Oral Dose Entry and Gastric-Emptying Boundary Conditions

Where an Oral Dose Enters a Segmental Model

A segmental intestinal model needs a precise description of how much drug is present at the start and how much arrives afterward. These are distinct modeling choices. A dose already present in the first modeled intestinal segment belongs in that segment’s initial condition. Drug that arrives later from the stomach belongs in an input-rate term at the segment’s inlet. Treating the same dose as both an initial amount and a continuing input counts it twice; treating it as neither omits it.

Here, amounts refer to the modeled drug mass, for example milligrams of active drug, on one consistent mass basis. Let \(A_i(t)\) denote the luminal drug amount in intestinal segment \(i\), and let \(u_1(t)\) denote the rate at which drug enters the first modeled segment. Thus \(A_i\) has units of amount and \(u_1\) has units of amount per time. Any losses or transfers within the intestinal model must be represented separately in its segment balances.

The starting condition and input are independent pieces of the model specification. A general first-segment balance can be written as

The exact transfer and loss terms depend on the segmental mass balances. The important boundary-condition point is that \(u_1(t)\) represents drug crossing into the modeled intestinal system; it is not itself the amount currently in the segment. If the oral dose is directly loaded into segment 1 at time zero, set \(A_1(0)=D\) and set the dose-related part of \(u_1(t)\) to zero. If the dose is in a modeled stomach compartment, set its initial amount there and let gastric emptying generate \(u_1(t)\).

Gastric Emptying as an Inlet Boundary

There are two common bookkeeping choices. A model may represent the stomach explicitly, with a gastric amount \(G(t)\), or may prescribe the amount entering the intestine as a time-dependent boundary input. These choices can describe the same emptying process, but they should not both be applied to the same material in a way that duplicates delivery.

For an explicit stomach with first-order emptying, assume the dose is initially in the stomach, emptying is the only gastric loss, and the emptying rate constant \(k_g\) is constant and nonnegative. Then the gastric amount and intestinal input satisfy

Here \(k_g\) has units of inverse time. The solution is \(G(t)=D e^{-k_gt}\), so \(u_1(t)=k_gD e^{-k_gt}\). This is a particular emptying assumption, not a requirement for every oral-dose model. If emptying is delayed or follows a different time course, the input function must reflect that alternative assumption.

A prescribed gastric-emptying profile can instead be expressed through \(E(t)\), the cumulative fraction of the gastric dose emptied by time \(t\). For a dose \(D\) initially in the stomach, assume \(E(0)=0\), \(E(t)\) is nondecreasing, and its eventual value is 1 if the entire dose empties. If \(E\) is absolutely continuous on \([0,T]\), its ordinary derivative defines the inlet rate almost everywhere, and

This form makes the amount delivered by any time \(T\) explicit. A profile that has a lag may have \(E(t)=0\) before the lag; a profile that leaves some material permanently in the stomach has a limiting value below 1, provided that the retained amount is accounted for in the model. If the profile has jumps or other non-absolutely-continuous increases, represent those portions as amount transfers or use an input measure that accounts for the full increase in \(E\), rather than silently approximating them by an ordinary finite rate.

Why the Boundary Condition Conserves Dose

For an explicit stomach and first intestinal segment, suppose the only transfer between them is gastric emptying. Write the first-segment balance schematically as \(dA_1/dt=k_gG-L_1(t)\), where \(L_1(t)\) collects the segment’s net losses, such as onward transit and absorption. Adding this equation to the gastric balance gives

The gastric outflow disappears from the sum because it is an internal transfer: it leaves the stomach and enters segment 1 in equal amounts. With no intestinal losses, \(G(t)+A_1(t)\) remains equal to its initial value \(D\). With losses, the total amount in the stomach and intestine falls only by those losses. This cancellation is the basic mass-conservation check for an emptying boundary.

For a prescribed inlet rate and no explicit stomach state, integration of the first-segment balance shows that the dose delivered through the boundary by time \(T\) is \(\int_0^T u_1(t)\,dt\). The model must then account for any administered dose not yet delivered, for example through an explicitly tracked gastric remainder or through the stated cumulative emptying profile. Prescribing an input does not make undelivered dose disappear from the dose accounting.

Diagnostics, Alternatives, and Limitations

A useful diagnostic is a dose ledger at selected times: initial amount, cumulative inlet delivery, amount remaining in the stomach, and any explicitly modeled losses or transfers. If no loss has yet occurred, the dose fractions in those locations should sum to the assigned dose. With losses, adding cumulative loss to the amounts still represented should recover the assigned dose, subject only to numerical integration error.

Common warning signs include an intestinal amount exceeding the administered dose before any external source or mass-generating process is specified, a nonzero input with units of amount rather than amount per time, or a cumulative input that never reaches its intended total. Another frequent error is to set \(A_1(0)=D\) and also use \(u_1(t)=k_gD e^{-k_gt}\). The latter input represents an additional dose; it is not a second way of expressing the same initial condition.

An explicit stomach compartment is useful when the remaining gastric amount matters or when its emptying law must interact with other modeled processes. A prescribed input is simpler when the gastric process is represented only through its delivery profile. Direct initial loading is appropriate when the model deliberately begins with the dose already in the first intestinal segment. These are alternative representations, not claims that the physiological process is identical in every setting. A fixed input cannot respond dynamically to changes in gastric amount unless that dependence is built into the input specification.

The first-order example assumes a constant emptying rate and an exponentially declining gastric remainder. It should not be used merely because it is mathematically convenient if the intended gastric delivery profile has a lag or another shape. Likewise, a prescribed profile is only as informative as its assumptions and supporting data. The introductory material on gastric emptying and intestinal transit-time distributions provides physiological context for choosing a profile; here the modeling requirement is to translate the chosen profile into a consistent boundary condition.