Professional Documents
Culture Documents
J Clinthera 2013 08 014
J Clinthera 2013 08 014
exceed a certain threshold.7 Dosing regimens are possible covariates: age, sex, weight, height, mechan-
therefore being re-evaluated to keep plasma concen- ical ventilation, serum creatinine, proteins, and blood
trations above certain thresholds over the whole urea nitrogen. Specific burn indices such as Baux,
treatment period, regardless of the administrat- UBS, Tobiasen, and burn area were also recorded. The
ion schedule.8 Because ceftazidime exhibits time- glomerular filtration rate was estimated from creati-
dependent killing of gram-negative bacteria in vitro nine clearance, calculated by using the Cockcroft-
or in critically ill patients, studies involving continu- Gault method21 and the Modification of Diet in
ous administration of cephalosporin confirm that the Renal Disease (MDRD).22
steady-state blood concentration should be in excess
of the bacterial MIC.7,9 Continuous infusion opti-
Drug Administration
mizes the pharmacodynamics of β-lactams by provid-
Antibiotics were prescribed for local infections or
ing adequate antibacterial activity over the 24-hour
for sepsis either empirically when a patient’s condition
dosing period with a reduction in the total daily dose
required immediate treatment or after a bacteriologic
of the antimicrobial agent.10–12
evaluation. Before inclusion, patients were random-
Some studies have reported an influence of cova-
ized to treatment according to the mode of admin-
riates such as age, creatinine, or creatinine clearance
istration: continuous infusion of ceftazidime with a
on ceftazidime disposition in burn patients.4,6,13 Two
loading dose or discontinuous administration. Seventy
studies used a population approach5,14 without any
patients entered into these studies between 2003 and
validation process. Therefore, a recommended cefta-
2009, and they received ceftazidime by 3 modes of
zidime dosage regimen based on a pharmacokinetic/
administration. One group received a ceftazidime dose
pharmacodynamic approach is not actually deter-
of 6 g/d in 3 separate 20-minute infusions of 2 g each
mined in such singular burn patients. The aim of the
every 8 hours (n = 25). The second group was given
present study was to propose an a priori–required
6 doses of 1 g each every 4 hours (n = 25). A trough
dosage of ceftazidime given by continuous infusion to
concentration was measured at the end of day 1; the
this population. Its estimation is based on clinical and
frequency and/or the dose of ceftazidime was then
biological parameters used in our practice at the
adjusted to obtain a target trough concentration if
bedside of burn patients.
necessary. The new dosing schedule could be increas-
ed to 1 g 8 times a day (ie, every 3 hours), or reduced to
MATERIALS AND METHODS
1 g 4 times a day (ie, every 6 hours) or to 1 g 3 times a
Two prospective, open-label, randomized population
day (ie, every 8 hours). The third group received 8 g on
studies of ceftazidime’s pharmacokinetics were con-
day 1, a 2-g loading dose over 30 minutes followed by
ducted in burn patients, adhering to the agreement of
6 g/d continuously administered (n = 20). This dosage
the Toulouse and Lyon Ethical Committees. The
was initially adapted in patients with an impaired renal
clinical parts of the studies were performed in the
function according to the usual recommendations
burn patient units of the University Hospital in
(summary of product characteristics).
Toulouse-Rangueil (n = 50) and of the Hospital “St
Joseph et St Luc” in Lyon (n = 20). The Toulouse
population’s data have been partially used in previous Blood Sampling and Measurements
studies.4–6 Written informed consent was obtained In the fractionated administration groups, blood
from the patients or their relatives. was sampled in dry tubes at 24, 24.33, 48, and 48.33
All the ceftazidime plasma concentration data were hours after the beginning of the treatment. In the
obtained and analyzed in the Laboratoire de Pharma- continuous infusion group, patients were randomly
cocinétique et Toxicologie Clinique of the Purpan distributed in 3 subgroups, and blood was sampled at
Hospital in Toulouse, France. 0, 0.25, 1.5, 6, and 24 hours or 0.5, 2, 12, and 24
hours or 1, 4, 18, and 24 hours. When the dosing
Patients regimen was changed for monitoring reasons, further
Patients were studied during the secondary phase of samples were taken to determine the new ceftazidime
their burn injuries. The following demographic, clin- concentrations. A total of 286 measurable serum
ical, and biological parameters were collected as concentration were available.
