Interpreting Concentrations Before Steady State

Why Dose History Matters

A concentration result is shaped not just by the most recent dose, but also by drug remaining from earlier doses. After treatment starts or a regimen changes, that residual drug is still accumulating toward a repeating pattern. A concentration obtained during this transition can therefore differ from the value expected at steady state, even when the dose, sampling time, and assay result are all correct.

This matters in therapeutic drug monitoring because many targets are defined for a particular sampling time at steady state. A concentration below such a target soon after treatment begins does not, by itself, show that the maintenance regimen is inadequate. The appropriate first question is whether the target applies to the patient's current dosing history and sampling phase.

Definition: A regimen is at steady state when the concentration-time pattern over a dosing interval repeats from one interval to the next. A concentration is pre-steady-state when it is measured before that repeating pattern has been reached closely enough for the intended decision. In practice, “close enough” depends on a stated tolerance and the clinical purpose; steady state is approached asymptotically, not reached at one exact dose.

A Simple Accumulation Model

To make the accumulation calculation explicit, consider a hypothetical one-compartment model with linear elimination and instantaneous intravenous doses. Let \(D\) be the same dose at each administration, \(V\) the apparent volume of distribution, \(k\) the elimination rate constant, and \(\tau\) the fixed interval between doses. The dose produces an immediate concentration increment \(D/V\), and between doses concentration declines exponentially. The half-life is \(t_{1/2}=\ln(2)/k\).

This model assumes the dose history and sample time are known; the interval and dose have been regular; \(k\) and \(V\) are constant; and concentration contributions from separate doses add. It is an illustrative model, not a universal description of every drug or route. In particular, it does not represent an absorption phase.

Model: Write \(q=e^{-k\tau}\) for the fraction of concentration remaining after one complete dosing interval. For \(n\) equal doses given at regular intervals, the concentration \(t\) time units after the most recent dose, where \(0\le t<\tau\), is

C_n(t)=\frac{D}{V}e^{-kt}\sum_{r=0}^{n-1}q^r =\frac{D}{V}e^{-kt}\frac{1-q^n}{1-q}.

At the same time \(t\) in the interval, the steady-state prediction is

C_{\mathrm{ss}}(t)=\frac{D}{V}e^{-kt}\frac{1}{1-q}.

Here \(r=0\) is the most recent dose; \(r=1\) is the preceding dose, whose contribution has undergone one full interval of elimination. The geometric sum accounts for each dose's remaining contribution at the sample time. A trough immediately before the next scheduled dose is represented by taking \(t\) toward \(\tau\), or equivalently by summing the dose contributions after they have each undergone at least one interval of elimination.

Why the Fraction of Steady State Has a Simple Form

At a fixed phase of the dosing interval, the pre-steady-state and steady-state expressions contain the same dose, elimination, and sampling-time factors. Their ratio therefore depends only on how many doses have contributed and how much remains from one interval to the next.

\frac{C_n(t)}{C_{\mathrm{ss}}(t)} = \frac{\frac{D}{V}e^{-kt}(1-q^n)/(1-q)} {\frac{D}{V}e^{-kt}/(1-q)} =1-q^n.

This derivation shows why a low early concentration can be expected even when the regimen would eventually produce a higher concentration. It also gives a practical measure of proximity to steady state under the model: the fractional shortfall is \(q^n\). If a tolerance of 5% is selected, the criterion is \(q^n\le0.05\), not a universal number of days or doses.

Worked Example: A Trough After Three Doses

Suppose a hypothetical patient receives equal instantaneous doses every 12 hours. The model half-life is 8 hours, and each dose produces an immediate increment \(D/V=4.000\ \mathrm{mg/L}\). Calculate the trough just before the fourth dose and compare it with the model's steady-state trough.

The interval retention factor is \(q=e^{-k\tau}=2^{-\tau/t_{1/2}}=2^{-12/8}=0.353553\). Three doses have been given. At the trough, the three dose contributions have undergone one, two, and three intervals of elimination:

C_{\mathrm{trough},3} =4.000(q+q^2+q^3) =4.000(0.353553+0.125000+0.044194) \approx2.091\ \mathrm{mg/L}.

The steady-state trough predicted by the same model is

C_{\mathrm{ss,trough}} =4.000\frac{q}{1-q} =4.000\frac{0.353553}{1-0.353553} \approx2.188\ \mathrm{mg/L}.

The exact model-based ratio is \(1-q^3=1-0.353553^3=0.955806\), so the three-dose trough is about 95.58% of the steady-state trough. The result is consistent with the calculated concentrations: \(2.091/2.188\) is approximately 0.9558, allowing for rounding. If a clinical target were defined for the steady-state trough, this calculation would indicate that the modeled concentration is close to, but not exactly at, that state. Whether this difference is acceptable depends on the chosen tolerance and clinical decision.

When the Regimen Has Changed

After a dose change, the concentration trajectory contains residual contributions from the old regimen as well as contributions from new doses. A shortcut that treats the patient as if the new dose had been used from the beginning discards that history. Under linear elimination, the general principle is to sum the contribution from every dose actually administered, with each contribution decayed from its administration time to the sample time.

C(t)=\sum_{j:\,t_j\le t}\frac{D_j}{V}e^{-k(t-t_j)}.

In this expression, \(D_j\) is the \(j\)th dose and \(t_j\) its administration time. A known concentration already present at the time of a regimen change can also be carried forward by exponential decay, provided the same elimination model remains applicable.

Worked Example: Interpreting the First Trough After a Dose Increase

Consider a hypothetical patient at steady state on a regimen whose dose increment is \(2.000\ \mathrm{mg/L}\). The dosing interval is 12 hours, and the half-life is also 12 hours, so \(q=0.5\). At the scheduled dose time, the dose is increased so that the new increment is \(4.000\ \mathrm{mg/L}\). What concentration is expected immediately before the second dose on the new regimen?

