Restraints and seclusion are used on people in institutions, children in schools, nursing home residents, general hospital patients, and other locations, but most often with people who have disabilities. Questions regarding legality, morality, and efficacy abound. These questions, compounded by the serious possible adverse consequences of restraints and seclusion, have commanded wide-ranging attention from legislatures, government agencies, human service professionals, direct care staff, advocates, clients and families, and the public. This article addresses the use of physical restraints and seclusion. It does not address the use of drugs as a behavior restraint, although much of the discussion applies in that context as well.Is the use of human services restraint therapeutic? Can restraint use be reduced or replaced with alternatives? Is it time to relinquish these practices, at least when incorporated in a treatment or habilitation plan?In this article I begin with a look at the early institutional use of restraints and seclusion and, as a reminder of what may ultimately be at stake, I note some worst-case results in the United States. I then consider efficacy and risks of “human services restraint.” I review efforts to reform and reduce the use of restraint and address legal liability questions that impact on agency policy and professional behavior. I conclude with some thoughts on the current state of knowledge, policies, and practices regarding human services restraint and on the future of these techniques.Dr. John Conolly accepted the judgment of Dr. Robert Hill, who had experimented with nonrestraint at the Lincoln asylum. Dr. Hill had stated, “In a properly constructed building with [enough attendants], restraint is never necessary, never justifiable, and always injurious” (Ozarin, 2001, p. 27). That was 170 years ago.Dr. Conolly was appointed superintendent of the Middlesex County asylum at Hanwell in England in June 1839. Over 40 of the 800 patients were restrained at the time. Within 3 months, by September 21, all forms of mechanical restraint were gone (Hunter & Macalpine, 1963). Dozens of other facilities followed suit. As a modern commentator explained (Saks 1986),Dr. Conolly epitomized the no-restraint policy within the “moral treatment” movement influenced by the Quakers in England and post–French Revolution reformers in France in the late 1700s and into the 1800s. England's 1854 Lunacy Acts prompted the reduction of restraint use as well. The names of Tuke, Pinel, and Kirkbride are a familiar part of this history (Ozarin, 2000; Tomes, 1984).In the United States, psychiatric hospital superintendents in the mid- to late 1800s were divided on the use of restraints but generally opposed the no-restraint, English position. Physical restraint was viewed as a form of therapeutic treatment and was an accepted practice for dealing with violent patients. American psychiatrists extolled the value of restraint and seclusion, with one noting that these practices are required by a specific American violence. Eugene Grissom of the North Carolina state asylum argued “that the moderate use of mechanical restraint was therapeutic and morally sound, that it was required by the peculiar violence of American Insanity; and it that prevented tragic accidents and injuries” (Tomes, 1988, p. 190). John Gray, the editor of the American Journal of Insanity wrote, “We look upon restraint and seclusion, directed and controlled by a conscientious and intelligent medical man, as among the valuable alleviating and remedial agents in the care and cure of the insane” (as cited in Tomes, 1988, p. 206).In 1875, Dr. (Lord) John Buckmill, a former superintendent of an asylum in England, visited American public and private asylums. He disagreed with the American viewpoint. At the superintendents' annual meeting that year, he invited any superintendent to visit England for a month, and he bet £100 that such a visitor could not find any form of restraint in British asylums. He had no takers. He later wrote, “[The American superintendents] will look back to their defense with the same wonderment … that has been said in defense of domestic slavery” (Ozarin, 2001, p. 27).Although the vast majority of cases of restraint and seclusion do not result in physical harm or death, and for the most part shackles have given way to personal physical restraint, it is sobering to keep in mind at least some of the instances in which death during restraint has occurred recently. Contemplating these deaths assists in reflecting on how alternative practices might have affected the situation, if at all.It is sometimes said that the use of restraints represents a treatment failure. Here are seven stories of such failure; they represent tragic results in the use of restraints.Restraint is the use of force to limit another person's movement. This may occur by physical contact among individuals, mechanically by devices to limit movement, or chemically by the use of drugs. Seclusion is the involuntary placement of a person in a room, exit from which is not a permitted choice.Human service restraint, a term taken from Tumeinski (2005), is used here. The term encompasses both restraints and seclusion. It refers to restraint of a client under the mandate of a program or agency, public or private, by staff who are taught specific restraint techniques. We distinguish human service restraint from actions among parents, friends, and others in freely given relationships.Human services restraint is used in response to, or to control, injury to others, self-injury, property damage, resistance to behavior control, inappropriate behavior, rule-breaking, and the like. It may or may not be used solely in emergency situations. It may or may not be used for treatment, as