Quantitative
Age, y 49 (19) 15–90 AGE
Weight, kg 74 (16) 43–120 WT
Height, cm 172 (8) 155–185 HEIG
Burned surface area (% of the total body surface) 32 (22) 10–98 BSA
Baux index 82 (24) 30–145 BAUX
UBS index 94 (76) 10–320 UBS
Tobiasen index 8 (3) 4–15 TOBI
Serum creatinine, mmol/L 77 (30) 24–234 CREATININE
Cockcroft creatinine clearance, mL/min/1.73 m2 118 (60) 33–394 CLCR
MDRD, mL/min/1.73 m2 114 (49) 29–295 MDRD
Proteinemia, g/L 57 (8) 38–75 PROT
Blood urea nitrogen, mmol/L 8 (4) 2–25 BUN
Qualitative variables
Sex 50, 20 SEX
Men ¼ 0; women ¼ 1
Mechanical ventilation 39, 31 VENT
Without ¼ 0; with ¼ 1
serum creatinine from 30 to 160 mmol/L, with a step 49 (19) years; they were badly injured (UBS, 94 [76];
of 10 mmol/L (12 cases); and (2) age from 20 to 90 Tobiasen, 8 [3]), with mechanical ventilation required
years, with a step of 10 years (8 cases). Ninety-six in 44% of cases. Mean serum creatinine was in the
(12 8) typical patients were studied, and 96,000 normal range (77 [30] mmol/L) as the mean Cockcroft
simulations were thus performed for each tested creatinine clearance (118 [60] mL/min) and as the
dosage. mean MDRD clearance (114 [49] mL/min). These
Ceftazidime concentrations were simulated over 72 parameters showed considerable interindividual
hours for dosage regimens, including a 2-g loading variability (%CV, 39%, 50%, and 43%, respectively).
dose followed by doses ranging from 3 to 16 g/d as a
continuous infusion with a 2-g step between 4 and Model Validation
16 g/d. The open 2-compartment pharmacokinetic model
with first-order elimination was chosen to describe
Interpretation the concentration–time data for ceftazidime in serum.
The concentrations generated by the previous A proportional error model was the most accurate
Monte Carlo process were used to explore the appro- for the residual variability. For the clearance, age,
priateness of different dosage regimens in burn pa- urea, serum creatinine, and glomerular filtration rate
tients by comparing these simulated concentrations estimated by using the Cockcroft and MDRD equa-
with a target interval. The concentrations have been tions, ventilation, Baux and Tobiasen indices, statisti-
arbitrarily simulated at 72 hours, ensuring they are at cally decreased the objective function and the inter-
steady state regardless of the patients’ ceftazidime t½ individual variability. For the central and peripheral
values. volumes of distribution, the statistically significant
The lower limit of the target interval was defined as covariates were, respectively, age, urea, creatinine
a steady-state concentration 5-fold higher than the clearance, ventilation status, and Tobiasen index.
MIC of the strain.7,15,16 The most frequently observed After independent deletion, the model was:
MIC for Pseudomonas aeruginosa was 1 to 2 mg/L; TVCL ¼ 21:2 – ð0:00683 creatinineÞ
therefore, the lower limit of 10 mg/L was studied. For – ð0:0115 ageÞ
strains with an MIC of 4 mg/L, the lower limit of TVV1 ¼ 13:4 þ ð0:00669 ageÞ
20 mg/L was also tested. The MIC 3-fold in empirical TVQ ¼ 1:41
situations is defined as 8 mg/L by the European TVV2 ¼ 28:9
Committee on Antimicrobial Susceptibility Testing; where TVCL, TVV1, TVQ, and TVV2 are typical
it leads to a lower limit of the target interval equal to values, respectively, for ceftazidime clearance (liters
40 mg/L.7,9,17 per hour), central volume of distribution (liters),
Conversely, the upper limit of the target interval intercompartmental clearance (liters per hour), and
was considered to be 100 mg/L to prevent adverse peripheral volume of distribution (liters).
events, regardless of the given dose. The linear regression between the individual
For each typical patient of the simulated popula- observed concentrations (Cobs) and the individual
tion, the target dosage regimen was defined as the predicted concentrations (IPred) determined by the
minimal dose leading to the highest percentile of conditional mode was as follows:
patients whose ceftazidime concentrations were in-
CIpred ðmg=LÞ ¼ 0:8454ðCobs Þ þ 5:9318ðr ¼ 0:9335Þ:
cluded between the chosen target and 100 mg/L. A
dose was considered as recommendable when 490% The objective function was reduced from 2274 to
of the population was in the target interval. 2215 (P o 0.0001; ddf ¼ 3). The ceftazidime clea-
rance interindividual variability was reduced from
RESULTS 65% to 52% and the central volume of distribution
Patients from 44% to 33%.