First calculate the old steady-state trough. The old trough is the residual from earlier doses, just before the next scheduled dose:

C_{\mathrm{old,trough}} =2.000\frac{q}{1-q} =2.000\frac{0.5}{0.5} =2.000\ \mathrm{mg/L}.

At the change time, this \(2.000\ \mathrm{mg/L}\) is present immediately before the new dose. The new dose adds \(4.000\ \mathrm{mg/L}\), giving \(6.000\ \mathrm{mg/L}\) immediately afterward. Over the next 12 hours, half remains:

C(12\ \mathrm{h})=(2.000+4.000)e^{-k(12)} =6.000(0.5) =3.000\ \mathrm{mg/L}.

The new regimen's steady-state trough would be \(4.000q/(1-q)=4.000\ \mathrm{mg/L}\). Thus this first trough is 75.00% of the new steady-state prediction. It is not appropriate to interpret the \(3.000\ \mathrm{mg/L}\) as though the new regimen had already reached its steady-state pattern. The calculation also depends on the stated starting condition: the patient was at the old regimen's steady state immediately before the dose change.

Matching the Target to the Sampling Phase

A concentration and a target are comparable only if their definitions match. A trough target calls for a sample taken at the appropriate pre-dose time; a postdose target calls for its specified time after a dose. Even then, a target defined at steady state does not automatically apply before steady state. If a time-matched model is credible, it can show what concentration is expected during the transition and how that value relates to the eventual repeating pattern.

Worked Example: A Postdose Target Before Steady State

Suppose a hypothetical drug follows the simple model above, with \(t_{1/2}=24\) hours, \(\tau=12\) hours, and \(D/V=5.000\ \mathrm{mg/L}\). A sample is collected 4 hours after the second dose. Calculate its predicted concentration and compare it with a hypothetical steady-state target of 12.0–16.0 mg/L defined specifically at 4 hours after a dose.

The interval retention factor is \(q=2^{-12/24}=0.707107\), and the 4-hour decay factor is \(e^{-k(4)}=2^{-4/24}=0.890899\). The most recent dose has decayed for 4 hours; the preceding dose has decayed for 16 hours, or \(4+12\) hours. Therefore,

C_2(4) =5.000(0.890899)(1+0.707107) \approx7.604\ \mathrm{mg/L}.

The steady-state prediction at that same phase is

C_{\mathrm{ss}}(4) =\frac{5.000(0.890899)}{1-0.707107} \approx15.21\ \mathrm{mg/L}.

The ratio is \(1-q^2=1-0.707107^2=0.5000\): after two doses, this model predicts half the steady-state concentration at the matched sampling time. Although \(7.604\ \mathrm{mg/L}\) is below the stated steady-state target range, that comparison alone does not establish that the eventual regimen will miss its target. It establishes that the current modeled concentration is below a range intended for a different dosing state.

Diagnostics, Alternatives, and Limitations

Before interpreting a pre-steady-state result, check whether the calculation's inputs describe what actually happened. Confirm the dose amounts and administration times, the sample time relative to the most recent dose, and whether the assumed dosing interval was followed. Check that the proposed elimination rate and dose-response model are appropriate for the decision. If those inputs are uncertain, the numerical prediction may be more precise than the evidence supports.

Incorrect dose history: The equal-dose geometric-series formula requires equal, regularly spaced doses. When dose amounts or times differ, use the individual-dose sum only if the actual history is known.
Wrong sampling phase: A concentration taken several hours after a dose is not a trough. Compare a measurement only with a target defined for the same phase, and record the actual sample time.
Unjustified steady-state assumption: A fixed number of days does not guarantee that a drug is close enough to steady state. The relevant accumulation depends on the elimination rate and the dosing interval.
Model mismatch: Changes in clearance, nonlinearity, or a route with a material absorption phase can invalidate this simple calculation. A more appropriate model or additional information may be needed.

The main alternatives are to defer a steady-state comparison until a suitable sampling occasion, or to compare the observation with a model prediction for the patient's actual dose history and sample time. The first is simpler but may delay a decision; the second can use information earlier, but depends on the model and its inputs. If a patient has an urgent clinical problem, a steady-state target should not be used as a substitute for clinical assessment.

A frequent mistake is to see a pre-steady-state concentration below the steady-state target and conclude, “the dose is too low.” That reasoning compares different states. The correction is to establish whether the target is intended for the current state, quantify expected accumulation when the model is credible, and interpret the measured value in the context of the clinical decision.

Key takeaway: Before applying a steady-state target, establish the dose history, sampling phase, and degree of accumulation. A pre-steady-state concentration is not automatically a failed target; it is an observation made during a changing concentration-time pattern.

Check Your Understanding

Answer each question using the assumptions stated. Solutions follow the questions.

A drug has \(q=0.5\). Under the equal-dose model, what fraction of its same-phase steady-state concentration is expected after four doses?
A steady-state trough target is defined just before a dose. Why is a sample taken 3 hours after dosing not directly comparable with that trough target?
A patient has received two equal doses at regular intervals, with \(q=0.8\). What fraction of the same-phase steady-state concentration is predicted after those two doses?
Under the simple model, what information is needed to predict the concentration after a dose change?

Solutions. (1) The fraction is \(1-q^n=1-0.5^4=0.9375\), or 93.75%. (2) The sample is at a different phase of the concentration-time pattern; it is not a trough, so the values do not represent the same metric. (3) The fraction is \(1-0.8^2=0.3600\), or 36.00%. (4) The relevant dose amounts and administration times, sample time, and applicable model parameters are needed; for this model, those include \(V\) and \(k\). Each dose's contribution must be decayed from its administration time to the sample time.