The clinical pharmacology of vancomycin in seriously ill preterm infants

Preterm Physiologically Based Pharmacokinetic Model. Part II: Applications of the Model to Predict Drug Pharmacokinetics in the Preterm Population

BACKGROUND

Preterm neonates are usually not part of a traditional drug development programme, however they are frequently administered medicines. Developing modelling and simulation tools, such as physiologically based pharmacokinetic (PBPK) models that incorporate developmental physiology and maturation of drug metabolism, can be used to predict drug exposure in this group of patients, and may help to optimize drug dose adjustment.

OBJECTIVE

The aim of this study was to assess and verify the predictability of a preterm PBPK model using compounds that undergo diverse renal and/or hepatic clearance based on the knowledge of their disposition in adults.

METHODS

A PBPK model was developed in the Simcyp Simulator V17 to predict the pharmacokinetics (PK) of drugs in preterm neonates. Drug parameters for alfentanil, midazolam, caffeine, ibuprofen, gentamicin and vancomycin were collated from the literature. Predicted PK parameters and profiles were compared against the observed data.

RESULTS

The preterm PBPK model predicted the PK changes of the six compounds using ontogeny functions for cytochrome P450 (CYP) 1A2, CYP2C9 and CYP3A4 after oral and intravenous administrations. For gentamicin and vancomycin, the maturation of renal function was able to predict the exposure of these two compounds after intravenous administration. All PK parameter predictions were within a twofold error criteria.

CONCLUSION

While the developed preterm model for the prediction of PK behaviour in preterm patients is not intended to replace clinical studies, it can potentially help with deciding on first-time dosing in this population and study design in the absence of clinical data.

How to use vancomycin optimally in neonates: remaining questions

In neonates, vancomycin, a narrow-spectrum antibiotic, is the first choice of treatment of late-onset sepsis predominantly caused by Gram-positive bacteria (coagulase-negative staphylococci and enterococci). Although it has been used for >50 years, prescribing the right dose and dosing regimen remains a challenge in neonatal intensive care units for many reasons including high pharmacokinetic variability, increase in the minimal inhibition concentration against staphylococci, lack of consensus on dosing regimen and way of administration (continuous or intermittent), duration of treatment, use of therapeutic drug monitoring, limited data on short- and long-term toxicity, risk of mutant selection and errors of administration linked to concentrated formulations. This article highlights and discusses future research directions, with specific attention given to dosing optimization of vancomycin, including the advantages of modeling and simulation approaches.

Neonatal vancomycin continuous infusion: still a confusion?

BACKGROUND

Continuous infusions of vancomycin over 24 hours have been shown in adults to reduce drug toxicity, lower treatment costs and require fewer blood samples for therapeutic drug monitoring. They may also improve clinical outcome through earlier attainment of target drug concentrations. In neonates, there is no consensus on vancomycin dosing. We reviewed the literature to assess the evidence for vancomycin dosing regimens for continuous infusion in neonates.

METHODS

Medline and Embase were searched for studies about continuous vancomycin dosing regimens in neonates that reported serum drug concentrations. The search identified 469 articles, of which 5 were relevant.

RESULTS

Five prospective studies were included; 2 studies used non-linear mixed effects modeling. Vancomycin was administered with parenteral nutrition or 5% dextrose. Target serum concentrations varied (range: 10-30 mg/L). Four studies used loading doses before continuous infusion; only 1 documented achievement of therapeutic concentrations after the load. The time to a therapeutic concentration was not reported for the other studies. Attainment of target concentrations ranged from 56% to 89% of measurements. Only 1 study compared intermittent to continuous infusions, reporting higher attainment of target concentrations with continuous dosing (82% vs. 46%). No adverse effects were reported, although 3 neonates developed a reversible raised serum creatinine in the setting of septicemia.

CONCLUSION

Continuous infusions of vancomycin in neonates are well tolerated, require less blood sampling and may result in improved attainment of target concentrations. Further prospective studies are needed in this population.