Seventy burn patients, with a total of 286 meas- The mean of the individual patients’ pharmacokinetic
urable serum ceftazidime concentrations, were in- parameters calculated by taking into account serum
cluded in the study. Table I summarizes patient creatinine level and age are shown in Table II. The
characteristics. Mean age of the patients w npde qualification (Figure 1) did not reveal any bias for
3 3 3
2 2 2
1 1 1
Y Quantiles
Y Quantiles
Y Quantiles
0 0 0
–1 –1 –1
–2 –2 –2
–3 –3 –3
–5 0 5 –5 0 5 –5 0 5
X Quantiles X Quantiles X Quantiles
Table III. Recommended ceftazidime dosage regimen after a 2-g loading dose required to reach a steady-state
concentration between 10 and 100 mg/L in the highest percentage of the typical burn patients as a
function of serum creatinine and age. All results correspond to recommended dosage regimen,
which lead to 490% of the population with a steady-state concentration in the defined interval.
30 10 g 10 g 10 g 8 g 6 g 6 g 6 g 6 g
40 10 g 10 g 8 g 8 g 6 g 6 g 4 g 4 g
50 10 g 10 g 8 g 8 g 6 g 6 g 4 g 4 g
60 10 g 8 g 8 g 6 g 6 g 6 g 4 g 4 g
70 8 g 6 g 6 g 6 g 6 g 4 g 4 g 3 g
80 8 g 6 g 6 g 6 g 4 g 4 g 4 g 3 g
90 8 g 6 g 6 g 6 g 4 g 4 g 4 g 3 g
100 6 g 6 g 4 g 4 g 4 g 4 g 4 g 3 g
120 6 g 6 g 4 g 4 g 4 g 3 g 3 g 3 g
140 6 g 4 g 4 g 4 g 4 g 3 g 3 g 3 g
160 4 g 4 g 4 g 3 g 3 g 3 g 3 g 3 g
Table IV. Recommended ceftazidime dosage regimen after a 2-g loading dose required to reach a steady-
state concentration between 20 and 100 mg/L in the highest percentage of typical burn patients
as a function of serum creatinine and age. All results correspond to recommended dosage
regimen, which lead to 490% of the population with a steady-state concentration in the defined
interval.
30 12 g 12 g 12 g 10 g 10 g 8 g 8 g 8 g
40 12 g 12 g 12 g 10 g 10 g 8 g 8 g 6 g
50 12 g 12 g 10 g 10 g 8 g 8 g 6 g 6 g
60 12 g 12 g 10 g 10 g 8 g 8 g 6 g 6 g
70 12 g 10 g 8 g 8 g 8 g 6 g 6 g 6 g
80 12 g 10 g 8 g 8 g 8 g 6 g 6 g 6 g
90 10 g 10 g 8 g 8 g 6 g 6 g 6 g 4 g
100 10 g 8 g 8 g 6 g 6 g 6 g 6 g 4 g
120 8 g 8 g 6 g 6 g 6 g 4 g 4 g 4 g
140 8 g 6 g 6 g 6 g 4 g 4 g 4 g 3 g
160 6 g 6 g 6 g 4 g 4 g 4 g 3 g 3 g
Table V. Recommended ceftazidime dosage regimen after a 2-g loading dose required to reach a steady-state
concentration between 40 and 100 mg/L in the highest percentage of typical burn patients
as a function of serum creatinine and age. These results correspond to recommended dosage
regimen, but do not lead to 490% of the population with a steady-state concentration in the
defined interval.
30 16 g 16 g 16 g 16 g 14 g 12 g 12 g 10 g
40 16 g 16 g 16 g 14 g 12 g 12 g 10 g 10 g
50 16 g 16 g 16 g 12 g 12 g 10 g 10 g 8 g
60 16 g 16 g 14 g 12 g 12 g 10 g 8 g 8 g
70 16 g 14 g 12 g 12 g 10 g 10 g 8 g 8 g
80 14 g 14 g 12 g 10 g 10 g 8 g 8 g 8 g
90 14 g 14 g 12 g 10 g 10 g 8 g 8 g 6 g
100 14 g 12 g 10 g 10 g 8 g 8 g 8 g 6 g
120 12 g 10 g 10 g 8 g 8 g 6 g 6 g 6 g
140 10 g 10 g 8 g 8 g 6 g 6 g 6 g 4 g
160 10 g 8 g 8 g 6 g 6 g 6 g 4 g 4 g
into account pharmacokinetic and pharmacodynamic was validated by using data splitting and the more
aspects. advanced model evaluation tools (npde).
