Beta-Lactam Therapeutic Drug Monitoring in the Critically Ill Children: Big Solution for Infections in Small People?*

    loading  Checking for direct PDF access through Ovid

Excerpt

Optimized antimicrobial therapy is a critical determinant of outcome in critically ill patients (1). For almost all antimicrobial classes, there are clinical data demonstrating the strong relationship between drug exposure (concentrations) and clinical cure (2–4), survival (5) and even toxicity (6). Specifically, for beta-lactam antimicrobials, optimal drug exposures also improve outcomes with data existing for cephalosporins (7) and carbapenems (8) and the beta-lactam family more generally (9). It follows that dosing of antimicrobials is intrinsically related to outcome of infection and that dose optimization should consider an essential intervention in all infected critically ill patients.
While all clinicians aim to provide maximally effective therapy for infections, how this can be provided is the principal challenge. Indeed, product information–guided dosing is associated with consistently nontherapeutic antimicrobial exposures due to profound changes in drug volume of distribution and clearance (10). Dose optimization interventions at an unit level (i.e., all patients receive the same intervention) such as prolonged infusion of beta-lactam antimicrobials have been shown to improve clinical cure in randomized studies and reduce in-hospital mortality in meta-analyses (11). Use of dosing nomograms have been shown to increase frequency of achievement of therapeutic exposures for vancomycin (12), although data describing clinical outcome benefits with this approach remain sparse. Dosing software is emerging as another potential tool for choosing better drug doses (10) but has not been widely studied in critical care units. A final method used for antimicrobial dose optimization is through use of therapeutic drug monitoring (TDM), the process of measuring drug concentrations in blood and then adjusting the dose to ensure attainment of an exposure (or concentration) within a therapeutic range that is associated with improved patient outcomes (10).
Beta-lactam TDM has seen an upsurge in reports over recent years with many hospitals and critical care units electing to establish this service as a means of improving drug dosing (13). Although no clinical outcome studies have evaluated the efficacy of TDM-guided therapy, the impetus has been driven by the studies described above, which demonstrate the clear association between optimized beta-lactam exposure and improved patient outcomes. This is unsurprising given that clinicians aim to provide optimized and personalized care for their patients. What is surprising is the number of hospitals that are now performing beta-lactam TDM given the expense of establishing a laboratory infrastructure to support it, an infrastructure that is far more costly than required for TDM of vancomycin and aminoglycosides. Furthermore, there is a fundamental lack of consistency with the way that sites are performing TDM in terms of timing of samples, target concentrations, and bioanalytical methods (13). Such variation raises concerns of whether some patients may not indeed be receiving optimal therapy because a consensus guideline for maximally effective TDM is not available. Despite this, beta-lactam TDM is now found in hospitals in Australia, Austria, Belgium, Denmark, Estonia, Finland, France, Germany, Italy, Malaysia, the Netherlands, United Kingdom, the United States, Singapore, Spain, Switzerland, and potentially other countries as well.
However, the use of beta-lactam TDM had not been previously reported in pediatric patients until the report from Cies et al (14) in this issue of Critical Care Medicine. In this observational study, the authors included 82 patients (median age 1 yr) to determine whether existing beta-lactam dosing regimens achieve predefined therapeutic exposures in critically ill children, including those receiving extracorporeal membrane oxygenation (n = 14) and/or renal replacement therapy (n = 2).
    loading  Loading Related Articles