State of Connecticut Child Fatality Review Panel’s
Investigation into the Death of Andrew M.
released: May 7,1998

http://ct.gov/oca/lib/oca/andrew_m.doc

State of Connecticut Child Fatality Review Panel’s
Investigation into the Death of Andrew M.
released: May 7,1998

Part I:  The Immediate Circumstances
 

Child Fatality Review Panel Members

Linda Pearce Prestley, Esq., Chairperson Child Advocate
John Bailey, Esq. Chief State's Attorney
Chief Leroy Bangham Farmington Police Department
H. Wayne Carver II, M.D.  Chief Medical Examiner
Gary Fitzherbert Executive Director  The Glenholme School
Leticia Lacomba, M.S.W. Regional Administrator  Department of Children and Families
Betty S. Spivack, M.D. Pediatrician

Consultants:

Suzanne M. Sgroi, M.D. Executive Director  New England Clinical Associates
Michael A. Nunno, D.S.W. Senior Extension Associate
Martha J. Holden, M.S. Senior Extension Associate  Family Life Development Center
College of Human Ecology, Cornell University

Office of the Child Advocate Staff:

Barbara J. Claire, Esq. Associate Child Advocate
Denise L. Scruggs Administrative Assistant

·         Under no circumstances, should the physical restraint of a child include compression of the child’s thorax by the weight of an adult.

·         Staff response at the Facility in which Andrew died reflected an inadequate behavior management program.

·         The Facility utilized an outdated physical restraint training program that did not conform to currently-accepted standards established by contemporary training programs.

·         Although not necessarily a contributing factor to Andrew’s death, the Facility’s staff response to this medical emergency was inadequate.

·         Although not a contributing factor to Andrew’s death, the treatment plan at the Facility lacked sufficient direction regarding the use of physical restraints on medically compromised children.

·         The Department of Children and Families should have conducted an assessment of behavior management programs and physical restraint policies affecting children under DCF’s care, after the death of Robert R.

·         The Department of Children and Families should promulgate regulations and policies that address the development of appropriate physical restraint policies for use in the facilities that they license and in the facilities in which children who are under the care and custody of DCF are placed.

·         The Department of Public Health should promulgate regulations designed to develop standards for behavior management programs and physical restraint policies in the children’s facilities that they license.

·         Neither the Facility nor the Department of Children and Families ensured that Andrew was advised of his right to a hearing and his right to an attorney upon involuntary admission to a psychiatric facility.

Table of Contents


 

INTRODUCTION

Pursuant to Connecticut General Statutes sections 46a-13l (b) and (c), the Connecticut Child Fatality Review Panel is mandated to review the circumstances of the death of a child who has received services from a state department or agency addressing child welfare, social or human services or juvenile justice. After a preliminary examination of the facts in this case, the Child Advocate, in her role as Chairperson of the Panel, convened a Fatality Review Panel meeting on March 26, 1998 to review the circumstances surrounding the death of Andrew M., a child who was legally committed to the care and custody of the Department of Children and Families (DCF) when he died at the "Facility" on March 22, 1998.

The purpose of this review is twofold: In Part I, the Panel seeks to identify the immediate circumstances surrounding, and particularly to isolate those factors playing the most prominent role in, the death of this child, with apposite recommendations. In Part II, the Panel assesses the less immediate circumstances surrounding the death of this child, such as the predicates for state involvement in Andrew’s case, the services and interventions provided, and the social work and therapeutic management of his case, and again provides relevant recommendations. Not only does this review of the broader circumstances put the first tier of inquiry into context, but also facilitates a better understanding of how this child might have been better served and protected by the system as a whole.

Part I includes an examination of the events which led up to the untimely death of this child; the institution’s policies and procedures on behavior management, including the use of therapeutic holds; and the clinical responses to this incident. It also addresses the role of state agencies in the protection of children in care in mental health and other facilities across the state. Part II, which will be released by the Fatality Review Panel on or before June 22, 1998, will include a consideration of the issues raised by Andrew's social, psychological and medical history; his history with DCF, Connecticut's child protection agency; and the efficacy of the therapeutic and medical management of his case.

In conducting its review of the above-described matters, Panel members took the sworn testimony of a number of witnesses (psychiatric facility employees, DCF employees, a medical doctor and an employee of a private social services provider), and invited them to provide information and their own recommendations for the Panel's consideration. Additionally, Panel members reviewed: all records and documents pertinent to this case, including the child protection records of Andrew M., his mother, and his siblings provided by the DCF; records provided by the Department of Public Health (DPH) pertaining to the "Facility;" Department of Social Services (DSS) records; Judicial Department court records regarding Andrew M. and his siblings; extensive inpatient and outpatient records of Andrew M. and his siblings provided by a number of hospitals and medical doctors; records from the Office of the Chief Medical Examiner; police reports and statements pertaining to the death of Andrew M.; records provided by numerous social service provider agencies regarding Andrew and members of his immediate family; and finally, the educational records of Andrew M. Additionally, Panel members interviewed Andrew’s court-appointed counsel by telephone, and made an on-site visit to the Facility as well. The Panel also requested and received extensive records from shelters, group homes, residential facilities, detention centers, and hospitals throughout the state on the policies and procedures pertaining to the issue of physical restraint of children in those facilities.

