RE: 42 CFR Parts 441 and 483
Basis for our opinions
Risks associated with procedures for authorization
Risks associated with mandatory notification procedures
Costs of the notification requirements
Costs of the training requirements
Basis for our opinions
The August Aichhorn Center operates the most secure residential treatment facility (RTF) in New York State. This program, the only RTF in Manhattan, and the only one in New York City that accepts normal IQ teenagers, was specifically intended to serve youngsters who could not be handled in any other facility in the foster care, mental health or juvenile justice systems except for State hospitals or correctional institutions.
The Aichhorn RTF has been in operation for 10 years, and has routinely accepted without screening (and retained, without use of outside psychiatric hospital services) any adolescent referred by the State-operated Pre-Admission Certification Committee for New York City. Despite the extreme selection of the population, the program has never experienced a patient fatality, and there has been only one major injury to a patient (which was self-inflicted). Despite its easily accessible urban location, the program also has a very low rate of "elopements," largely because most residents feel fairly safe and comfortable there.
We have, thus, developed and demonstrated considerable expertise in safe management of very large, very behaviorally disturbed teenagers in an RTF environment.
We do not have, and have never had, a "seclusion" or "time-out" room or area, and we have never utilized mechanical restraint devices of any kind. We do not believe such interventions can ever be described as therapeutic, and we have successfully demonstrated that, provided adequate numbers of adequately trained staff are available, they are not necessary. We have no question, therefore, about the effect of the present regulatory proposal which appears intended to minimize if not eliminate the use of such approaches; ultimately, we would support an effort to eliminate them entirely, provided this was backed with fiscal provision for adequate staffing.
We believe it is appropriate that the regulatory proposal limits any application of physical force to the minimum necessary to insure the safety of all concerned, and that the autonomy and dignity of all patients be respected as far as possible at all times. Because we support this point of view, however, we are greatly concerned about the interim proposal's approach to regulation of "personal restraint," largely because we believe that the actual effects of this section of the regulation will in fact be the opposite of the stated goals of improved patient protection (viz the title of section 483.356).
Large, aggressive, impulsive and sometimes irrational children and adolescents can be extremely dangerous to themselves and to each other. It is common that some level of physical interaction is required to prevent residents from seriously injuring themselves and/or each other. While there is some inherent risk in such interventions, and while we agree fully that it is essential to carefully control and monitor their use to prevent overreliance on such techniques, the present regulatory proposal appears to ignore the risks involved in failure to appropriately institute "personal safety interventions" when they are required to avoid imminent risk of injury to the patient or others. Many readers of the proposal among our staff have commented that if our procedures were significantly altered in response to these regulations, the level of physical risk to residents (primarily from each other) would increase substantially.
The rationale for the original regulatory proposal [HCFA-2065-IFC], cites the Hartford Courant as the source of purportedly alarming estimates of as many as 50 deaths per year nationally of juveniles in psychiatric facilities of all types. However, the cost analysis section of the proposal estimates the national incidence of restraint/seclusion in residential treatment facilities alone at 282,000 per year [p.7156]. Thus, accepting this highest "statistical projection" of mortality (which is offered with minimal justification by the Courant), and even assuming that all of the deaths occurred in Medicaid-funded RTFs (which is obviously false), still only suggests a mortality rate from all types of restraint and seclusion of less than .00018 (.018%), which is extremely low for almost any physical intervention (i.e. administration of antibiotics or high school contact sports).
In fact, however, the actual rate is at least an order of magnitude smaller. The Courant's actual statistics reveal that the number of documented deaths of individuals under 21 nationally, in an 11 month period, in all types of facilities including group homes, summer camps, detention programs, psychiatric hospitals, and mental retardation facilities as well as residential treatment programs (accredited or otherwise) was twelve; of these, only five were in facilities that might have been RTFs. [10/11/98] Even assuming that all of these deaths occurred in the Medicaid-funded RTFs included in the HCFA-estimated 282,000 restraint/seclusions per year, the mortality rate for all episodes of all types of restraint and seclusion in these facilities would have been .000019 (.002%). Further, several of the five fatal events involved interventions other than "personal restraint," and some undoubtedly occured in non-Medicaid treatment settings, so the actual rate is even lower.
