Report of the Council on Scientific Affairs
Use of Restraints for Patients in Nursing Homes

Rosalie Guttman, PhD; Roy D. Altman, MD; Mitchell S. Karlan, MD;
for the Council on Scientific Affairs, American Medical Association

Arch Fam Med. 1999;8:101­105

Objective: To examine issues related to the use of re­ straints on nursing home patients, including regula­ tions and guidelines, risks and benefits of restraint use, system problems, and measures to reduce restraint use, to determine when the use of restraints results in clini­ cally desirable outcomes.

Methods: Sources of information included a review of published articles and reports, a survey of federal and state regulations and guidelines relating to restraint use in nurs­ ing homes, review of current legislative initiatives, and consultation with experts in the field.

Results: The data reveal that restraint use imposes more risk of falls and other undesirable outcomes than it prevents. In response to legislative initiatives and regulatory activities and by implementing alternatives, the prevalence of restraint use has decreased by 20% in recent years. In many states, facilities have created restraint­free environments or restraint­free policies and goals. The Council on Scientific Affairs finds that current federal and state regulations on the use of restraints have benefited the vast majority of nursing home patients.

Conclusions: While guidelines are in place for the use of restraints when clinically necessary, the Council on Scientific Affairs recommends increased research to de­ termine when the use of restraints results in desirable out­ comes. Extraregulatory initiatives, such as widespread educational programs, are needed for professionals and consumers to improve awareness of the risks and ben­ efits of restraints, as well as the rights of residents with respect to restraint use.

In June 1989, the Council on Ethical and Judicial Affairs of the American Medical Association (AMA) published a report, "Guidelines for the Use of Restraints in Long­term Care Facilities." [1] Among the ethical and legal issues addressed were the right of the nursing home resident to be free from needless bodily restraint and to informed consent when restraints are deemed medically necessary. The Council on Ethical and Judicial Affairs report included recommendations that advocate judicious caution in the use of restraints, which should only be applied when explicitly ordered by physicians, that restraints not be used as punishment or for the convenience of staff, that the right to informed consent be incorporated into institutional policy, that maximum bodily mobility be permitted whenever restraints are applied, and that restraints be used only in accordance with appropriate clinical indications. In light of recent changes in the regulations governing restraint use, there is a need for an updated report.

This Council on Scientific Affairs report reviews subsequent policy or regulations and research findings relative to restraint use in nursing homes and proposes additional AMA policy on this subject.


Published studies from the years 1990 to 1997 were identified through a MEDLINE search of English­language articles, using the key phrase "physical restraints or chemical restraints and nursing homes." Additional publications were identified by review of references cited in the above noted articles and in textbooks and by consultation with experts in the field. Approximately 120 articles from the medical or scientific literature and other documents on the subjects of restraints and nursing homeswere analyzed and the most relevant were selected; secondary sources and meta­analyses were excluded. In addition, existing federal and state regulations and guidelines relating to restraint use in nursing homes and current legislative initiatives were reviewed. The report underwent peer review by experts and by the AMA Council on Scientific Affairs. This report was transmitted by the AMA Board of Trustees to the House of Delegates, where the recommendations were adopted as AMA policy in June 1997.


Among the legal changes incorporated into the Omnibus Budget Reconciliation Provisions of 1987 (OBRA) were strict controls governing the use of restraints in nursing facilities and the communication of rights of nursing home residents to be free from restraints. The Nursing Home Reform Provisions [2] were included in OBRA and implemented in 1990, and it stipulated that restraints were to be imposed only to ensure the physical safety of the resident or that of other residents and only on the written order of a physician. Moreover, physician orders for restraints were required to be specific with regard to the duration and circumstances of their use.

Since the implementation of the Nursing Home Reform Provisions, the prevalence of restraint use in the United States decreased from an estimated 41% nationwide to about 20%. [3] Despite this reduction, one fifth, or more than 300,000 of the elderly population in nursing homes remain physically restrained. Although the actual law did not change, the policy of the Health Care Financing Administration toward restraints continued to evolve, as what was considered "medically necessary" restraint use became more narrowly defined by medical professionals.

