Одноразове на добу дозування аміноглікозиду для тяжкохворого — how sepsis-driven volume expansion and augmented renal clearance break the standard extended-interval nomogram
The general once-daily aminoglycoside nomogram (Hartford-style) is built on population pharmacokinetics from relatively stable patients, assuming a volume of distribution near 0.25 L/kg. Critical illness invalidates that assumption. Sepsis triggers systemic capillary leak — endothelial glycocalyx injury lets fluid and small hydrophilic molecules like gentamicin, tobramycin, and amikacin shift from plasma into a swollen interstitial space. Add in the tens of liters of crystalloid many septic patients receive in the first 24–48 hours, and the same weight-based mg/kg dose is now distributed across a dramatically larger volume — producing a lower, potentially subtherapeutic peak concentration.
Aminoglycosides are highly hydrophilic, poorly protein-bound, and largely confined to extracellular fluid — they do not meaningfully penetrate cells or fat. Under normal conditions this keeps their volume of distribution predictable and closely tied to lean body weight.
Sepsis disrupts the endothelial glycocalyx, the gel-like layer lining capillaries that normally restricts fluid and small-molecule egress. Pro-inflammatory mediators (TNF-α, IL-1, IL-6) degrade this barrier, and capillaries become "leaky." Plasma water — carrying dissolved aminoglycoside — shifts into the interstitium, clinically manifesting as edema, third-spacing, and a positive fluid balance.
Superimposed on this physiologic leak is the therapy itself: septic shock resuscitation protocols call for large-volume crystalloid boluses, and many ICU patients accumulate 5–10+ liters of positive fluid balance within the first day. Both mechanisms — pathologic leak and iatrogenic fluid loading — expand the same extracellular space the aminoglycoside distributes into.
The net pharmacokinetic effect: C(peak) = Dose / Vd. If Vd roughly doubles while dose stays fixed, peak concentration is roughly halved — often dropping the crucial Cmax:MIC ratio below the threshold needed for reliable bactericidal killing, even though total drug clearance may be unchanged.
A standard 5–7 mg/kg gentamicin dose calculated for a "textbook" 0.25 L/kg Vd can land 30–50% below target peak in a septic patient whose effective Vd has expanded to 0.4–0.5 L/kg — the dose was never wrong on paper, only for the patient's actual physiology.
The broader once-daily / extended-interval dosing concept — high peak, prolonged trough-free interval exploiting concentration-dependent killing and the post-antibiotic effect — still applies in the ICU. What changes is the assumed starting physiology feeding that nomogram.
A general nomogram (as used for a stable ward patient with community-acquired pyelonephritis, for example) is calibrated to typical Vd and typical renal clearance. Critical illness perturbs both inputs simultaneously and often in opposite directions: Vd expands (favoring a higher dose) while — as covered in the next stage — clearance may also increase (favoring a shorter interval or unchanged interval, but never a lower dose). Applying the general nomogram unmodified in a septic ICU patient systematically under-doses the peak.
A distinct and often under-recognized phenomenon in critical illness is augmented renal clearance (ARC): a subset of ICU patients — typically younger, with preserved cardiovascular reserve and a hyperdynamic, high-cardiac-output physiology driven by the systemic inflammatory response — filter and clear renally-eliminated drugs substantially faster than population equations (Cockcroft-Gault, MDRD, CKD-EPI) predict. Because aminoglycosides are cleared almost entirely unchanged by glomerular filtration, ARC directly threatens to leave patients underdosed between administrations, even when the initial peak was adequate.
Early sepsis and the systemic inflammatory response can produce a hyperdynamic circulatory state: increased cardiac output, reduced systemic vascular resistance, and — critically for drug clearance — increased renal blood flow and glomerular filtration rate. This is the physiologic mirror image of the acute kidney injury clinicians instinctively worry about in the ICU.
ARC is most frequently reported in younger patients with few comorbidities, trauma, burns, or early sepsis, and often coincides with the same large-volume resuscitation discussed in Stage 1 — increased intravascular volume and cardiac output both expand Vd and augment glomerular filtration at once. This is why critical-illness aminoglycoside dosing cannot be tuned along a single axis: the patient in front of you may simultaneously need a higher dose (expanded Vd) and a shorter interval or closer monitoring (augmented clearance).