From a pharmacokinetic point of view, we chose to In this population, the mean ceftazidime clearance
only perform simulation by continuous route even if (7.6 [2.7] L/h) was higher than that reported in burn
our model is based on continuous and discontinuous patients (2.72 L/h)14 or in ICU patients (5.48 [2.20]
administrations. These different administration schemes L/h). This observed ceftazidime clearance is in agreement
allow us to better estimate pharmacokinetic parameters, with that reported (6.88 L/h) in a previous study.5
particularly the volume of distribution, which could not Compared with healthy volunteers,5,13,14 our study
be assessed with continuous infusion data alone. By found an increase in the mean total volume of distribu-
adopting the Monolix method, we investigated the tion (57.3 [23.7] L).
quantitative relationships between pharmacokinetic pa- From a pharmacodynamic point of view, the lower
rameters and physiologic features in burn patients limit of the target concentration interval has been defin-
treated with ceftazidime. The data collected were best ed as a steady-state concentration 5-fold higher than the
described by using a 2-compartment pharmacokinetic MIC of the strain, as recently suggested.23 There is no
model with a usual high interindividual variability fitted consensus about the number of times the MIC must be
by a proportional error model.4,5,14,18,19 exceeded. Some studies have based their target on 1
After the independent deletion step, age and serum MIC.11 Mouton et al showed that administration by
creatinine were the most significant variables, and the continuous infusion achieving concentrations near the
interindividual variability was significantly reduced as MIC does not provide bacteriologic advantages com-
previously demonstrated.4,14 Age and serum creatinine pared with bolus dosing. Other in vitro and in vivo
fitted the results better than creatinine clearance, studies7,24 have confirmed that concentrations given by
regardless of the estimation formula. Their implication continuous infusion should be maintained at 4- to 5-
is also in accordance with the renal elimination of fold the MIC to have a better cure rate. In a clinical
ceftazidime20 because they are used in all equations to study using continuous infusion of ceftazidime in ICU
estimate creatinine clearance or glomerular filtration patients, Lipman et al17 proposed a target concentra-
rate.21,22 The present population pharmacokinetic tion of 40 mg/L. More recently, Aubert et al9 used the
model was built on data from 70 burn patients and same target concentration.
80 3 g/d
4 g/d
6 g/d
70 8 g/d
10 g/d
60
50
0 20 40 60 80 100 120 140 160 180
Serum Creatinine (µmol/L)
90
80 3 g/d
4 g/d
6 g/d
8 g/d
70 10 g/d
12 g/d
60
50
0 20 40 60 80 100 120 140 160 180
Serum Creatinine (µmol/L)
90
80 6 g/d
8 g/d
10 g/d
12 g/d
70 14 g/d
16 g/d
60
50
0 20 40 60 80 100 120 140 160 180
Serum Creatinine (µmol/L)
Figure 2. Percentage of 50-year-old patients reaching the (A) 10- to 100-mg/L, (B) 20- to 100-mg/L, and (C)
40- to 100-mg/L steady-state concentration target intervals after a 2-g loading dose followed by a
4- to 16-g/d continuous infusion of ceftazidime in function of serum creatinine.
Even if the relationship between ceftazidime con- use of ceftazidime in burn patients. The recommended
centration and toxicity is poorly described,25 the dosage regimens are higher than in other patients, and
maximum target concentration has been fixed to doses between 3 and 16 g/d are proposed, taking into
100 mg/L by assimilation to the described toxicity of account the MICs of pathogens. However, for sepsis
cefepime.26 This molecule is chemically and pharmaco- caused by a pathogen with an MIC Z8 mg/L, an
kinetically very close to ceftazidime.4 insufficient percentage of burn patients will reach the
We conducted 3 simulation trials taking into therapeutic target with the recommended dosages.
account 3 typical MICs (2, 4, and 8 mg/L). All these
simulations found that in a burn population, a stand- ACKNOWLEDGMENT
ard 6-g dose of ceftazidime given as a continuous Drs. Conil, Ravat, and Saivin were responsible for the
infusion in most cases is not sufficient to reach the conception, design, pharmacokinetic/dynamic analy-
defined target concentration and consequently to sis, and drafting of the manuscript. Dr. Georges was
ensure efficacy of this β-lactam similar to other clinical responsible for the collection of the clinical data,
situations. High ceftazidime doses are required due to interpretation of data and drafting of the manuscript.
the hyperfiltration glomerular rate that is frequently Drs. Ruiz and Seguin were responsible for the collec-
observed in the youngest patients with low serum tion of the clinical data. Dr. Metsu was responsible for
creatinine concentrations.27 This phenomenon had the collection of biological data. Dr. Fourcade was
already been highlighted by Loirat et al28 with the responsible for the coordination of the study and
use of tobramycin in burn patients. helped to draft the manuscript.
