No image so epitomizes a humane approach to the treatment of people with mental illness as does the painting by Robert-Fleury that shows Philippe Pinel ordering the shackles removed from the inmates at the Bicêtre in 1793 (1). Physical restraint of the mentally ill then and since has been considered— at times with justification— synonymous with brutal custodial care. Lost in the visceral reaction that many people have to the use of restraint and seclusion, however, is the unhappy reality that there may be instances in which they are essential for the protection of severely disordered patients and of others with whom they come into contact.

Periodically an issue of contention, seclusion and restraint practices in psychiatric facilities have been thrust back into the limelight by a series of articles that appeared last fall in the Hartford Courant, Connecticut's leading newspaper (2,3,4,5,6). An investigative reporting team from the paper canvassed health care and licensing officials, patient advocates, and others in all 50 states to identify cases in which patients had died, allegedly because of the use of seclusion or restraint. For the period from 1988 to 1998, the team identified 142 deaths in psychiatric wards, group homes and residential facilities for troubled youths, and treatment centers and group homes for persons with mental retardation (7).

Because New York was said to be the only state to require reporting of all deaths in facilities—64 people died in New York during or shortly after restraint or seclusion from 1988 to 1997—the paper expressed the belief that the actual numbers were much higher. A statistician hired by the Courant, using an unspecified methodology, estimated that the annual rate of deaths could range from 50 to 150 in the country as a whole.

These troubling statistics and the dramatic case examples cited in the Courant's report have attracted the attention of other media around the country and of a number of members of Congress. Thus, for the first time, the prospect looms of federal legislation regulating seclusion and restraint practices, previously— except for restraint in nursing homes— the exclusive preserve of the states. At this writing, three bills have been introduced in Congress to monitor or restrict the use of seclusion and restraint in psychiatric and other settings.

Perhaps not surprisingly, two of the bills have been introduced by Connecticut's senators, Joseph Lieberman (D.) and Christopher Dodd (D.) (8,9). Taken together, their proposed statutes would grant all patients in facilities receiving federal funds "the right to be free from physical or mental abuse, corporal punishment, involuntary seclusion, and any physical or chemical restraints imposed for purposes of discipline or convenience." Restraints could be used only to ensure the physical safety of the patient or others and would, except in emergencies, require the written order of a physician. For a subset of facilities, the Secretary of Health and Human Services would be empowered to develop regulations to ensure appropriate staffing levels, training in restraint use, and notification of death.

The Senate bills emphasize reporting of deaths or "serious physical or psychological injury" to a variety of agencies, including national accrediting bodies, state licensure boards, the federally funded protection and advocacy agency in each state, and the federal government. One of the bills would limit reporting only to "unexpected occurrence[s]— unrelated to the natural course of the individual's illness or underlying condition," thus granting facilities some discretion as to when reports must be made.

A somewhat more aggressive approach is taken in the bill introduced in the House of Representatives by Diana DeGette (D.-Colo) (10). Titled the "Patient Freedom From Restraint Act of 1999," its criteria for when seclusion and restraint can be used are similar to those in Lieberman's and Dodd's proposals, with the added requirement that they constitute a "last resort" to be employed only when other less restrictive approaches have failed. A log of every instance of seclusion and restraint, its rationale, and the plans developed to avoid future use would be made available to the state's protection and advocacy agency, and aggregate statistics would be reported to the federal government. "Sentinel events," that is, events involving injury, would also be reported to the protection and advocacy organization. Substantial penalties for violation, including loss of federal funding and civil fines, are provided.

How should psychiatrists and other mental health professionals view these pieces of proposed legislation? All of these bills bear evidence of the haste with which they were prepared in reaction to reports in the media. One of the Senate bills, for example, limits the use of restraint to the protection of patients, but not of staff, visitors, or others; it also appears to abolish the use of seclusion altogether. The other effort from the Connecticut senators defines all restraint and seclusion as procedures "imposed for the purpose of discipline or convenience," having earlier precluded any use of restraint or seclusion for these purposes. But these imperfections in drafting will, it is hoped, be corrected as the bills move through the legislative process.