In the course of its investigation, the Panel retained the services of Suzanne Sgroi, M.D., the Executive Director of New England Clinical Associates, for her expertise in residential reviews and physical restraints; and Michael Nunno, D.S.W. and Martha J. Holden, M.S., Senior Extension Associates at the Family Life Development Center, School of Human Ecology, Cornell University, who are experts in therapeutic physical restraints and training techniques. Their combined expertise and experience have been invaluable in assisting the Panel members to understand the theoretical, practical and technical aspects of the therapeutic physical restraint of children.
 

The Fatality Review Panel wishes to note publicly that it received complete cooperation in its investigation from the DCF, the DPH, the DSS, the Department of Mental Retardation (DMR), the Department of Mental Health and Addiction Services (DMHAS), the Judicial Department, and private individuals and entities (including the Facility’s administrators and its employees) involved with Andrew M. and his family. The Panel is most appreciative of this cooperation.

After intake of the above-described body of material, and after extended discussions with experts, and extensive testimony, the Panel members shared their findings with one another and drafted this report. The review, the findings, and the recommendations are the gravamen of this public document. The names of specific individuals, specific service providers, private agencies and hospitals have been omitted from this report for reasons of confidentiality. It is not the intent of the Fatality Review Panel or the Office of the Child Advocate to assess guilt or find negligence by or of any individual, institution or agency. It was the Panel's strong sense that the focus of its investigation should be on Andrew, his involvement with state agencies, and on the systemic issues raised by this child’s tragic death.

Since recollections and memories are not completely reliable, the facts as set forth below represent the Panel’s best efforts at piecing together the history of the case as a whole, as well as what happened on the day of the child’s death. Although the reliability and content of accounts and records may vary to some degree, the Panel, after a comparison of sources where possible, believes that this is a reasonably accurate account.

SUMMARY STATEMENT OF FACTS

Andrew M. was born on December 6,1986, and first came to the attention of DCF (then known as the Department of Children and Youth Services) prior to his third birthday, as a result of medical neglect referrals. His family has had sporadic involvement with DCF since that time over concerns of chronic abuse as well as medical, physical and educational neglect. While in the care of his mother and grandmother, Andrew suffered three separate eye injuries resulting in the complete blinding of his left eye at the age of three.

At the time of his admission into the Facility, Andrew was eleven years old and had experienced three changes in his legal guardianship, a host of serious injuries and illnesses, a period of commitment and extensions of commitment to DCF, repeated inpatient psychiatric hospitalizations, placement in three traditional foster homes, placement in one therapeutic foster home, a day treatment program, and partial hospitalization programs. Andrew was a child described by many as "sweet" and "endearing," who was eager to please and wanted to learn to read. He was also a child who had an extensive history of acting out, sometimes violently, and of planning harm to others, of acting on those plans on at least one occasion, of threatening suicide, of having to be physically restrained in school and in psychiatric facilities (on at least twelve and four occasions respectively), and of running away. He had been diagnosed on several occasions with "oppositional defiant disorder," "conduct disorder: and "intermittent explosive disorder," disorders in which strongly imposed authority is frequently met with aggressive behavior and uncontrollable rage.

On March, 19, 1998, Andrew was admitted to the Facility under a Physician’s Emergency Certificate (PEC) from Hospital A, after he exhibited threatening behaviors against another child in Foster Home A. At the Facility, Andrew was examined by a psychiatrist and admitted to the inpatient program on the S Unit. A staff therapist was assigned to Andrew's case, and he began to gather background information on Andrew. The therapist worked with Andrew two times in group therapy, but had not engaged in individual therapy with Andrew up to the time of his death.

On the morning of Sunday, March 22, 1998, Andrew became involved in a series of escalating exchanges with Mental Health Worker 1 (MHW 1) which led to Andrew’s removal to the time-out room. Once in the time-out room, another series of escalating confrontations took place leading to Andrew being placed in a physical restraint by MHW 1, who was then assisted by MHW 2. This physical intervention resulted in Andrew’s untimely death.

The Chief Medical Examiner has ruled that the cause of Andrew's death was traumatic asphyxia, resulting from compression of the chest due to the weight of an adult individual applied during a so-called "therapeutic restraint hold." The manner of death has been ruled accidental. The police investigation is continuing. The DCF and DPH investigations are reportedly complete but have not yet been released to the public.

DISCUSSION OF ISSUES

Behavior Management and Therapeutic Intervention

The primary focus of this report is to assess the events that occurred surrounding the death of Andrew by looking at the policies and procedures relative to behavior management and therapeutic intervention in place during his time at the Facility in order to determine whether those elements played a part in his death. A retrospective view of the events in this case allowed the Panel to form conclusions which are the basis for its recommendations.

Background Information

On March 17, 1998, Foster Mother A informed DCF Social Worker A that she had learned that, three weeks prior, Andrew had put cleaning disinfectant on his younger foster brother's toothbrush because he "wanted him dead." On March 19, 1998, Andrew informed his mother by telephone that he was going to kill that same foster brother. Concerned, Andrew's mother immediately informed the foster family and Andrew was taken by Foster Mother A to Hospital A. A physician at Hospital A signed a Physician's Emergency Certificate (PEC) based on "dangerous behaviors," authorizing Andrew's immediate involuntary admission for inpatient psychiatric care. Because there were no beds available at Hospital A, Andrew was transported by ambulance to the Facility, which is owned by Hospital A.

