Skip to content

Within-Window Dose and Cadence Titration

Dosing protocol — instantiates Critical-Window Intervention Timing

Sets and adjusts how much exposure to deliver and how often within the open window, climbing toward effect while staying under a safety ceiling that prevents overload or harm.

Being inside the window tells you nothing about the right amount. Within-Window Dose and Cadence Titration governs the intensity and spacing of exposure once delivery is underway — how large each dose is, how frequently it repeats, and how quickly to climb — and it does so between two constraints: enough to drive acquisition, but never so much that the receiving system is overloaded, distressed, or coerced. Its defining move is closed-loop climbing under a ceiling: it steps the dose up in response to how the system is actually absorbing it, while a hard safety bound caps the ascent so that the pressure to "use the window" can never tip into harm. It assumes the schedule of when is already set elsewhere; it owns only how much, how often, and how safely.

Example

A patient begins allergen immunotherapy, whose early build-up phase is a bounded window in which the immune system can be pushed toward tolerance by escalating exposure to the allergen. The protocol governs that climb. It starts from a low dose, sets a cadence (weekly injections through build-up), and titrates upward on a schedule — but every step is checked against a safety bound: if a dose provokes more than a mild local reaction, the next step is held or reduced rather than advanced.[^therapeutic_window]

For one patient the climb proceeds smoothly to the maintenance dose; for another, a strong reaction at an intermediate step caps the ascent and the cadence stretches out. The output is a per-patient exposure plan that reaches the effective dose as fast as the individual safely tolerates — no faster. The protocol never decides whether the window is open or when the injections happen on the calendar; it decides only the size and spacing of each one, bounded by what is safe.

How it works

The protocol maintains a bounded exposure plan — a starting dose, a cadence, and a climb rule — and runs it as a feedback loop: after each exposure it reads the system's response and decides whether to advance, hold, or step back before the next. Governing the whole loop is a safety bound: an absolute ceiling and per-step limits that the climb may never breach regardless of how much window remains. That ceiling is what distinguishes titration from a fixed schedule — the plan is adjusted to the system's tolerance in flight, and the urgency of a closing window is never allowed to override the cap.

Tuning parameters

  • Climb rate — how aggressively the dose escalates between steps. Faster reaches effect sooner but raises overload and drop-out risk; slower is gentler but may not finish before the window closes.
  • Cadence — the spacing between exposures. Tighter spacing concentrates the effect but crowds recovery; wider spacing is safer but slower.
  • Safety margin — how much headroom sits between the working dose and the hard ceiling. A wide margin is conservative; a narrow one extracts more effect at higher risk.
  • Response sensitivity — how strong an adverse signal triggers a hold or step-down. A twitchy setting stalls easily; a dull one lets harm accumulate before it reacts.

When it helps, and when it misleads

It is most valuable when the effect is dose-dependent and the receiving system can be overloaded — when both too little and too much fail, and the safe path is a monitored climb. Its honest failure mode is a mis-set ceiling: too low and the exposure never reaches effect before the window closes; too high and the climb harms the very system it is trying to configure. The classic misuse is letting a closing window justify breaching the safety bound — escalating dose or cadence past what is tolerable because "we're running out of time," which is exactly how a critical-window rationale slides into coercion or injury. The discipline is to treat the safety bound as inviolable, titrate on observed response rather than on the clock, and accept an incomplete acquisition over an unsafe one.

How it implements the components

  • bounded_exposure_and_support_plan — it is the concrete plan of dose size, cadence, and climb that delivers exposure inside the window.
  • safety_bound — it holds the ceiling and per-step limits that cap the climb and force holds or step-downs.

It does not set the schedule that phases delivery to the window edges (intervention_window, timing_rule — see Time-Locked Exposure Protocol), does not provide the fading learner-support scaffolds (stabilization_support — see Scaffolded Acquisition and Fade), and does not verify durable uptake afterward (durability_and_transfer_check — see Stabilization and Consolidation Schedule).

Notes

The safety bound is not a tuning knob to be relaxed under time pressure — it is the guardrail that keeps "use the window" from becoming "overload the system." When a window is closing and the dose still hasn't reached effect within safe limits, that is a signal to consider a Missed-Window Remediation Plan, not to breach the ceiling.