavulanate.
Clinical cure (treatment exclusions) Fourteen patients in each group had a beta-lactam suscepti- Participants where the study drug, excluding beta-lactam ble organism identified as the primary causative organism antibiotic de-escalation, was changed due to nonresolution ofinfection are defined as nonresolution (regardless of clinical (Table Four patients in the intervention group had a non- response at test of cure date)—otherwise as above susceptible organism identified (Enterococcus species in 3 238 • CID 2013:56 (15 January) • Dulhunty et al

CONSORT flow diagram. Abbreviations: CRP, C-reactive protein; CRRT, continuous renal replacement therapy; CVC, central venous catheter; ICU, intensive care unit. aSub-group analysis.
patients and human metapneumovirus in a fourth). Four pa- The ratio of plasma antibiotic concentration to MIC for the tients in the control group had a nonsusceptible organism intervention and control group is displayed in Figure for all identified: methicillin-resistant Staphylococcus aureus in 2 pa- 3 samples: 3.3 [1.9–4.8] μg/mL vs 1.7 [0.4–3.8] μg/mL for tients, Coxiella burnetii (Q fever) in one, and Stenotrophomo- sample 1, 3.0 [1.6–4.1] μg/mL vs 1.1 [0.5–6.8] μg/mL for nas maltophilia in a fourth. Baseline characteristics of the 2 sample 2, and 2.8 [1.5–4.8] μg/mL vs 1.0 [0.3–2.2] μg/mL groups are reported in Table .
for sample 3, respectively.
Study endpoints are displayed in Table and survival analysis is shown in Figure . For patients receiving meropenem, plasma Plasma antibiotic concentration measured in the first sample antibiotic concentration was greater than MIC for all samples in was significantly higher in the intervention group compared 8 of 8 patients (100%) in the intervention group, compared with with the control group for meropenem (9.2 [7.9–12.9] μg/mL 2 of 9 (22%) in the control group; for patients receiving pipera- vs 3.3 [0.8–4.2] μg/mL), but not for piperacillin (35.6 [21.4– cillin-tazobactam, group differences in plasma antibiotic concen- 52.0] μg/mL vs 36.4 [6.2–142.2] μg/mL) or ticarcillin (9.1 μg/ tration above MIC were 9 of 12 (75%) vs 4 of 11 (36%), and for mL and 130.9 μg/mL vs 14.1 μg/mL, respectively; Figure ticarcillin-clavulanate 1 of 2 (50%) vs 0 of 1, respectively.
Continuous Infusion of Beta-Lactams • CID 2013:56 (15 January) • 239 Organisms Identified on Blood Culture Baseline and Study Characteristics Intervention Group Enterococcus sppa Chronic health evaluation Pseudomonas aeruginosa Serratia marascens Proteus mirabilis Aeromonas hydrophilia Burkholderia cepacia Enterobacter cloacae Haemophilus influenzae Klebsiella oxytoca Morganella morganii Salmonella typhimurium Duration of study treatment (days) Organism identified Streptococcus milleri Site of infection Streptococcus pneumonia Streptococcus pyogenes Vibrio vulnificus Abbreviations: ABC, Acinetobacter baumanii-calcoaceticus complex; ESBL, Skin or skin structure extended-spectrum beta-lactamase; MRSA, methicillin-resistant Staphylococcusaureus; MSSA, methicillin-sensitive Staphylococcus aureus.
aEnterocococcus faecalis in 2 cases and Enterococcus spp (unidentified) in 1 Central nervous system b–e Indicate multiple organisms identified in 4 cases.
Postrandomization CRP Adequacy of Blinding Nursing and medical staff completed a blinding questionnaire for 56 (93.3%) and 51 study participants (85.0%), respectively.
