A nonprofit division of Sevadar Foundation Inc.

For clinicians  ·  For facilities  ·  For families  ·  For staff

What deprescribing trials actually measure

The evidence base

Deprescribing is a well-studied field that does not answer the question most people are asking of it. That is not a scandal; it is a consequence of what the trials chose to measure, and it is worth understanding before quoting any of them.

Almost every trial measures a count

Look at the primary endpoints of the field’s best-known randomized trials.

TrialPopulationPrimary outcomeResult
OPTIMIZE (2022)3,012 adults 65+ with dementia or mild cognitive impairment, ≥1 other chronic condition, ≥5 long-term medications; 18 primary care clinicsNumber of prescribed long-term medications; percentage on ≥1 potentially inappropriate medicationNull. 6.4 vs 6.5 medications (p = 0.14); 17.8% vs 20.9% on a PIM (p = 0.08)
Shed-MEDS (2023)372 randomized, mean age 76.2, age ≥50 with ≥5 pre-hospital medications; academic medical center plus 22 post-acute care facilitiesTotal medication count at hospital discharge and post-acute discharge14% fewer medications at post-acute discharge (ratio 0.86, 95% CI 0.80–0.93); 15% fewer at 90 days
TAPERAdults 70+ on ≥5 medicationsNumber of medicationsMain trial results not published. A feasibility study randomized 39 patients and makes no efficacy claim.
Primary endpoints of three well-known deprescribing trials. Two of three measure a count; the third has not reported.

A count is a legitimate endpoint. It is also a process measure, and a trial powered to detect a change in it is usually not powered to detect a change in falls, hospitalization or death — which are rarer, noisier and further downstream.

What that means when you read a headline

“Deprescribing trial succeeds” almost always means “fewer medications were prescribed.” “Deprescribing trial fails” almost always means the same measure did not move. Neither headline is about whether anybody was better off, because in most cases that was not measured, or was measured as a secondary endpoint in a trial too small to answer it.

The claims that did not survive replication

Mortality

A 2016 meta-analysis of 132 papers covering 34,143 participants reported a mortality reduction with deprescribing in non-randomized studies (odds ratio 0.32, 95% CI 0.17 to 0.60) and a non-significant one in randomized studies (0.82, 95% CI 0.61 to 1.11). The subgroup finding traveled far beyond that qualification.

The same group published an update in 2024 covering 259 studies. Deprescribing polypharmacy did not significantly reduce mortality in randomized studies (odds ratio 0.96, 95% CI 0.84 to 1.09) or in non-randomized ones (0.70, 95% CI 0.36 to 1.38). Signals persist in subgroups — adults aged 65 to 79, and patient-specific interventions — but the main effect is gone.

Worth noticing: that 2024 paper’s own abstract conclusion is more optimistic than its own primary result, leaning on the subgroups. If you cite it, cite the numbers.

Falls

Covered in full on its own page. Five independent syntheses, no demonstrated effect. Deprescribing and falls.

Appropriateness

A Cochrane review of 32 studies of interventions to improve appropriate polypharmacy in adults aged 65 and over on four or more medications found a reduction in the number of potentially inappropriate medications (standardized mean difference −0.22, 95% CI −0.38 to −0.05; 7 studies, 1,832 participants) at very low certainty, and little or no difference in hospital admissions or quality of life at low certainty. Its conclusion: “It is unclear whether interventions to improve appropriate polypharmacy… resulted in clinically significant improvement.”

What the evidence does support, without qualification

Deprescribing is safe. This is the most defensible positive claim in the field and it is well supported from several directions.

  • The 118-trial, 417,412-patient synthesis found no signal of harm across seven outcomes — adverse drug reactions, injurious falls, quality of life, medical visits, emergency admissions, hospitalizations or mortality.
  • Shed-MEDS found adverse drug event rates similar between arms (hazard ratio 0.83, 95% CI 0.52 to 1.30).
  • The OPTIMIZE safety analysis found adjusted relative risks of 0.92 (95% CI 0.72 to 1.16) for hospitalization and 1.19 (95% CI 0.67 to 2.11) for mortality, with zero deaths in either group adjudicated as likely attributable to a medication change.
  • A 2026 umbrella review of 27 systematic reviews concluded that focused deprescribing “appears feasible and generally safe.”

And most patients agree to it when asked. In the Shed-MEDS trial, participants or their surrogates agreed with 63% of clinician deprescribing recommendations across 883 medications. The barrier is less often refusal than the conversation never happening.

Two trials worth watching

The over-75 statin question is being tested directly. STAREE has completed enrollment of 9,971 community-dwelling adults aged 70 and over without cardiovascular disease, dementia or diabetes; PREVENTABLE is recruiting toward 20,000 adults aged 75 and over. Neither has posted results, and nothing on this site attributes a finding to either. Statins in the very old.

Common questions

If the trials mostly measure pill counts, is deprescribing worth doing?

