Physical and Chemical Restraints in Nursing Homes

Nursing home residents have an explicit federal right to be free from physical or chemical restraints imposed for discipline or convenience, and not required to treat the resident's medical symptoms. That right is written into 42 CFR Part 483, enforced through CMS survey tags, and supported by decades of clinical evidence showing that unnecessary restraints cause measurable harm without producing the safety benefits staff sometimes assume they will.

This guide walks through the federal definitions, the F-tags surveyors use to cite violations, the documented risks of both physical devices and psychotropic drugs, and the steps families can take when they believe a resident is being restrained improperly. It is a companion to our broader overview of nursing home neglect.

Federal Right to Be Free From Unnecessary Restraints

The core rule appears in 42 CFR § 483.12(a)(2). According to the regulation, a facility must ensure that residents are free from physical or chemical restraints imposed for purposes of discipline or convenience that are not required to treat the resident's medical symptoms. When a restraint is medically indicated, the regulation requires the facility to use the least restrictive alternative, limit duration to the minimum necessary, and document ongoing re-evaluation of need.

Those protections sit inside a broader framework of resident rights under 42 CFR § 483.10, which gives every resident the right to refuse treatment and to participate in comprehensive care planning. According to Cornell Legal Information Institute's mirror of the Code of Federal Regulations, care plans must describe the services the facility will provide and any services refused because the resident exercised that right. Restraint decisions that bypass the care plan or ignore a resident's refusal can themselves be a violation, independent of whether the restraint causes physical injury.

The restraint-free standard is not new. It was established by the Nursing Home Reform Act of 1987 (OBRA-87), which crystallized the consensus that restraints should not be used to manage behavior or staffing pressures. A PubMed-indexed analysis of pre- and post-reform data (PMID 9256852) found that physical restraint prevalence declined from 37.4% to 28.1% after OBRA implementation, a statistically significant shift (P < .001) that reflected the intended impact of the reform across long-term care.

What Counts as a Physical Restraint

CMS defines a physical restraint in the State Operations Manual Appendix PP as any manual method or physical or mechanical device, material, or equipment attached or adjacent to the resident's body that the individual cannot remove easily and that restricts freedom of movement or normal access to one's body. The definition is function-based: what matters is whether the resident can release the device independently and whether it limits movement, not what the device is called on the purchase order.

Common devices that can meet the definition include:

• Bed rails that a resident cannot lower independently and that prevent voluntary exit from bed.
• Lap trays, lap belts, and wheelchair bars that the resident cannot release.
• Hand mitts, soft ties, and vest or jacket restraints used in chairs or beds.
• Tucked-in sheets, recliners, or furniture arrangements that prevent a resident from standing or moving on their own.

The FDA's bed-safety guidance is especially relevant here. According to FDA reports on hospital bed entrapment, between 1985 and 2013 the agency received 901 incidents of patients caught, trapped, entangled, or strangled in hospital beds, including 531 deaths and 151 nonfatal injuries. FDA guidance notes that most victims were frail, elderly, or cognitively impaired, and that the absence of timely toileting, position changes, and nursing care increases entrapment risk. Bed rails are not neutral safety equipment; they are a mechanical device that requires a risk-benefit assessment under federal guidance.

What Counts as a Chemical Restraint

A chemical restraint, according to CMS guidance in Appendix PP, is the inappropriate use of psychotropic medications, including antipsychotics, benzodiazepines, and sedatives, for the primary purpose of controlling or restricting behavior or movement rather than treating a diagnosed medical symptom. The question is not whether the resident is taking a psychoactive drug; it is whether the drug has a valid medical indication and whether it is being used to replace staff attention, engagement, or care.

Peer-reviewed prevalence data (PMC article 3009456) found that roughly 26% of nursing home residents use antipsychotics and 40% of that use has no appropriate clinical indication. Benzodiazepine use is documented in about 13% of residents, with 42% lacking an appropriate indication. Together, the literature shows that a large share of psychotropic prescribing in long-term care cannot be justified by a diagnosed medical symptom.