Determination of vancomycin pharmacokinetics in neonates to develop practical initial dosing recommendations

Variability in neonatal vancomycin pharmacokinetics and the lack of consensus for optimal trough concentrations in neonatal intensive care units pose challenges to dosing vancomycin in neonates. Our objective was to determine vancomycin pharmacokinetics in neonates and evaluate dosing regimens to identify whether practical initial recommendations that targeted trough concentrations most commonly used in neonatal intensive care units could be determined. Fifty neonates who received vancomycin with at least one set of steady-state levels were evaluated retrospectively. Mean pharmacokinetic values were determined using first-order pharmacokinetic equations, and Monte Carlo simulation was used to evaluate initial dosing recommendations for target trough concentrations of 15 to 20 mg/liter, 5 to 20 mg/liter, and ≤20 mg/liter. Monte Carlo simulation revealed that dosing by mg/kg of body weight was optimal where intermittent dosing of 9 to 12 mg/kg intravenously (i.v.) every 8 h (q8h) had the highest probability of attaining a target trough concentration of 15 to 20 mg/liter. However, continuous infusion with a loading dose of 10 mg/kg followed by 25 to 30 mg/kg per day infused over 24 h had the best overall probability of target attainment. Initial intermittent dosing of 9 to 15 mg/kg i.v. q12h was optimal for target trough concentrations of 5 to 20 mg/liter and ≤20 mg/liter. In conclusion, we determined that the practical initial vancomycin dose of 10 mg/kg vancomycin i.v. q12h was optimal for vancomycin trough concentrations of either 5 to 20 mg/liter or ≤20 mg/liter and that the same initial dose q8h was optimal for target trough concentrations of 15 to 20 mg/liter. However, due to large interpatient vancomycin pharmacokinetic variability in neonates, monitoring of serum concentrations is recommended when trough concentrations between 15 and 20 mg/liter or 5 and 20 mg/liter are desired.

Clinical pharmacokinetics of vancomycin in the neonate: a review

Neonatal sepsis is common and is a major cause of morbidity and mortality. Vancomycin is the preferred treatment of several neonatal staphylococcal infections. The aim of this study was to review published data on vancomycin pharmacokinetics in neonates and to provide a critical analysis of the literature. A bibliographic search was performed using PubMed and Embase, and articles with a publication date of August 2011 or earlier were included in the analysis. Vancomycin pharmacokinetic estimates, which are different in neonates compared with adults, also exhibit extensive inter-neonatal variability. In neonates, several vancomycin dosing schedules have been proposed, mainly based on age (i.e., postmenstrual and postnatal), body weight or serum creatinine level. Other covariates [e.g., extracorporeal membrane oxygenation (ECMO), indomethacin or ibuprofen, and growth restriction] of vancomycin pharmacokinetics have been reported in neonates. Finally, vancomycin penetrates cerebrospinal fluid (range = 7-42%). Renal function drives vancomycin pharmacokinetics. Because either age or weight is the most relevant covariate of renal maturation, these covariates should be considered first in neonatal vancomycin dosing guidelines and further adjusted by renal dysfunction indicators (e.g., ECMO and ibuprofen/indomethacin). In addition to the prospective validation of available dosing guidelines, future studies should focus on the relevance of therapeutic drug monitoring and on the value of continuous vancomycin administration in neonates.

Reversing the Myths Obstructing the Determination of Optimal Age- and Disease-Based Drug Dosing in Pediatrics

The need for critical, well-designed comprehensive clinical pharmacology research in pediatrics that encompasses the age continuum, from the most premature infant through adolescence, may be more important today than ever. New drug regimens often require greater adherence to specific dose guidelines to maximize efficacy and minimize toxic potential. The climate that allowed the propagation of the “therapeutic orphan” concept is now mostly of historical perspective. Nevertheless, the negative impact of this concept continues to linger due to continued propagation of many, now outdated myths surrounding the effective study of optimal drug dosing in pediatrics. Advances in clinical medicine combined with the advances in study design, sampling, and analysis has dramatically improved the paradigm for clinical pharmacology research in infants and children. Capitalizing upon and thoughtfully using these many advances while dispelling these myths will result in greater research focused on optimal drug therapy in pediatric practice.

