Joint Titration of Dose Amount and Interval

Why Adjust Both Dose and Interval?

A dose amount and a dosing interval jointly define a regimen. Increasing the dose while keeping the interval fixed raises the amount administered per unit time. Shortening the interval while keeping the dose fixed also raises that amount per unit time. But the two changes do not produce identical concentration profiles: for a given average exposure, more frequent, smaller doses generally reduce peak–trough fluctuation.

This distinction matters when a titration objective involves more than one exposure feature. A dose-only change can raise a trough, for example, but may raise the peak more than necessary. A joint change can increase average exposure while also making administration more frequent and smoothing the profile. There is no universally best split between dose and interval: the appropriate choice depends on the exposure target, tolerability constraints, feasible schedules, and how well the pharmacokinetic model represents the drug.

Model and Steady-State Relationships

For explicit calculations, assume repeated instantaneous intravenous bolus doses in a one-compartment model with first-order elimination. Let \(D\) be the dose amount in mg, \(\tau\) the fixed interval between doses in hours, \(V\) the apparent volume of distribution in L, and \(CL\) clearance in L/h. Assume \(V\) and \(CL\) are constant, so \(k=CL/V\) is the elimination rate constant in h\(^{-1}\). The doses are identical, intervals are regular, and the system has reached steady state. There is no absorption phase, nonlinear elimination, or missed dose in this model.

Immediately after a dose at steady state, concentration is at its maximum for the interval, \(C_{\max,\mathrm{ss}}\). Immediately before the next dose it is at its minimum, \(C_{\min,\mathrm{ss}}\). Between doses, concentration declines exponentially. The instantaneous increase from one dose is \(D/V\), so accumulation from preceding doses gives:

C_{\max,\mathrm{ss}}=\frac{D/V}{1-e^{-k\tau}}, \qquad C_{\min,\mathrm{ss}}=C_{\max,\mathrm{ss}}e^{-k\tau} =\frac{D/V}{e^{k\tau}-1}.

The average steady-state concentration over one interval is \(C_{\mathrm{avg,ss}}=AUC_{\tau,\mathrm{ss}}/\tau\), where \(AUC_{\tau,\mathrm{ss}}\) is the area under the concentration curve during that interval. Integrating the exponential decline shows why average exposure depends on dose rate, while peak and trough also depend on interval.

AUC_{\tau,\mathrm{ss}} =\int_0^\tau C_{\max,\mathrm{ss}}e^{-kt}\,dt =\frac{C_{\max,\mathrm{ss}}(1-e^{-k\tau})}{k} =\frac{D}{Vk} =\frac{D}{CL}, \qquad C_{\mathrm{avg,ss}}=\frac{D}{CL\,\tau}.

The derivation uses \(k=CL/V\) and the steady-state peak formula. If bioavailability \(F\) is included for a linear model with the same simplified input assumptions, replace \(D\) in these exposure expressions by \(FD\). For clarity, the worked examples below use intravenous dosing, so \(F=1\).

Key relationship: Under this model, the dose rate \(D/\tau\) determines average steady-state concentration through \(C_{\mathrm{avg,ss}}=D/(CL\,\tau)\). The interval also affects peak and trough: shortening \(\tau\) at a fixed dose rate generally narrows peak–trough fluctuation.

Worked Example: Establishing a Baseline Profile

Suppose a hypothetical patient receives 100 mg intravenously every 12 h. Assume \(V=50\) L and \(CL=5\) L/h. Calculate the steady-state peak, trough, and average concentration. These parameters are illustrative, not data from a clinical study.

First, \(k=CL/V=5/50=0.1000\ \mathrm{h}^{-1}\), and the single-dose concentration increase is \(D/V=100/50=2.000\ \mathrm{mg/L}\). With \(e^{-1.2}=0.3011942\):

C_{\max,\mathrm{ss}} =\frac{2.000}{1-e^{-(0.1000)(12)}} =\frac{2.000}{1-0.3011942} =2.862\ \mathrm{mg/L}.

Using the unrounded exponential value in the calculation, the trough is:

C_{\min,\mathrm{ss}} =2.862026(0.3011942) =0.8620\ \mathrm{mg/L}.