This is the backdrop for the CMS National Partnership to Improve Dementia Care, launched in 2012. According to CMS, the long-stay antipsychotic prevalence rate was 23.9% at the 2011Q4 baseline, and the Partnership's initial goal of a 30% reduction was met ahead of schedule. By 2018Q4 the long-stay rate had fallen to 14.6%, a 39% total reduction from baseline, although state-level performance varied and some states continued to exceed 26%. CMS has framed the Partnership as evidence that deprescribing is feasible when facilities, surveyors, and public reporting pull in the same direction.

The FDA has also been unambiguous about antipsychotics in older adults with dementia. An April 2005 advisory placed a black box warning on atypical antipsychotics after controlled trials showed a drug-treated death rate of 4.5% versus 2.6% on placebo over roughly 10 weeks, a relative risk 1.6 to 1.7 times higher. The warning was extended to conventional (typical) antipsychotics in June 2008 because the safety profile was similar. According to FDA and NIH-indexed analyses, reported causes of death included congestive heart failure, sudden cardiac death, infections such as pneumonia, and cerebrovascular events. Antipsychotics are not approved for treating behavioral symptoms in elderly patients with dementia.

CMS Survey Tags That Enforce the Rule

CMS surveyors translate the regulations into specific F-tags. Three are central to restraints:

• F-604 — Right to Be Free from Physical Restraints. Addresses unlawful use of physical restraints, including bed rails that restrict voluntary movement when a resident cannot self-lower them.
• F-605 — Right to Be Free from Chemical Restraints. The investigative protocol examines inappropriate use of psychotropic medications to control behavior without a clinical indication or informed consent.
• F-758 — Unnecessary Use of Psychotropic Medications. Requires a physician examination and clinical assessment before prescribing PRN antipsychotic medications. According to CMS's list of revised F-tags, F-758 was consolidated into the F-605 investigative protocol effective 2020 to increase consistency and reduce overlap.

These tags are visible in the deficiency history a state surveys publishes when a facility is cited. Families can look for restraint-related F-tags when they review a facility's public inspection record alongside related neglect indicators such as medication errors and understaffing.

When Restraints Are Permitted

A restraint is not automatically a violation. Under 42 CFR § 483.12(a)(2), a restraint may be used when it is required to treat the resident's medical symptoms. Three conditions define the narrow permitted use:

Documented medical symptom

The restraint must be tied to a specific, assessed medical symptom, not to behavior that staff find inconvenient. CMS guidance in Appendix PP makes clear that agitation, wandering, or calling out do not by themselves constitute a medical symptom that justifies restraint.

Least-restrictive alternative

Before a restraint is used, the facility must attempt non-restraint alternatives and document why those alternatives were not sufficient. The regulation requires the least restrictive intervention and the shortest necessary duration.

Informed consent and ongoing re-evaluation

Care planning under 42 CFR § 483.10 requires the resident (or the resident's representative when consent has been given) to participate in decisions, including restraint decisions. Ongoing re-evaluation must be documented; a restraint that was justified on day one is not automatically justified on day thirty.

Harms Caused by Improper Physical Restraints

Research does not support the idea that physical restraints keep residents safer. A propensity-matched analysis of 264,068 nursing home residents using 2004 to 2005 data (PubMed 19786918) compared restrained and unrestrained residents and found that restrained residents experienced significantly worse outcomes across nearly every measure studied.

• Increased falls and injuries, despite fall prevention often being cited as the reason for the restraint.
• Greater walking dependence and decline in activities of daily living.
• Higher prevalence and severity of pressure injuries.
• Joint contractures from prolonged immobility.
• Worsened cognitive performance and increased problem behaviors.

The only outcome where the analysis found no significant difference was depression rates, suggesting that restraints do not improve mood either. The authors concluded that physical restraint use represents poor clinical practice. Bed rail entrapment adds a separate, catastrophic category of harm: the FDA's incident data are the reason clinical guidance now treats bed rails as a mechanical restraint rather than a passive safety device.