The facility is a psychiatric hospital for children and adults with branch programs around the state. Its purpose is to provide the highest level of care to individuals with psychiatric and chemical dependency problems. There are two adult units and two children's units at the main campus. The children's units are known as the P Unit and the S Unit. There are a total of fifty-four beds available in these two units. The P Unit generally serves the adolescent population, ages 13-18, and has thirty beds available. The S Unit serves children between the ages of 5 and 12 and has twenty-four beds. When the S Unit is full to capacity, children may be admitted to the S Unit, but be assigned rooms on the P Unit.

The facility primarily serves a population of children who exhibit depression or behavior that is harmful to themselves or others, who have been diagnosed as psychotic, who have been sexually abused, or who exhibit conduct disorders. The average length of stay for children at this Facility is approximately eight days unless there is a problem with transition to another placement. Over 80% of the population on the children's units are children involved with DCF, who are receiving Title XIX benefits.

The time-out room where the incident occurred is roughly triangular, measures ten feet and twelve feet at its greatest dimensions and, at the time of Andrew’s death, was covered on the walls and floors with blue foam padding, approximately three inches thick. Prior to Andrew's death, there was no lock on the door to this room. There is a round outside skylight in the ceiling, and a small window in the door. There is a mirror positioned in the room to allow for observation of any part of the room by someone sitting outside the room. The time-out room is used in this unit for circumstances ranging from children voluntarily taking breaks from stressful situations, to children being involuntarily restrained and secluded.

On the day of Andrew's death, the S Unit was staffed by five nursing personnel (four mental health workers and one nurse), some of whom were full-time and some part-time staff. On weekends, a nursing supervisor oversees both the children's units and the adult units at this facility and was supervising on the weekend of Andrew’s death. On the day of Andrew's death, there were 26 children on the S Unit, with two of those housed on the P Unit.

On that Sunday, both Nurse 1, who was responsible for the S Unit, and Nurse 2, the supervisor, were weekend staff; accordingly, they each had only one day’s experience with Andrew prior to the incident. MHW 1 is a part-time employee who principally works on an adult unit. He had no prior experience with Andrew at all. In general, mental health workers who "float" to S Unit are not given direct patient care duties. However, MHW 1 had spent a great deal of time on S Unit in the preceding months and was regarded as "one of the staff." He had not been specifically assigned to Andrew, and therefore he had not reviewed his chart.

Mental health workers at the Facility must have a minimum of three years of college education in a related field, three years of experience in a related field or an associate's degree in a related field, combined with one year of related experience. Nurses at the facility are required to be Registered Nurses.

At least one physician is available in the Facility at all times. On the day of the incident, the sole physician who was present was a psychiatrist. Clinical staff, which consists of psychologists and social workers, are available during weekend days, but were not present on S Unit when the incident occurred.

Staff training for mental health workers at the Facility consists of a two week orientation at the start of employment, followed by a six week probationary period during which the employee gets a "multitude of different levels of training." There are also required expectations for individuals in certain areas with some mandatory in-service training in subjects such as growth and development, infection control, fire and safety, and protective intervention techniques. There are also requirements for the amount of in-service training that needs to be accomplished on an annual basis. If an employee is assigned to the children’s unit, there is more focused training on children. Individuals who are only occasionally assigned to the children’s units do not get the same degree of training on children’s issues as the regularly-assigned staff. MHW 1 had received three of the four development training modules available.

Review of the Interactions Leading to Andrew’s Restraint and Death

On March 21, 1998, the evening before the incident, Andrew got into an argument with another child while they were playing in the gym. He was given a choice of consequences and chose to go quietly to his room. Although assigned to the S Unit, Andrew slept on the P Unit because of a lack of beds. The morning of March 22, 1998 was Andrew's fourth morning at the facility.

On weekends, breakfast is brought to the unit and the children get their trays and sit at tables to eat the meal. Andrew was brought to the S Unit at 7:30 a.m. Presumably, he had awakened earlier and had probably already had been through a morning routine of washing and getting dressed on the P Unit. When he arrived at the S Unit he was "not talking but he didn't seem upset." Unlike the other children on the unit, Andrew had no room in which to wait for the staff to begin the weekend morning program.

Consequently, it is possible that he felt more like an outsider than the other children who actually slept on the unit. Andrew was asked to rearrange some furniture on the unit, "which he did with no problem." Then, he watched television until MHW 2 called him and two other boys to breakfast first "as a reward for being quiet."

After getting his food, Andrew sat down next to another child. MHW 1 circulated around the room making conversation with the children. He then sat at Andrew’s table and made some attempts to engage the children in conversation but Andrew ignored him and "kept staring at his bowl". MHW 1, receiving no response, interpreted Andrew's silence to his conversational approaches and questions as hostile and challenging. The child next to Andrew then told MHW 1 about the incident in the gym the previous night. Andrew’s response was to put his open hand in front of the other child's face and punch it with his fist. MHW 1 conveyed to Andrew that such a gesture was inappropriate, and Andrew responded that the other child should "stay out of his business". MHW 1 felt that Andrew was angry and, wishing to avoid a physical altercation between the children, asked Andrew to move to another table. When Andrew refused to move, he was ordered to move while MHW 1 counted to three. Andrew again refused and, after another verbal exchange, Andrew then stood up and MHW 1 held his right wrist and right elbow and employed an "escort hold" to take him to the time-out room. There are conflicting reports regarding whether or not Andrew struggled on the way to the time-out room, or whether he appeared angry at that point.