Vancomycin

This review describes the use of vancomycin in neonates over the last three decades. Given the relation of late-onset neonatal septicaemia to outcome and the increase in coagulase-negative staphylococcal infection as causative organism, vancomycin remains an important antibacterial in the neonatal intensive care unit. The pharmacokinetic behaviour of vancomycin in neonates can be adequately described by a one- or two-compartment model and is mainly determined by postconceptional age and renal function. In neonates, a patent ductus arteriosus as well as treatment with indomethacin or extracorporeal membrane oxygenation (ECMO) leads to an increase in volume of distribution and a decrease in clearance. Microbiological studies in vitro have shown that an increase in vancomycin concentrations above the minimum inhibitory concentration does not result in more effective killing. The microbiological and clinical efficacy of vancomycin in neonates has only been studied explicitly in a restricted number of patients. There are no definitive data relating serum concentrations to effect in this patient group. Vancomycin-related nephrotoxicity and ototoxicity in neonates is rare, and no clear relation to serum concentrations has been demonstrated. Based on the pharmacokinetic profile of vancomycin in neonates, several administration regimens have been constructed. Recent guidelines have suggested that dosage can be independent of gestational age or postconceptional age in neonates without renal failure. In patients with renal failure, therapy can be adequately tailored by using a regimen based on serum creatinine. The usefulness of routine monitoring of peak serum concentrations is doubtful based on the current literature. Recent research demonstrates a shift towards taking only routine trough serum concentrations in order to optimise efficacy. Patients with renal failure and other special subpopulations, such as patients exposed to ECMO or indomethacin, need to be monitored more closely.

Dose regimen for vancomycin not needing serum peak levels?

AIM

To determine the safety, efficacy, and need to measure peak serum vancomycin concentrations in a neonatal population using a standard vancomycin dosage regimen.

METHOD

A total of 101 infants who were admitted to a regional neonatal intensive care unit and received vancomycin (15 mg/kg every 12 or 18 hours depending on postnatal age) were studied retrospectively. Infants who had been started on vancomycin before they were transferred to the unit were excluded. The proportion of infants was measured whose serum vancomycin concentrations were within a conservative therapeutic range of trough 5-10 mg/l, peak 20-40 mg/l, and a less conservative, but still safe, range of trough 5-12 mg/l, peak 15-60 mg/l.

RESULTS

Trough concentrations of 5-10 mg/l were achieved by 46.5% of infants, and 5-12 mg/l by 55.4%. Peak concentrations of 20-40 mg/l were found in 83.2% of infants, and 15-60 mg/l in 99.0%. Highest peak concentration was 47.2 mg/l. Some 89.4% of infants with trough concentrations of 5-10 mg/l had a peak concentration of 20-40 mg/l.

CONCLUSIONS

The vancomycin dosage regimen used in this study produces acceptable therapeutic serum vancomycin concentrations. Peak serum vancomycin concentrations do not need to be measured in neonates using this dosage regimen.

Monitoring the treatment of sepsis with vancomycin in term newborn infants

A prospective study was conducted to determine if standardized vancomycin doses could produce adequate serum concentrations in 25 term newborn infants with sepsis.

PURPOSE

The therapeutic response of neonatal sepsis by Staphylococcus sp. treated with vancomycin was evaluated through serum concentrations of vancomycin, serum bactericidal titers (SBT), and minimum inhibitory concentration (MIC).

METHOD

Vancomycin serum concentrations were determined by the fluorescence polarization immunoassay technique, SBT by the macro-broth dilution method, and MIC by diffusion test in agar.

RESULTS

Thirteen newborn infants (59.1%) had adequate peak vancomycin serum concentrations (20 - 40 mg/mL) and one had peak concentration with potential ototoxicity risk (>40 microg/mL). Only 48% had adequate trough concentrations (5 - 10 mg/mL), and seven (28%) had a potential nephrotoxicity risk (>10 microg/mL). There was no significant agreement regarding normality for peak and trough vancomycin method (McNemar test : p = 0.7905). Peak serum vancomycin concentrations were compared with the clinical evaluation (good or bad clinical evolution) of the infants, with no significant difference found (U=51.5; p=0.1947). There was also no significant difference between the patients' trough concentrations and good or bad clinical evolution (U = 77.0; p=0.1710). All Staphylococcus isolates were sensitive to vancomycin according to the MIC. Half of the patients with adequate trough SBT (1/8), also had adequate trough vancomycin concentrations and satisfactory clinical evolution.

CONCLUSIONS

Recommended vancomycin schedules for term newborn infants with neonatal sepsis should be based on the weight and postconceptual age only to start antimicrobial therapy. There is no ideal pattern of vancomycin dosing; vancomycin dosages must be individualized. SBT interpretation should be made in conjunction with the patient's clinical presentation and vancomycin serum concentrations. Those laboratory and clinical data favor elucidation of the probable cause of patient's bad evolution, which would facilitate drug adjustment and reduce the risk of toxicity or failing to achieve therapeutic doses.

