Nursing home antipsychotic drugs: spot off-label use

If your parent moved into a nursing home this summer and came home from the first care conference on a new medication ending in -pine, -done or -zole, you are not alone. About one in five long-stay residents in the United States is prescribed an antipsychotic drug, and CMS Care Compare data through the first quarter of 2026 shows the national rate has barely moved in three years. The Food and Drug Administration has never approved any antipsychotic for the behavioural symptoms of dementia. The agency added a boxed warning almost twenty years ago noting that these drugs raise the risk of stroke and death in older adults with dementia, yet they remain the single most common off-label prescription written inside nursing homes. This guide walks through why the practice persists in 2026, what CMS's tightened enforcement expects of every facility, the specific documents families can ask to see, and the non-drug care steps that resolve most agitation before a psychiatric medication is warranted.

What counts as off-label use inside a nursing home

A drug is being used off-label when it is prescribed for a condition, an age group or a dosage that the FDA has not reviewed for safety and effectiveness. Off-label prescribing itself is legal and sometimes appropriate; a cardiologist may use a beta blocker for a rhythm the label does not name, and the evidence base supports it. The problem in nursing homes is narrower. The most common antipsychotics prescribed to long-stay residents, quetiapine, risperidone, olanzapine and aripiprazole, are FDA-approved for schizophrenia and bipolar disorder. They are not approved for the behavioural and psychological symptoms of dementia (BPSD), which includes agitation, wandering, resistance to care, calling out and disturbed sleep. Yet BPSD is the reason most of these drugs are started inside a facility. A 2011 HHS Office of Inspector General audit found that 83% of atypical antipsychotic claims for elderly nursing home residents were for off-label indications, and follow-up work through the OIG's ongoing oversight reports has documented only partial improvement since.

Why the practice persists in 2026

Three forces keep the numbers high. Staffing is the first. A behavioural care plan for agitation in dementia is labour-intensive and requires a consistent aide who knows the resident, and the national aide turnover rate is still above 50% a year. A one-time prescription is faster, cheaper and requires no continuity of staff. Diagnostic drift is the second. Because Care Compare's public quality measure excludes residents with a documented schizophrenia, Tourette or Huntington diagnosis, some facilities have quietly added those diagnoses to the record so the antipsychotic no longer counts against the measure. The Government Accountability Office flagged this pattern almost a decade ago and CMS has since sharpened its audit of new schizophrenia diagnoses that appear only after a resident enters a facility. Prescribing culture is the third. When the medical director, the consulting psychiatrist and the pharmacy have used the same short list of agents for a decade, the default order set is hard to unwind, and family requests to review a medication are the most reliable trigger for change.

What the drugs actually do to a frail resident

The clinical case against these drugs in dementia is not theoretical. A pooled analysis behind the FDA's drug safety communications on antipsychotics showed a 1.6 to 1.7-fold increased risk of death within ten weeks of starting an antipsychotic in older adults with dementia, driven by cardiovascular events and pneumonia. Falls rise sharply because the drugs sedate, drop blood pressure and blunt the reflexes that catch a stumble. Aspiration risk rises because swallowing coordination slows. Confusion often worsens rather than improves because the anticholinergic side effects of some agents cloud cognition further. The National Institute on Aging's guidance on managing personality and behaviour changes in dementia is explicit that non-drug approaches should be tried first, and that any antipsychotic use should be short, monitored and reviewed. In the real building, that review is the piece most often missing.

What CMS now expects, and how families can read the signal

The CMS State Operations Manual Appendix PP defines tag F758, Free from Unnecessary Psychotropic Medications and Psychotropic Medications, as the enforcement lever inspectors use during survey. A 2023 memo, QSO-23-05-NH, sharpened the definition of an unnecessary antipsychotic and added an audit for schizophrenia diagnoses added after admission. In 2026, F758 citations are one of the most common deficiency tags nationally, and they show up on the CMS 2567 inspection report that families can pull from Medicare.gov Care Compare. Two Care Compare numbers matter. First, the long-stay antipsychotic quality measure; a home in the top quartile has under 12%, a home in the bottom quartile is above 24%. Second, any F758 deficiency in the last three surveys. The national partnership to improve dementia care in nursing homes still coordinates the state-level targets, and a well-run facility can name its rate to within a percentage point.

