Constant-Rate IV Infusion Models

Why infusion rate, clearance, and volume matter

A constant-rate intravenous infusion supplies drug continuously while elimination removes it. In a one-compartment model, the amount of drug in the body therefore reflects the balance between a constant input and an amount-dependent output. The resulting concentration rises toward a limiting value rather than increasing without bound. That limiting concentration depends on the infusion rate and clearance; the speed of approach also depends on volume.

This tutorial considers a constant infusion rate while the infusion is being delivered. It focuses on the concentration profile and its steady-state value, not on handling changes in infusion input. The model is useful as a transparent structural description and as a way to relate dosing rate to exposure, provided its assumptions are plausible.

Definition: In a one-compartment constant-rate IV infusion model, drug enters the compartment at rate \(R_0\) (amount per unit time), and drug is eliminated with first-order rate constant \(k=CL/V\). Here \(CL\) is clearance (volume per unit time), \(V\) is the apparent volume of distribution (volume), and \(C(t)\) is the compartment concentration at time \(t\). The model assumes the compartment is well mixed, the infusion rate is constant, and clearance and volume are constant over the concentration range considered.

Deriving the concentration profile

Let \(A(t)\) be the amount of drug in the compartment. The infusion supplies \(R_0\) amount per unit time. Under first-order elimination, the elimination rate is \(CL\,C(t)\), an amount per unit time, because \(CL\) has units of volume per time and concentration has units of amount per volume. With \(A(t)=V C(t)\), the amount balance is

\frac{dA(t)}{dt}=R_0-CL\,C(t)

If \(V\) is constant, dividing through by \(V\) gives a differential equation directly for concentration:

\frac{dC(t)}{dt}=\frac{R_0}{V}-\frac{CL}{V}C(t) =\frac{R_0}{V}-kC(t)

The steady-state concentration \(C_{ss}\) is the value at which the concentration no longer changes, so \(dC/dt=0\). Setting the right-hand side to zero yields \(R_0/V-kC_{ss}=0\). Since \(k=CL/V\), this gives \(C_{ss}=R_0/CL\). To obtain the full profile, subtract this steady-state value from the differential equation. The difference \(C(t)-C_{ss}\) decays exponentially at rate \(k\), giving the general solution for an infusion at constant rate:

C(t)=C_{ss}+\left(C_0-C_{ss}\right)e^{-kt} =C_0e^{-kt}+\frac{R_0}{CL}\left(1-e^{-kt}\right)

Here \(C_0\) is the concentration at the time origin used for the calculation. If the compartment is drug-free at that time, \(C_0=0\), and the equation simplifies to

C(t)=\frac{R_0}{CL}\left(1-e^{-(CL/V)t}\right) =C_{ss}\left(1-e^{-kt}\right)

Key result: For a constant-rate infusion in this model, \(C_{ss}=R_0/CL\). The concentration approaches that value exponentially, with rate constant \(k=CL/V\). At any finite time, the fraction of the rise from \(C_0\) to \(C_{ss}\) that has occurred is \(1-e^{-kt}\).

This derivation assumes a linear, amount-proportional elimination rate and constant model parameters. It does not require that concentration already be at steady state. Under the model, steady state is the limiting concentration as time becomes large; a concentration is not exactly equal to that limit at a finite time unless the initial concentration was already at that value.

Worked calculations

Worked Example: Concentration During an Infusion From a Drug-Free State

Suppose a hypothetical patient receives a constant infusion at \(R_0=20\) mg/h. Assume \(CL=4\) L/h and \(V=40\) L, with \(C_0=0\). Calculate the concentration after 3 h and the steady-state concentration.

First, \(k=CL/V=(4\ \text{L/h})/(40\ \text{L})=0.100\ \text{h}^{-1}\). The steady-state value is \(C_{ss}=R_0/CL=(20\ \text{mg/h})/(4\ \text{L/h})=5.00\) mg/L. After 3 h,

C(3)=5.00\left(1-e^{-(0.100)(3)}\right) =5.00(1-e^{-0.300}) =5.00(1-0.7408) =1.296\ \text{mg/L}

Thus the predicted concentration is approximately 1.30 mg/L after 3 h, compared with a model steady-state concentration of 5.00 mg/L. The units are consistent: \(kt\) is dimensionless, and \(R_0/CL\) has units mg/L.