The objective conclusion is inescapable that, based on the information supplied by the regulatory agency itself, "personal restraint," whatever its limitations, actually carries with it a real but exceedingly small risk of death. On the other hand, it is also clear, that "personal restraint" is often the only way of avoiding serious physical harm to patients (including death).
A rational, ethically defensible risk-benefit analysis of any treatment intervention requires that the estimated risks caused by unnecessary use of the procedure be weighed against a corresponding estimate of the risks incurred by withholding the intervention when it would be beneficial. This balance determines the point at which we conclude that the failure to treat carries more risk than the treatment. Such assessments are the mainstay of clinical care, whether we are considering major surgery or over-the-counter medication. Thus we expect a fairly high level of unnecessary appendectomies, because the real but relatively small risks of this procedure are deemed less grave than the risks of ruptured appendix. We recognize that either choice may lead to a fatal outcome in some instances, but the risk of withholding surgery is greater than the risk of performing it. Similarly, appropriate planning to regulate use of "personal restraint" requires some estimate of the patient mortality (and morbidity) associated with underutilization of this intervention, to balance against the stated risks of the procedure itself. Nowhere in the regulatory statement do we find any information on the dangers of failure to intervene physically to protect patients, although it is (or should be) clear to all concerned that substantial risks of this kind do exist.
Rather than developing from an appropriate risk-benefit discussion, the present regulatory proposal appears to be based on three unsupported assumptions:
1) The primary source of dangerous and/or inappropriately violent treatment of patients is non-professional staff, while professional staff and parents are a bulwark against such abuse.
2) There are no "risks" to patients from underutilization of physical intervention in psychiatric treatment facilities.
3) The financial and organizational "costs" associated with increased surveillance, reporting and intervention, and modified training, will not divert important resources from patient care.
We are unaware of the basis for the first assumption, which appears to underlie the others, and we are at a loss to explain it. It is not supported by the regulatory document or by our own experience. The tone of the interim regulations suggests a belief that family members and certain credentialled staff will necessarily act as a check on abusive action by other staff (child care workers). In fact, the opposite is often the case. Most residents of our RTF have been severely neglected and/or abused by their families. Many parents remain exceedingly punitive, and sometimes brutal, in their attitudes; some have asked to be notified when their youngster misbehaves so they can come and "deal with him myself." We do not find it surprising that a recent nationally-reported case of patient death during a "therapeutic" holding ("re-birthing") session took place in the presence of the patient's family; families of children in private facilities have repeatedly defended the institutions' use of various behavior modification techniques that many would consider abusive. The belief that involvement of certain professional staff will automatically control inappropriate use of force in patient management is also without empirical basis, as discussed below. The Hartford Courant series, which seems to offer the only empirical basis for the regulatory effort, indicates, in its first and most sensational example, that the youngster involved was restrained (in a hospital, not an RTF) by three staff members, one of whom was a nurse [10/14/98]. This apparently did not influence the fatal outcome.
If the first assumption is at best controversial, the second and third are demonstrably false. Embracing them uncritically will result in significant harm to patients. We discuss them specifically below.
Risk-Benefit Issues in the Interim Proposal
A. Risks associated with procedures for authorization
Essentially, the present formulation of the proposal relies on mandated authorization by medically credentialled staff (defined in New York as physicians, nurses, psychologists, and social workers) to limit inappropriate or dangerous use of physical force by child care workers. This approach is unworkable and, worse, counterproductive.
1. There is no evidence that professional credentials of this type are based on any training, skills, or interest in predicting, avoiding, "deescalating," or safely controlling physical confrontations. In fact, training and experience in education is probably the single professional credential best correlated with experience in maintaining order among groups of adolescents.
The regulation offers no guidance at all about what criteria the professional is to use in assessing the need for restraint or seclusion. The assumption appears to be that, merely on the basis of a degree or credential, the individual will be equipped to determine how dangerous is the situation, and how best to ameliorate it.