Research findings suggest that the prevalence rates for restraint use of 0% to 5% are achievable for the proper care of the nursing home population. [4,5] Based on these conclusions, the Health Care Financing Administration, while not committing to a given percentage of restraint use as a criterion, has agreed that the standard of care that is currently widely accepted in the medical community permits less justification for restraint use than was foreseen whenOBRAwas enacted in 1987. It is also likely that the scope of what is currently considered to be medically appropriate use of restraints will continue to narrow even further in the future. [6]


Regulations contained in the OBRA governing requirements of participation for long­term care facilities were revised in 1996. The modifications included operational instructions to state surveyors. The Interpretive Guidelines, [7] the primary federal guide to surveyors on the application of the regulations, were expanded to help eliminate some of the confusion that persists regarding the definition of a restraint and the contexts in which restraints may and may not be used.

The restraint and seclusion standards issued by the Joint Commission on Accreditation of Healthcare Organizations require facilities to limit and, if feasible, reduce the use of restraints and seclusion. Additionally, they encourage administrators to actively take steps toward using preventive strategies, develop protocols for consistency, and institute time­limited orders for all residents so as to permit more frequent assessments and early release from restraints when they are used in cases of clinical necessity. [8]

Physical restraint is defined as:

any manual method or physical or mechanical device, material, or equipment attached or adjacent to the individual's body that the individual cannot remove easily which restricts freedom of movement or normal access to one's body. [9]

Chemical restraint is defined as any psychopharmacologic drug that is used for convenience or to control undesirable behaviors and not required to treat medical symptoms. [9]


Physical restraints include, but are not limited to, leg restraints, arm restraints, hand mitts, soft ties or vests, lap cushions, and lap trays the resident cannot remove. Also included as restraints are facility practices that meet the definition of a restraint, such as using bed rails to keep a resident from voluntarily getting out of bed as opposed to enhancing mobility while in bed; tucking in a sheet so tightly that a bed­bound resident cannot move; using wheelchair safety bars to prevent a resident from rising out of a chair; placing a resident in a chair that prevents rising; and placing a resident who uses a wheelchair so close to a wall that the wall prevents the resident from rising. [7 (p53)]

Orthotic body devices are to be used only for therapeutic purposes to enhance overall functional capacity of the patient.

When a restraint is used, it must serve as an enabler (ie, to increase functional ability) and it must be the least restrictive alternative. For example, bed rails may be used as restraints (to prevent residents from getting out of bed) or they may be used to assist in mobility and transfer. Thus, the use of bed rails as restraints is prohibited unless they are necessary to treat a resident's medical symptoms. Bed rails used as restraints add risk to the resident by increasing the possibility of more significant injury caused by a fall from a bed with bed rails as opposed to a fall from a bed without rails. [7]

The standards are based on the goals of preserving and protecting residents' rights, dignity, and physical and emotional well­being when restraints are applied. To this end, the resident has the right to participate in the plan of care and to refuse or accept restraints. So the resident can make an informed choice about the use of restraints, the facility is required to explain the negative outcomes of restraint use. Potential and multiple undesirable outcomes include incontinence, decreased range of motion and ability to ambulate, symptoms of withdrawal or depression, or reduced social contact. [10]

If a resident is not competent to make a decision, a surrogate or agent may exercise this right based on the same information that would have been given to the resident. However, the regulation that restraints may be used only to treat the resident's medical condition applies even when permission is given for restraint use by a representative or surrogate. Restraint use is never justifiable solely for the convenience of facility staff, for disciplinary purposes, or because a surrogate requests them. [9]

All restraint use must be justified: that is, the facility must demonstrate the presence of medical symptom(s) and a written explanation of how the restraint would treat the cause of the symptom(s) and assist the resident in reaching his or her highest level of physical and psychosocial well­being must be provided. [7]


Evans and Strumpf [11] pointed out the paradox in the willingness of health professionals to use restraints on the elderly, given the existing knowledge about the range of serious effects and consequences related to restraint and immobility in this population. Among the consequences of restraint application are physical risks, such as avoidable decline in the ability to ambulate, contractures, decreased muscle tone, increased risk of pressure sores and infections, constipation, and urinary incontinence or retention, and psychological risks, such as agitation, increase in disorganized behavior, depression, humiliation, fear of being abandoned, impaired selfimage, and (in the presence of dementia) catastrophic reaction. [11­13]