In patients aged o50 years, the minimal required
dosage given by continuous infusion is 4 g/d, and it
CONFLICTS OF INTEREST
could be 8 to 16 g/d for patients with the lowest serum
The authors have indicated that they have no conflicts
creatinine; the goal is to reach a minimal target
of interest regarding the content of this article.
concentration of 10 to 40 mg/L. In the oldest patients,
whenever serum creatinine is o160 mmol/L, no dose
r3 g/d would lead to the target steady-state concen- REFERENCES
tration. The use of such high doses is not in accord- 1. Richard P, Le Floch R, Chamoux C, et al. Pseudomonas
ance with the summary of the product characteristics, aeruginosa outbreak in a burn unit: role of antimicrobials
but it has been reported that the initial doses of in the emergence of multiply resistant strains. J Infect Dis.
1994;170:377–383.
ceftazidime are insufficient to ensure efficacy against
2. Georges B, Conil JM, Dubouix A, et al. Risk of emergence
P aeruginosa in various clinical situations.29,30
of Pseudomonas aeruginosa resistance to beta-lactam
For the 2 lowest-target MICs, it is possible to
antibiotics in intensive care units. Crit Care Med.
propose a ceftazidime dose leading to 490% of the 2006;34:1636–1641.
population reaching a steady-state concentration in the 3. Rains CP, Bryson HM, Peters DH. Ceftazidime. An update
target interval. For MICs equal to 8 mg/L, a maximum of its antibacterial activity, pharmacokinetic properties
of 76% of the population will reach the target interval and therapeutic efficacy. Drugs. 1995;49:577–617.
with the recommended doses. In such cases, combina- 4. Conil JM, Georges B, Lavit M, et al. Pharmacokinetics of
tion and/or use of other antibiotics may be discussed. ceftazidime and cefepime in burn patients: the impor-
tance of age and creatinine clearance. Int J Clin Pharmacol
CONCLUSIONS Ther. 2007;45:529–538.
Ceftazidime clearance and volume of distribution are 5. Conil JM, Georges B, Lavit M, et al. A population
pharmacokinetic approach to ceftazidime use in burn
increased in burn patients, exposing them to the risk
patients: influence of glomerular filtration, gender and
of inefficient plasma concentrations. Age and serum
mechanical ventilation. Br J Clin Pharmacol. 2007;64:27–35.
creatinine level significantly influenced the ceftazidime 6. Conil JM, Georges B, Fourcade O, et al. Intermittent
disposition in these burn patients. These covariates, administration of ceftazidime to burns patients: influence
which are readily usable at the bedside of burn of glomerular filtration. Int J Clin Pharmacol Ther.
patients, were taken into account to propose the 2007;45:133–142.
initial doses of ceftazidime. Our Monte Carlo simu- 7. Mouton JW, den Hollander JG. Killing of Pseudomonas
lations found that there should be a reappraisal of the aeruginosa during continuous and intermittent infusion of
ceftazidime in an in vitro pharma- 16. Cappelletty DM. Evaluation of several 24. Mouton JW, van Ogtrop ML, An-
cokinetic model. Antimicrob Agents dosing regimens of cefepime, with des D, Craig WA. Use of pharma-
Chemother. 1994;38:931–936. various simulations of renal function, codynamic indices to predict
8. Lipman J, Wallis SC, Rickard C. against clinical isolates of Pseudomo- efficacy of combination therapy
Low plasma cefepime levels in nas aeruginosa in a pharmacody- in vivo. Antimicrob Agents Chemo-
critically ill septic patients: phar- namic infection model. Antimicrob ther. 1999;43:2473–2478.
macokinetic modeling indicates Agents Chemother. 1999;43:129–133. 25. Jackson GD, Berkovic SF. Ceftazi-
improved troughs with revised dos- 17. Lipman J, Gomersall CD, Gin T, dime encephalopathy: absence
ing. Antimicrob Agents Chemother. et al. Continuous infusion ceftazi- status and toxic hallucinations.