The average is \(D/(CL\,\tau)=100/(5\times12)=1.667\ \mathrm{mg/L}\). Equivalently, the area in each interval is \(100/5=20.00\ \mathrm{mg\cdot h/L}\), and \(20.00/12=1.667\ \mathrm{mg/L}\). The peak and trough are not the average; they describe the extremes within the interval.

Formulating a Joint Adjustment

A useful way to specify a joint titration is to first choose the intended change in dose rate, then decide how to divide that change between dose amount and interval. Let \(q>0\) be the factor by which the dose rate should change, and let \(w\) be a chosen allocation parameter between 0 and 1. A continuous update is:

D_{j+1}=D_jq^w, \qquad \tau_{j+1}=\tau_jq^{w-1}.

Here \(j\) indexes the current regimen. The update has the intended rate effect because:

\frac{D_{j+1}}{\tau_{j+1}} =\frac{D_jq^w}{\tau_jq^{w-1}} =\frac{D_j}{\tau_j}q.

For \(0<w<1\), if \(q>1\), the dose amount increases and the interval shortens. When \(w=1\), the rule changes dose only; when \(w=0\), it changes interval only. This parameterization is a way to describe a policy, not a guarantee that the chosen regimen meets every clinical target. A dose-rate change controls average concentration only under the stated linear model; the selected interval determines how peaks and troughs respond.

Real regimens must also respect available dose strengths, permitted dosing times, minimum and maximum doses, and any interval restrictions. After translating a continuous update into a feasible schedule, calculate the exposure consequences of the actual rounded regimen rather than assuming rounding leaves the intended change unchanged.

Worked Example: Increasing Dose Rate While Smoothing the Profile

Use the baseline model above and target a 20% increase in dose rate. Consider a hypothetical joint regimen of 110 mg every 11 h. Its dose rate is \(110/11=10.00\ \mathrm{mg/h}\), compared with \(100/12=8.333\ \mathrm{mg/h}\), a factor of \(10.00/8.333=1.200\). The predicted average concentration is therefore \(10.00/5=2.000\ \mathrm{mg/L}\).

For the joint regimen, \(k\tau=1.100\), \(e^{-1.1}=0.3328711\), and \(D/V=110/50=2.200\ \mathrm{mg/L}\). Thus:

C_{\max,\mathrm{ss}} =\frac{2.200}{1-0.3328711} =3.298\ \mathrm{mg/L}, \qquad C_{\min,\mathrm{ss}} =3.2977(0.3328711) =1.098\ \mathrm{mg/L}.

Compare this with a dose-only change designed to produce the same dose rate: 120 mg every 12 h. Its average is also \(120/(5\times12)=2.000\ \mathrm{mg/L}\). With \(D/V=2.400\ \mathrm{mg/L}\) and \(e^{-1.2}=0.3011942\), its peak is \(2.400/(1-0.3011942)=3.434\ \mathrm{mg/L}\), and its trough is \(3.4344(0.3011942)=1.034\ \mathrm{mg/L}\).

The joint regimen has a lower peak and a higher trough than the dose-only regimen at the same predicted average concentration. This is the expected smoothing effect of a shorter interval in this model. It also requires a more frequent dosing schedule; the calculation alone does not establish that the schedule is preferable for a particular patient.

Targeting a Trough: Two Different Routes

A target trough does not determine a unique regimen. At fixed interval, dose is proportional to both peak and trough in this linear model. Alternatively, at fixed dose, changing the interval changes accumulation and the dose rate simultaneously. These routes may reach the same trough but imply different average exposure and peak concentration.

Worked Example: Reaching a Trough Target by Shortening the Interval

Using the baseline parameters, suppose the illustrative trough target is \(1.200\ \mathrm{mg/L}\). Keep the dose at 100 mg and solve for the interval that produces this trough. From \(C_{\min,\mathrm{ss}}=(D/V)/(e^{k\tau}-1)\):

1.200=\frac{2.000}{e^{0.1000\tau}-1}, \qquad e^{0.1000\tau}=1+\frac{2.000}{1.200}=2.6667, \qquad \tau=\frac{\ln(2.6667)}{0.1000}=9.808\ \mathrm{h}.