Once in the time-out room, Andrew was told that he needed to remove his shoes, a standard policy at this facility when a child is in the time-out room. He kicked off first one shoe and then the other, each of which hit the opposite wall, narrowly missing MHW 1. When ordered to move from the left wall to the right corner, a command which was based not on policy but rather on practice at the Facility, Andrew refused. At this point, MHW 1 told Andrew that he would count to three while Andrew complied with the order. When Andrew failed to comply, MHW 1 approached Andrew again and applied the same type of escort hold as used previously. Andrew erupted into an angry outburst and began to struggle violently with MHW 1.

MHW 1 put Andrew's right arm down to the front of Andrew’s waist and, as he was attempting to grab Andrew's left arm, MHW 2 came into the room. She had been on the telephone for less than ten minutes and hurried to the room when she heard a scream. On her way to the time-out room, she passed Nurse 1 who was eating cereal at the staff table and who appeared not to hear the sounds coming from the room, although she was only footsteps away. MHW 1 got Andrew's left arm down to his waist and was behind Andrew in an effort to do a "one-man takedown." MHW 2 grabbed Andrew's legs and MHW 1 went down to his knees. Then, all three went to the floor with Andrew positioned on his right side, with his arms crossed in front of him, underneath his body. MHW 1 extended his own body over Andrew applying pressure with his chest on Andrew’s left side. MHW 2 crossed Andrew’s ankles and held down his legs. Andrew continued to scream and struggle for the next several minutes, saying "Get off! Get off!" About a minute into the hold, Andrew expelled gas. Approximately two to three minutes after MHW 2 had entered the room, Nurse 1 came in and observed the scene, noting that Andrew’s face was turned toward the door and away from the two people who were restraining him. She stated that she would check on Andrew's medications to see if a p.r.n. (as needed) medication was ordered for chemical purposes, and left the room. MHW 1 and 2 attempted to calm Andrew down by talking to him. Andrew's head was going back and forth from left to right and he continued to scream and struggle.

After a few more minutes, Andrew stopped screaming but continued struggling. The mental health workers thought he was starting to calm down and MHW 1 let some pressure off. The workers discussed aloud that Andrew was a "Level 3" (pursuant to the Facility's behavior system), and should not be behaving in this manner, and that he would probably be dropped down to a "Level 2." Andrew’s face was then turned so that his right cheek was on the floor. MHW 1 asked MHW 2 what Andrew needed to say to get out of the hold. MHW 2 responded that Andrew needed to "commit to safety." Andrew was asked some questions, but he did not respond. The workers continued to talk to him for one to two minutes after they smelled urine, attempts were made to rouse him and they turned him over onto his back. They observed urine on his pants, his eyes nearly closed, and his mouth was open. He did not respond to questions or commands. When they sat him up, his head flopped. MHW 2 attempted to find a pulse, was unsuccessful, and then called the nurse.

During the next few minutes, Nurse 1 attempted to find a pulse, asked for a stethoscope, checked for a heartbeat, and began mouth-to-mouth resuscitation. An ambulance from the Facility’s private service was called. Subsequently, 911 was also called. Other medical personnel including a doctor and Nurse 2, who was the nursing supervisor for the entire hospital that day, were summoned. Upon her arrival, Nurse 2 began chest compressions on Andrew. Upon the arrival of other CPR providers, Nurse 2 then coordinated the ambulance response. The ambulance arrived, and paramedics took over the emergency medical treatment, and transported Andrew to the local general hospital where he was pronounced dead.

Use of Physical Restraint

1.    Introduction
Contrary to popular belief, the use of physical restraint in caring for persons with mental illness or emotional disturbances is not an obscure topic. Indeed, an abundance of recent professional literature exists on the use of seclusion and restraint in treating children, adolescents and adults in psychiatric facilities. Numerous articles and studies about these topics have been published in the last two decades in journals devoted to psychiatry, psychology, child mental health, child welfare, developmental disabilities, education and juvenile correction. Most of the articles that are referenced reflect data and opinions about the psychiatric care of children and adolescents. However, the Panel also relied on a selection of landmark articles dealing with seclusion and restraint of adult psychiatric patients or persons with mental retardation. A bibliography may be found in Appendix D.

2.    Restraint in the context of a behavior management program 

A review of best practices indicates that physical restraint should never be used as retributive measure. It has a place as a last resort to prevent patients from injuring themselves or others. The purpose of the therapeutic program should be directed to preventing the need for physical restraint from arising; accordingly, incidents involving physical restraint should be rare. In addition to an acknowledgment that use should be infrequent, personnel who are interacting with potentially violent or suicidal children must also have adequate training to ensure that these rare events are conducted in a manner which is maximally safe for children and staff. When restraint is utilized, it should end as soon as safety of the child and others has been assured. Assuring safety does not require the child to utter a verbal formula; the staff members involved must judge the safety of the situation from the entire range of behavioral and verbal interactions which are occurring.

In some contrast to physical restraint, Miller, Walker and Friedman (1989) have described a reactive treatment technique called "therapeutic holding," which involves having three to four trained staff members contain a violent child by taking hold of the individual and forcing her or him to the floor (this is known as a "take down"). The staff members then immobilize the child’s limbs, giving "careful attention to the patient’s position and movement to avoid injury, " offer verbal reassurance and comfort, and may support the child’s head and neck with a small pillow. This technique is used until the child has calmed down. The child is then released and allowed to resume regular activities. The child is not sedated or put into seclusion.