Standard estimating equations for renal function were derived and validated in stable, largely non-critically-ill populations. In ARC, measured creatinine clearance (ideally from a timed urine collection) frequently exceeds 130–150 mL/min — well above what Cockcroft-Gault predicts from a normal or even low serum creatinine, because a hyperdynamic patient can generate that GFR while still producing (and clearing) creatinine efficiently.
Serum creatinine can look reassuringly normal — or even be mildly low from fluid dilution and low muscle mass — while the true glomerular filtration rate is running well above normal. A "normal" creatinine in a young septic patient is not proof of normal renal drug clearance; it can be the opposite.
Because aminoglycosides are eliminated almost entirely by glomerular filtration with minimal tubular reabsorption, drug clearance tracks renal clearance closely. When true CrCl runs 130–150+ mL/min instead of the ~80–120 mL/min a standard equation might estimate, the aminoglycoside elimination half-life shortens meaningfully.
Applied to a fixed nomogram interval (commonly 24, 36, or 48 hours based on estimated renal function), ARC can cause: (1) troughs that clear faster than expected — generally a safety margin for nephrotoxicity, but (2) more importantly, sub-therapeutic drug exposure for a larger fraction of the dosing interval if the interval is not shortened or the regimen otherwise re-optimized, undermining the pharmacodynamic target that justified using an aminoglycoside in the first place.
Because ARC is a clearance phenomenon, its impact is on trough clearance and time-dependent exposure — not on the peak itself. This is why the corrective lever for suspected ARC is interval adjustment (consider a shorter interval) rather than dose adjustment, which instead responds to Vd.
Given the expanded distribution volume characteristic of septic and critically ill patients, the initial aminoglycoside dose frequently needs to be higher — on a per-kilogram basis — than the general once-daily extended-interval nomogram would specify for a stable patient. The dosing goal is unchanged (a peak concentration well above the pathogen's MIC, since aminoglycoside killing is concentration-dependent), but the starting dose required to hit that target must account for the larger volume the drug is diluted into.
The basic loading-dose relationship, Dose = Target Cmax × Vd, makes the critical-illness adjustment mechanistic rather than empirical. If the target peak concentration is unchanged (it is set by the pathogen's MIC and the desired Cmax:MIC ratio, not by the patient's illness severity) but Vd is 50–100% higher than the general nomogram assumes, the dose required to reach that same peak must rise proportionally.
Many ICU-specific protocols therefore recommend higher weight-based starting doses for critically ill or septic patients — for example, gentamicin/tobramycin doses in the 7 mg/kg range (versus 5–7 mg/kg generally), with some patients with marked capillary leak or massive resuscitation volumes needing even more. Amikacin, with its lower per-mg potency, sees analogous escalation.
Crucially, this is a starting-dose correction layered on top of — not a replacement for — the underlying extended-interval strategy: exploit concentration-dependent killing and a prolonged post-antibiotic effect by giving one large dose per interval rather than smaller, more frequent doses.
Because aminoglycosides distribute in extracellular/lean body water rather than fat, dosing weight itself matters: actual body weight overestimates the volume in obese patients, while adjusted or ideal body weight formulas were derived outside the ICU setting. In critically ill patients, superimposing sepsis-driven Vd expansion onto an already complex weight-dosing decision is exactly why empiric higher mg/kg starting doses are paired with early confirmatory level measurement — the calculated dose is a well-reasoned starting estimate, not a guarantee, and must be verified against a measured peak.
The higher initial dose is not "more aggressive antibiotic therapy" in a general sense — it is dose correction for a diluted starting volume, aimed at hitting the exact same pharmacodynamic peak target used in stable patients. Under-dosing for Vd, not over-dosing, is the real risk critical illness introduces.
The general once-daily nomogram assumes pharmacokinetics that hold reasonably steady from dose to dose, allowing a single trough/random level and a nomogram lookup to guide subsequent interval selection. Critical illness breaks that assumption: renal function, fluid balance, and cardiac output can all shift meaningfully within 24 hours. Level monitoring in the ICU therefore needs to start earlier — often on day one rather than after 2–3 nomogram-guided doses — and repeat more frequently than in a stable outpatient or ward setting.
