IDEAS (Imaging Dementia-Evidence for Amyloid Scanning)

IDEAS was a single-group, multisite, longitudinal observational study run by the American College of Radiology under Medicare's Coverage with Evidence Development programme, designed to determine whether amyloid PET changes clinical management in Medicare beneficiaries with mild cognitive impairment (MCI) or dementia of uncertain aetiology who met published appropriate-use criteria. Dementia specialists recorded an intended management plan before amyloid PET and again 90 days after, covering Alzheimer's disease drug therapy, other drug therapy, and safety/future-planning counselling; the study was powered to detect a 30% or greater change in each of the MCI and dementia groups.

Of 18,488 registered participants, 11,409 were included in the primary analysis (6,905 with MCI, 4,504 with dementia). The composite management endpoint changed in 60.2% of the MCI group and 63.5% of the dementia group, both far exceeding the 30% target; Alzheimer's disease drug use changed most often. The aetiological diagnosis flipped from Alzheimer's to non-Alzheimer's disease in 25.1% of patients and the reverse in 10.5%. A later health-utilisation analysis found a smaller-than-hoped 4.5% relative reduction in 12-month hospitalisations, short of the pre-specified 10% effect size, though patients with positive scans were hospitalised less often than those with negative scans.

IDEAS is a management-impact study, not a trial of clinical outcomes from a therapeutic intervention, and it predates the era of anti-amyloid disease-modifying drugs (lecanemab, donanemab), which have since made amyloid status a treatment-eligibility gate rather than only a diagnostic aid. Its results, together with the follow-on New IDEAS study (NCT04426539), underpinned the 2023 removal of the Medicare national coverage determination that had required amyloid PET to be performed only within an approved research study.