Worked Example: Selecting a Rate for a Target Steady-State Concentration

For a hypothetical drug with \(CL=6.0\) L/h, what constant infusion rate corresponds to a target steady-state concentration of 2.5 mg/L? If \(V=30\) L, how long does the model predict it takes to reach 75% of that concentration from \(C_0=0\)?

Rearrange \(C_{ss}=R_0/CL\) to obtain \(R_0=CL\,C_{ss}\). Substitution gives \(R_0=(6.0\ \text{L/h})(2.5\ \text{mg/L})=15\) mg/h. The rate has units of amount per time, as required.

Next, \(k=CL/V=(6.0\ \text{L/h})/(30\ \text{L})=0.200\ \text{h}^{-1}\). From a drug-free state, the fraction of steady state reached at time \(t\) is \(1-e^{-kt}\). Set this equal to 0.75 and solve:

0.75=1-e^{-kt} \quad\Longrightarrow\quad e^{-kt}=0.25 \quad\Longrightarrow\quad t=\frac{-\ln(0.25)}{0.200}=6.931\ \text{h}

The model predicts about 6.93 h to reach 75% of the target concentration, or 1.875 mg/L. This time depends on \(k\), not on the numerical target concentration: the target determines the rate, while \(CL/V\) determines the fraction-of-steady-state time course.

Worked Example: Accounting for a Nonzero Initial Concentration

Consider a hypothetical ongoing constant infusion with \(R_0=15\) mg/h, \(CL=5.0\) L/h, and \(V=25\) L. At the chosen time origin, the concentration is \(C_0=1.0\) mg/L. Calculate the concentration 4 h later.

The elimination rate constant is \(k=(5.0\ \text{L/h})/(25\ \text{L})=0.200\ \text{h}^{-1}\), and \(C_{ss}=(15\ \text{mg/h})/(5.0\ \text{L/h})=3.0\) mg/L. Use the general solution, since the initial concentration is not zero:

C(4)=C_0e^{-kt}+C_{ss}(1-e^{-kt})

Substituting the values gives \(C(4)=(1.0)e^{-(0.200)(4)}+(3.0)(1-e^{-(0.200)(4)})\) mg/L. Since \(e^{-0.800}=0.4493\),

C(4)=(1.0)(0.4493)+(3.0)(1-0.4493) =0.4493+1.652 =2.101\ \text{mg/L}

The predicted concentration is 2.10 mg/L. It lies between the initial concentration of 1.0 mg/L and the steady-state concentration of 3.0 mg/L, as expected for this constant-rate infusion and parameter set.

Interpreting the parameters and checking the model

The steady-state equation separates two roles. Increasing \(R_0\) increases \(C_{ss}\) in direct proportion, while increasing \(CL\) decreases it in inverse proportion. Volume does not appear in \(C_{ss}\): at steady state, input rate and elimination rate balance, and clearance determines the concentration needed for elimination to match the input. Volume does affect the time course through \(k=CL/V\). For fixed clearance, a larger volume means a smaller \(k\) and a slower approach to steady state.

The exponential form also provides a useful check. Starting below \(C_{ss}\), the predicted concentration increases monotonically toward it; it does not overshoot. The proportion of the remaining difference from steady state is \(e^{-kt}\). For example, at \(t=1/k\), the fraction of the rise completed is \(1-e^{-1}\), approximately 0.632. This is a model-based description, not a claim that every observed concentration must follow the curve exactly.

In practice, check that the recorded infusion rate and concentration units agree, that \(CL\) and \(V\) refer to compatible units and the same modeled subject or population quantity, and that the concentration is interpreted in the model's compartment context. A rate reported in mg/min cannot be used as mg/h without conversion. Likewise, \(R_0/CL\) cannot be reported as an amount: its units are concentration.

Check the input and units.
Confirm that the input is a constant amount-per-time rate and that \(R_0/CL\) produces concentration units.