part of a planned behavioral intervention, or as an aversive consequence for a target behavior.Restraints do not include orthopedically prescribed devices, protective helmets, holding someone to conduct routine physical examinations or tests, protections against one falling out of bed, or assistance to permit someone to participate in activities without risk of physical harm [42 C.F.R. 482.13 (e)(1)(i)(C)].In hospitals, restraint appears to be allowed in nonemergency situations; in intermediate care facilities–mental retardation (ICF/MR) programs, physical restraint and time-out are specifically permitted in nonemergency situations and, when prescribed, must be included in a client's individual program plan for active treatment. The United States provides some regulatory definitions. These differ depending on the type of program and people served. Two examples are provided here.For Medicare and Medicaid participating hospitals, including psychiatric facilities, the regulations are new. Restraint is “Any manual method, physical mechanical device, material, or equipment that immobilizes or reduces the ability of a patient to move his or her arms, legs, body, or head freely” [42 C.F.R. 482.13 (e) 2006]. Restraint is allowed in nonemergency situations. Seclusion is “The involuntary confinement of a patient alone in a room or area from which the patient is physically prevented from leaving. Seclusion may only be used for the management of violent or self-destructive behavior. Seclusion may only be used for the management of violent or self- destructive behavior.”The older ICF/MR regulations forbade “unnecessary” restraints and required that clients must be “provided active treatment to reduce dependency on drugs and physical restraints” [42 C.F.R. 483.420(a)(6)]. The ICF/MR regulations permit the use of a time-out room as part of “an approved systematic time-out program” incorporated into the client's individual program plan. [42 C.F.R. 483.450(c)]. Physical restraints are permitted only:The use of these human services restraint techniques is not simply a response to client behavior. There is an interplay among staff, setting, the characteristics of the individual, and the individual's behavior, which is perhaps best conceptualized, in the words of one researcher, within an “ecobehavioral” perspective (Day, 2000). Day noted, for example, that age and nature of disability may affect restraint use; there is a greater use of restraints for children with “lower intelligence or neurological impairments.” (Day, 2000, p. i). As discussed below, policy, training, and staff behavior greatly affect the extent of restraint use.Fisher, in a review of the existing literature published in 1994, reached the conclusion that “[l]ocal non-clinical factors, such as cultural bias, staff role perceptions and the attitudes of hospital administrators, have a greater influence on the use of these practices than any clinical factors” (Fisher, 1994, p. 1590).There is a vast amount of literature on these practices, much of it simply descriptive, policy oriented, and useful in training staff in techniques. There is a also what might be called a “negative literature” and a “positive literature.” Critiques of restraint use have been made on multiple grounds:The negative literature includes a small number of self-described “voices of protest” who defend restraint and seclusion and seek to “slow down the locomotive of opinions and pressure tactics that may lead mental health treatment in the wrong direction:The discussion here focuses on what the author believes are two fundamental questions that are most likely to influence governmental, agency, and judicial decisions on restraint: (a) efficacy and (b) risk of harm. There is relatively little scientific investigation of the techniques' efficacy, and no proof of their efficacy. There is, however, much evidence of their risks.Rapid intervention limited to protecting someone from immediate harm is sometimes necessary in an emergency. In such cases, the intervention is limited to the least duration and to the least risky method and must be accomplished by specially trained personnel. “Emergency restraint” is not planned and is not for the purpose of treatment or reduction of harmful behavior.Planned human services restraint for treatment, to support positive behavior, or to reduce negative behavior has not been shown to be effective. For example, a review of 109 articles spanning 35 years between 1965 and 2000 on restraints and seclusion on children and adolescents found that the techniques have only “questionable efficacy” (Day, 2000, p. 28). Research on human services restraint is characterized as sketchy and inconclusive. Both governmental and professional reviews have found no therapeutic value in the practices. For example, the state of Wisconsin concluded,The National Alliance on Mental Illness (NAMI) stated:In a training manual, A Roadmap to Seclusion and Restraint Free Mental Health Services for Persons of All Ages, the U.S. Department of Health & Human Services' Substance Abuse and Mental Health Services Administration (SAMHSA; 2006) set a goalDespite the absence of evidence of efficacy, there are volumes on proper procedures and criteria, minivolumes on documentation, and innumerable dollars spent annually on programs for staff training in techniques that have not been found to be effective.It is difficult to evaluate the relative risks in the use of human services restraint. The absence of data on restraint use (including data on routine use as well as the frequency of untoward events) makes it extremely difficult to comment on the relative risks involved in restraint, the comparative risks involved across a wide range of individual procedures, and the relative risks involved