A more significant issue is the extent to which federal legislation is needed to regulate seclusion and restraint in psychiatric and related facilities. After all, almost all states have regulations governing these practices. In our system of delegated powers, we usually avoid federal intrusion on traditional areas of state activity, since state regulation is thought to be more flexible and better adapted to local conditions. Exceptions do arise, of course, when the states cannot or will not act to correct problems that have become apparent. Are the deaths and improper restraint practices reported in the Hartford Courant's series sufficient grounds to warrant federal intervention?

Some observers have questioned the data on which the estimate of deaths due to seclusion and restraint are based. For example, for the 125 cases in which reason for death could be specified, 26 percent were due to cardiac-related causes. Many cardiac-related deaths, as well as many of the other reported deaths, may reflect the consequences of extreme patient agitation rather than the improper use of seclusion or restraint. Moreover, although we know the estimated number of patient deaths, we do not know the denominator that provides the context— that is, the total number of restraint episodes, including the vast majority that did not result in death.

Still, even if the numbers are somewhat off the mark, it remains disturbing that so many patient deaths may relate to restraint practices. A reasonable federal initiative at this point might be to gather the data necessary to understand the full dimensions of the problem. This task could be accomplished by requiring reporting of patient deaths or serious physical harm attributable to seclusion or restraint and combining that data with aggregate data on the incidence of use of these techniques. Means must be found to protect the integrity of information generated by the usually confidential medical peer review process. In addition, focal studies in several jurisdictions could identify the problems that result in patient injury. On their own, the states would be unable to provide such national data, which is needed to establish a sound basis for determining if other federal action is required and what shape it should take.

Are further federal interventions required at this point? Uniform criteria for when seclusion or restraint may be applied run the risk of limiting the states' flexibility in this area. For example, it is not unreasonable for a maximum-security forensic facility, to which persons with histories of extreme violence are committed, to have a lower threshold for seclusion and restraint than would an ordinary psychiatric unit or group home. A single national standard would eliminate the states' power to decide that such differentiation in their approach to seclusion or restraint made sense to them.

Errors in the formulation of national standards also tend to be magnified 50-fold. The DeGette bill, for instance, would require that less restrictive approaches fail before seclusion or restraint is employed. However, in emergency situations, when violence threatens, there is no time to attempt other courses of action if patients and staff are to be appropriately protected. Moreover, many of the reported abuses occurred when existing rules were ignored; increased training and supervision may be the appropriate remedy here.

Congress has other options than to legislate these issues directly for the states. Operating less intrusively, it can instruct the states, as a condition of federal funding, to develop their own clear criteria for the use of seclusion and restraint, reporting mechanisms by which use and consequent problems can be tracked, and requirements for staff training. This approach would further the protection of patients and the proper use of these techniques while preserving the virtues of local diversity and adaptation to unique circumstances.

Of all the provisions in the bills, the one of greatest concern to hospital and facility administrators is the section of one of Senate proposals that would allow the Secretary of Health and Human Services to set national standards for staffing levels in covered facilities. In an era in which psychiatric units are being squeezed by cuts in payments from managed care companies and from the federal government itself, the prospect of such mandates— without the funding to implement them— is understandably perturbing. Flexibility in designing staffing levels would be utterly foreclosed. To curtail this flexibility would be a step too far, uncalled for by the current concerns about seclusion and restraint.

The media often perform a critical service in calling our attention to social problems others have neglected. No one believes that patients should be injured when restraints are applied or seclusion occurs, and the review provided by the Hartford Courant series may have a salutary impact, especially in states with lax regulation of these practices. But the media also gravitate toward sensational anecdotes that have a way of distorting the bigger picture and encouraging precipitous actions based on inadequate data. One hopes that Congress avoids the temptation for such actions in addressing this complex issue and instead takes a measured and sequential approach.

Dr. Appelbaum, the editor of this column, is professor and chair of the department of psychiatry at the University of Massachusetts Medical School. Address correspondence to him at the Department of Psychiatry, University of Massachusetts Medical School, 55 Lake Avenue North, Worcester, Massachusetts 01655.