Check the structural assumptions.
Ask whether a single well-mixed compartment and first-order elimination are adequate over the concentrations and times of interest.

Compare predictions with observations.
Systematic curvature, a persistent mismatch in the apparent plateau, or different concentration patterns across time can indicate that the assumed structure or parameter values are inadequate.

A one-compartment model may be a poor description when distribution between central and peripheral spaces creates distinct concentration phases, when clearance changes materially with concentration or time, or when the actual input does not match the assumed rate. The model's \(C(t)\) is the concentration assigned to its compartment; it is not automatically a reliable prediction for every sampled matrix, tissue, or sampling time. A curve that fits one interval also does not establish that the assumptions hold over a wider range.

How this model differs from alternatives

The one-compartment IV bolus model represents input as an instantaneous dose, whereas a constant-rate infusion represents input spread continuously over time. The repeated IV bolus model represents successive instantaneous inputs. These are different input histories, so their concentration profiles differ even if the same one-compartment elimination parameters are used. The infusion model's steady-state relationship \(R_0/CL\) is especially direct because its continuous input rate can be balanced against the clearance-mediated elimination rate.

A more complex distribution model may be needed if one concentration compartment cannot describe the observed profile. That alternative adds structural parameters and may better represent multiple distribution processes, but it also requires data that can inform the additional structure. Conversely, a one-compartment model is parsimonious and interpretable when its predictions are adequate. Model choice should be guided by the design, observed concentration-time patterns, and intended predictions, rather than by complexity alone.

Common mistakes and limitations

Using dose instead of rate. The steady-state formula uses \(R_0\), an amount per time, not the total infused amount. A total amount must be divided by its infusion duration to obtain a rate.
Putting volume in the steady-state formula. \(R_0/V\) is not \(C_{ss}\). Volume affects the rate of approach; \(C_{ss}=R_0/CL\).
Calling a finite-time value steady state. The predicted profile approaches \(C_{ss}\) asymptotically. At a finite time from a different initial concentration, it remains separated from \(C_{ss}\) by \((C_0-C_{ss})e^{-kt}\).
Ignoring the initial condition. The simplified drug-free formula is valid only when \(C_0=0\) at the time origin. Otherwise use the general solution.
Treating a structural prediction as guaranteed. Parameter uncertainty, measurement error, input records, and structural misspecification can all make observed concentrations differ from the curve.

Key takeaway: Under constant-rate input and linear one-compartment elimination, \(C_{ss}=R_0/CL\), while \(k=CL/V\) governs how quickly the concentration approaches that limit. The equations are useful only to the extent that their input and structural assumptions describe the situation being analyzed.

Check Your Understanding

Answer each question, then compare with the solutions below.

A hypothetical infusion runs at 12 mg/h with \(CL=3.0\) L/h. What is the model steady-state concentration?
For \(CL=4.0\) L/h and \(V=20\) L, calculate \(k\). What fraction of the rise from a drug-free state is completed at \(t=5\) h?
Why does increasing \(V\) while holding \(CL\) and \(R_0\) fixed change the time course but not the steady-state concentration?
A learner uses \(C(t)=C_{ss}(1-e^{-kt})\) even though the concentration at the time origin is 2 mg/L. What is wrong, and which equation should be used?
Give one observation or assumption violation that could make this one-compartment prediction unreliable.

Solutions. (1) \(C_{ss}=R_0/CL=(12\ \text{mg/h})/(3.0\ \text{L/h})=4.0\) mg/L. (2) \(k=(4.0\ \text{L/h})/(20\ \text{L})=0.200\ \text{h}^{-1}\). The completed fraction is \(1-e^{-(0.200)(5)}=1-e^{-1}=0.6321\), or 63.2%. (3) \(C_{ss}=R_0/CL\) contains no volume term, but \(k=CL/V\) decreases as \(V\) increases, making the approach slower. (4) The simplified equation assumes \(C_0=0\). Use \(C(t)=C_0e^{-kt}+C_{ss}(1-e^{-kt})\). (5) Examples include systematic curvature inconsistent with a single exponential, concentration-dependent clearance, or an input rate that was not actually constant.