Automation in Busulfan Therapeutic Drug Monitoring: Evaluation of an Immunoassay on two Routine Chemistry Analyzers.

BACKGROUND: Therapeutic drug monitoring (TDM) of busulfan is recommended for hematopoietic stem cell transplant recipients. Timely reporting of these TDM results is essential given the short administration period and the planned dose adjustments on day 2. The authors evaluated the performance of a new nanoparticle-based competitive immunoassay on two routine clinical chemistry analyzers and compared its performance to that of an in-house high-resolution mass spectrometry (HRMS) method. METHODS: The MyCare Oncology Busulfan Assay Kit (Saladax Biomedical) was applied on two routine clinical chemistry analyzers (Abbott Architect c8000 and Roche Cobas c502) with a linearity range of 187-2000 ng/mL. The study evaluation measured imprecision and accuracy, sample probe carry-over, and dilution integrity. Method comparison with liquid chromatography (LC)-HRMS was performed using samples from patients undergoing busulfan treatment. RESULTS: Within- and between-run coefficient of variations for both analyzers were ≤5.23% and ≤8.45%, respectively, across the busulfan concentration range. The obtained biases were ≤10.3%. Both analyzers met the acceptance criteria for sample probe carry-over and dilution integrity. Agreement between the immunoassay and LC-HRMS was high: 92% and 89% of the samples measured on Architect and Cobas, respectively, were within the ±15% limit compared to the corresponding LC-HRMS results. CONCLUSIONS: Overall, good analytical performance and high agreement with LC-HRMS results were obtained for the immunoassay installed on both routine clinical chemistry analyzers. Therefore, this assay could be implemented as a valid alternative for LC methods in clinical laboratories on different open-channel clinical chemistry analyzers, resulting in shorter turn-around times for reporting busulfan TDM results with subsequent faster dosage adjustments.

Suboptimal Beta-Lactam Therapy in Critically Ill Children: Risk Factors and Outcome.

OBJECTIVES: In critically ill children, severely altered pharmacokinetics may result in subtherapeutic ß-lactam antibiotic concentrations when standard pediatric dosing regimens are applied. However, it remains unclear how to recognize patients most at risk for suboptimal exposure and their outcome. This study aimed to: 1) describe target attainment for ß-lactam antibiotics in critically ill children, 2) identify risk factors for suboptimal exposure, and 3) study the association between target nonattainment and clinical outcome. DESIGN: Post hoc analysis of the "Antibiotic Dosing in Pediatric Intensive Care" study (NCT02456974, 2012-2019). Steady-state trough plasma concentrations were classified as therapeutic if greater than or equal to the minimum inhibitory concentration of the (suspected) pathogen. Factors associated with subtherapeutic concentrations and clinical outcome were identified by logistic regression analysis. SETTING: The pediatric and cardiac surgery ICU of a Belgian tertiary-care hospital. PATIENTS: One hundred fifty-seven patients (aged 1 mo to 15 yr) treated intravenously with amoxicillin-clavulanic acid, piperacillin-tazobactam, or meropenem. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Three hundred eighty-two trough concentrations were obtained from 157 patients (median age, 1.25 yr; interquartile range, 0.4-4.2 yr). Subtherapeutic concentrations were measured in 39 of 60 (65%), 43 of 48 (90%), and 35 of 49 (71%) of patients treated with amoxicillin-clavulanic acid, piperacillin-tazobactam, and meropenem, respectively. Estimates of glomerular filtration rate (eGFR; 54% increase in odds for each sd increase in value, 95% CI, 0.287-0.736; p = 0.001) and the absence of vasopressor treatment (2.8-fold greater odds, 95% CI, 1.079-7.253; p = 0.034) were independently associated with target nonattainment. We failed to identify an association between antibiotic concentrations and clinical failure. CONCLUSIONS: Subtherapeutic ß-lactam concentrations are common in critically ill children and correlate with renal function. eGFR equations may be helpful in identifying patients who may require higher dosing. Future studies should focus on the impact of subtherapeutic concentrations on clinical outcome.

Development and evaluation of uncertainty quantifying machine learning models to predict piperacillin plasma concentrations in critically ill patients.