Other undesirable outcomes associated with restraint use are falls and injuries that ensue when individual residents attempt to personally remove restraints or try to climb over bed rails. It is estimated that as many as 200 deaths occur annually as a result of strangulation or suffocation caused by restraints, even when t hey are correctly applied. [14­18] Several recent studies [13,15,19­21] found that serious fall-related injuries were much more common in restrained residents than in those who were not restrained (17% vs 5%, respectively). Tinetti et al [15] found that among residents of skilled nursing facilities residents who were restrained were 3 times as likely to sustain an injury in a fall or related incident than those who were not restrained. In their study of 397 nursing home residents, Schnelle and Smith [19] noted that 10% of the falls occurred while the individual was restrained or immediately after restraints were removed. Many falls occur after restraints are removed, which suggests that if restraints cause falls and related injuries they do so indirectly. The authors [13,15,19] believed that restraints worsen deconditioning and gait and balance abnormalities, thereby increasing the risk of falls and injuries. Other studies comparing fall rates i n restra i ned and unrestrained residents of nursing homes showed similar results: the authors [20,21] reported a lack of association between restraint use and a decrease in falls.

Similar risks also are associated with the use of psychoactive drugs as chemical restraints. Toxic reactions to these drugs, especially in the elderly, are well documented in the literature. Examples of these reactions are dizziness, tremors, tardive dyskinesia, increased agitation and confusion, dehydration, constipation, and urinary incontinence. In addition, studies [22,23] showed that an increased risk for falls and hip fractures is associated with psychoactive drug use among the elderly. Moreover, the results of research showed that the sequelae of psychotropic drug use (dizziness, agitation, and confusion, for example) lead to further restraint with mechanical appliances. [24]

The benefits of using restraints, according to facility personnel, are the prevention of falls and fall­related injuries, better management of therapeutic regimens and difficult behaviors, aid in maintenance of body alignment , increased resident feeling of security and safety, and the prevention of wandering. [11,15] The perceived benefits of restraints are anecdotal, and no clinical trials or outcome studies have been conducted yet to assess the efficacy of physical or chemical restraint interventions.

At present, there is no scientific basis of support that demonstrates the efficacy of restraints in preventing injury to nursing home patients. Rather, the scientific literature overwhelmingly demonstrates the disproportionate risk-benefit ratio in the use of restraints.


Several factors contribute to the problems associated with restraint use in long­term care facilities. Providers indicate that there is a lack of congruence between surveyors' and providers' use of restraints and/or alternatives. Additionally, providers cite family pressure for restraint use as a majorbarrier to developing care plans in some cases. Some family members believe that a person has to be restrained to prevent a fall, and providers are concerned about litigation if they do not comply with family demands. However, providers have been successfully sued when injury or death has resulted from restraint application. [14] Conversely, according to findings from a national project, more than 90% of families find the use of restraints acceptable. [25] Furthermore, the rights of residents are often perceived to be antithetical to the rights of professionals. Thus, a tension or adversarial relationship is created between residents and families and health care professionals, which intensifies concerns about malpractice suits. [26,27]

There is an additional relationship between the propensity of nursing home staff to use restraints and the level of education and employment characteristics. There is a greater reluctance on the part of less well­trained staff to try alternatives; their attitudes and beliefs favor restraint use. [28] Other burdens imposed on facilities are financial concerns. Many administrators perceive that alternatives to restraint use are costly and time­consuming and that they haveneither the financial nor the personnel resources to implement less restrictive interventions. [29,30]

System factors that create dilemmas with regard to restraint use can be summarized as follows: administrative pressure to avoid litigation, family demands, the availability of restraint devices, staff attitudes, and insufficient staffing.


The extent to which nursing home residents require restraints to prevent or control accidental falls and other injuries has not been established. Because of the known undesirable effects of restraint use and its limitations as a preventive strategy, in conformity with federal and state regulations, there has been a nationwide endeavor to reduce restraint use in long­term care facilities. Several prospective randomized trials have been conducted using alternatives to both physical and chemical restraints, with salutary outcomes. These studies [31­35] demonstrated that effective programming can eliminate a need for restraints.