When a child is admitted to a psychiatric institution, the psychiatric service of a general hospital or a residential facility, a physician must promptly review the pertinent medical history and conduct physical and mental status examinations. That historical review and examination will provide the initial information necessary to begin to formulate a treatment plan, including any limitations of standard practice. The process of assessment should continue with full input from the members of the team caring for that child. Every portion of the activities engaged by the child should contribute to the therapeutic plan. As the child becomes better known to the staff, strategies that work in fostering appropriate behavioral self-control should be identified and passed along from shift to shift. If a crisis situation develops which has resulted in the use of physical or mechanical restraint, the team should review the events and discuss what triggers existed and possible ways in which the situation might have been de-escalated, thereby avoiding the need for restraint and promoting the therapeutic advancement of the child.

The Facility has expressed the goal of becoming a "restraint-free institution." This goal has been achieved in several pediatric psychiatric programs cited in the pertinent literature. Irwin (1987) described an eight-bed psychiatric unit for children aged 4-11 years that never utilized seclusion and had a 2:1 child-to-staff ratio; they did occasionally use "safe, gentle holding". Masters and Devany (1992) reported on their experience in a unit for children ages 2-12 years that used physical restraint only once in twelve years; instead, the staff (at a 3:1 child-to-staff ratio) employed milieu therapy, extensive "talking down," and a graded series of time-out options, with a locked seclusion room as a last resort.

At best, the use of such restraint techniques in a proactive manner requires sufficient personnel to identify situations before they escalate and to intervene before a crisis has developed. Needless to say, these personnel will have more success if they are extensively trained in the use of these techniques and if they know the child well. The Facility had a 5.2:1 child to staff ratio on the day of the incident. MHW 1 had never seen Andrew before and the nurse in charge was also weekend staff. This ratio of children to staff appears high for a psychiatric facility that is expected to care for children who have serious degrees of emotional disturbance. It is unlikely that any program that deals with a population of emotionally disturbed children can be truly "restraint-free" with staff attention spread over a large number of children.

3. Training in de-escalation and restraint techniques

Any program committed to restraint as a rare intervention must devote considerable training time in teaching de-escalation strategies which enable a child to regain self-control. Such a program of training must involve role-playing, with the trainer demonstrating the effects of productive and counter-productive approaches to a difficult patient. The trainer must then critique responses of the trainees and hone their abilities to identify and de-escalate a situation. At least as much time should be devoted to de-escalation strategies as to teaching physical restraint techniques. Formal re-education should occur at least annually, but case-based teaching should occur much more frequently, with the program director and other clinical staff taking the lead in reviewing good and bad examples of interventions which have occurred on the unit.

Teaching of physical restraint techniques is predicated upon using a well-defined program, solidly based on current, up-to-date research which takes into account minimization of risks to both patient and staff members. It is not plausible for each institution to generate such a program on their own. There are several national programs which incorporate "train-the-trainer" as well as training courses. These programs allow the institution to develop a well-trained, externally certified cadre of trainers, who can then teach in their own setting. These trainers must update their knowledge periodically, so that their teaching continues to reflect current standards of practice, and also to prevent an evolution of individual practice in the institution which may arise without research based on a solid foundation.

The largest general program is Crisis Prevention Institute (CPI), which provides a twenty-four hour "train-the-trainer" program and a twelve-hour introductory program. This organization authorizes certification and re-certification of both trainers and providers. Another respected program is the Therapeutic Crisis Intervention (TCI) program developed at Cornell University. TCI differs from CPI in that there is more instruction in child development issues affecting de-escalation strategies. The nature of the restraints is also somewhat different. Both programs provide a consistent approach with emphasis on both de-escalation and safety. There are other programs available as well.

At the time of the incident involving Andrew, the Facility was using a training technique which was brought into the institution more than ten years ago (Protective Intervention Technique, or PIT). Since the program was obtained, all new trainers at the Facility have been trained by the existing in-house training staff. As a result, the program as used at the Facility has not changed appreciably over the years, except that some informal variations have crept into the routine practice within the Facility. There is no separate "train-the-trainer" program; new trainers observe and participate in at least three in-house training programs before they are certified, within the institutional structure, to train others. Safety considerations (including evaluation of the child during application of restraint and pitfalls for each of the restraint holds) are not emphasized. The expressed perception of at least one experienced trainer in this institution was that these techniques were safe; he was unaware of any previous injury more serious than bruises or rug burns.

Within the last several years, the Facility’s commitment to becoming a "restraint free" institution has resulted in an increased amount of training time devoted to de-escalation strategies. This has been accomplished by introducing a Management of Aggression training module and incorporating portions of this into the PIT training. Nevertheless, this still represents a small portion of the ten hours devoted to the initial training.

In fact, restraints are used frequently on these units. In the first seventy-five days of 1998, 132 incidents occurred which resulted in use of some sort of restraint, an average rate of 1.76 incidents per day. (This did represent a 33% decrease from the equivalent period in 1997 when there were 199 episodes of restraints, an average rate of 2.65 incidents per day).

The Facility began tracking mechanical restraint, such as use of a safety jacket, on the adult and adolescent units over four years ago, and they have been tracking the use of physical restraints on the pediatric unit for the past fifteen months. The intervention team is debriefed on the adult and adolescent units whenever a mechanical restraint is used; no such debriefing exists on the pediatric unit at this time.

Prior to Andrew's death, the Facility required all mental health workers to receive a yearly one-hour period of retraining in management of aggression and physical restraint. This is insufficient to afford an opportunity for staff members to practice de-escalation skills (negotiation, talking down, relaxation skills) or to correct any flaws in physical restraint technique which have developed since the time of the original training.