BACKGROUND: Beta-lactam antimicrobial concentrations are frequently suboptimal in critically ill patients. Population pharmacokinetic (PopPK) modeling is the golden standard to predict drug concentrations. However, currently available PopPK models often lack predictive accuracy, making them less suited to guide dosing regimen adaptations. Furthermore, many currently developed models for clinical applications often lack uncertainty quantification. We, therefore, aimed to develop machine learning (ML) models for the prediction of piperacillin plasma concentrations while also providing uncertainty quantification with the aim of clinical practice. METHODS: Blood samples for piperacillin analysis were prospectively collected from critically ill patients receiving continuous infusion of piperacillin/tazobactam. Interpretable ML models for the prediction of piperacillin concentrations were designed using CatBoost and Gaussian processes. Distribution-based Uncertainty Quantification was added to the CatBoost model using a proposed Quantile Ensemble method, useable for any model optimizing a quantile function. These models are subsequently evaluated using the distribution coverage error, a proposed interpretable uncertainty quantification calibration metric. Development and internal evaluation of the ML models were performed on the Ghent University Hospital database (752 piperacillin concentrations from 282 patients). Ensuing, ML models were compared with a published PopPK model on a database from the University Medical Centre of Groningen where a different dosing regimen is used (46 piperacillin concentrations from 15 patients.). RESULTS: The best performing model was the Catboost model with an RMSE and [Formula: see text] of 31.94-0.64 and 33.53-0.60 for internal evaluation with and without previous concentration. Furthermore, the results prove the added value of the proposed Quantile Ensemble model in providing clinically useful individualized uncertainty predictions and show the limits of homoscedastic methods like Gaussian Processes in clinical applications. CONCLUSIONS: Our results show that ML models can consistently estimate piperacillin concentrations with acceptable and high predictive accuracy when identical dosing regimens as in the training data are used while providing highly relevant uncertainty predictions. However, generalization capabilities to other dosing schemes are limited. Notwithstanding, incorporating ML models in therapeutic drug monitoring programs seems definitely promising and the current work provides a basis for validating the model in clinical practice.

Population pharmacokinetics and evaluation of the predictive performance of pharmacokinetic models in critically ill patients receiving continuous infusion meropenem: a comparison of eight pharmacokinetic models.

Background: Several population pharmacokinetic (PopPK) models for meropenem dosing in ICU patients are available. It is not known to what extent these models can predict meropenem concentrations in an independent validation dataset when meropenem is infused continuously. Patients and methods: A PopPK model was developed with concentration-time data collected from routine care of 21 ICU patients (38 samples) receiving continuous infusion meropenem. The predictability of this model and seven other published PopPK models was studied using an independent dataset that consisted of 47 ICU patients (161 samples) receiving continuous infusion meropenem. A statistical comparison of imprecision (mean square prediction error) and bias (mean prediction error) was conducted. Results: A one-compartment model with linear elimination and creatinine clearance as a covariate of clearance best described our data. The mean ± SD parameter estimate for CL was 9.89 ±âŸ3.71 L/h. The estimated volume of distribution was 48.1 L. The different PopPK models showed a bias in predicting serum concentrations from the validation dataset that ranged from -8.76 to 7.06 mg/L. Imprecision ranged from 9.90 to 42.1 mg/L. Conclusions: Published PopPK models for meropenem vary considerably in their predictive performance when validated in an external dataset of ICU patients receiving continuous infusion meropenem. It is necessary to validate PopPK models in a target population before implementing them in a therapeutic drug monitoring program aimed at optimizing meropenem dosing.

Measuring Unbound Versus Total Piperacillin Concentrations in Plasma of Critically Ill Patients: Methodological Issues and Relevance.

BACKGROUND: Piperacillin is considered a moderately protein-bound antibiotic (20%-40%), with albumin being an important binding protein. Although infrequently used in practice, different methods to measure the fraction unbound (fu) are available, but uncertainty remains as to what the most appropriate method is. The main goal of this study was to estimate the impact of the methodology used to measure unbound piperacillin in plasma on the fu of piperacillin; we compared ultrafiltration (UF) at 4°C and 37°C with the reference method, equilibrium dialysis. In addition, we analyzed the impact of other proteins on the fu. METHODS: Anonymized left-over Li-heparin plasma samples (n = 41) from 30 critically ill patients who were treated with piperacillin were used for the analyses. RESULTS: We found that the piperacillin fu, determined by UF, is on average 8% higher at 37°C (91%) than at 4°C (83%). There were no systematic or proportional differences between UF at 4°C and equilibrium dialysis at 4°C. This emphasizes the importance of the temperature during UF, which should therefore be clearly stated in publications that report on the methodology of UF. No significant impact of the albumin-, IgA-, total protein-, or α1-acid glycoprotein concentration on the fu was found. The fu found in this study was higher than the generally assumed fu value of 60%-80%. A possible explanation lies in the studied population or in the temperature used. Based on our results, routine monitoring of unbound piperacillin in intensive care unit patients is not recommended. CONCLUSIONS: Based on the prediction model, we can state that in intensive care patients the fu of piperacillin is 91% (SD 7%), determined with UF at 37°C.