The peak is the trough plus the concentration increase from the 100 mg dose: \(1.200+100/50=3.200\ \mathrm{mg/L}\). The average is \(100/(5\times9.808)=2.039\ \mathrm{mg/L}\). The dose rate has increased because the same dose is given more frequently.

For comparison, keep the interval at 12 h and increase dose to reach the same trough. Since trough is proportional to dose at fixed interval, the required dose is \(100(1.200/0.862026)=139.2\ \mathrm{mg}\). Its peak is the target trough plus the dose increment, \(1.200+139.2/50=3.984\ \mathrm{mg/L}\), and its average is \(139.2/(5\times12)=2.320\ \mathrm{mg/L}\). Under these assumptions, shortening the interval with dose fixed reaches the same trough with lower predicted average concentration and peak than this dose-only route. The comparison is model-dependent and is not, by itself, a clinical dosing recommendation.

Assumptions, Diagnostics, and Alternatives

The formulas and comparisons above rely on linear, time-invariant pharmacokinetics, constant clearance and volume, regular administration, and steady state. They also assume instantaneous input and a one-compartment concentration decline. If absorption is slow, distribution is multi-compartmental, clearance changes with concentration or time, or the regimen has not reached steady state, these equations may not predict the relevant peak, trough, or average accurately. For such cases, use the appropriate pharmacokinetic model and simulate the candidate regimens under the intended conditions; do not apply the bolus equations as though their assumptions were satisfied.

Useful checks include verifying that the implemented dose and interval produce the intended \(D/\tau\), confirming that peaks and troughs are calculated at the correct points in the interval, and checking that the predicted concentration profile is consistent with the assumed elimination rate. Unexpectedly large changes in observed concentrations, poor fit to the assumed decay, or systematic disagreement between predicted and measured profiles suggest that the model or its parameters may be inadequate. Measurement timing, missed doses, and departure from steady state can also explain discrepancies, so a mismatch should prompt investigation rather than automatic further titration.

Approach	What it directly changes	Main exposure consequence under this model	Important limitation
Dose-only adjustment	Dose amount at fixed interval	Average, peak, and trough scale with dose	May raise peak substantially when the goal is primarily to raise trough
Interval-only adjustment	Time between fixed-size doses	Changes dose rate and fluctuation together	May create an impractical schedule or unintended average exposure
Joint dose-and-interval adjustment	Both dose amount and interval	Can set dose rate while changing fluctuation	Requires explicit constraints and model-based checking; it is not uniquely determined by one target

A frequent pitfall is to reason, “The average concentration should rise by 20%, so increasing the dose by 20% is the only appropriate adjustment.” That conclusion confuses the required dose-rate change with the choice of regimen. Increasing dose by 20% at fixed interval does produce a 20% average increase under this model, but a joint change can produce the same dose rate with a different peak–trough profile. Conversely, shortening the interval without accounting for the resulting dose rate may increase average exposure more than intended. Always check both \(D/\tau\) and the profile features relevant to the objective.

Key takeaway: Joint titration is a deliberate choice about both average dose rate and concentration fluctuation. State the target, specify the constraints and model assumptions, calculate the resulting profile, and reassess the actual regimen after practical rounding.

Check Your Understanding

Answer each question, then compare your reasoning with the solutions below.

In the repeated-bolus model, what determines average steady-state concentration, and what additional role does \(\tau\) play?
A regimen changes from 80 mg every 8 h to 120 mg every 12 h. Does its dose rate change? Explain.
Why can two regimens with the same average concentration have different peaks and troughs?
What important model assumption fails if drug absorption is slow rather than instantaneous?

Solutions. (1) Average concentration is \(D/(CL\,\tau)\), so dose rate and clearance determine it. The interval also affects accumulation and peak–trough fluctuation. (2) Both rates are \(10.00\ \mathrm{mg/h}\): \(80/8=10.00\) and \(120/12=10.00\). With unchanged clearance, the model predicts the same average concentration, although the profiles can differ because the intervals differ. (3) The average is determined by dose per unit time under the linear model, while fluctuation depends on the interval relative to the elimination rate; more frequent dosing generally smooths the profile. (4) The instantaneous-input assumption fails. A model that represents absorption is needed to predict the resulting concentration-time profile appropriately.