The objective evidence that does exist suggests that experienced correctional officers and police officers are more successful in predicting violence than mental health professionals. This is not surprising since developing these clinical skills is a central element in the career success of these individuals. The same is, of course, true for child care workers in an RTF for seriously disruptive youngsters, and probably for successful teachers in urban high schools.
In fact, mental health professionals, who typically find violent physical interaction distasteful, are much more likely to prematurely authorize use of force by (other) staff against patients than are competent and experienced child care workers who generally exercise maximum efforts to avoid becoming involved in potentially dangerous interactions, and who often pride themselves on their relationships with and ability to "talk down" very disturbed residents.
The regulation appears to contemplate an increase in "professional" medical staffing to allow oversight and authorization of restraint at all hours of the day and night. The above observation on the knowledge, skills, and motivation of the designated professionals is particularly true with respect to members of these professions who, with attractive jobs readily available in all areas, would choose specifically to work quiet (i.e. night) shifts with the primary job purpose of authorizing restraint and seclusion. It is exceedingly unlikely that the members of the mental health professions who choose to take positions where they will have minimal opportunity for patient contact, and where their primary job responsibility will be authorizing use of force by other staff, will exercise any meaningful role in moderating such use of force. On the contrary, their presence is likely, if anything, to increase and legitimize these interventions.
2. In real emergency situations, which the regulation appropriately designates as the only ones in which force should be used, people "on the spot" must be able to react immediately. Requirements for prior supervisory approval cannot work. They will either be ignored, or they will result in grave danger to patients. Consider two routine examples:
A young woman flings a tape cassette player to the floor, stamps on the case, picks up a battery, swallows it, and begins cutting her wrist with a broken fragment. This entire sequence of events may take 30 seconds. At present, our incident review of such a case would focus on what the child care worker did, or failed to do, when the player hit the floor. We would expect immediate, physical intervention to prevent the later steps. Under the regulation, the child care worker's responsibility would apparently be limited to seeking authorization to intervene. Conscientious staff will ignore the regulation. Other staff will allow serious self-injury to occur--and demonstrate complete callousness to the patient.
Two young men become involved in a verbal argument. Staff intervene and move them apart. Suddenly, one turns back, runs towards the other, and punches him in the face. At present, before the second boy can retaliate, staff pull them apart. Under the regulation, staff would apparently watch them fight while seeking permission to intervene. Under the regulation, the child care worker's responsibility would apparently be limited to seeking authorization to intervene. Conscientious staff will ignore the regulation. Others will allow a serious assault to proceed--and demonstrate complete callousness to the patients.
The comments on the relevant section of the regulation (483.358) appear to obfuscate the issue raised here by stating that an order must be obtained, "at the time," that an intervention is initiated. This is intellecutally dishonest, and provides no guidance at all to staff. Are they to intervene and then seek an order, or to seek an order and wait to intervene? Supposing they initiate intervention and the order is subsequently refused? Will they then be disciplined, or charged with assault? If this is the case, what is the likelihood staff will risk their own welfare and safety to intervene promptly in dangerous situations?
Or will the order automatically be written post hoc, in which case there is no point requiring it? If the response is that the order will always be written if the intervention was "reasonable", then, in fact, the judgement of line staff about the "reasonable" need for intervention is re-affirmed, and the entire exercise is pointless.
Perhaps the intent of requiring line staff to obtain a restraint order is not to require prior substitution of "professional" judgement, but to insure that qualified supervisory personnel are involved as promptly as possible in overseeing and de-escalating an emergency situation. That is a reasonable objective, but it is addressed separately by the regulatory requirement for documented "face to face" evaluation.
As it stands, the requirement for an order "at the time" of intervention will create a substantial risk of harm to residents due to delay in staff action. It also inevitably suggests (to residents as well as staff) a complete lack of confidence in the people who are trained, paid, and dedicated to providing on-going care.
3. The regulation implicitly encourages transfer and diversion of psychiatric patients out of RTFs and into more restrictive facilities, particularly those of the correctional system, which routinely utilize physical coercion far beyond the scope of "personal restraint," and are outside the jurisdiction of the regulators. This is a potentially very grave threat to the welfare of very large numbers of severely ill psychiatric patients.