Vancomycin pharmacokinetics and Bayesian estimation in pediatric patients

The vancomycin pharmacokinetic profile was characterized in six pediatric patients and the potential of nonlinear mixed effects modeling and Bayesian forecasting for vancomycin monitoring was explored using NONMEM V (1.1). Based on steady state serial vancomycin concentrations, the estimates of mean t1/2, Vd, and Cl derived by the Sawchuk and Zaske method (1) were 3.52 hours, 0.57 L/kg, and 0.12 L/h per kg, respectively. NONMEM analysis demonstrated that a weight-adjusted two-compartment model described individual patients' data better than a comparable one-compartment model. The two-compartment estimates of mean t1/2alpha, t1/2beta, Vss, and Cl were 0.80 hour, 5.63 hours, 0.63 L/kg, and 0.11 L/h per kg, respectively. The relatively long mean t1/2alpha suggests that peak vancomycin concentrations measured earlier than 4 hours postdose do not reflect postdistributional serum concentrations. NONMEM population modeling revealed that a weight-adjusted two-compartment model provided a better fit than a comparable one-compartment model. The resulting population parameters and variances were fixed in NONMEM to obtain Bayesian predictions of individual vancomycin serum concentrations. Bayesian estimation with either a single midinterval or trough sample has the potential to provide accurate and precise predictions of vancomycin concentrations. This should be evaluated using a vancomycin population pharmacokinetic model based on a larger sample of pediatric patients.

Pharmacokinetics and dose requirements of vancomycin in neonates

AIMS

To design and evaluate dosing guidelines for vancomycin based on data collected during routine use of the drug.

METHODS

Following the observation that 66% of neonatal vancomycin trough concentrations were outside the target range, new dose guidelines were developed using a population pharmacokinetic approach. NONMEM (non-linear mixed effects model) was used to analyse dose histories and 347 concentration measurements collected during routine therapeutic drug monitoring in 59 neonates.

RESULTS

Postconceptual ages in the patient group ranged from 26-45 weeks, weights from 0. 57-4.23 kg, and creatinine concentrations from 18-172 micromol/l. The population estimate of vancomycin clearance (l/h/kg) was 3. 56/creatinine concentration (micromol/l) with an interpatient coefficient of variation (CV) of 22% and volume of distribution 0.67 l/kg with a CV of 18%. Residual error was 4.5 mg/l. When the new recommendations on dosing were used prospectively in a separate group of neonates the proportion of acceptable troughs increased from 33% to 72%.

CONCLUSIONS

The pharmacokinetics of vancomycin in neonates and young infants depend on weight and serum creatinine. Preliminary results from the new guidelines indicate an improvement on previous practice, but also an ongoing need to monitor concentrations.

Lack of vancomycin-associated nephrotoxicity in newborn infants: a case-control study

OBJECTIVE

The purpose of this study was to compare the incidence of nephrotoxicity, defined as doubling of baseline serum creatinine concentration, in newborn infants with peak vancomycin serum concentrations </=40 microg/mL at steady state to infants with peak vancomycin serum concentrations >40 microg/mL. A secondary objective was to correlate concomitant disease states and potentially nephrotoxic drug therapy with rises in serum creatinine in vancomycin recipients.

METHODS

Newborn infants with culture-proven Staphylococcus aureus or coagulase-negative staphylococcal septicemia who received vancomycin therapy for >3 days between 1985 and 1995 were identified from an existing database and a review of medical record. All 69 patients included in the study had serial serum creatinine determinations, including a baseline value within 48 hours of starting treatment with vancomycin, and serum vancomycin concentrations determined after at least three doses, with peak and trough concentrations determined 1 hour after a 60-minute infusion and 15 to 30 minutes before a dose, respectively. Infants with congenital renal or cardiac anomalies were excluded. Demographic characteristics, vancomycin dosing regimen, serum vancomycin concentrations and sample times, concomitant drug therapy, and disease states were recorded. Patients were divided into group A (peak vancomycin concentration </=40 microg/mL) and group B (peak vancomycin concentration >40 microg/mL). The change in serum creatinine concentration between the start and end of vancomycin therapy was determined. Nephrotoxicity was identified if serum creatinine doubled at any time from the start to the end of vancomycin therapy. Alternative definitions of nephrotoxicity (any rise in serum creatinine to >0.6 mg/dL or new abnormalities of urine sediment) were used in additional analyses.

