Midazolam Clearance in Children: Why the Average Does Not Identify the Individual

A published average clearance is not a measurement of the child in front of you. That distinction matters when interpreting midazolam concentrations in pediatric intensive care. A population estimate describes the patients and methods that produced it; applying it to another patient requires evidence that the comparison fits.

What the uploaded review found

Altamimi and colleagues reviewed 25 reports of intravenous midazolam pharmacokinetics in children. Only 13 supplied the full range of individual clearance values. The reported coefficients of variation spanned 18% to 170% in critically ill patients and 13% to 54% in patients who were not critically ill. These are ranges across studies, not a single pooled estimate or a prediction interval for a particular child.

The review also separated age groups. Large variability appeared throughout the critically ill pediatric groups, while average clearance tended to be lower in neonates than in older children. A weight-normalized value therefore does not, by itself, resolve differences associated with developmental stage and illness.

Why the underlying study population matters

One primary study illustrates the limitation. De Wildt and colleagues studied 21 intensive care patients, from newborns to adolescents, during and after continuous infusion. Its final pharmacokinetic analysis excluded three patients with renal failure, hepatic failure, or concomitant erythromycin-fentanyl treatment. The remaining 18 had a mean clearance of 5.0 mL/kg/min with a standard deviation of 3.9 mL/kg/min. Those exclusions matter when deciding whether the reported average represents a patient with comparable complications. Read the primary study record.

A large range is evidence of uncertainty

The review’s studies differed in age, illness, sampling and pharmacokinetic analysis. Some coefficients of variation were reported directly, others were calculated from individual data, and others were estimated from a mean and standard deviation under a distributional assumption. Treating all of those numbers as interchangeable would hide methodological differences.

A high coefficient of variation describes substantial dispersion relative to the mean. It does not identify the clearance of an unmeasured patient. Nor does the largest observed range establish how frequently extreme values occur in another hospital or population.

What a defensible interpretation requires

Before using a literature clearance value to interpret a child’s concentration history, identify the administration route, infusion and sampling times, age group, illness, relevant co-medication, and the population used to estimate the parameter. State whether the value was measured, individually estimated through a model, or borrowed from a group.

This is a pharmacokinetic question about drug disposition. The clearance literature alone does not establish an individual child’s sedation, injury, or clinical outcome. The review is also historical: its search extended through 2012, so it should not be presented as a current dosing protocol. Its durable lesson is narrower and useful: an average does not erase patient variability.

Source

Altamimi MI, Sammons H, Choonara I. Inter-individual variation in midazolam clearance in children. Archives of Disease in Childhood. 2015;100:95-100. doi:10.1136/archdischild-2013-305720. Original analysis based on the supplied paper; no figures or tables reproduced.