Yes, for what it demonstrably does: it safely reduces the number of medications and the number of potentially inappropriate ones, and it removes drugs that are no longer treating anything. What it should not be sold as is a falls or mortality intervention. The deprescribing step.

Why do trials not measure outcomes that matter more?

Because those outcomes are rarer and noisier, so detecting a change in them requires far larger and longer trials than most of this research has been funded for — and because the population is frail, hard to follow, and frequently excluded from trials altogether. The largest synthesis to date was assembled by pooling 118 small trials rather than by running one big one. Deprescribing and falls.

Back to the evidence base.

What the evidence does and does not show

Sedating medications impair gait, balance and reaction time, and their labels say so. That part is not in doubt. What has not been shown is that a deprescribing program reduces fall rates — the trials that tested it were small, short, and mostly measured a medication count rather than a fracture. Those are two different questions, and absence of evidence from underpowered trials is not evidence that the drugs do not cause falls. What the evidence actually shows about deprescribing and falls.

References

  1. Bayliss EA, Shetterly SM, Drace ML, et al. Deprescribing Education vs Usual Care for Patients With Cognitive Impairment and Primary Care Clinicians: The OPTIMIZE Pragmatic Cluster Randomized Trial. JAMA Internal Medicine. 2022;182(5):534–542. PMID 35343999 · DOI 10.1001/jamainternmed.2022.0502.
  2. Boyd CM, Shetterly SM, Powers JD, et al. Evaluating the Safety of an Educational Deprescribing Intervention: Lessons from the OPTIMIZE Trial. Drugs & Aging. 2024;41(1):45–54. PMID 37982982 · DOI 10.1007/s40266-023-01080-y.
  3. Vasilevskis EE, Shah AS, Hollingsworth EK, et al. Deprescribing Medications Among Older Adults From End of Hospitalization Through Postacute Care: A Shed-MEDS Randomized Clinical Trial. JAMA Internal Medicine. 2023;183(3):223–231. PMID 36745422 · DOI 10.1001/jamainternmed.2022.6545.
  4. Kim JL, Lewallen KM, Hollingsworth EK, Shah AS, Simmons SF, Vasilevskis EE. Patient-Reported Barriers and Enablers to Deprescribing Recommendations During a Clinical Trial (Shed-MEDS). The Gerontologist. 2023;63(3):523–533. PMID 35881109 · DOI 10.1093/geront/gnac100.
  5. Quek HW, Page A, Lee K, et al. The effect of deprescribing interventions on mortality and health outcomes in older people: An updated systematic review and meta-analysis. British Journal of Clinical Pharmacology. 2024;90(10):2409–2482. PMID 39164070 · DOI 10.1111/bcp.16200.
  6. Page AT, Clifford RM, Potter K, Schwartz D, Etherton-Beer CD. The feasibility and effect of deprescribing in older adults on mortality and health: a systematic review and meta-analysis. British Journal of Clinical Pharmacology. 2016;82(3):583–623. PMID 27077231 · DOI 10.1111/bcp.12975.
  7. Rankin A, Cadogan CA, Patterson SM, et al. Interventions to improve the appropriate use of polypharmacy for older people. Cochrane Database of Systematic Reviews. 2018;9(9):CD008165. PMID 30175841 · DOI 10.1002/14651858.CD008165.pub4.
  8. Persaud N, Workentin A, Rizvi A, et al. Interventions to Address Potentially Inappropriate Prescribing for Older Primary Care Patients: A Systematic Review and Meta-Analysis. JAMA Network Open. 2025;8(6):e2517965. PMID 40577011 · DOI 10.1001/jamanetworkopen.2025.17965.
  9. García Álvarez Á, Riera Serra P, Ricci-Cabello I, et al. Effectiveness and safety of focused deprescribing interventions targeting specific drug classes in older adults across clinical settings: an umbrella systematic literature review. Age and Ageing. 2026;55(8):afag243. PMID 42607278 · DOI 10.1093/ageing/afag243.
  10. Monash University. A Clinical Trial of STAtin Therapy for Reducing Events in the Elderly (STAREE). ClinicalTrials.gov identifier NCT02099123. Active, not recruiting; no results posted. ClinicalTrials.gov. 2026. https://clinicaltrials.gov/study/NCT02099123.
  11. Duke University. PRagmatic EValuation of evENTs And Benefits of Lipid-lowering in oldEr Adults (PREVENTABLE). ClinicalTrials.gov identifier NCT04262206. Recruiting; no results posted. ClinicalTrials.gov. 2026. https://clinicaltrials.gov/study/NCT04262206.

Bibliographic records on this page were retrieved from PubMed. Every reference links to its DOI or PubMed record.

Written and medically reviewed by Dr. Gurpreet Singh Padda, MD, MBA, MHP, founder and medical director of DWARAA. Last reviewed . This page is educational. DWARAA does not prescribe, deprescribe, diagnose or treat.