Continuous Renal Replacement Therapy and Drug Clearance

Why CRRT Requires Its Own Clearance Estimate

Continuous renal replacement therapy (CRRT) removes drug over many hours, but its contribution to total drug clearance depends on the treatment actually delivered and on the drug’s ability to cross the filter. A prescribed effluent rate alone is not a drug clearance: two drugs exposed to the same circuit can have different clearances because their sieving or saturation coefficients differ. Residual kidney function also remains relevant, as does clearance by nonrenal pathways.

The practical calculation is therefore to estimate extracorporeal clearance from the delivered flows and drug-specific coefficients, add residual renal and nonrenal clearance when these are available on a compatible basis, and then calculate the maintenance input needed for the exposure target. This approach is an estimate, not a substitute for evaluating the patient’s clinical response, drug concentrations when useful, or the reliability of the underlying assumptions.

From Effluent Flow to Extracorporeal Clearance

In continuous venovenous hemofiltration (CVVH), drug is removed mainly with ultrafiltrate. If \(Q_{\mathrm{conv}}\) is the convective ultrafiltration flow and \(SC\) is the applicable sieving coefficient, the estimated convective clearance is \(Q_{\mathrm{conv}}SC\). In continuous venovenous hemodialysis (CVVHD), drug is transferred into dialysate, so dialytic clearance is estimated as \(Q_{\mathrm{dial}}SA\). Here, \(Q_{\mathrm{dial}}\) is the dialysate flow rate.

For continuous venovenous hemodiafiltration (CVVHDF), both routes may contribute. A practical flow-based estimate is the sum of the dialytic and convective components, provided that the coefficients correspond to those components and treatment conditions:

Flow rates must be expressed in volume per time to obtain clearance in the same volume per time. For example, a flow of 30 mL/kg/h in a 70 kg patient corresponds to \(30 \times 70 = 2{,}100\) mL/h, or 2.1 L/h. The convective flow is the total ultrafiltration across the membrane, not simply the net fluid removed from the patient. Replacement-fluid settings affect how this flow is implemented; confirm how the machine reports the relevant effluent components before applying the calculation.

In practice, effluent flow is often used as a convenient summary of delivered treatment intensity. If a single effective effluent-to-plasma concentration ratio is available, the corresponding estimate is \(CL_{\mathrm{CRRT}} \approx Q_{\mathrm{eff}} \times\) that ratio. Do not multiply total effluent by separate coefficients and then also add component clearances: that would count removal twice. When component flows are used, add the component clearances once.

Combining CRRT With Other Clearance Pathways

When clearance pathways operate in parallel and the drug follows linear pharmacokinetics over the relevant concentration range, their clearances can be added:

Here, \(CL_{\mathrm{RK}}\) is residual kidney clearance of the drug, and \(CL_{\mathrm{nonrenal}}\) represents clearance by other pathways. These terms must refer to the same drug and compatible units. An estimated glomerular filtration rate or creatinine clearance is not automatically equal to the drug’s renal clearance. Use a drug-specific relationship, an appropriate measured clearance, or a justified model estimate rather than adding a kidney-function marker directly as though it were drug clearance.

A useful justification for the addition is that, at a given plasma concentration, each independent pathway removes drug at a rate equal to its clearance multiplied by that concentration. The total elimination rate is therefore the sum of the pathway-specific rates. Dividing that total rate by the common plasma concentration gives the sum of the clearances. This assumes the pathways are adequately represented as independent, approximately linear processes over the concentration range of interest.

Calculating a Maintenance Regimen

For a drug with linear pharmacokinetics and stable clearance, the average steady-state concentration is determined by the systemic input rate divided by total clearance. For a repeated dose \(D\) given every \(\tau\) hours, with bioavailability \(F\), the average systemic input rate is \(FD/\tau\). Equating average input and average elimination gives:

This equation calculates a maintenance dose for a target average steady-state concentration \(C_{\mathrm{ss,avg}}\). For intravenous administration, \(F=1\). It assumes that the target is an average concentration, the dose and clearance are expressed on compatible bases, and clearance does not change substantially over the dosing interval. It does not calculate a loading dose, which depends primarily on the target concentration and volume of distribution.

Delivered Treatment, Diagnostics, and Failure Modes

CRRT clearance can change when the treatment changes or is interrupted. Consequently, a calculation based only on the prescribed settings may overstate the clearance delivered over a dosing interval. Review the actual machine record, including downtime, changes in dialysate or replacement rates, filter performance, and fluid balance. If the delivered intensity varies, a time-averaged estimate may be more appropriate than treating a single prescribed flow as constant.

A rate given in mL/kg/h should be converted using the relevant body weight and then converted to L/h before calculating clearance. For instance, 25 mL/kg/h in a 72 kg patient is \(25 \times 72 = 1{,}800\) mL/h, or 1.8 L/h. Mixing mL/h with L/h in the clearance equation creates a thousand-fold error.

Flow-based estimates are also less reliable when the coefficients are uncertain or poorly matched to the patient’s conditions. Drug transfer may be limited by protein binding, molecular size, membrane characteristics, or changes in the circuit. Predilution replacement can lower the drug concentration entering the filter relative to systemic plasma; an effective coefficient referenced to systemic plasma should reflect that dilution. Do not apply a separate dilution correction if the coefficient already incorporates it. Check the basis of the coefficient before using it.

A useful diagnostic is to compare the estimated clearance and predicted exposure with available concentration measurements or the observed clinical response, while accounting for sampling time and the expected concentration profile. A large mismatch can indicate inaccurate flow data, inappropriate coefficients, changing residual kidney function, nonlinearity, or an incorrect estimate of nonrenal clearance. It is a reason to revisit the assumptions, not to assume that the machine’s prescribed effluent rate is the drug’s true clearance.

Choosing and Interpreting the Estimate

The flow-and-coefficient method is useful when treatment settings and an appropriate drug-specific coefficient are available. A measured extracorporeal extraction or clearance may be preferable when reliable samples can be obtained and interpreted. A population pharmacokinetic model can incorporate additional patient and treatment information, but its predictions depend on the model’s applicability and inputs. A kidney-function-only dose adjustment is less informative when CRRT is a substantial clearance pathway, because it omits extracorporeal removal.

No single approach removes uncertainty. CRRT settings can vary, coefficients may not be known for every drug, and the usual assumption of linear, stable clearance may fail. In particular, this calculation does not cover concentration-dependent elimination, major changes in distribution, or a rapidly changing clinical state. Use it as a transparent estimate within a broader dosing assessment, and update it when the delivered treatment or the patient’s residual kidney function changes.