ICU length of stay Perceptions of randomization status are displayed in Table Of the 33 respondents (30.8%) who believed they knew which Abbreviations: APACHE, Acute Physiology and Chronic Health Evaluation; treatment arm the participant was in, 13 made a judgment CRP, C-reactive protein; ICU, intensive care unit.
based on physical characteristics of the infusion bag or a Five participants in each group had an additional site of identified infection syringe, and 9 made the judgment with reference to improve- (lung, urinary tract, and intra-abdominal).
b One participant had an additional site of infection (lung).
ment or nonimprovement in the patients' condition, withvarious reasons provided for the remaining judgments. Blind-ing indices are reported in Table .
No adverse events occurred as a result of study participation.
Two patients died during study enrolment: one patient deteri- This is the first multicenter ICU trial to our knowledge com- orated following consent but prior to commencement of the paring the effects of continuous and intermittent administra- blinded medication with the cause of death septic shock due tion of beta-lactam antibiotics. Our results showed that to aspiration pneumonitis, and one patient with deteriorating continuous infusion of beta-lactam antibiotics achieved signif- respiratory failure and septic shock died 3 days after ICU icant pharmacokinetic separation in T> MIC and higher rates admission due to pneumonia. Both events were assessed as of clinical cure compared with intermittent administration in unlikely to be related to the study drug or intervention.
critically ill patients with severe sepsis. Our study is the only 240 • CID 2013:56 (15 January) • Dulhunty et al

Study Endpoints by Treatment Group Plasma antibiotic concentration >MIC Clinical cure (test of Clinical cure (test of cure date withtreatment exclusions) Clinical cure (last day 11 (6.75–24.25)b 16.5 (7–28)b resolution (days) Free plasma antibiotic concentration between treatment Time to resolution of groups on the first sample. Abbreviations: CI, continuous infusion; IB, intermittent bolus.
ICU length of stay 19.5 (12.75–24) continuous vs intermittent beta-lactam dosing trial that has 18 (12.75–22)d .22 been conducted in a blinded fashion with allocation conceal- ment [and the largest of a limited number of studies con- Hospital survival ducted exclusively in an ICU setting ]. This Abbreviations: CRP, C-reactive protein; ICU, intensive care unit; MIC, multicenter study demonstrated the feasibility of randomizing minimum inhibitory concentration.
patients following commencement of 3 commonly prescribed a Plasma samples were available for 22 and 21 patients in the intervention beta-lactam antibiotics for severe sepsis and the ability to ad- and control groups, respectively (subgroup analysis).
b Time to clinical resolution was set at 28 d for 7 and 13 patients in the minister concealed medications in the ICU in a safe manner.
intervention and control groups, respectively, as clinical resolution did not Continuous infusion has shown to produce higher blood occur during this period.
and interstitial fluid concentrations and more rapid bacterial c Postrandomization CRP levels were available for 25 and 24 patients in theintervention and control groups, respectively (subgroup analysis); time to killing, particularly for bacteria with high MIC values in im- resolution of CRP was set at 28 d for 6 patients in each group as CRP was munodeficient ex vivo and animal models A ret- not measured below 100 mg/L during this period.
d rospective study by Lodise and colleagues in critically ill Subgroup analysis (28 and 26 patients in intervention and control groups, patients with P. aeruginosa found that using extended infu-sions of piperacillin-tazobactam to increase T>MIC resulted in improved 14-day survival (12.2% vs 31.6%, P = .04) in a sub-population of patients with high levels of sickness severity(APACHE II score >17) compared with a historical cohort[Another retrospective review of 359 patients treated forgram-negative infections across 14 hospitals in the UnitedStates found that extended infusion of piperacillin-tazobactamprolonged survival by 2.8 days (P < .01) compared with nonex-tended infusion of beta-lactam antibiotics []. However,apart from a single center ICU study by Roberts and col-leagues, which observed a 27% higher cure rate with continu-ous infusion of ceftriaxone (P = .06) [our study is the onlytrial to our knowledge to report a significant difference in clin-ical cure rates for continuous versus intermittent administra-tion of beta-lactam antibiotics. This may in part by beexplained by a focus on patients with a higher acuity of illness Free plasma antibiotic concentration to minimum inhibitory concentration ratio for 3 samples. Abbreviations: CI, continuous infusion; and dosing that was independent of treatment arm. Given pre- IB, intermittent bolus.