1999;43:2559–2561. dime in intensive care: a random- J Neurol Neurosurg Psychiatry. 1992;
9. Aubert G, Carricajo A, Coudrot M, ized controlled trial. J Antimicrob 55:333–334.
et al. Prospective determination of Chemother. 1999;43:309–311. 26. Jallon P, Fankhauser L, Du Pasqu-
serum ceftazidime concentrations 18. Weinbren MJ. Pharmacokinetics of ier R, et al. Severe but reversible
in intensive care units. Ther Drug antibiotics in burn patients. J Anti- encephalopathy associated with
Monit. 2010;32:517–519. microb Chemother. 1999;44:319–327. cefepime. Neurophysiol Clin. 2000;
10. Alou L, Aguilar L, Sevillano D, 19. Weinbren MJ. Pharmacokinetics of 30:383–386.
et al. Is there a pharmacodynamic antibiotics in burns patients. J 27. Conil JM, Georges B, Fourcade O,
need for the use of continuous Antimicrob Chemother. 2001;47:720. et al. Assessment of renal function
versus intermittent infusion with 20. Ljungberg B, Nilsson-Ehle I. Com- in clinical practice at the bedside
ceftazidime against Pseudomonas parative pharmacokinetics of cef- of burn patients. Br J Clin Pharma-
aeruginosa? An in vitro pharmaco- tazidime in young, healthy and col. 2007;63:583–594.
dynamic model. J Antimicrob Che- elderly, acutely ill males. Eur J Clin 28. Loirat P, Rohan J, Baillet A, et al.
mother. 2005;55:209–213. Pharmacol. 1988;34:179–186. Increased glomerular filtration rate
11. Nicolau DP, Nightingale CH, 21. Cockcroft DW, Gault MH. Prediction in patients with major burns and
Banevicius MA, et al. Serum bac- of creatinine clearance from serum its effect on the pharmacokinetics
tericidal activity of ceftazidime: creatinine. Nephron. 1976;16:31–41. of tobramycin. N Engl J Med.
continuous infusion versus inter- 22. Levey AS, Greene T, Kusek JW, Beck 1978;299:915–919.
mittent injections. Antimicrob Agents GJ. A simplified equation to predict 29. Taccone FS, Laterre PF, Dugernier
Chemother. 1996;40:61–64. glomerular filtration rate from se- T, et al. Insufficient beta-lactam
12. Mouton JW, Horrevorts AM, rum creatinine. J Am Soc Nephrol. concentrations in the early phase
Mulder PG, et al. Pharmacoki- 2000;11:155A. of severe sepsis and septic shock.
netics of ceftazidime in serum 23. Sinnollareddy MG, Roberts MS, Lip- Crit Care. 2010;14:R126.
and suction blister fluid during man J, Roberts JA. Beta-lactam 30. Seyler L, Cotton F, Taccone FS,
continuous and intermittent infu- pharmacokinetics and pharmacody- et al. Recommended beta-lactam
sions in healthy volunteers. Antimi- namics in critically ill patients and regimens are inadequate in septic
crob Agents Chemother. 1990;34: strategies for dose optimization: a patients treated with continuous
2307–2311. structured review. Clin Exp Pharmacol renal replacement therapy. Crit
13. Walstad RA, Aanderud L, Physiol. 2012;39:489–496. Care. 2011;15:R137.
Thurmann-Nielsen E. Pharmacoki-
netics and tissue concentrations of
ceftazidime in burn patients. Eur J
Clin Pharmacol. 1988;35:543–549.
14. Dailly E, Pannier M, Jolliet P, Bourin
M. Population pharmacokinetics of
ceftazidime in burn patients. Br J
Clin Pharmacol. 2003;56:629–634.
15. Fantin B, Farinotti R, Thabaut A,
Carbon C. Conditions for the emer-
gence of resistance to cefpirome
and ceftazidime in experimental Address correspondence to: Sylvie Saivin, PharmD, MD, PhD, Labora-
endocarditis due to Pseudomonas toire de Pharmacocinétique et Toxicologie Clinique, Institut Fédératif de
aeruginosa. J Antimicrob Chemother. Biologie, 330 Avenue de Grande Bretagne, TSA 40031, 31059 Toulouse
1994;33:563–569. Cedex 9, France. E-mail: saivin.s@chu-toulouse.fr