Cohen and colleagues [25] developed the national demonstration and research project to determine whether physical restraints could be safely removed from nursing home residents. They studied a random sample of facilities with a high level of restraint use. The intervention emphasized individualized care, multidisciplinary team assessment, and rehabilitation to enhance functional independence and education of staff and families. The goal was to teach a process that would reduce restraint use to 5% or less. Restraints were classified into categories, and alternatives were developed for each category. At the end of 2 years, restraint use decreased from 41% (858 restrained residents) to 4.05% (89). Ultimately, 13 of 16 facilities had achieved a restraint rate of 5%or less. The remainder achieved a reduction to more than 5% but less than 10%. [25] Neufeld et al [36] reported that, in their restraint reduction study of 16 facilities with 2075 beds and a high level of restraint use, their educational intervention effected a decrease in restraint use to 5%. A major finding was an increase in staff morale associated with the reduction in restraint use.

Other studies [37,38] similarly demonstrated the effectiveness of education and expert nurse consultation on the reduction of restraint use. Reduction was achieved without increasing falls or the use of psychotropic medications or chemical restraints.

It is estimated that 20% or more of nursing home residents receive antipsychotic drugs, primarily for the behavioral manifestations of dementia. Many of these drugs have high toxic effects, especially in the elderly. By teaching the use of behavioral techniques to physicians, nurses, and other nursing home personnel for the management of behavioral symptoms and encouraging the withdrawal of psychoactive drugs, drug use decreased by 72% in an intervention sample vs 13% in nursing homes serving as controls. [30]

In response to the federal regulations on the use of psychotropic drugs, nursing homes in many states have been successful in decreasing the rates of antipsychotic drug use. A survey [23] of all nursing homes in Minnesota found that antipsychotic drug use had decreased by 33% during the 4­year study period. Similar studies [39,40] in Baltimore, Md, and Tennessee reported decreases of 37% and 27%, respectively, in antipsychotic drug use. Finally, Rader and Donius [41] describe a 3­level restraint reduction intervention in Oregon. These researchers found that multiple factors, including resident characteristics, staff education, and environmental manipulation, affect restraint use. The care of level 1 residents is easily maintained without the use of restraint with alternatives. Level 2 residents have more complex problems and require structural changes in the environment and specialized equipment to facilitate restraint release. For level 3 residents, who are the most difficult to manage and who generally do not do well without restraints (they may be receiving life-sustaining treatment, fall on every attempt to stand or walk, or engage in harmful behaviors), the goal is to provide maximum freedom by using the least restrictive devices and minimizing the time they are used. [41]

Numerous restraint reduction initiatives and alternatives to restraints have been described in the literature. These include educational programs for facility staff and families, involving residents in activities and exercise, low beds and chairs and recliners, pillows and positioning aids, monitoring caffeine and sugar intake, behavioral mapping as an aid in assessment of new residents and those with dementia, alarm devices, and the establishment of timely toileting schedules and regular rest periods. As Schnelle and Smith [19] pointed out, restraint alternatives need to be described with sufficient specificity for them to be more readily implemented.

Overall, the restraint reduction movement and the implementation of alternative strategies have been beneficial with regard to decreasing the risks inherent in restraint use. Perceived benefits of restraint removal are improvement in quality of life, enhanced functional status, fewer falls and other injuries, decreased use of antipsychotic medications, fewer behavioral symptoms, and increased staff morale. [33,34]


Most study findings indicate that the use of restraints does not reduce the risk or incidence of falls, other accidents, or disruption of medical care when appropriate alternative interventions are provided. In fact, research suggests that the use of restraints causes more problems than it prevents. The existing regulations governing the use of restraints in long­term care facilities, while encouraging providers to use alternatives to the extent possible in an attempt to reduce restraint use, are quite explicit in permitting their use in cases of documented medical necessity. However, the regulations are subject to variable interpretation by surveyors and providers. When restraints are used, there must be ongoing assessment, intermittent relief from use, and conformity to the resident's (or surrogate's) right to informed consent and self­determination. Because of the constraints and limitations imposed on facilities, as well as the nonspecific guidelines and fear of penalty related to disjunction between some surveyor expectations and family practices, there has been an exaggeration toward non­use of restraints.

