The sequence is familiar. A referral produces a diagnosis and a letter that stops short of an executable plan. The specialist has answered the diagnostic question. Primary care retains responsibility for longitudinal care; the risk-bearing organization retains accountability for outcomes and total cost. Yet no one is assigned the next step: translating the diagnosis into a plan the patient and care partner can carry out.
The post-diagnostic period is when estimated incremental traditional Medicare spending concentrates. The Alzheimer’s Association’s 2026 report places nearly half of five-year incremental costs in the year after diagnosis and 87% within two years. Claims-based diagnosis may lag symptom onset. These figures identify concentrated spending; they do not establish its causes.
Consider this first case composite: an older adult with diabetes, heart failure, and newly diagnosed dementia. Her regimen assumes she can fill a pillbox, manage insulin, track weights, and recognize symptoms warranting a call. No one confirms those abilities. Her daughter handles the bills but remains outside the care plan; no one has clarified the patient’s preferences, decision-making capacity, or her daughter’s role. When the regimen fails, the record says “nonadherence.” More accurately, it depended on abilities already changing.
In a second case composite, an older adult continues driving and taking sedating medications after diagnosis. Neither the medication review nor the driving review has an owner. A fall prompts coordination. It does not establish causation; it exposes foreseeable work that remained unowned until crisis.
Cognitive status is a feasibility variable across the care plan. Objective cognitive assessment provides a baseline; comparable serial assessments help quantify change. Both support, but do not replace, clinical synthesis across collateral history, function, medications, mood, delirium, and medical contributors. Management must reconcile medications, address safety and driving, support caregivers, adapt chronic-disease plans, and define follow-up. The diagnosis creates each task and assigns none.
Primary care should lead the longitudinal clinical plan; the risk-bearing organization should own handoff reliability. One named clinician or team must remain accountable for integration and follow-through. Work may be distributed across disciplines, but accountability cannot disappear. Referral is not the failure; unowned implementation is.
The cause is structural: clinical and financial accountability travel on different maps. Depending on the arrangement, accountability for quality and some or all of total cost remains with a Medicare Advantage plan, ACO, or delegated provider organization. Data, incentives, and operational obligations do not always cross that boundary to convert consultation into longitudinal management. Specialists may lack payment or infrastructure for ongoing care; plans may learn of the diagnosis through claims. Contracts may define consultation completion while leaving a completed handoff undefined.
Where I practice, PACE shows how the maps can converge. Capitated, integrated financing places comprehensive medical and social services under one organization’s accountability. Specialty services may still be contracted out, but responsibility for integrating recommendations remains with the PACE organization. Capitation creates the condition; an interdisciplinary team operationalizes it by receiving findings, updating the care plan, assigning the work, and verifying follow-through.
Through GUIDE, CMS is testing tiered per-patient monthly payments for defined dementia care management, including coordination, navigation, caregiver support, and 24/7 access, with separate respite payments for eligible patients. GUIDE serves eligible beneficiaries in traditional Medicare; Medicare Advantage and PACE enrollees are excluded. PACE holds this responsibility through capitation. GUIDE creates no parallel requirement for Medicare Advantage; plans and delegated entities seeking comparable support must build it through their own payment and contracting arrangements.
The lever should match the arrangement. Medicare Advantage plans can set contracted-network standards; delegated entities can act within delegated authority; ACOs can use participating-specialist agreements and closed-loop referral protocols; integrated systems can require internal handoffs. I would propose prompt transmission of a structured diagnostic summary, documented acceptance by the receiving team, and, within 30 days, an actionable plan naming the accountable clinician or team, required actions, and follow-up date. Measure diagnosis-to-handoff, handoff-to-plan, and overall diagnosis-to-plan. Was the answer sent or converted into care?
For the patient with diabetes and heart failure, that standard could distinguish a regimen built on abilities she has lost from one built on the abilities she has. A diagnosis creates a promise. The question is who is named to keep it, and the organizations that hold the outcome are positioned to answer.
About the author:
Anthony Zizza, MD, is a Harvard-trained geriatrician with over a decade of experience leading clinical care for older adults across New England. As Chief Medical Officer at Element Care PACE, he leads interdisciplinary teams serving more than 1,100 PACE participants and 1,850 Senior Care Options members across Essex County, the Merrimack Valley, and Greater Boston. He also serves as an advisory board member at Creyos, a digital cognitive assessment platform. Dr. Zizza previously held senior clinical leadership roles at Optum at Home, Landmark Health, and the PACE Program at Cambridge Health Alliance/Harvard Medical School. He is committed to advancing coordinated, value-based care for older adults so they can remain safely in their homes and communities.