in alternative interventions, including seclusion, mechanical restraint, or medication (Busch & Shore, 2000; Patterson, 2003).Clients and staff may be injured during the imposition of restraint. The ultimate risk is death. In the professional literature, there has been discussion of the risk of death associated with the use of physical restraints since at least the 1980s. More recent research has focused on deaths and other adverse consequences in restraint (Fisher, 1994; Milliken, 1998; Mohr, 2003; Patterson, 2003). Although much of the research concerns psychiatric restraint use, the physiological effects of restraint are no different for individuals with intellectual and developmental disabilities. These risks exist with both mechanical restraint and with so-called personal restraint, where a staff person's body is used to impose the restraint.Upset clients, when restrained, are held down or held tight, often with bodily organs and chest compressed; the heart begins to beat faster or out of rhythm, as the body attempts to obtain more oxygen to support itself (DiDino & Zaccardi, 2007). Restraints involving neck holds or obstruction of the nose or mouth have a high risk of fatality, as do mechanical restraints or prone tying, including “hobble tying.” “Hobble tying is the term used to describe the prone positioning of a patient, following which their wrists are secured behind their back, their ankles are tied, and their wrists and ankles are subsequently secured together by pulling the shoulders back and bending the legs towards them” (Horsburgh, 2004, p. 8). In a series of 214 cases of hobble tying in agitated delirium, death occurred in nearly 12% of the cases (Stratton, 2001). Seated restraint is also risky; preexisting physical conditions, such as obesity, heart disease, general physical ill health, or exhaustion, are additional risks (Horsburgh, 2004).It is fair to say that there is no way to predict who may die due to the use of physical restraint or who may be seriously injured. Almost 10 years ago, an editorial in the Canadian Medical Association Journal noted the asserted benefits of human services restraint and then reminded us, “However, restraint is not itself harmless; some proportion of those who are restrained may die. We do not know what this proportion is, or how many others will come near death and need to be revived” (Milliken, 1998, p. 1611).Children appear to be particularly likely to be subjected to restraints and to die while restrained (Cotton, 1989; Delaney & Fogg, 2005; Earle & Forquer, 2005; Fassler & Cotton, 1992; Nunno, 2006). The use of restraints in schools is of increasing concern and is now the subject of specific research attention; restraint use in schools is often not subject to accreditation or regulatory control (Ryan & Peterson, 1992).There is a great deal of evidence that the use of human services restraint can be dramatically reduced and, as some programs have shown, eliminated. One significant technique for achieving this result is the use of positive behavioral supports, which harkens back, in spirit at least, to the approaches pioneered by Pinel, Tuke, Conolly, and the moral treatment movement, and American institutional reformers such as Dorothea Dix. Dedicated leadership, improved policies, acquisition and distribution of restraint data, explicit goals, careful debriefing of incidents, and specially designed staff training have all contributed to successful reduction efforts.Singh (1999) taught staff on an adolescent unit about behavioral principles and a treatment approach that focused on the patient's strengths. The investigators gradually reduced the number of acceptable hours of restraint and seclusion, setting progressively lower quarterly criterion levels. Staff met these levels, eventually reaching zero use. They maintained the gain at 1-year follow up (Singh 1999).A study published in 2005 regarding Pennsylvania's reduction program examined the use of seclusion and mechanical restraint from 1990 to 2000 and the rate of staff injuries from patient assaults from 1998 to 2000 in Pennsylvania's state hospital system. From 1990 to 2000, the rate of seclusion decreased from 4.2 to 0.3 episodes per 1,000 patient-days. The average duration of seclusion decreased from 10.8 to 1.3 hours. The rate of restraint decreased from 3.5 to 1.2 episodes per 1,000 patient-days. The average duration of restraint decreased from 11.9 to 1.9 hours. No significant changes were seen in rates of staff injuries from 1998 to 2000 (Smith, 2005).The National Association of State Mental Health Program Directors has conducted training on human services restraint reduction and has reported that seclusion/restraint hours were reduced by as much as 79%, the proportion of individuals in seclusion/restraint was reduced by as much as 62%, and the incidents of seclusion/restraint events in a month were reduced by as much as 68%. These data are based on 3- to 6-month posttraining evaluations of the first 12 states trained, with 8 hospitals sending data (Conley, 2004; Huckshorn, 2006).As indicated, alternative approaches often encompass (a) organizational and policy changes, (b) quality assurance techniques, and (c) changes to clinical programming, such as prevention of the behavior or situation leading to restraints or seclusion, or the use of alternative procedures. Formulas for reduction typically include such elements as the following (American Academy of Child & Adolescent Psychiatry, 2002; Day, 2000; Delaney, 2006; Miller, 2006; Milliken, 1998; Petti, 2001; Sullivan, 2005):There are other approaches to limiting human services restraint. These may be imposed by authorities outside the entity that is serving clients. These include the approaches discussed below.Proposed legislation