The Facility has indicated to the Panel that it is beginning a transition to CPI training. The "train-the-trainers" program will begin in May 1998.

4.     Dynamics of the incident leading to Andrew’s restraint

Our research reflects that most incidents of seclusion and restraint of children in psychiatric facilities occur at times when children are being asked to "shift gears" and move from one activity to another, e.g., after school, around 4:00 p.m.; at bedtime, around 9:00 p.m.; and first thing in the morning. This is the case on S Unit as well. Various authors postulate that these are times of increased staff-child interaction, accompanied by the stress of getting the children to cooperate with commands or directions from their caretakers as they move from one activity to another. It is also likely that children may act out or challenge the authority of their caretakers during periods of "downtime," times when there is no specific programming for them and they have been expected to amuse or soothe themselves. Numerous authorities have noted that incidents of violent behavior between staff members and children most often involve male staff persons and male children, particularly adolescents, at times of high staff-patient contact and low programming (Earle and Forquer 1987).

The scenario on the morning of Andrew's death is an example of the type of "downtime" that frequently is experienced by children in most institutions: hours may be spent each day waiting for an activity to begin, often with an expectation that children watch television as a way to pass the time. Unfortunately, that "downtime" often appears to be a contributing factor when children act out or have confrontations with the staff.

After Andrew made an aggressive gesture toward the other child and refused to move to another table, proper training could have supplied MHW 1 with alternatives to the measures that were employed. Unfortunately, each of MHW 1's interventions escalated the interaction in what amounted to a power struggle between him and the child. The interventions utilized with Andrew that morning contributed to a progressive escalation of aggression and counter-aggression. Rather than utilizing specialized approaches to de-escalate the situation at any one of a number of points, the responses by MHW 1 only served to escalate the situation. The Facility’s training does not provide sufficient modeling of appropriate de-escalation techniques to ensure their appropriate utilization in the incident described. Such techniques are not likely to be used unless there is adequate training and emphasis upon their utilization. It is not reasonable to expect such sophisticated interventions from psychiatric aides in the absence of intensive training and modeling by the clinical staff.

The sequence of actions taken by MHW 1 resulted in escalating rather than de-escalating aggressive behavior. Yet the techniques described in the Facility’s "Management of Aggression" training module note that reducing the number of commands issued, respecting "personal space," avoiding power struggles and remaining calm can enable the patient to assume control over his or her own behavior. These techniques are especially important in patients with conduct disorder or oppositional- defiant disorder.

Children with oppositional defiant disorder are likely to have an increased "personal space," and to become aggressive when that zone is "invaded." Their antagonism often escalates when given a direct order. They are much more responsive to a calm environment which does not encroach upon them. Close proximity at the breakfast table with the other child and an authoritative adult, coupled with the apparent alliance of that adult with the other child, were likely triggers for Andrew's actions. Alternative responses such as removing the other child from Andrew's environment (without appearing to penalize that child) could have restored Andrew's self-control. Alternatively, Andrew'could have been told, "You can start to calm down by going on your own for a time out in (an area other than the seclusion room). If you need help with this, I can help you by taking your arm and walking with you. You don’t have to apologize now if you don’t feel up to it; you can apologize after you feel calmer about this." In fact, a review of Andrew’s mental health records confirmed that he responded positively to such alternatives. Cotton (1989) makes the point that punishment within the context of appropriate discipline is an acceptable alternative as long as it is educational in nature. It follows that seclusion and physical restraint should never be imposed in a retaliatory or purely reactive fashion.

Once in the padded time-out room, with his shoes off and standing quietly against the far wall, Andrew represented little danger to himself and no danger to others. At worst, he might have injured himself by banging his head or limbs against the walls (which were padded) or against the window in the door. The apparently arbitrary command to "get into the corner," accompanied by an escort hold, escalated the situation to the point where Andrew "went ballistic" and the therapeutic restraint was initiated.

If the policy at this facility requires that children who are "going ballistic" be restrained in the time-out room, as opposed to being left alone to calm down, this policy should be re-examined. If the time-out room is not designed to minimize the potential for injury to unrestrained children of Andrew's age, the room should be re-designed for greater safety. If the policy directs or permits a single staff member to "take down" a violent child in the time-out room, this portion of the policy should be re-examined as well. In many psychiatric facilities, policy under such circumstances dictates that the staff member leave the seclusion room and call for assistance. Help should arrive in less than one minute and a violently acting out child is restrained by as many as four or five staff members. 

In order for any of the foregoing alternatives to be viable, the hospital needs to make significant changes in its policies, procedures and culture. For a psychiatric care facility to offer a true therapeutic milieu, the administration and medical staff need to model an attitude that makes the least intrusive and most educational methods of patient care a priority. Constructive change will require intensive staff training and supervision in non-confrontational methods of interacting with children. A variety of other changes probably will be necessary including lower child to staff ratios, greater availability of other management options and implementations of individualized care plans for all children seven days a week.

5.     Safety issues in the use of physical restraint

Safety issues become paramount when one or more adults are exerting physical restraint on a child. Further, even in one-on-one situations, as in this case, disparity in size between adult and child must be taken into account. In all restraint holds, a crucial element of safety involves the avoidance of chest compression leading to disability of the lungs to expand and get oxygen in to the body. Use of a prone restraint hold which permits the restrainer's weight to be placed upon the patient's chest is dangerous. This danger is magnified when the restrainer is twice the weight of the child. Under no circumstances should a staff member be allowed to lie on top of a child in order to restrain her or him.

