Skip to content

Minimum Effective Dose Review

De-escalation review — instantiates Effective-Input Delivery Assurance

Periodically re-examines a standing input to find the lowest level that still works, deliberately shedding dose to reduce off-target burden without losing the effect.

Minimum Effective Dose Review is a recurring, qualified re-examination of a standing input whose bias is deliberately downward. It asks "can this be done with less?" and de-escalates toward the lowest level that still produces the effect, treating any dose above that minimum as off-target burden to be trimmed. Its defining move is minimization over time: where calibration finds a working level and holds it, the review keeps re-challenging that level, using the off-target burden boundary as the standing reason to push the input lower. It exists because inputs get escalated in a crisis and then quietly left there while their burden accumulates.

Example

A patient with an autoimmune condition was started on a high dose of prednisone to bring a flare under control. The corticosteroid works — and the standing risk is precisely that it is never revisited while cumulative burden mounts: bone loss, elevated glucose, weight gain. The Minimum Effective Dose Review, on a schedule, tapers the dose in steps, checking at each step that disease control holds. It keeps lowering until symptoms threaten to return, then stops — that step is the current minimum effective input. The patient ends up maintained at, say, ≈5 mg rather than 20 mg: the same control, a fraction of the burden. The review's whole value is that it turned "it's working, leave it" into "what is the least that still works?" — the steroid-sparing instinct made into a standing procedure.

How it works

The review runs scheduled downward trials: step the input down by a set amount, verify the effect is still maintained, and stop at the lowest level that holds. Against that, it weighs the off-target burden — the accumulating harm that justifies going lower in the first place. What distinguishes it from its siblings is the direction and the recurrence: it is not tuning up to a target or specifying a band, it is repeatedly testing whether the standing level can be reduced.

Tuning parameters

  • Taper step and interval — how much to cut and how long to observe before the next cut. Aggressive tapers shed burden fast but risk relapse.
  • Efficacy floor — the minimum acceptable effect that stops further lowering. A stricter floor keeps more headroom but minimizes less.
  • Review cadence — how often the standing input is re-challenged. Rare reviews let inertia win and burden accumulate.
  • Burden weighting — how heavily off-target burden counts against a working-but-high level, which sets how hard the review pushes down.
  • Relapse-recovery rule — the pre-agreed step-back if a reduction triggers loss of control, so de-escalation is reversible rather than a gamble.

When it helps, and when it misleads

Its strength is that it counters therapeutic inertia — the tendency to leave an input high simply because it is working — and turns "it works" into "this is the least that works," which is where the burden savings live.

Its failure modes are the mirror image of its virtue. Over-aggressive minimization can trigger relapse, and a minimum found today may only hold until the next stressor moves it. The classic misuse is to cut for cost or optics rather than evidence, or to ratchet the input down with no relapse-recovery plan, so a reduction that fails leaves the patient worse off than before. The discipline that keeps it honest is to require an explicit efficacy floor and a pre-agreed step-back before any reduction, and to lower on evidence rather than zeal — the structured practice of deprescribing.[1]

How it implements the components

Minimum Effective Dose Review fills the floor-and-burden slots — the lowest workable level and the harm that motivates finding it:

  • minimum_effective_input — its central product: the lowest input level that still achieves the intended effect.
  • off_target_burden_boundary — the accumulating-harm boundary it watches as the dose comes down, and the reason to keep de-escalating.

It does not define the standing band (target_range / adjustment_rule, handled by Dosage Window Protocol), tune the input up to a target (response_monitoring / side_effect_signal, handled by Medication Dose Calibration), or map the response shape (sensitivity_profile, handled by Exposure Dose Curve); it drives the input toward its floor.

Notes

The minimum effective input is not a fixed number — it drifts with the subject's state, so a floor found this year may be too low next year, which is exactly why the mechanism is a recurring review rather than a one-time search. Its opposite-facing neighbour is Medication Dose Calibration: calibration hunts a target and holds; this review keeps trying to go lower.

References

[1] Deprescribing is the structured, supervised reduction or cessation of a medication to lower burden while preserving benefit — the clinical counter-practice to escalation-and-forget. ↩