Thus, there is a need for more effective education about the use of restraints for families, patients, and professionals, as well as better outcome studies on restraint use that meets regulatory requirements. Ultimately, there needs to be a satisfactory resolution of the ethical dilemma of beneficence vs autonomy and consonance with the uppermost goal of all AMA policy, delivery of care that is in the best interest of the patient's or resident's quality of life provided within the least restrictive environment.


The following statements, recommended by the Council on Scientific Affairs, were adopted by the AMA House of Delegates as AMA policy in June 1997.

  1. The AMA reaffirms the opinion of the Council on Ethical and Judicial Affairs "that all individuals have a fundamental right to be free from unreasonable bodily restraint" (Policy 280.987). [42]
  2. The AMA recommends further research to support or refute the findings that physical restraints in nursing homes tend to be more harmful than beneficial.
  3. The AMA supports the position that there must be compelling reasons to justify the use of restraints.
  4. The AMA encourages widespread dissemination of information and educational initiatives for the public as well as health care professionals on the risks and uncertain benefits of restraints.
  5. The AMA encourages physicians to communicate the consequences, risks, and potential benefits of restraint use with family members of residents who ask for restraints.
  6. The AMA encourages research to determine precisely when the use of restraints results in improved outcomes.
  7. The AMA encourages the long­term evaluation of effects of the restraint regulations on the health and well­being of nursing home residents.