Development and validation of an LC tandem MS assay for the quantification of ß-lactam antibiotics in the sputum of cystic fibrosis patients.

OBJECTIVES: Antibiotic therapy is of vital importance for the control of infectious exacerbations in cystic fibrosis (CF) patients. However, very little is known regarding the fraction of systemically administered antibiotics reaching the lower respiratory tract secretions. We developed and validated a method to measure the concentrations of piperacillin, ceftazidime, meropenem and aztreonam in CF sputum, and present the validation data. METHODS: Ultra-performance LC coupled to tandem MS was used. A single sample can be measured in 2.5 min with multiple reaction monitoring in positive electrospray ionization mode. Deuterated internal standards were used and a concentration range of 0.7-160 mg/L was covered. The method was validated according to the EMA guideline on analytical method validation. RESULTS: The boundaries within which a reliable measurement in CF sputum can be performed were determined. A few constraints are linked to the instability of the antibiotics in sputum. Piperacillin showed limited stability at room temperature and during freeze-thaw cycles. Autosampler instability was observed after 15 h for aztreonam at low concentrations. CONCLUSIONS: The method allows a reliable measurement of the selected antibiotics, if precautions are taken regarding the limited stability of piperacillin at room temperature. Due to freeze-thaw instability, piperacillin should always be analysed on the day of sampling. Quick review of the analytical data and reanalysis are needed as low concentrations of aztreonam are not stable in the autosampler.

Quantitative urine test strip reading for leukocyte esterase and hemoglobin peroxidase.

BACKGROUND: Recently, urine test strip readers have become available for automated test strip analysis. We explored the possibilities of the Sysmex UC-3500 automated urine chemistry analyzer based on complementary metal oxide semiconductor (CMOS) sensor technology with regard to accuracy of leukocyte esterase and hemoglobin peroxidase results. We studied the influence of possible confounders on these measurements. METHODS: Reflectance data of leukocyte esterase and hemoglobin peroxidase were measured using CMOS technology on the Sysmex UC-3500 automated urine chemistry analyzer. Analytical performance (imprecision, LOQ) as well as the correlation with white blood cell (WBC) and red blood cell (RBC) counts (Sysmex UF-5000) were studied. Furthermore, the influence of urinary dilution, haptoglobin, pH and ascorbic acid as confounders was determined. RESULTS: Within- and between-run imprecision (reflectance signal) ranged from 1.1% to 3.6% and 0.9% to 4.2% for peroxidase and 0.4% to 2.5% and 0.4% to 3.3% for leukocyte esterase. Good agreement was obtained between the UF-5000 for RBCs and peroxidase reflectance (r=0.843) and for WBCs and leukocyte esterase (r=0.821). Specific esterase activity decreased for WBC counts exceeding 100 cells/µL. Haptoglobin influenced the peroxidase activity, whereas leukocyte esterase and peroxidase activities showed a pH optimum between 5.0 and 6.5. A sigmoidal correlation was observed between urinary osmolality and peroxidase activity. CONCLUSIONS: CMOS technology allows to obtain high quality test strip results for assessing WBC and RBC in urine. Quantitative peroxidase and leukocyte esterase are complementary with flow cytometry and have an added value in urinalysis, which may form a basis for expert system development.

Measurement of sirolimus concentrations in human blood using an automated electrochemiluminescence immunoassay (ECLIA): a multicenter evaluation.

