Administration-Time Constraints and Regimen Feasibility

Why Schedule Feasibility Needs Its Own Check

A regimen can satisfy a daily dose limit and a frequency limit yet still be impossible to carry out as written. A clinic may administer doses only during specified hours; an interval between administrations may need to fall within a permitted range; and an infusion must fit within the time a facility or device is available. These are schedule constraints. They determine whether the proposed times and durations can be implemented, not whether the resulting exposure or response meets a pharmacologic target.

The practical task is to translate each operational rule into a condition on administration times or infusion intervals, then check all conditions together. A candidate is feasible only if there is a set of times and durations that satisfies every applicable condition. A schedule that violates even one hard constraint is infeasible, even if it performs well on a separate dose or exposure objective.

Unless otherwise stated, time windows in this tutorial are closed: an administration at either endpoint is permitted. Times are measured on a continuous clock, with units stated explicitly. If a real protocol excludes an endpoint, use an open inequality instead. Also specify whether an interval is measured from the start or end of an administration; examples below measure dose intervals from one administration start to the next.

Representing Time Windows and Interval Constraints

Write a permitted window for administration \(i\) as \([a_i,b_i]\). A proposed administration time \(t_i\) is allowed when \(a_i\leq t_i\leq b_i\). If the time between consecutive administrations must be between \(L\) and \(U\), inclusive, then \(L\leq t_2-t_1\leq U\). These conditions must hold simultaneously: a gap that is acceptable by itself does not make a schedule feasible if either administration falls outside its permitted window.

For two windows on the same day, one can check whether any pair of times can achieve a required interval without prematurely assuming which times will be selected. Consider all pairwise differences \(t_2-t_1\), where \(t_1\in[a_1,b_1]\) and \(t_2\in[a_2,b_2]\). The smallest such difference is \(a_2-b_1\), and the largest is \(b_2-a_1\). Thus the full set of differences over all pairs is \([a_2-b_1,b_2-a_1]\). But that range includes differences from pairs in which the second administration is earlier than the first. If the second administration must be later on the same day, only positive differences are eligible.

This is an existence test, not a guarantee that every pair of permitted times works. If it passes, a specific pair still has to be selected and checked. If the regimen fixes exact times rather than allowing a choice within windows, check those times directly against each constraint.

When windows overlap, the ordering check matters. For example, if both administrations are allowed from 08:00 to 10:00, the pairwise difference range is \([-2,2]\) hours. That does not mean a later administration can be separated by a negative interval. For a strictly later second administration, the eligible differences are \((0,2]\): a positive gap up to two hours can be achieved, but zero and negative gaps cannot. The general intersection test handles this case without requiring the windows to be ordered in advance.

Repeating Schedules and Midnight

A daily regimen repeats its clock-time windows on successive days. To check an interval that crosses midnight, represent times on an unwrapped timeline rather than subtracting clock readings as if the day did not change. If an administration occurs at time \(t\) hours after midnight, the same clock time the next day is \(24+t\). For a schedule with two administrations each day, check both the within-day gap and the overnight gap from the later administration to the next day's earlier administration.

For more than two administrations, apply the required interval condition to every successive pair, including the last administration of one day and the first of the next. A repeating schedule is not feasible merely because all within-day gaps pass; its overnight gap must also pass. If some doses are intentionally omitted on certain days, specify that pattern explicitly before checking the relevant gaps.

Infusion Availability Includes the Full Duration

An infusion's start time can be permitted while the infusion itself cannot be completed in the available period. Let an infusion start at \(s\), last for \(d\) hours, and use an availability block \([u,v]\). To fit entirely within that block, it must satisfy \(s\geq u\) and \(s+d\leq v\). Equivalently, \(s\in[u,v-d]\), provided \(d\leq v-u\). These conditions check continuous availability for the whole infusion, not just at its start.

If an infusion also has its own permitted start window \([a,b]\), a feasible start exists in that availability block exactly when \([a,b]\cap[u,v-d]\neq\varnothing\). If the infusion must follow another administration by an allowed interval, that condition must be checked as well. When availability consists of multiple separate blocks, test the infusion against each block; an infusion cannot be treated as continuous across a period when the resource is unavailable.

A Systematic Feasibility Check

For a candidate schedule, collect the constraints before judging feasibility. This avoids treating each rule as if it were the only one that matters.

A short derivation explains the interval-range test. Since \(t_2-t_1\) increases as \(t_2\) increases and decreases as \(t_1\) increases, its minimum over the two closed windows occurs at \(t_2=a_2\), \(t_1=b_1\), giving \(a_2-b_1\). Its maximum occurs at \(t_2=b_2\), \(t_1=a_1\), giving \(b_2-a_1\). Every intermediate difference is attainable because the windows are intervals and the difference changes continuously as either time varies. Therefore all pairwise differences form \([a_2-b_1,b_2-a_1]\). Requiring the second administration to be later removes zero and negative differences, which is why the positive-time condition must be applied when the required interval does not already enforce it.

Diagnostics, Alternatives, and Limits

A useful diagnostic is to identify which constraint causes infeasibility. An empty intersection between the permitted start window and the duration-adjusted availability block indicates a resource conflict. A required interval that does not intersect the positive pairwise-difference range indicates a timing conflict. A candidate that passes an existence test but fails when its actual times are substituted indicates that a feasible choice may exist, but the proposed choice is not that choice.

Common warning signs include intervals calculated from clock readings without accounting for midnight, checking an infusion's start but not its end, and using a difference range that includes pairs in the wrong temporal order. Also verify that time units match: a duration in minutes cannot be added directly to a time expressed in hours unless it is converted first.

One alternative to testing a single proposed schedule is to check whether any feasible schedule exists within all the windows and interval bounds. The intersection tests above are useful for simple two-time cases. For a larger schedule, constraints can be written together as inequalities and checked jointly; candidate schedules can then be screened against those inequalities. This is more reliable than adjusting one time at a time without rechecking previously satisfied constraints. Conversely, when exact times are prescribed, direct substitution is simpler than an existence calculation.

Operational feasibility is necessary for implementation, but it is not sufficient for a pharmacologically suitable regimen. A schedule can satisfy every time and facility rule yet fail a concentration or response target; it can also satisfy a model-based target while being impossible to administer under the stated operational rules. Keep these conclusions separate. This tutorial checks timing and availability constraints, not whether a feasible schedule is clinically appropriate or optimal.