in New Jersey would forbid the use of restraints or seclusion as planned interventions or treatment, recognizing only its emergency use. This is Matthew's Law, named after Matthew Goodman, a child with autism who died after use of restraints (New Jersey Bills No. A948, S71, 2006–2007). Various states have passed or are considering legislation limiting restraint use. Other legislation on restraints is in place (Children's Health Act, 2000).Federal law regulates restraint use. Some examples are as follows: 42 U.S.C. §290jj (2002), regulating public or private, nonmedical, community-based facilities, guarantees a “right to be free from … any restraints or involuntary seclusions imposed for purposes of discipline or convenience.” The Developmental Disabilities Assistance and Bill of Rights Act of 2000 [42 U.S.C. §15009(3)(A)(3)] includes congressional findings that public funds should be for programs that meet minimum standards relating to “prohibition of the use of physical restraint and seclusion for such an individual unless absolutely necessary to ensure the immediate safety of the individual or others, and prohibition of the use of such restraint and seclusion as a punishment or as a substitute for a habilitation program.”Stefan (2006) reported that, after a state mental health department medical director spent the day in ambulatory restraints, he prohibited their use in state facilities. Court-appointed monitors and expert witnesses have also been affected by having institutional staff administer restraints to them.The U.S. Department of Health and Human Services, Substance Abuse and Mental Health Services Administration (2006) seeks to “ultimately eliminate the use of seclusion and restraint in behavioral healthcare.” President George W. Bush's New Freedom Commission on Mental Health 2003 final report recommended that seclusion and restraint be used “only as safety interventions of last resort, not as treatment interventions” (U.S. Mental Health Commission, p. 9). The New York State Psychological Association's task force on the issue “recommends that aversive behavior interventions be prohibited, without exception, as part of a behavior intervention plan” (New York State Psychological Association, 2006, p. 6). A broader ban is urged by The Alliance to Prevent Restraint, Aversive Interventions and Seclusion (APRAIS). APRAIS (2005) was created by a number of professional and advocacy organizations.Legal decisions may have a significant effect. I turn to the law in the next section.Much contemporary discussion on the use of human services restraint focuses on legal liability. Professional standards of care are integral to grounds on which legal liability is based. When will staff, teachers, or an agency be liable for damages for using restraint at all, or for failing to use it? When is there liability for a restraint gone awry? When does it violate treatment rights to use, or not to use, restraint?Because liability considerations may affect agency and professional behavior, it is appropriate to consider the current state of the law in this regard.On the basis of how the cases have been tried and litigated thus far, the courts have generally been supportive of restraint use and have not established significant barriers to the use of programmatic restraint. It does not appear, however, that the courts (at least in their published opinions) have taken into account the professional research on efficacy and risks or the policies adopted by professional organizations and governmental agencies.There is a constitutional right to be free from bodily restraint, with the right circumscribed by the “professional judgment” standard announced by the Supreme Court 25 years ago in a case originating in the Pennhurst State School and Hospital restraint practices (Youngberg v Romeo, 1982). Legal discussions of restraint have been framed by the Romeo standard under the 14th Amendment, related decisions, and other legal principles (i.e., 4th Amendment, 8th Amendment) and statutes such as the Americans With Disabilities Act and education laws.The extreme is clear. Where a patient is subjected to kicking, stomping, strangling, and twisting in the process of a takedown for restraint, the constitutional limits are breached, as a Minnesota federal court found recently (Nicolaison v Brown, 2007). A fair statement of the state of the law is found in that decision. “Constitutionally infirm practices are those that are punitive in intent, those that are not rationally related to a legitimate purpose or those that are rationally related but are excessive in light of their purpose.”Under current court decisions, the legal standard cited above would likely tolerate restraint use when treatment professionals or teachers decided to restrain or seclude someone for safety or behavior control purposes, and the action is taken in at least arguable good faith. Even severe and injurious actions have not been condemned by the courts.Arguably, some of these courts avoided specific rulings on the restraint techniques by making decisions based on immunity and other substance-avoiding principles. At the same time, it is noteworthy that the court decisions in these cases found nothing especially troubling in the particular use of restraint.The courts have not been provided a full accounting of the history and nature of restraint, the dearth of evidence of efficacy, the high risk, or the weight of professional opinion. The law will likely evolve from its current state as courts receive that accounting. At that point, I expect the decisions to take another direction. Courts will recognize the fragility of the prior decisions upholding restraint and will begin to examine claims that restraints are therapeutic with great skepticism.Restraints presented