BACKGROUND: Therapeutic drug monitoring (TDM) of sirolimus is essential in transplant recipients. We evaluated the performance of a new electrochemiluminescence immunoassay (ECLIA) for measuring sirolimus concentrations in whole blood at five European laboratories. METHODS: Study assessments included repeatability, intermediate precision and functional sensitivity (concentration at coefficient of variation [CV] of 20%) experiments. Method comparisons with liquid chromatography-tandem mass spectrometry (LC-MS/MS; reference method) and two immunoassays (chemiluminescent microparticle immunoassay [CMIA] and antibody-conjugated magnetic immunoassay [ACMIA]) were performed using native samples from patients with kidney transplants. RESULTS: Imprecision testing CVs were ≤6.4% and ≤10.7% across the sirolimus concentration range for both repeatability and intermediate precision, respectively. The ECLIA showed excellent functional sensitivity: the CV did not reach 20%; the CV at the assay's limit of quantitation (1.5 µg/L) was 7.0%. Agreement between the ECLIA and LC-MS/MS using native kidney samples was close, with weighted Deming regression analysis yielding a slope of 1.05, an intercept of 0.154 µg/L and a Pearson's correlation coefficient (r) of 0.94, while Bland-Altman analysis showed a combined mean bias of 0.41 µg/L (±2 standard deviation [SD], -1.96 to 2.68). The ECLIA also showed good correlation with the two other immunoassays: the CMIA (slope=0.91, intercept=0.112 µg/L and r=0.89) and the ACMIA (slope=0.99, intercept=0.319 µg/L and r=0.97). CONCLUSIONS: The ECLIA showed good precision, functional sensitivity and agreement with other methods of sirolimus measurement used in clinical practice, suggesting that the assay is suitable for TDM in transplant recipients and provides an alternative to LC-MS/MS.

Impact of vancomycin protein binding on target attainment in critically ill children: back to the drawing board?

Objectives: The objectives of this observational study were to investigate plasma protein binding and to evaluate target attainment rates of vancomycin therapy in critically ill children. Patients and methods: Paediatric ICU patients, in whom intravenous intermittent dosing (ID) or continuous dosing (CD) with vancomycin was indicated, were included. Covariates on unbound vancomycin fraction and concentration were tested using a linear mixed model analysis and attainment of currently used pharmacokinetic/pharmacodynamic (PK/PD) targets was evaluated. Clinicaltrials.gov: NCT02456974. Results: One hundred and eighty-eight plasma samples were collected from 32 patients. The unbound vancomycin fraction (median = 71.1%; IQR = 65.4%-79.7%) was highly variable within and between patients and significantly correlated with total protein and albumin concentration, which were both decreased in our population. Total trough concentration (ID) and total concentration (CD) were within the aimed target concentrations in 8% of patients. The targets of AUC/MIC ≥400 and f AUC/MIC ≥200 were achieved in 54% and 83% of patients, respectively. Unbound vancomycin concentrations were adequately predicted using the following equation: unbound vancomycin concentration (mg/L) = 5.38 + [0.71 × total vancomycin concentration (mg/L)] - [0.085 × total protein concentration (g/L)]. This final model was externally validated using 51 samples from another six patients. Conclusions: The protein binding of vancomycin in our paediatric population was lower than reported in non-critically ill adults and exhibited large variability. Higher target attainment rates were achieved when using PK/PD indices based on unbound concentrations, when compared with total concentrations. These results highlight the need for protein binding assessment in future vancomycin PK/PD research.

Volumetric absorptive microsampling at home as an alternative tool for the monitoring of HbA1c in diabetes patients.

BACKGROUND: Microsampling techniques have several advantages over traditional blood collection. Dried blood spot (DBS) sampling and blood collection with heparinized capillaries are the standard techniques. Volumetric absorptive microsampling (VAMS) is a novel technique that collects a fixed volume of blood by applying an absorbent tip to a blood drop. In the present study we explored the feasibility of HbA1c monitoring with VAMS sampling at home and analysis in the laboratory. METHODS: Diabetic patients were enrolled in this study during consultation with the endocrinologist. A venous (adults) or capillary (children) sample was taken for immediate HbA1c analysis. DBS (n=1) and dried VAMS (n=2) were collected at home and sent to the laboratory. For 25 pediatric patients one VAMS was collected during consultation for immediate analysis (without drying), referred to as "wet VAMS". HbA1c analyses were performed on a Tosoh HLC-723 G8 high-performance liquid chromatography (HPLC) analyzer. RESULTS: The median time between sampling at home and analysis was 3 days. Results of HbA1c in dried VAMS showed a poor agreement with venous/capillary blood collected in hospital (concordance correlation coefficient CCC=0.72). Similar observations were found with standard DBS. An excellent agreement was obtained between HbA1c results on wet VAMS (CCC=0.996) and standard blood samples. Patients experienced VAMS and DBS as easy and convenient to use. CONCLUSIONS: Utilizing equipment standard available in the clinical laboratory, the use of home-sampled dried VAMS and DBS is not a reliable tool for the monitoring of HbA1c. However, perfect agreement between HbA1c measured on wet VAMS and capillary microsamples was obtained.