Harms Caused by Improper Chemical Restraints

The peer-reviewed literature on inappropriate psychotropic use documents a consistent pattern of harm. According to PMC article 3009456 and related reviews, chemical restraint use is associated with decreased functional and cognitive performance, increased falls and fractures, excess sedation and respiratory depression, stroke risk (cerebrovascular events), and increased mortality. The FDA black box warnings on atypical and conventional antipsychotics were issued precisely because the mortality signal was reproducible across controlled trials.

According to analyses indexed in PMC, typical agents such as haloperidol show comparable or higher mortality rates than atypical alternatives while carrying greater movement-disorder risks. In other words, swapping one antipsychotic class for another does not create a safe pharmacologic option for behavioral control in elderly patients with dementia. The correct clinical conclusion, reflected in the Partnership's deprescribing goal, is that the drug should come off whenever the medical indication does not support it. Improper chemical restraint can also overlap with elopement and wrongful death concerns when sedation masks deterioration or increases fall and stroke risk.

Non-Pharmacologic Alternatives

A substantial body of randomized and controlled research, indexed in PMC (articles 6140014, 1420764, and 3711645), supports non-pharmacologic approaches for managing the behaviors that restraints are typically used to suppress. These approaches address behavioral drivers rather than overriding them.

• Structured activities and regular, meaningful engagement throughout the day.
• Caregiver education and staff training in dementia communication.
• Adult day services and small-group therapeutic engagement.
• Cognitive Stimulation Therapy (CST), a group intervention combining cognitive stimulation, reality orientation, reminiscence, and validation therapy, supported by randomized trials in mild to moderate dementia.
• Pain management and optimized support for activities of daily living, because untreated pain and unmet ADL needs often drive the behaviors that staff misinterpret as aggression.

Implementation research cited in those PMC reviews emphasizes that non-pharmacologic programs work most effective when there is strong facility policy, individualized assessment of the source of each resident's behavior, regularization of daily schedule, and reduced environmental and sensory stress. Person-centered care that prioritizes these elements is correlated with reduced use of both physical and chemical restraints.

Warning Signs Families Can Recognize

Improper restraint is not always obvious at first glance. Families visiting a nursing home resident can look for patterns that suggest the care team is substituting restraint for engagement.

• New bruises or skin marks that match the pattern of a belt, vest, mitt, or rail.
• Bed rails that the resident cannot lower on their own and that are routinely up during quiet waking hours.
• Unexplained sedation, disengagement, or a sudden drop in alertness that does not match the resident's baseline.
• A new antipsychotic or benzodiazepine prescription without a clear diagnosis or an explanation on the care plan.
• A wheelchair with a lap tray or bar that the resident cannot remove and that is used as a default during the day.
• Care-plan meetings that do not discuss restraints, behaviors, or psychotropic medications even though the resident is receiving them.

What Families Can Do

Residents have the right to refuse treatment under 42 CFR § 483.10 and the right to participate in care planning. A family member who believes a restraint is being used improperly can start with the facility and escalate outward.

• Request the medication administration record, the care plan, and any physician orders authorizing a specific restraint or psychotropic medication.
• Ask the director of nursing to convene a care-plan meeting and discuss documented medical symptoms, least-restrictive alternatives that were tried, and the timeline for re-evaluation.
• File a grievance with the facility's designated grievance official, a right protected under 42 CFR § 483.10(j).
• Contact the state Long-Term Care Ombudsman, who is authorized under the federal Older Americans Act to investigate and advocate on the resident's behalf.
• File a complaint with the State Survey Agency, which can cite the facility under F-604, F-605, or the consolidated psychotropic investigative protocol.

State reporting procedures vary. Our state-level reporting guides for California, Illinois, Pennsylvania, and Texas explain where to file locally.

If regulatory reporting does not resolve the concern and you want to understand legal options outside the survey process, you can speak with an attorney about a civil claim. That conversation is independent of the reporting channels described above. For context on adjacent issues that often appear alongside restraint violations, see our overviews of staffing requirements and the broader category of nursing home injuries.