RESULTS

A total of 69 evaluable patients (gestational age, 28.9 +/- 3.0 weeks; birth weight, 1219 +/- 516 g) were identified, 61 in group A and 8 in group B. Six patients in group A underwent doubling of serum creatinine concentration during vancomycin therapy, whereas none in group B did so. Serum creatinine doubled to >0.6 mg/dL in only 3 infants (all in group A). Any increase in serum creatinine to >0.6 mg/dL was seen in 10 infants, 9 of whom were in group A. No confounding variable, including previous or concomitant underlying disease states associated with renal dysfunction or treatment with other potentially nephrotoxic agents, were associated with a significant rise in serum creatinine.

CONCLUSION

Vancomycin-associated nephrotoxicity is rare in neonates, even with serum peak concentrations >40 microg/mL.

Constant rate infusion of vancomycin in premature neonates: a new dosage schedule

AIMS

Since vancomycin's bactericidal action has been shown to be time-dependent, a constant rate infusion over 24 h might result in a better bactericidal efficacy. The purpose of this study was to define a new dosage schedule in prematures.

METHODS

Two vancomycin 24 h constant rate infusion schedules were tested in two groups of neonates. Postconceptional age (PCA) was 27 to 41 weeks in group 1 (n=24) and 28 to 51.5 weeks in group 2 (n=29). Group 1 neonates received continuous infusion of 10 to 30 mgkg(-1) day(-1), adjusted for PCA and weight. Group 2 was designed to take into account the significant relationship observed in group 1 between vancomycin clearance standardized on weight and PCA and consisted of a constant loading dose of 7 mg kg(-1) followed by continuous infusion of 10 to 40 mg kg(-1) day(-1) adjusted for PCA and weight.

RESULTS

Mean vancomycin serum concentration at steady state was 11+/-3.1 mg1(-1) in group 1 and 15.4+/-6.2 mg1(-1) in group 2. Fifty-six percent of group 1 values vs 88% of group 2 values were between 10 and 30 mg at steady state (P<0.01). Both regimens were well tolerated.

CONCLUSIONS

A loading dose of vancomycin followed by constant rate infusion of the appropriate dose adjusted for PCA and weight might improve vancomycin concentrations in neonates.

Vancomycin pharmacokinetics in neonates and infants: a retrospective evaluation

OBJECTIVE

To evaluate the frequency with which current loading and maintenance vancomycin dosages achieve target serum concentrations based on pharmacokinetic parameters obtained after the initial dose. Also, to identify the daily vancomycin dosage necessary to achieve target serum concentrations at steady-state and to determine if any relationships exist between vancomycin pharmacokinetic parameters and various patient characteristics.

SETTING

Neonatal intensive care unit (NICU) at Georgia Baptist Medical Center.

PATIENTS/METHODS

Twenty-three infants with suspected or documented gram-positive infection who received intravenous vancomycin between July 1990 and November 1991 were included in this retrospective analysis. Gestational age range from 23 to 41 weeks and postconceptional age (PCA) at the time of the study ranged from 26 to 46 weeks. Vancomycin therapy was initiated with a loading dose of 15 mg/kg, followed by a maintenance dosage of 20-30 mg/kg/d, which was usually given as 10 mg/kg q8-12h. All vancomycin doses were administered using a syringe pump. Peak and trough serum concentrations were obtained following the first dose. Vancomycin pharmacokinetic parameters were determined using a one-compartment model. Infants receiving indomethacin within two weeks prior to study were analyzed separately (group 2, n = 4). All other infants were included in group 1 (n = 19).