  1. Guidelines for the use of restraints in long­term care facilities. In: 1989­1990 Code of Medical Ethics Reports of the Council on Ethical and Judicial Affairs of the American Medical Association. Vol 1. Chicago, Ill: American Medical Association; 1992.
  2. The Nursing Home Reform Provisions of the Omnibus Reconciliation Act of 1987 (as amended), 42 USC §1395I­3 and 1396r (1992).
  3. Health Care Financing Administration. Medicare/ Medicaid Nursing Home Information, 1987­ 1988. Washington, DC: Health Care Financing Administration; 1989.
  4. Ejaz F, Folmar S, Kaufman M, Rose M, Goldman B. Restraint reduction: can it be achieved? Gerontologist. 1994;34:694­699.
  5. Werner P, Koroknay V, Braun J, Cohen­Mansfield J. Individualized care alternatives used in the process of removing physical restraints in the nursing home. J Geriatr Soc. 1994;42:321­325.
  6. Health Care Financing Administration. The national restraint reduction initiative. Presented at: HealthCare Financing Administration Restraint Reduction Conference; December 3­4, 1996; Philadelphia, Pa.
  7. The Newly Updated OBRA for Long­term Care Facilities. Springfield, Ill: Life Services Network of Illinois; 1996.
  8. Joint Commission on Accreditation of Healthcare Organizations. 1996 Standards for Longterm Care. Oakbrook, Ill: Joint Commission on Accreditation of Healthcare Organizations; 1996.
  9. Healthcare Financing Administration. Interpretive guidelines. In: State Operations Manual. Washington, DC: Health Care Financing Administration; 1992. Transmittal 250.
  10. Joint Commission on Accreditation of Health Care Organizations. 1996 Comprehensive Accreditation Manual for Long­term Care. Oakbrook, Ill: Joint Commission on Accreditation of Health Care Organizations; 1996.
  11. Evans LK, Strumpf NE. Tying down the elderly: a review of the literature on physical restraint. J Am Geriatr Soc. 1989;37:65­74.
  12. Williams C. Long­term care and the human spirit. Generations. 1990;14:25­28.
  13. Tinetti M, Liu W, Marottolo R, Ginter S. Mechanical restraint use among residents of skilled nursing facilities: prevalence, patterns, and predictors. JAMA. 1991;265:468­471.
  14. Kapp MB. Nursing home restraints and legal liability. J Leg Med. 1992;17:22­25.
  15. Tinetti M, Liu W, Ginter S. Mechanical restraint use and fall­related injuries among residents of skilled nursing facilities. Ann Intern Med. 1992; 116:369­374.
  16. Miles SH, Irvine P. Deaths caused by physical restraints. Gerontologist. 1992;32:762­766.
  17. Miles SH. Restraints and sudden death. J Am Geriatr Soc. 1993;41:1013.
  18. Pedal I, Mattern R, Reibold R, et al. Sudden fatalities in mechanically restrained patients [in German]. Z Gerontol Geriatr. 1996;29:180­184.
  19. Schnelle TF, Smith RL. To use physical restraints or not? J Am Geriatr Soc. 1996;44:727­728.
  20. Schnelle JF, MacRae PG, Simmons SF. Safety assessment for the frail elderly: a comparison of restrained and unrestrained nursing home residents. J Am Geriatr Soc. 1994;42:586­592.
  21. Capezuti E, Evans L, Strumpf N. Physical restraint use and falls in nursing home residents. J Am Geriatr Soc. 1996;44:1043­1048.
  22. Toenniessen LM, Casey DE, McFarland BH. Tardive dyskinesia in the aged. Arch Gen Psychiatry. 1985;42:278­284.
  23. Garrard J, Chen V, DowdB. The impact of the 1987 federal regulations on the use of psychotropic drugs in Minnesota nursing homes. Am J Public Health. 1995;25:771­776.
  24. Mion LC, Minnick A, Palmer R, Kapp MB, Lamb K. Physical restraint use in the hospital setting: unresolved issues and directions for research. Millbank Q. 1996;74:3­20.
  25. Cohen C, Neufeld R, Dunbar J, Pflug L, Breuer B. Old problem, different approach: alternatives to physical restraints. J Gerontol Nurs. 1996;22:23­29.
  26. Johnson SH. The fear of liability and the use of restraints in nursing homes. LawMedHealth Care. 1990;18:263­273.
  27. Kirschbaum L, O'Connor SJ. Legal impact of restraining the elderly in nursing homes. Top Geriatr Rehabil. 1992;8:29­34.
  28. Hill J, Schirm V. Attitudes of nursing staff toward restraint use in long­term care. J Appl Gerontol. 1996;15:314­324.
  29. Health Care Financing Administration. Restraints and the fear of injury. HCFA Natl Restraint Reduction Newslett. 1996;4:1­7.
  30. Ray WA, Taylor JA, Meador K, et al. Reducing antipsychotic drug use in nursing homes: a controlled trial of provider intervention. Arch Intern Med. 1993;153:713­721.
  31. Evans LK, Strumpf NE, Williams C. Redefining a standard of care for frail older people: alternatives to routine physical restraint. Adv Longterm Care. 1991;1:81­108.
  32. Werner P, Cohen­Mansfield J, Korokney V, Braun J. The impact of a restraint reduction program on nursing home residents. Geriatr Nurs. 1994;15: 142­156.
  33. Dunbar JM ,Neufeld RR ,White HC ,Libow L. Retrain, don't restrain. Gerontologist. 1996;36:539­542.
  34. Mahoney DF. Analysis of restraint­free nursing homes. Image J Nurs Sch. 1995;27:155­160.
  35. Werner P, Cohen­Mansfield J, Koroknay V, Braun J. Reducing restraints: impact on staff attitudes. J Gerontol Nurs. 1994;20:19­24.
  36. Neufeld RR, Libow LS, Foley WF, White H. Can physical restraints in nursing home residents be untied safely? intervention and evaluation design. J Am Geriatr Soc. 1995;43:1264­1268.
  37. Evans LK, Strumpf NE, Allen­Taylor SL, Capezuti E, Maislin G, Jacobsen B. A clinical trial to reduce restraints in nursing homes. J Am Geriatr Soc. 1997;45:675­681.
  38. Stratmann D, Vinson MH, Magee R, Hardin SB. The effects of research on clinical practice: use of restraints. Appl Nurs Res. 1997;10:39­43.
  39. Rovner BW, Edelman BA, Cox MP, Shmuely Y. The impact of antipsychotic drug regulations on psychotropic prescribing practices in nursing homes. Am J Psychiatry. 1992;149:1390­1392.
  40. Shorr RI, Fought RL, Ray WA. Changes in antipsychotic drug use in nursing homesduring implementation of OBRA­97 regulations. JAMA. 1994; 271:358­362.
  41. Rader J, Donius M. Leveling off restraints. Geriatr Nurs. 1991;12:71­73.
  42. AMA Council on Long Range Planning and Development. Policy Compendium. Chicago, Ill: American Medical Association; 1997.