The prone restraint hold illustrated in the Facility's PIT manual does not advocate weight across the patient's chest. Rather, the child’s arms are crossed in front with the hands below the waist. If this is performed correctly, the restrainer’s hands are situated below the waist, and the arms are not exerting pressure on the chest beneath the child. A side position restraint is also taught. In this situation, too, the hands are supposed to be low. A practice evolved at the Facility which allows the restrainer to extend his body over the child, using the trainer’s chest as a "ceiling" over the child’s body. Trainers at the Facility advised trainees to use only enough weight to maintain the hold. Small variations in this practice may lead to a restrainer lying on the top or side of the child’s back with the child’s face to the floor. As in this case, this variation may lead to a restrainer’s weifht compressing the child’s chest.

Neither the CPI nor TCI programs permit a technique like the prone PIT hold described in the Facility’s manual, or the variation which has emerged. The CPI program does not include any prone restraints. The TCI program incorporates a prone restraint which does not involve crossing the child’s arms in front. This restraint also does not allow the restrainer to put any pressure on the child’s chest or back. Neither program advocates that restraints other than escort holds be performed by one person.

It is noteworthy that the clinical staff at the Facility did not, apparently, receive training that alerted them to the possibility that a child might sustain a serious injury or even die as a result of physical restraint. In Andrew’s case, the staff interpreted his struggles and complaints behaviorally rather than medically. Whenever a child who is being restrained yells "Stop!" or "I can’t breathe!" or "That hurts!", the staff who are performing the restraint need to do something differently (release the child or change the hold or the child’s position and reassess the situation). While it is true that the child simply may be trying to escape or evade restraint and may not be experiencing significant physical distress, it is not safe to ignore the signal and assume that the child is being manipulative. Some may believe that, if the child is screaming or talking, she or he is not in respiratory distress. Unfortunately, it is possible that what the child now is screaming or saying will be the last vocalization that she or he is able to make.

There are other issues as well which may affect child safety during application of physical restraint. Many medical conditions may make one or more methods of restraint more dangerous than usual. Patients with chronic, severe cardiac or respiratory conditions are at higher risk of arrhythmia, respiratory failure, myocardial infarction or cardiac arrest. Patients with Down's Syndrome are at high risk for atlanto-occipital instability. In this condition, the head can "slip off" the spinal column, leading to a severing of the spinal cord causing death or quadriplegia. Patients with any condition that decreases bone density (e.g., osteoporosis, chronic renal failure, osteogenesis imperfecta) are at increased risk of fracture from physical restraint. Patients with bleeding disorders such as hemophilia, or who are taking anticoagulant medication, are at increased risk of severe bruising and hematoma formation resulting from physical restraint. Only a knowledgeable physician can gauge the relative risk of physical restraint of an individual child versus the risk that the child will behave in a way that will cause self-injury or harm to others. For these reasons, information regarding a child's personal medical status is vital to the treatment and planning for each individual child. There are children with certain medical conditions who are at risk of serious injury or death by even the most innocuous forms of restraint. Those conditions must be considered in determining whether any form of restraint poses a risk to the safety and well-being of the child.

The Panel's review of the record in Andrew's case indicates that he had a history of asthma dating back to 1993. While we note one brief hospitalization and two other emergency department visits for asthma in Andrew's past medical history, the hospital admission occurred four and one-half years prior to Andrew's death. There is no indication that Andrew M.'s asthma played any role in his death. The autopsy found that the cause of death was traumatic asphyxia, a condition resulting from mechanical causes.  

There was no evidence at autopsy of the pulmonary or vascular changes symptomatic of severe asthma. This information is, however, significant to our findings. While Andrew's medical condition appears to have played no part in his death, there was evidence that the Facility, despite having this child's medical history at intake, failed to consider this information in determining the appropriateness of using physical restraints on him. Furthermore, there is no indication that this information was passed down to the direct care mental health workers who provided oversight and supervision to him.

In addition, while in restraint, Andrew was initially positioned on his right side. This left his only functional eye down in contact with the padded floor of the time-out room. His ability to see MHW 1 and 2 was seriously impaired and may have contributed to his continued struggles, since they stayed on his left side throughout the entire period of his restraint.

This suggests to the Panel that differential treatment and planning is not provided at the Facility for every child who has a compromised medical condition. 

6.  Statewide policy and standards on the use of physical restraints

There are no current national or statewide standards regarding restraints or restraint programs. A review of documents including policies, procedures, operating statements, and restraint training materials from fifty-four facilities in Connecticut was conducted. The types of facilities reviewed included shelters, group homes, residential treatment centers, hospitals, the three juvenile detention centers operated by the Judicial Department, and Long Lane School. The facilities provided material ranging from a simple statement of "no physical restraint used" to policies and procedures regarding formal prevention/ intervention programs.

Individual facilities have developed their own approaches to restraint, and have incorporated these approaches into their treatment philosophies. There is considerable variation in the formality of programs throughout the state, in part due to differences in severity of behaviors, but also because of the ages and sizes of child populations. The choice of which restraint program to use is made at the facility level. Most facilities using a formal training program utilized one generated outside of their facility. However, it was not clear how often outside experts are used to refresh the in-house trainers or, in fact, how diligent in-house training efforts are.