vious data showing that, in critically ill patients in the ICU, Continuous Infusion of Beta-Lactams • CID 2013:56 (15 January) • 241

.13 (−.011, .27)b −.036 (−.11, .035)b −.096 (−.026, .071)b Abbreviations: BI, blinding index; CI, confidence interval.
James' BI reference range (0 to 1): 0 = complete unblinding, .5 = randomguessing, 1 = complete blinding. Bang's BI reference range (−1 to 1): −1 = complete blinding, 0 = random guessing, 1 = complete unblinding.
a 95% CIs that are >.5 indicate adequate blinding.
b 95% CIs that include 0 indicate adequate blinding.
greater concentration range in the piperacillin-tazobactambolus group, including a greater number of patients with low Survival curve for patients in both treatment groups (data have been censored for patients discharged from hospital). Abbreviations: The study was not powered to evaluate any effect on surviv- CI, continuous infusion; IB, intermittent bolus.
al and suggests a clinical signal for the surrogate endpoint ofclinical cure at 7–14 days after study drug cessation (27%higher in the intervention group), even after adjusting for maintaining 100% T>MIC for beta-lactam antibiotics is associ- treatment changes. Additionally, a number of other surrogate ated with greater clinical cure than dosing that results in any- clinical endpoints, including ICU-free days at day 28 moved thing <100% (82% vs 33%, P = .002) [the nonequivalent in a favorable direction but did not achieve statistical signifi- dosing between treatment arms (lower in the continuous arm) cance. The progression to achieving a definitive clinical in 13 of the 14 previous trials may be a significant confound- answer via a stepwise research program is well described in ing factor ]. Our study demonstrated that clinician-deter- the literature [Our study provides an important step in mined dosing by continuous infusion might alone be establishing suitable endpoints for a large well-designed pro- sufficient to improve clinical cure. Although differences in spective phase II multicenter study of continuous administra- plasma antibiotic concentration between groups were most tion of beta-lactam antibiotics in critically ill patients with prominent in patients receiving meropenem, higher rates of severe sepsis.
100% T>MIC in measured samples were also present for pa- The potential benefits to patients and the health system by tients on piperacillin-tazobactam. This was evidenced by the improved methods of antibiotic delivery of beta-lactam antibi-otics are considerable. If a 4% absolute reduction in hospital Perception of Blinding Status mortality is achievable (with point estimates of 6.6%–10.0%observed in this study), then this intervention has the poten- tial to save over 800 lives each year in Australia and New Zealand [and over 37 000 lives in the United Sates [In addition, in an era of increasingly expensive therapies, admin- istration of beta-lactam antibiotics via continuous infusion compared with intermittent dosing represents greater cost-effi- ciency in terms of workload and labor costs, while remaining believe —continuous cost neutral in terms of drug costs This study has a number of limitations. Despite treatment groups being largely well balanced, differences existed for some baseline characteristics, such as 6 years younger mean age, 13% more males, 13% higher comorbidity, and a 13% higher proportion of pre-ICU infections in the intervention group. A modest sample size in each group may have similarly 242 • CID 2013:56 (15 January) • Dulhunty et al resulted in potential confounding by unmeasured variables. In Potential conflicts of interest. J. A. R. has served as a consultant for terms of plasma antibiotic concentrations, only trough con- AstraZeneca, Pfizer, Gilead and Janssen-Cilag. S. A. R. W. has attendedAdvisory Boards and acted as a consultant to Janssen-Cilag and centrations were measured. Therefore, concentrations at 40%– AstraZeneca. C. G. has served as a consultant for Janssen-Cilag and 70% T>MIC could only be inferred to be greater than the MIC.