The movement of severely psychiatrically ill individuals from inpatient psychiatric facilities to prisons is by now well documented. Since the regulatory proposal relies heavily on reports in the press, reference is made to Fox Butterfield's essays, "Concern Rising Over Use of Juvenile Prisons to 'Warehouse' the Mentally Ill" [New York Times, 12/5/2000], and "Prisons Brim With Mentally Ill, Study Finds," [7/12/99]. The latter reports that the U.S. Department of Justice estimated there were 283,800 inmates with mental illness, about 16% of the nation's prison population. These individuals have essentially no protections against punitive, coercive, arbitrary and/or negligent use of seclusion, restraint, and significant physical force.
There are already strong incentives for mental health agencies to avoid the costly and relatively unrewarding task of providing intensive inpatient supervision to behaviorally disruptive patients, turning limited resources instead toward the much larger, more lucrative, and more politically active group of patients who can manage well and productively with relatively cheap and simple outpatient treatment. We presume it was not the intention of this regulation to accelerate the trend of moving the severely ill out of regulated treatment facilities into institutions where oversight and control of restraint and seclusion is minimal. However, the tone and content of the regulation strongly suggest that the regulators believe any effort to hold back an agitated youngster is highly controversial (viz. the sections on notification of rights, notification of family, etc.), and therefore that individuals who behave dangerously without physical intervention are generally not appropriate for the population of a psychiatric residential treatment facility. We believe, on the contrary, that outpatient rather than RTF care is the approach of choice for almost all youngsters who do not require at least occasional physical interventions, and that only individuals who require substantial external controls should remain in inpatient psychiatric facilities.
Our program is currently involved in extended negotiations with a Rockland County criminal court to avoid the transfer of one of our patients to the county jail. The young man is unquestionably chronically and severely psychotic, requires large doses of medication, and is extremely contrite about his crime: "assaulting" a staff member while he was a patient in a State Children's Psychiatric Center, before transfer to our RTF. The interim regulation on personal restraint, as written, is an open invitation to psychiatric facilities to use police and the criminal justice system (which do not require an order from any oversight agency or credentialled staff to use punitive mechanical restraints, seclusion, and even deadly force) to contain assaultive behavior in psychiatric treatment facilities. It is also an incentive to transfer long-term care of these individuals to the criminal justice system.
Additionally RTFs may simply expand the common pattern of refusing to admit individuals who may become assaultive or physically agitated. Since it is clear such patients are no longer retained in long term hospitals, they will enter the general population briefly until they are arrested, sometimes after causing substantial harm to others, after which they will move directly into the criminal justice system. This is already a common sequence of events.
B. Risks associated with mandatory notification procedures
Although the regulatory comments indicate that it is inappropriate for the agency to practice medicine, the patient care mandates in sections 483.366 and 483.370 are, in fact, prescriptions for specific clinical interventions which substitute mandated actions for clinical judgement. Like any such blanket mandates, they will cause significant harm to some patients if applied across the board, but the regulation completely fails to recognize or acknowledge this, presumably because the formulators are themselves unaware of the risks. The underlying assumption that physical intervention is always or almost always tantamount to abuse is most inappropriately apparent in this section. The regulators should recall that whenever a "personal safety intervention" occurs appropriately, it reflects severely inappropriate behavior by the patient. Once they have calmed, many patients are seriously embarrassed and ashamed by such behavior, and emphatically do not want it reported to their families.
1) The regulations (483.366) require notification of family, without the consent of a minor patient, whenever a "personal safety intervention" occurs. In fact, family notification may or may not be appropriate. Patients are to be allowed no discretion in this matter.
Many patients do not want their parents notified when they are involved in problematic behavior, fearing, sometimes correctly, that this will only further alienate the families, who not infrequently retaliate by cancelling planned visits, etc. While we ordinarily attempt to facilitate communication between residents and their families about all troublesome matters, we also try to be respectful of patient autonomy and privacy. Where the outburst has been brief, and no harm has occurred, it may be appropriate not to involve a parent who is already firmly convinced that the child is "no good."