The smaller facilities such as the shelters and group homes, which operate on the less-restrictive end of the treatment scale, generally state a policy of "no use of physical restraint." Therefore, the staff usually have no formal training in safe and appropriate physical restraint techniques. Should extreme behavior problems develop, the policy is to call 911 for police and/or paramedic assistance. Until recently, these less-restrictive facilities did not admit children deemed to require intensive treatment that their staff was not trained to provide. A repeated concern of these facilities in the materials reviewed, however, is that the new "no reject/no eject" language included in the current contract with DCF has resulted in more disturbed and/or violent children being placed by DCF in these "less-restrictive" settings. This practice has resulted in more injuries to the clients and to the staff, and increased incidents of property damage. Facilities are also reporting that DCF is not moving quickly enough to remove a child when he or she becomes a risk to him or herself or others, perhaps because of a lack of readily available specialized resources.

The use of more formal restraint training generally is used at those facilities which are designed to serve populations with more severe behavior difficulties and in the larger treatment facilities. Most programs utilize either the CPI or TCI programs. All Department of Children and Families-operated facilities and the Judicial Department's juvenile detention centers currently use CPI. Many private facilities use TCI. No placement or treatment program reviewed by this Panel, except the Facility, utilized PIT or the variation of PIT that evolved at the Facility.

The philosophy that physical restraint is to be used only as a last resort is a common theme throughout the policies and programs reviewed. In all facilities, the use of preventative interventions are incorporated, taught and emphasized to some degree. Many facilities stated that more time is spent in training on prevention and intervention techniques than on physical restraint training. Facilities also generally train their staffs that physical restraints are only to be maintained for the time needed. It was not necessarily clear from the materials how that is determined.

Only a few facilities were able to provide good data on their critical incidents, injuries to clients, and injuries to the staff. These generally were the larger facilities, which maintain this information as a requirement of outside accreditation bodies. Of those facilities that reported such data, serious injuries such as broken bones were very rare. Most injuries involved rug burns, abrasions, and mildly swollen wrists.

It appears that virtually all facilities in Connecticut with physical restraint programs use one or more types of face-down restraints. However, in many cases, full training manuals were not submitted, and it was not possible to determine how many use the specific hold that was employed on Andrew. The use of restraints or non-use of restraints throughout the state varies because of different philosophies of treatment, availability of staff, availability of funding to purchase outside training, the physical plant, and the availability of immediate response from law enforcement.

7.   Recommendations

·         The Department of Public Health and the Department of Children and Families should formulate regulations that address the development of appropriate physical restraint policies for use in the facilities they license, and for ensuring that staff is properly trained (and retrained on a yearly basis). Such regulations should include the development of policies matched to age, size, and disability characteristics of the facility's population, the training requirements for each program, the need to maintain training records, lists of approved formal training programs, lists of non-approved techniques, the requirement that an outside trainer train the facilities’ trainers, the requirement of outside recertification of trainers, and the recording of injuries to clients and staff while using restraints. Failure of a facility to comply should be deemed a regulatory violation and require a written corrective action plan. Repeated violations should result in a suspension/termination of licensing.

·         Physical restraint of children should be performed only when two or more staff persons are present  available except under extraordinary circumstances.

·         Under no circumstances should physical restraint of a child include compression of the thorax of a child by the weight of an adult.

·         All facilities need to establish a supervisory hierarchy for the behavior management program they utilize, including the restraint policy and the use of time-out or seclusion rooms, with one identified individual (preferably a psychiatrist or clinical psychologist) ultimately responsible. All programs should be integrated into the facility-wide treatment approach, and disseminated to each individual in the program.

·         Without one responsible individual, programs tend to be modified over time and unofficial policies and procedures develop. Commands like "sit in the corner" and "commit to safety" are typical of day-to-day practices that are passed on to new staff who, in turn, believe those modified procedures to be correct. The individual responsible for a facility’s behavior management program must protect the integrity of the program by allowing no changes or additions without going through a formal process. This person must also stay current in the field and update the program through a best practices approach. The goal should always be to have the program in the field match the program that is on paper, thus field reviews must be conducted often.

·         DCF should establish procedures for safety for those facilities, such as many shelters, that do not utilize physical restraint or train their staff to use it. For example, prior to admission, each child should be assessed for severity of behavior problems, and the information should be provided in written form to the facility.  Additionally, DCF should have a plan for the immediate removal of a violent child from a facility that is unable safely to manage her or him.

·         Whenever physical restraints are utilized on children in Connecticut, notification should be made, on a quarterly basis, to the Office of the Child Advocate, detailing the type of restraint used and the circumstances surrounding the need for restraint, for purposes of tracking trends in the practice.

·         Every psychiatric hospital, psychiatric unit of a general hospital and residential facility for psychiatric patients should promulgate a policy delineating its standard for use of physical restraints in patients with medical conditions or on medication which may affect their response to physical restraint. This policy should be reviewed annually and amended as necessary. Each child should be thoroughly evaluated at the outset of his or her admission and appropriate interventions should be part of the treatment plan. The admitting orders should reflect any modification of the institution’s standard physical restraint policies due to the medical condition or history of the child. Such orders should indicate the required modification and the medical condition or history leading to the modification of the physical restraint policy. These orders may be amended at any subsequent time that the child’s condition changes or more history becomes available. Such orders should be reviewed at each shift change so that responsible personnel are aware of each child’s status with regard to the use of physical restraint.

·         Any situation requiring the use of locked seclusion or physical restraint on a pediatric unit must lead to a follow-up review of the incident. