Approaches for dosage individualisation in critically ill patients

To discuss the altered pharmacokinetic properties of selected antibiotics in critically ill patients and to develop basic dose adjustment principles for this patient population. PubMed, EMBASE, and the Cochrane-Controlled Trial Register. Relevant papers that reported pharmacokinetics of selected antibiotic classes in critically ill patients and antibiotic pharmacodynamic properties were reviewed. Antibiotics and/or antibiotic classes reviewed included aminoglycosides, beta-lactams (including carbapenems), glycopeptides, fluoroquinolones, tigecycline, linezolid, lincosamides, and colistin. Antibiotics can be broadly categorized according to their solubility characteristics which can, in turn, help describe possible altered pharmacokinetics that can be caused by the pathophysiological changes common to critical illness. Hydrophilic antibiotics (e.g., aminoglycosides, beta-lactams, glycopeptides, and colistin) are mostly affected with the pathphysiological changes observed in critically ill patients with increased volumes of distribution and altered drug clearance (related to changes in creatinine clearance). Lipophilic antibiotics (e.g., fluoroquinolones, macrolides, tigecycline, and lincosamides) have lesser volume of distribution alterations, but may develop altered drug clearances. Using antibiotic pharmacodynamic bacterial kill characteristics, altered dosing regimens can be devised that also account for such pharmacokinetic changes. Knowledge of antibiotic pharmacodynamic properties and the potential altered antibiotic pharmacokinetics in critically ill patients can allow the intensivist to develop individualized dosing regimens. Specifically, for renally cleared drugs, measured creatinine clearance can be used to drive many dose adjustments. Maximizing clinical outcomes and minimizing antibiotic resistance using individualized doses may be best achieved with therapeutic drug monitoring.

The liver plays a central role in the pharmacokinetics of the majority of drugs. Liver dysfunction may not only reduce the blood/plasma clearance of drugs eliminated by hepatic metabolism or biliary excretion, it can also affect plasma protein binding, which in turn could influence the processes of distribution and elimination. Portal-systemic shunting, which is common in advanced liver cirrhosis, may substantially decrease the presystemic elimination (i.e., first-pass effect) of high extraction drugs following their oral administration, thus leading to a significant increase in the extent of absorption. Chronic liver diseases are associated with variable and non-uniform reductions in drug-metabolizing activities. For example, the activity of the various CYP450 enzymes seems to be differentially affected in patients with cirrhosis. Glucuronidation is often considered to be affected to a lesser extent than CYP450-mediated reactions in mild to moderate cirrhosis but can also be substantially impaired in patients with advanced cirrhosis. Patients with advanced cirrhosis often have impaired renal function and dose adjustment may, therefore, also be necessary for drugs eliminated by renal exctretion. In addition, patients with liver cirrhosis are more sensitive to the central adverse effects of opioid analgesics and the renal adverse effects of NSAIDs. In contrast, a decreased therapeutic effect has been noted in cirrhotic patients with beta-adrenoceptor antagonists and certain diuretics. Unfortunately, there is no simple endogenous marker to predict hepatic function with respect to the elimination capacity of specific drugs. Several quantitative liver tests that measure the elimination of marker substrates such as galactose, sorbitol, antipyrine, caffeine, erythromycin, and midazolam, have been developed and evaluated, but no single test has gained widespread clinical use to adjust dosage regimens for drugs in patients with hepatic dysfunction. The semi-quantitative Child-Pugh score is frequently used to assess the severity of liver function impairment, but only offers the clinician rough guidance for dosage adjustment because it lacks the sensitivity to quantitate the specific ability of the liver to metabolize individual drugs. The recommendations of the Food and Drug Administration (FDA) and the European Medicines Evaluation Agency (EMEA) to study the effect of liver disease on the pharmacokinetics of drugs under development is clearly aimed at generating, if possible, specific dosage recommendations for patients with hepatic dysfunction. However, the limitations of the Child-Pugh score are acknowledged, and further research is needed to develop more sensitive liver function tests to guide drug dosage adjustment in patients with hepatic dysfunction.

We performed a simulation study to investigate the accuracy of bootstrap estimates of optimism (internal validation) and the precision of performance estimates in independent validation samples (external validation). We combined two data sets containing children presenting with fever without source (n=376+179=555; 120 bacterial infections). Random samples were drawn from this combined data set for the development (n=376) and validation (n=179) of logistic regression models. The models included statistically significant predictors for infection selected from a set of 57 candidate predictors. Model development, including the selection of predictors, and validation were repeated in a bootstrapping procedure. The resulting expected optimism estimate in the receiver operating characteristic (ROC) area was compared with the observed optimism according to independent validation samples. The average apparent ROC area was 0.74, which was expected (based on bootstrapping) to decrease by 0.07 to 0.67, whereas the observed decrease in the validation samples was 0.09 to 0.65. Omitting the selection of predictors from the bootstrap procedure led to a severe underestimation of the optimism (decrease 0.006). The standard error of the observed ROC area in the independent validation samples was large (0.05). We recommend bootstrapping for internal validation because it gives reasonably valid estimates of the expected optimism in predictive performance provided that any selection of predictors is taken into account. For external validation, substantial sample sizes should be used for sufficient power to detect clinically important changes in performance as compared with the internally validated estimate.