Pfizer. J. M. has received travel and speaker fees in relation to investigator- A limited number of extreme concentration values in the in- initiated research projects from Fresenius Kabi. D. L. P. has received re-search grants from AstraZeneca and has attended Advisory Boards, acted termittent group suggested the presence of some sample as a consultant to, or given lectures with honoraria from Three Rivers timing error.
Pharmaceuticals, Merck, AstraZeneca, SanofiAventis, Pfizer, Johnson & Clinician blinding is important for surrogate outcomes, Johnson, and Leo Pharmaceuticals. J. L. has received research grants fromAstraZeneca and has attended Advisory Boards, acted as a consultant to, such as ICU-free days, which can be influenced by discharge or given lectures with honoraria from AstraZeneca, Janssen-Cilag, Merck decisions and clinician ratings of clinical cure. Although a mi- Sharp & Dohme, Pfizer, and Wyeth Australia. All other authors report no nority of staff was able to determine treatment arm by subtle physical indicators, we demonstrated that concealed adminis- All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the tration achieved satisfactory levels of blinding in a multicenter content of the manuscript have been disclosed.
context. In particular, compounding of antibiotic medicationsin infusion bags and labeling of syringes to obscure content for intermittent dosing was sufficient to achieve blinding without the need for more costly and labor-intensive mea- 1. Finfer S, Bellomo R, Lipman J, French C, Dobb G, Myburgh J. Adult- sures, such as colored tubing and covered infusion bags. The population incidence of severe sepsis in Australian and New Zealand finding that medical staff identified the intermittent arm at a intensive care units. Intensive Care Med 2004; 30:589–96.
2. Department of Infrastructure, Transport, Regional Development and significantly higher rate than chance may relate to a smaller Local Government. Road deaths Australia: 2008 statistical summary, sample size, given that a similar identification rate for nursing 2009. Available at: staff in the intermittent group was nonsignificant.
Accessed 11 October 2012.
3. Angus DC, Linde-Zwirble WT, Lidicker J, Clermont G, Carcillo J, Pinsky MR. Epidemiology of severe sepsis in the United States: analy- sis of incidence, outcome, and associated costs of care. Crit Care Med2001; 29:1303–10.
4. Vincent JL, Rello J, Marshall J, et al. International study of the preva- This is the first multicenter ICU trial that we are aware of that lence and outcomes of infection in intensive care units. JAMA 2009; compares continuous and intermittent administration of beta- 5. Martin CM, Priestap F, Fisher H, et al. A prospective, observational lactam antibiotics. The results provide evidence of the phar- registry of patients with severe sepsis: the Canadian sepsis treatment macokinetic separation of continuous infusions against bolus and response registry. Crit Care Med 2009; 37:81–8.
dosing, higher rates of clinical cure associated with continuous 6. Dellinger RP, Levy MM, Carlet JM, et al. Surviving sepsis campaign: international guidelines for management of severe sepsis and septic infusion, and the feasibility of blinding study medications in a shock: 2008. Intensive Care Med 2008; 34:17–60.
multicenter study. We believe evaluating continuous infusion 7. Dulhunty JM, Paterson DL, Webb SAR, Lipman J. Antimicrobial uti- in a severe sepsis cohort via a phase II randomized controlled lisation in 37 Australian and New Zealand intensive care units.
trial is both justified and feasible.
Anaesth Intensive Care 2011; 39:231–7.
8. Lodise TP Jr, Lomaestro B, Drusano GL. Piperacillin-tazobactam for Pseudomonas aeruginosa infection: clinical implications of an extend- ed-infusion dosing strategy. Clin Infect Dis 2007; 44:357–63.
9. McKinnon PS, Paladino JA, Schentag JJ. Evaluation of area under the Associate Professor Graham Reece (Blacktown inhibitory curve (AUIC) and time above the minimum inhibitory Hospital) and Associate Professor Dianne Stephens (Royal Darwin Hospi- concentration (T>MIC) as predictors of outcome for cefepime and tal) provided site coordination as Associate Investigators on this study. We ceftazidime in serious bacterial infections. Int J Antimicrob Agents thank ICU research coordinators Leah Peck and Helen Young (Austin 2008; 31:345–51.
Hospital), Kiran Nand and Treena Sara (Blacktown Hospital), Patricia Leung (Prince of Wales Hospital), Renae Deans, Paul Jarrett, Melissa resistance–what's dosing got to do with it? Crit Care Med 2008; 36: Lassig-Smith, Therese Starr and Janine Stuart (Royal Brisbane and Women's Hospital), Jane Thomas (Royal Darwin Hospital); Steven 11. Roberts JA, Webb S, Paterson D, Ho KM, Lipman J. A systematic Fowler, ICU pharmacist at Royal Darwin Hospital; Steven Wallis, Suzanne review on clinical benefits of continuous administration of beta-lactam Parker-Scott and Xin Liu at The Burns, Trauma and Critical Care Re- antibiotics. Crit Care Med 2009; 37:2071–8.
search Centre, The University of Queensland, for laboratory analysis; and 12. Roberts JA, Kirkpatrick CM, Roberts MS, Robertson TA, Dalley AJ, Lee Jones and Louise Marquart at the Queensland Institute of Medical Lipman J. Meropenem dosing in critically ill patients with sepsis and Research for statistical analysis of blinding.
without renal dysfunction: intermittent bolus versus continuous ad- Financial support.
This work was supported by the Intensive Care ministration? Monte Carlo dosing simulations and subcutaneous Foundation (2010); a Royal Brisbane and Women's Hospital Foundation tissue distribution. J Antimicrob Chemother 2009; 64:142–50.
grant to J. M. D. (2011); and National Health and Medical Research 13. Roberts JA, Paratz J, Paratz E, Krueger WA, Lipman J. Continuous Council Training Research Fellowships to J. A. R. and J. S. D. (569917 and infusion of beta-lactam antibiotics in severe infections: a review of its role. Int J Antimicrob Agents 2007; 30:11–8.
Continuous Infusion of Beta-Lactams • CID 2013:56 (15 January) • 243 14. Kasiakou SK, Lawrence KR, Choulis N, Falagas ME. Continuous 26. MIMS Online (Australia). Full product information: Timentin, 2012.
versus intermittent intravenous administration of antibacterials with Available at: . Accessed 11 October time-dependent action: a systematic review of pharmacokinetic and pharmacodynamic parameters. Drugs 2005; 65:2499–511.
27. European Committee on Antimicrobial Susceptibility Testing. Break- 15. Kasiakou SK, Sermaides GJ, Michalopoulos A, Soteriades ES, Falagas point tables for interpretation of MICs and zone diameters, 2012. Avail- ME. Continuous versus intermittent intravenous administration of an- tibiotics: a meta-analysis of randomised controlled trials. Lancet Infect Dis 2005; 5:581–9.
11 October 2012.
16. Bernard GR, Vincent JL, Laterre PF, et al. Efficacy and safety of re- 28. Roberts JA, Boots R, Rickard CM, et al. Is continuous infusion ceftri- combinant human activated protein C for severe sepsis. N Engl J Med axone better than once-a-day dosing in intensive care? A randomized controlled pilot study. J Antimicrob Chemother 2007; 59:285–91.
17. Viaene E, Chanteux H, Servais H, Mingeot-Leclercq MP, Tulkens PM.
29. Vincent JL, Moreno R, Takala J, et al. The SOFA (Sepsis-related Comparative stability studies of antipseudomonal beta-lactams for po- Organ Failure Assessment) score to describe organ dysfunction/ tential administration through portable elastomeric pumps (home failure. Intensive Care Med 1996; 22:707–10.
therapy for cystic fibrosis patients) and motor-operated syringes (in- 30. Kolahi J, Bang H, Park J. Towards a proposal for assessment of blind- tensive care units). Antimicrob Agents Chemother 2002; 46:2327–32.
ing success in clinical trials: up-to-date review. Community Dent Oral 18. Arlicot N, Rochefort GY, Schlecht D, Lamoureux F, Marchand S, Epidemiol 2009; 37:477–84.
Antier D. Stability of antibiotics in portable pumps used for bronchial 31. Bang H, Ni L, Davis CE. Assessment of blinding in clinical trials.
superinfection: guidelines for prescribers. Pediatrics 2007; 120:1255–9.
Control Clin Trials 2004; 25:143–56.
19. Zhang Y, Trissel LA. Stability of piperacillin and ticarcillin in Auto- 32. Nicolau DP, McNabb J, Lacy MK, Quintiliani R, Nightingale CH.
Dose infusion system bags. Ann Pharmacother 2001; 35:1360–3.
Continuous versus intermittent administration of ceftazidime in inten- 20. Smith DL, Bauer SM, Nicolau DP. Stability of meropenem in polyvi- sive care unit patients with nosocomial pneumonia. Int J Antimicrob nyl chloride bags and an elastomeric infusion device. Am J Health Agents 2001; 17:497–504.
Syst Pharm 2004; 61:1682–5.
33. Rafati MR, Rouini MR, Mojtahedzadeh M, et al. Clinical efficacy of 21. Patel PR, Cook SE. Stability of meropenem in intravenous solutions.
continuous infusion of piperacillin compared with intermittent dosing Am J Health Syst Pharm 1997; 54:412–21.
in septic critically ill patients. Int J Antimicrob Agents 2006; 22. McWhinney BC, Wallis SC, Hillister T, Roberts JA, Lipman J, Ungerer JP. Analysis of 12 beta-lactam antibiotics in human plasma by HPLC 34. Sakka SG, Glauner AK, Bulitta JB, et al. Population pharmacokinetics with ultraviolet detection. J Chromatogr B Analyt Technol Biomed and pharmacodynamics of continuous versus short-term infusion of Life Sci 2010; 878:2039–43.
imipenem-cilastatin in critically ill patients in a randomized, con- 23. US Department of Health and Human Services Food and Drug Ad- trolled trial. Antimicrob Agents Chemother 2007; 51:3304–10.
ministration, Centre for Drug Evaluation and Research, and Center for 35. Georges B, Conil JM, Cougot P, et al. Cefepime in critically ill patients: Veterinary Medicine. Guidance for industry: bioanalytical method val- continuous infusion vs an intermittent dosing regimen. Int J Clin idation, 2001. Available at: Pharmacol Ther 2005; 43:360–9.
36. Mouton JW, Vinks AA. Continuous infusion of beta-lactams. Curr 11 October 2012.
Opin Crit Care 2007; 13:598–606.
24. MIMS Online (Australia). Full product information: Merrem IV, 37. Yost RJ, Cappelletty DM, group RS. The retrospective cohort of ex- 2012. Available at: . Accessed 11 tended-infusion piperacillin-tazobactam (RECEIPT) study: a multi- October 2012.
center study. Pharmacotherapy 2011; 31:767–75.
25. MIMS Online (Australia). Full product information: Tazocin EF, 2012.
38. McAuley DF, O'Kane C, Griffiths MJD. A stepwise approach to justify Available at: Accessed 11 October phase III randomized clinical trials and enhance the likelihood of a positive result. Crit Care Med 2010; 38:S523–27.
244 • CID 2013:56 (15 January) • Dulhunty et al

Source: http://criticalcare.queensu.ca/assets/CID_2012_DulhuntyContinuous_Infusion_of_Beta-Lactam_Antibiotics_in_Severe_Sepsis_A_Multicenter_Double-Blind_Randomized_Controlled_Trial.pdf