The human immunodeficiency virus (HIV) Rev-binding protein (HRB) is a co-factor for HIV-1 Nef-mediated CD4 downregulation.

Human immunodeficiency virus type 1 (HIV-1)-mediated CD4 downregulation is an important determinant of viral replication in vivo. Research on cellular co-factors involved in this process could lead to the identification of potential therapeutic targets. We found that CD4 surface levels were significantly higher in HIV-1-infected cells knocked-down for the HIV Rev-binding protein (HRB) compared with control cells. HRB knock-down affected CD4 downregulation induced by Nef but not by HIV-1 Vpu. Interestingly, the knock-down of the related protein HRBL (HRB-like), but not of the HRB interaction partner EPS15 (epidermal growth factor receptor pathway substrate 15), increased CD4 levels in Vpu-expressing cells significantly. Both of these proteins are known to be involved in HIV-1-mediated CD4 downregulation as co-factors of HIV-1 Nef. These results identify HRB as a previously unknown co-factor for HIV-1 Nef-mediated CD4 downregulation and highlight differences with the related protein HRBL, which affects the CD4 downregulation in a dual role as co-factor of both HIV-1 Nef and Vpu.

How to Solve the Underestimated Problem of Overestimated Sodium Results in the Hypoproteinemic Patient.

OBJECTIVES: The availability of a fast and reliable sodium result is a prerequisite for the appropriate correction of a patient's fluid balance. Blood gas analyzers and core laboratory chemistry analyzers measure electrolytes via different ion-selective electrode methodology, that is, direct and indirect ion-selective electrodes, respectively. Sodium concentrations obtained via both methods are not always concordant. A comparison of results between both methods was performed, and the impact of the total protein concentration on the sodium concentration was investigated. Furthermore, we sought to develop an adjustment equation to correct between both ion-selective electrode methods. DESIGN: A model was developed using a pilot study cohort (n = 290) and a retrospective patient cohort (n = 690), which was validated using a prospective patient cohort (4,006 samples). SETTING: ICU and emergency department at Ghent University Hospital. PATIENTS: Patient selection was based on the concurrent availability of routine blood gas Na⁺(direct) as well as core laboratory Na⁺(indirect) results. INTERVENTIONS: In the pilot study, left-over blood gas syringes were collected for further laboratory analysis. MEASUREMENT AND MAIN RESULTS: There was a significant negative linear correlation between Na⁺(indirect) and Na⁺(direct) relative to changes in total protein concentration (Pearson r = -0.69; p < 0.0001). In our setting, for each change of 10 g/L in total protein concentration, a deviation of ~1.3 mmol/L is observed with the Na⁺(indirect) result. Validity of our adjustment equation protein-corrected Na⁺(indirect) = Na⁺(indirect) - 10.53 + (0.1316 × total protein) was demonstrated on a prospective patient cohort. CONCLUSIONS: As Na⁺(direct) measurements on a blood gas analyzer are not influenced by the total protein concentration in the sample, they should be preferentially used in patients with abnormal protein concentrations. However, as blood gas analyzers are not available at all clinical wards, the implementation of a protein-corrected sodium result might provide an acceptable alternative.

An Unusual Type of Kidney Stone.

A very rare case of acetylsulfapyridine nephrolithiasis is presented in a 54-year-old female patient who had been prescribed sulfasalazine (6 x 500 mg/day) because of psoriatic arthritis for the last 9 years. The patient's renal function was only slightly impaired. Reflectance infrared spectroscopy and gas chromatography-mass spectrometry allowed the identification of the chemical nature of the stone. As acetylsulfapyridine is a metabolite of sulfasalazine, administration of the latter drug was the cause of the nephrolithiasis.

A Simulation Study Reveals Lack of Pharmacokinetic/Pharmacodynamic Target Attainment in De-escalated Antibiotic Therapy in Critically Ill Patients.