The infringement on patient autonomy in this section is particularly troubling not only because it is inherently inappropriate, but also because the apparent regulatory objective (not allowing the facility to "conceal" episodes of potentially abusive behavior) can so easily be addressed in a completely unintrusive and uncontroversial manner by specifying that all RTF patients should have reasonably frequent opportunity for unrestricted communication with their families or other interested persons (i.e. law guardians, etc.). Residents of our RTF have access to the facility telephones at least daily (often much more frequently), and in our experience those who wish to involve their parents or others in complaints about their treatment are able to do so usually within hours of the event. (Sometimes an offer to make telephone contact with a trusted adult is a useful way of calming an agitated resident.) Additionally, all residents have access to the mail.
It would be appropriate to specify, as a regulatory safeguard, that such opportunities for uncensored communication with family members must be afforded to all RTF patients. However, we see no basis for believing that it is either useful or ethical for a regulatory policy to override patient rights to autonomy and privacy in the matter of communication with families by forcing disclosure of events that patients would prefer to keep confidential. Over-riding patient requests in this matter requires a careful clinical judgement made on an individual basis, and should not be mandated by blanket regulation.
2) Almost equally troubling are the requirements set forth for post intervention meetings (483.370). In general, the inclusion of this section implies that the regulators have no confidence in the ability of RTF clinical leadership to recognize and organize basic staff and patient communication, and this section is an attempt to substitute regulatory mandate for clinical judgement. In addition, the mandate to hold the meetings within a specific time frame is inconsistent with the practices expected by New York State investigatory authorities which specifically advise against organized discussions among the participants in alleged instances of abuse until they have been individually interviewed by an independent investigator. Since, as an essential element of patient protection, any patient can trigger an investigation at any time by simply stating that s/he is alleging abuse, and all such complaints, even those made in the heat of a restraint situation, must be documented and independently investigated, the provisions of section 483.370 appear to undermine a significant existing patient protection procedure.
Part A, requiring participation of the patient in a meeting with all the individuals involved in his/her restraint, is an excursion into the prescription of clinical practice that cannot but inflict serious psychological suffering and possibly harm on many youngsters.
The regulation appears to assume that every physical interaction must become a major problem for all concerned. In many instances, where the loss of control is brief and no harm is done, youngsters may appropriately be allowed to simply let the incident pass, for later discussion with a therapist or unit leader, but without being forced into an extended and humiliating immediate apology to all concerned.
The regulation also seems to assume that some kind of harm has always been inflicted on a youngster who is restrained, and that some kind of apology or gesture of reconciliation by staff is required. If the restraint has been warranted by the regulatory criteria, the opposite is true. Frequently, staff who attempt to prevent a child from injuring herself are rewarded with an outburst of violent rage. The regulation mandates that an actively psychotic youngster who has ripped a handful of hair from the head of a staff member, torn off the top of her blouse and bra, and lacerated and bitten her chest, should meet with her and a group of her colleagues within 24 hours for a "debriefing". What will be the content of this debriefing? Since the regulators seem to assume that the staff involved will be insensitve and perhaps brutal, and they will certainly be hurt and angry, it would be safe to predict that this meeting will be an extended exercise in verbal humiliation, which may well end in another physical attack.
Obviously, competent treatment leaders will arrange for youngsters who have lost control and become involved in physical confrontations with staff (or with each other) to discuss the situation, first with trusted adults who may not have been involved (often a primary therapist or unit leader), and then, often along with the trusted adults, with the object of the conflict. The hope is to effect some mutual understanding, and, if possible, reconciliation, and to teach youngsters that verbal as opposed to physical expressions of their feelings are possible and effective. If the regulatory agency does not believe that RTF clinical leadership is capable of recognizing this need, and of formulating individually appropriate plans for who should meet with a patient in what order on what schedule, the regulation should not permit the facilities to function at all.
Further, with respect to section B, requiring documentation of staff meetings to discuss restraint episodes, it is difficult to see how the agency can entrust RTF clinical leadership with any aspect of patient care if they do not trust this leadership to arrange appropriate staff discussion of clinical problems. The mandated meetings are an infinitesimal part of the challenge of creating and maintaining a viable treatment team and a therapeutic milieu. We already hold several types of meetings of this kind on a regular basis. Multiple similar provisions of the regulatory proposal, such as 483.360 mandating discussion with the treating physician when an incident occurs, and 483.372, mandating prompt medical treatment for injuries, appear similarly gratuitous and/or demeaning in view of the ordinary "standard of care" requirements for any accredited medical facility. Part B of this section is pointless.
Medical and psychological follow-up management of a dangerous act by a patient is not intrinsically different from management of any other behavior problem (i.e. refusal to do homework), and it is inimical to patient welfare to specify by regulation the details of clinical practice (i.e. who should attend a meeting at what time).
Cost-Benefit Issues in the Interim Proposal
The same uncredentialled staff whom the regulators will not entrust with judgement about whether to institute a simple, low-risk physical restraint without consultation, are apparently being entrusted with judgement about whether to independently institute CPR, an intervention carrying substantial physical risk--even though there is a separate mandate that appropriate medical staffing be available!
The regulatory requirement for involvement of nursing staff (who are required to maintain certification in CPR), the requirements for use of only safe methods of restraint, and for staff training in these methods, and the overall regulatory intent of minimizing episodes of restraint altogether, provide more than adequate protection against any accidental cardivascular death that would be preventable by non-medical staff use of CPR. Inappropriate attempts by inexperienced, minimally trained persons to begin unnecessary CPR carry substantial risks of morbidity and mortality, and are unjustifiable when qualified medical staff are available. We believe that universal staff certification in CPR will not benefit a single RTF resident, and challenge the agency to produce any data to the contrary. The complete absence of rational support for increased CPR training of non-medical staff in RTFs is dramatically underscored by a self-laudatory editorial in which the Courant congratulated itself on the "improvements" its own reports had prompted at Elmcrest Hospital: "...thanks to the bitter lessons of his death, one life has been saved and another has been improved.. Two months later, staffers who had just undergone state-ordered CPR training saved an adult patient whose heart was failing." [10/15/98]. This fortuitous event, if true, was, of course, irrelevant to the RTF setting, which has no adult patients, unrelated to the issue of restraint and seclusion, which was not involved, and, in fact, had nothing whatsoever to do with psychiatric treatment of any kind. Resort to such chicanery reflects the level of intellectual integrity of the entire discussion.
Further, the cost analysis of the training requirement does not recognize the alternative, relevant training functions that will be supplanted by the on-going CPR certification. If the costly staff time involved in training is devoted instead to understanding of psychiatric diagnoses, purposes and effects of psychotropic medication, normal child development, understanding of community issues, methods of avoiding confrontation, etc. etc. (as ours now is), the immediate and long-range benefit to patients and staff is substantial. The loss of this training is the real unstated cost of the mandate for CPR certification.
We recognize that the regulatory agency is under substantial political pressure to produce evidence of its concern about abuse of patients in psychiatric facilities. However, responsible government action, like responsible medical treatment, should reflect a reasoned assessment of the severity of the problem, and of the risks and costs as well as the benefits of any proposed intervention. While it is easy and tempting to make demagogic use of press surveys, the documented scope of the problem of abuse of children in psychiatric facilities is actually very small. It is dwarfed by the unquestioned frequency and severity of abusive incidents (including fatalities) occurring in various congregate foster care institutions, in non-Medicaid private treatment centers or schools, in juvenile justice facilities, and in homes, all of which are beyond HCFA's jurisdiction. Transferring children to more dangerous settings would be an ironic outcome for an effort to assure them of greater protection.
It is certainly appropriate for the regulatory
agency to explore ways of addressing the problems of restraint and seclusion.
As indicated in our introduction, we are sympathetic to the idea of eliminating
all seclusion and mechanical restraint because we believe they can never
meet the appropriate criteria for emergency intervention. However, we are
gravely concerned that elements of the present regulatory proposal are
counterproductive and will substantially endanger RTF patients, erode RTF
patient rights, reduce the general quality of RTF care, and ultimately
displace the most vulnerable RTF patients into much less therapeutic and
much more restrictive and intrusive types of unregulated care.
Michael A. Pawel, M.D.