A general extended-interval nomogram is a population-derived shortcut: it substitutes a single creatinine-clearance-based lookup for individualized pharmacokinetic calculation, on the reasonable assumption that renal function and volume status are relatively stable over the following 24–48 hours. That assumption is frequently wrong in the ICU.
Augmented renal clearance can develop or resolve within a day. Fluid resuscitation continues to expand Vd hour by hour during early shock and then reverses during diuresis. Vasopressor titration, nephrotoxin exposure, and evolving organ dysfunction can push renal function in either direction unpredictably. A nomogram interval chosen from an admission creatinine may already be wrong by the time the second dose is due.
Early, direct measurement — rather than population estimation — closes this gap. A peak (or near-peak) level confirms whether the higher initial dose actually achieved the target Cmax:MIC ratio; an early trough or mid-interval level confirms whether clearance matches the assumed interval, catching both under-dosing (subtherapeutic exposure) and drug accumulation (nephrotoxicity/ototoxicity risk) before they compound over several doses.
Because the goal is to catch rapidly evolving pharmacokinetics rather than confirm a stable steady state, ICU-focused protocols generally favor: a level obtained within the first 24 hours of therapy (rather than waiting for multiple doses), and repeat levels every 1–2 days for the duration of therapy rather than the every-48–72-hour cadence common in stable patients. Any major clinical change — a new vasopressor requirement, large volume resuscitation, initiation of renal replacement therapy, or a rising/falling creatinine — is itself a trigger for an unscheduled level check, independent of the routine schedule.
In the ICU, a level check is not a box to tick after the nomogram interval elapses — it is the primary feedback signal correcting for pharmacokinetics that the nomogram's population assumptions cannot capture on a day-to-day basis.
Perhaps the single most important operational lesson in critically ill aminoglycoside dosing is that renal function cannot be assessed once at ICU admission and then assumed fixed for the remainder of therapy. The same patient can swing from augmented renal clearance early in resuscitation to acute kidney injury days later as septic shock, nephrotoxin exposure, or hemodynamic instability evolve — and the dosing regimen has to be continuously re-derived from current status, not the admission snapshot.
If the dosing regimen is fixed at admission and renal function is never reassessed, two opposite failure modes become likely over the course of therapy:
Failure mode 1 — undetected transition into AKI: a regimen calculated for augmented clearance (shorter interval, or an interval assuming brisk elimination) becomes dangerously frequent once creatinine rises and true clearance falls. Drug accumulates, driving nephrotoxicity and ototoxicity risk — ironically often from the aminoglycoside itself contributing to the renal injury, compounding the problem.
Failure mode 2 — undetected resolution of AKI or unrecognized ARC: a regimen extended for presumed reduced clearance (or simply left at a standard interval) under-doses a patient whose renal function has since recovered, or who was augmented-clearance from the outset but never identified as such, leaving subtherapeutic troughs and inadequate cumulative exposure against the infection.
Both failure modes are avoided the same way: renal function (ideally corroborated by measured/estimated clearance, urine output trends, and drug levels — not creatinine alone) is reassessed at least daily, and the interval and/or dose is explicitly reconsidered against that current status rather than carried forward unchanged.
"Set the interval once at admission" is the single most common conceptual error in critically ill aminoglycoside dosing. The correct mental model is a closed feedback loop: reassess renal function → reassess Vd/fluid status → re-derive dose and interval → confirm with a level → repeat, for as long as therapy continues.
The general once-daily / extended-interval aminoglycoside nomogram remains the right conceptual foundation: concentration-dependent bactericidal killing, a high single peak, and a prolonged interval exploiting the post-antibiotic effect and minimizing nephrotoxic/ototoxic trough exposure. What critical illness adds is not a different drug strategy, but a set of corrections layered on top of that foundation — a higher starting dose to compensate for expanded Vd, closer attention to the possibility of augmented renal clearance shortening the effective interval, earlier and more frequent level checks to catch rapid pharmacokinetic drift, and a standing commitment to reassess renal function continuously rather than at a single admission timepoint. None of these corrections replace the nomogram; they are the ICU-specific vigilance the nomogram alone cannot provide.