A correspondence in The Lancet Healthy Longevity argues that the clinical rationale for stopping a medication — absence of benefit, accumulating harm, or treatment burden — should determine trial design in deprescribing research, rather than defaulting to non-inferiority frameworks. Non-inferiority designs appropriately shift the burden of proof when continuation is the standing default, but they are not uniformly suited to every deprescribing question. The distinction carries direct implications for how polypharmacy is safely reduced in older adults.
Key Points
- Clinical rationale should precede trial design choice in deprescribing research
- Non-inferiority designs shift burden of proof toward continued prescribing
- Attrition bias and dual-analysis requirements shape valid deprescribing inference
Longevity Analysis
Removing what no longer serves the patient is the first and most consequential intervention in later-life medicine, yet the evidence base for stopping drugs remains far thinner than the evidence for starting them. Each retained medication draws on the same clearance pathways, alters cognition and postural stability, and can generate symptoms that are then misread as new disease and treated with further prescribing. Methodologically sound deprescribing trials are what allow clinicians to distinguish a genuine therapeutic signal from iatrogenic noise, and to act on that distinction with confidence rather than inertia.
Original published by The Lancet Healthy Longevity, by Lorenz Roger Van der Linden.