De-escalation of empirical antibiotic therapy is often included in antimicrobial stewardship programs in critically ill patients, but differences in target attainment when antibiotics are switched are rarely considered. The primary objective of this study was to compare the fractional target attainments of contemporary dosing of empirical broad-spectrum ß-lactam antibiotics and narrower-spectrum antibiotics for a number pathogens for which de-escalation may be considered. The secondary objective was to determine whether alternative dosing strategies improve target attainment. We performed a simulation study using published population pharmacokinetic (PK) studies in critically ill patients for a number of broad-spectrum ß-lactam antibiotics and narrower-spectrum antibiotics. Simulations were undertaken using a data set obtained from critically ill patients with sepsis without absolute renal failure (n = 49). The probability of target attainment of antibiotic therapy for different microorganisms for which de-escalation was applied was analyzed. EUCAST MIC distribution data were used to calculate fractional target attainment. The probability that therapeutic exposure will be achieved was lower for the narrower-spectrum antibiotics with conventional dosing than for the broad-spectrum alternatives and could drastically be improved with higher dosages and different modes of administrations. For a selection of microorganisms, the probability that therapeutic exposure will be achieved was overall lower for the narrower-spectrum antibiotics using conventional dosing than for the broad-spectrum antibiotics.

Measuring unbound versus total vancomycin concentrations in serum and plasma: methodological issues and relevance.

BACKGROUND: Studies on the unbound fraction (fu) of vancomycin report highly variable results. Great controversy also exists about the correlation between unbound and total vancomycin concentrations. As differences in (pre-)analytic techniques may explain these findings, we investigated the impact of the procedure used to isolate unbound vancomycin in serum/plasma on fu and the correlation between total and unbound concentrations. METHODS: Patient samples (n = 39) were analyzed for total and unbound vancomycin concentrations after ultrafiltration (UF, Centrifree at 4°C and 37°C) or equilibrium dialysis (ED, using a Fast Micro-Equilibrium Dialyzer at 37°C) on an Architect i2000SR. To investigate correlations with potential binding proteins, total protein, albumin, alpha-1-acid glycoprotein, and IgA concentrations were also measured. RESULTS: The median fu after ED was 72.5% [interquartile range (IQR), 68.7%-75.0%]. Ultrafiltration at 4°C and 37°C resulted in a median fu of 51.6% (IQR, 48.6%-54.8%) and 75.2% (IQR, 69.3%-78.6%), respectively, with no significant difference between unbound vancomycin concentrations after ED and UF at 37°C (P = 0.13). Unbound concentrations obtained through ED and UF correlated linearly (4°C: r = 0.9457; 37°C: r = 0.9478; both P < 0.0001). Linear mixed-model regression showed that total vancomycin as such was the predominant determinant for the unbound concentration, allowing a reliable prediction (mean bias ± SD, 5.0% ± 7.6%). The studied protein concentrations were of no added value in predicting the unbound concentration. CONCLUSIONS: Vancomycin fu after UF at 4°C was on average 30.6% lower than that after UF at 37°C, demonstrating the importance of temperature during UF. Ultrafiltration at 37°C resulted in unbound vancomycin concentrations equivalent with ED. As the unbound concentration could be reliably predicted based on total vancomycin concentrations as such, measurement of unbound vancomycin concentrations has little added value over measurements of total vancomycin concentrations.

Population pharmacokinetics and dosing simulations of cefuroxime in critically ill patients: non-standard dosing approaches are required to achieve therapeutic exposures.

OBJECTIVES: To investigate the population pharmacokinetics of cefuroxime in critically ill patients. METHODS: In this observational pharmacokinetic study, multiple blood samples were taken over one dosing interval of intravenous cefuroxime. Blood samples were analysed using a validated ultra HPLC tandem mass spectrometry technique. Population pharmacokinetic analysis and dosing simulations were performed using non-linear mixed-effects modelling. RESULTS: One hundred and sixty blood samples were collected from 20 patients. CL(CR) ranged between 10 and 304 mL/min. A two-compartment model with between-subject variability on CL, V of the central compartment and V of the peripheral compartment described the data adequately. Twenty-four hour urinary CL(CR) was supported as a descriptor of drug CL. The population model for CL was CL = θ(1) × CL(CR)/100, where θ(1) is the typical cefuroxime CL in the population, which is 9.0 L/h. The mean V was 22.5 L. Dosing simulations showed failure to achieve the pharmacokinetic/pharmacodynamic target of 65% fT(>MIC) for an MIC of 8 mg/L with standard dosing regimens for patients with CL(CR) ≥50 mL/min. CONCLUSIONS: Administration of standard doses by intermittent bolus is likely to result in underdosing for many critically ill patients. Continuous infusion of higher than normal doses after a loading dose is more likely to achieve pharmacokinetic/pharmacodynamic targets. However, even continuous infusion of high doses (up to 9 g per day) does not guarantee adequate levels for all patients with a CL(CR) of ≥300 mL/min if the MIC is 8 mg/L.