The three documents families are allowed to see

Under 42 CFR 483.10, the federal resident rights rule, a resident or the resident's legal representative may examine the medical record within 24 hours of a request and receive copies within two working days. Three documents cover the antipsychotic question. First, the Medication Administration Record (MAR), which lists every scheduled dose, every as-needed (PRN) dose actually given, and the reason recorded. A quiet resident whose MAR shows a PRN antipsychotic every afternoon at 3 pm for three straight weeks is a resident whose behavioural care plan is not working. Second, the physician order sheet, which shows who wrote the order, when, and for what indication. An order for quetiapine with a diagnosis line that reads only 'agitation' or 'behaviour' is not a diagnosis the FDA has approved for that drug. Third, the interdisciplinary care plan, which should name the target behaviour, the non-drug interventions tried first, and the criteria for stopping the drug. If the care plan is silent on stopping criteria, that is a legitimate concern to raise with the director of nursing.

Non-drug approaches that resolve most agitation first

The evidence base for non-drug approaches is stronger than the evidence base for antipsychotics in dementia, and the interventions are practical inside a normal building. Pain is the first thing to rule out; the Agency for Healthcare Research and Quality's long-term care patient safety resources list untreated pain, urinary tract infection and constipation as the three medical causes most often mistaken for a psychiatric symptom. Sensory correction is second; a resident who cannot hear the aide is a resident who resists care, and a working hearing aid resolves that faster than any pill. Structured routines are third; consistent staff assignments, a predictable meal and toileting schedule, and a bedroom that keeps the same layout are the environmental controls most closely linked to lower behavioural symptoms in the peer-reviewed literature. Person-centred activity is fourth; music, walking, family visits and simple task engagement reduce afternoon and evening agitation without a prescription. The Alzheimer's Association guide to medications for memory covers when a drug is warranted and when it is not, and it is a useful document to bring to the next quarterly care conference.

The exact questions to ask at the next care conference

Quarterly care conferences are the family's best opening. Bring five questions in writing and take notes. One: what is the specific target behaviour this medication is treating, and how is progress measured. Two: what non-drug approaches were tried first, for how long, and with what result. Three: what is the plan and timeline for a gradual dose reduction (GDR), which the CMS rules expect at least every six months unless a physician documents a clinical reason to hold. Four: what is the drug's FDA-approved indication for our parent's diagnosis, and if there is none, what is the specific rationale for off-label use documented in the record. Five: who is the consulting pharmacist and when is the next monthly regimen review. A good facility answers all five in a single sitting. A facility that becomes defensive at question three is a facility whose review process has slowed, and the next call should be to the long-term care ombudsman or to the state survey agency.

When a prescription is genuinely appropriate

Antipsychotics are not always wrong. A resident with a clearly documented schizophrenia diagnosis that predates admission, a resident with a psychotic depression that failed a full antidepressant trial, or a resident in the last weeks of life with severe distress that non-drug measures have not eased may be a resident for whom a short, closely monitored course is the right call. The distinguishing features of an appropriate prescription are the ones missing in most off-label use: a specific target symptom, a defined trial length, a written stopping criterion, and a documented plan for a gradual dose reduction. If those four elements are in the chart, the medication is being used within the standard of care. If any one is missing, the same care conference questions apply, and the same rights to review, request a reduction, and escalate to the ombudsman remain available. Our involuntary discharge guide and signs of elder abuse guide cover the escalation options when a facility responds to family advocacy with pressure to move the resident out.

More from the blog

Industry analysis

The 2026 nursing home staffing crisis: what families actually need to know

Costs & funding

The hidden costs of nursing home care: 9 fees nobody mentions on the tour