RESULTS

For group 1, vancomycin clearance (Cl), volume of distribution (Vd), and half-life were (mean +/- 1 SD) 0.072 +/- 0.032 L/kg/h, 0.52 +/- 0.08 L/kg, and 5.6 +/- 1.6 hours, respectively. For both groups, loading doses provided 1-hour postinfusion peak concentrations of 25-35 mg/L in one of every two infants studied, whereas only three percent of initial maintenance doses were projected to provide desired peak and trough concentrations at steady-state. For group 1, the mean daily dosage necessary to provide target peak (25-35 mg/L) and trough (5-10 mg/L) concentrations at steady-state was larger than that initially prescribed (29.6 +/- 13.1 vs. 22.2 +/- 4.7 mg/kg/d). For group 2, the mean daily dosage required to achieve target peak and trough concentrations at steady-state was smaller than that initially prescribed (14.8 +/- 4.3 vs. 20.0 +/- 0.1 mg/kg/d) and was exactly half of that required for group 1. Excellent correlations were observed between PCA and vancomycin Cl (L/h) (r = 0.92; p < 0.0001), body weight and Vd(L) (r = 0.94; p < 0.0001), body weight and vancomycin Cl (L/h) (r = 0.85; p < 0.0001), PCA and Vd (L) (r = 0.89; p < 0.0001), and body surface area and Vd (L) (r = 0.93; p < 0.0001) for group 1. Moderate correlations were also noted between PCA and Cl relative to body weight (L/kg/h), postnatal age and Cl (L/kg/h), and PCA and vancomycin dosage requirements (mg/kg/d). No linear correlation was observed between any patient characteristic and Vd standardized for body weight.

CONCLUSIONS

Our data demonstrate the need for a more accurate method of estimating initial vancomycin dosage requirements in this NICU population. Although some of the relationships revealed in this study could be used to determine vancomycin dosage for infants in the range of approximately 30-36 weeks PCA, we hesitate to suggest this approach presently because of the potential limitations of our study design. Further prospective study is needed to confirm these observations. In addition, further study is necessary to describe the time course of the interaction between vancomycin and indomethacin in infants with successful and unsuccessful closure of their patent ductus arteriosus.

Clinical pharmacology of imipenem and cilastatin in premature infants during the first week of life

The first-dose and multidose pharmacokinetics of imipenem and cilastatin were evaluated in 41 premature infants during their first week of life. Premature infants (gestational age, less than or equal to 37 weeks) were assigned to receive 10-, 15-, 20-, or 25-mg/kg doses of imipenem-cilastatin (1:1) as a single- or multiple-dose regimen. A total of 39 infants received a single dose, whereas 18 infants received multiple doses. No differences were observed in pharmacokinetic parameter estimates for either agent relative to the dose administered or infant body weight; thus, the data were pooled. Elimination half-life, steady-state volume of distribution, and body clearance averaged 2.5 h, 0.5 liter/kg, and 2.5 ml/min per kg, respectively, for imipenem and 9.1 h, 0.4 liter/kg, and 0.5 ml/min per kg, respectively, for cilastatin. Similar values for these parameter estimates were observed after multidose administration, although substantial accumulation of cilastatin in serum was observed. A total of 21% of the imipenem and 43% of the cilastatin were excreted unchanged in the urine over a 12-h collection period. Corresponding renal clearances averaged 0.4 and 0.2 ml/min per kg for imipenem and cilastatin, respectively. Substantial differences were observed in the route by which imipenem was cleared from the body compared with data from adult volunteers. These data suggest that infants should receive an imipenem dose of 20 mg/kg administered every 12 h for the treatment of bacterial infections outside the central nervous system.

Antimicrobial therapy in neonates, infants and children

Many antimicrobial medications may be administered to paediatric patients with a degree of impunity because they are relatively non-toxic and have a wide therapeutic margin. However, because of different pharmacokinetics from those in adults, the potential for toxicity exists with the use of some of these agents. Drug absorption in paediatric patients, either orally or parenterally, is generally similar to that in adults, except among neonates and, particularly, premature neonates. Similarly, in neonates, drug distribution is altered, plasma protein binding is decreased and hepatic metabolism and renal excretory capacity are limited by physiological immaturity. Thus, in neonates, only drugs that have pharmacokinetically derived dosage schedules should be used, and therapeutic monitoring of plasma drug concentrations is recommended during therapy with aminoglycosides, vancomycin and chloramphenicol. In older infants and children, the pharmacokinetics of antimicrobial drugs generally approximate those in adults, and recommended dosages have been determined relative to bodyweight. Therapeutic monitoring of plasma drug concentrations may be important in certain patients, such as those with major organ failure, and may be useful in cases of suspected noncompliance. Additional pharmacokinetic considerations concerning antimicrobial medication and paediatric patients are the extent to which drug therapy penetrates the cerebrospinal fluid in meningitis, and the potential for and implications of exposure of infants to antimicrobial medications excreted in breast milk.