by a school or facility as an efficacious treatment for inclusion in a treatment or as an educational technique will likely be rejected by courts. The efficacy evidence is lacking. The specific ICF/MR regulations that permit time-out and restraint as part of an individual active treatment plan will likewise be rejected, initially as applied in individual cases and then more broadly.The use of restraint is increasingly subject to scrutiny and disfavor. Federal agencies and professional organizations in recent years have become very critical of any use of restraint and seclusion.As already noted, President George W. Bush's New Freedom Commission on Mental Health final report (2003) advised, “Seclusion and restraint will be used only as safety interventions of last resort, not as treatment interventions” (U.S. Mental Health Commission, 2003, p. 9). The National Institute on Disability and Rehabilitation Research (NIDRR) provides research grants on positive behavioral supports as a way to reduce the use of restraint and seclusion. NIDRR has stated, “Unnecessary institutionalization remains a problem, as do the practices of seclusion, restraint, and forced treatment” (NIDRR, 2006, 71 Fed.Reg. 8183 [Feb. 15, 2006]). In 2003, SAMHSA reported established seclusion and restraint as a priority area and developed a National Action Plan to reach its vision of seclusion and restraint free mental health services” (U.S. Department of Health and Human Services, 2003).Public attention has focused on the issue, both in the news and in response to government reports. The U.S. General Accounting Office (GAO) created a report, almost 10 years ago, titled Improper Restraint or Seclusion Use Places People At Risk (U.S. GAO, 1999). The National Technical Assistance Center for State Mental Health Planning recently issued a detailed white paper on the subject (Haimowitz, 2006).Advocates, treatment professionals, and other organizations have pressed the issue, taking positions and issuing white papers. Among the groups who have done so are the American Academy of Child and Adolescent Psychiatry, American Academy of Pediatrics, American Psychiatric Association, American Association of Child and Adolescent Psychiatry, International Society of Psychiatoric-Mental Health Nurses, NAMI, Autism National Committee, Child Welfare League, National Mental Health Association, and the New Jersey Council on Developmental Disabilities.There are now entire Web sites devoted to this issue. There is an organization named Children Injured by Restraint and Aversives and one sponsored by multiple national groups to end the use of seclusion and restraints (APRAIS). Citizens Against Restraint (Toronto, Canada) seeks to ban all restraint use.In my view, the intense level of attention on restraint and seclusion stems from social and personal discomfort with the imposition of these involuntary measures on persons with disabilities and the great risks entailed by these techniques. Human services restraint negatively affects not just the individuals subjected to restraint but also staff, leadership, agency culture, families, and the public. Those wide-ranging effects have a substantial social cost that is often overlooked both in practice and in the published research.It is valuable to consider whether the use of unproven human services restraint within planned treatment interventions is consistent with the values of our society, our treatment programs, and our personal ethics and morality. Many believe that it is immoral to impose restraint on people who are vulnerable and who have a devalued status. Some have gone further and urged that all use of human services restraint be prohibited simply on moral grounds.If the justification for restraints is treatment, then its advocates cannot fairly show either that restraints are therapeutic or that the life-threatening risks can be credibly minimized.What does the future hold? The use of human services restraint is moving into history. We now know that emergency restraint can be avoided in many instances through policy and organizational changes, adequate training, deescalation techniques, positive supports, debriefing, and other mechanisms. We know that restraint as part of a treatment program does not work and is extremely dangerous. There is a professional and governmental consensus against the use of restraints.I believe that, eventually, all programmatic restraint will be prohibited, and that restraint will not be permitted in an habilitation or treatment plan or as an acceptable aversive consequence for a target behavior. I believe this will occur for several reasons:Where there is a need for momentary restraint or seclusion in an emergency to prevent immediate harm to a person, the techniques may be allowed, subject to extensive limitations such as those in recent federal regulations applicable to hospitals. Even in these cases, only vertical, person-to-person restraint should be permitted (due to the special death risk of prone restraint) and that restraint for a very limited duration and under extensive safeguards.These conclusions are neither new nor revolutionary. They date back to discussion in the field close to 200 years ago. What is new is that we now have research on efficacy and risk, and we have perhaps a more refined sense of the dignity due to whose who might be at risk of restraint.The renewal of a nonrestraint policy in human services will occur in an environment I call the “new moral treatment,” spearheaded by contemporary “John Conollys,” together with allied administrators, professionals, advocates, and clients. With the knowledge and skills gained in a 150 years of continued service to people with disabilities, hopefully we will do no worse than Dr. Conolly. Surely, we can do better.