Deprescribing is commonly judged by whether medicines are stopped, doses are reduced, or potentially inappropriate prescribing declines. These outcomes demonstrate medication change but do not necessarily establish that an older adult is clinically safer, functions better, experiences less treatment burden, or regards the intervention as beneficial. This umbrella review examines systematic reviews and meta-analyses published between 2017 and 2026 to determine what constitutes successful deprescribing when success is assessed across clinical, functional, safety, treatment-burden, and patient-reported domains. Review-level evidence was synthesized while preserving distinctions between medication-process outcomes and patient-important outcomes, considering methodological quality, certainty, population and setting, intervention type, and possible duplication of primary studies across reviews. The evidence pattern indicates greater consistency for reductions in medication exposure, potentially inappropriate prescribing, and related medication-burden measures than for mortality, hospitalization, falls, function, cognition, quality of life, or other patient-reported outcomes. Absence of demonstrated benefit in sparsely measured domains was not interpreted as evidence of no benefit. We therefore propose that deprescribing success should be evaluated as a multidomain construct in which medication reduction is an intermediate outcome rather than a sufficient endpoint. Interpretation remains constrained by heterogeneous interventions, variable outcome definitions, limited patient-centered measurement, primary-study overlap, and uneven review quality. More informative deprescribing evaluation requires explicit separation of medication change, clinical consequences, safety, function, treatment burden, and outcomes that matter directly to patients.
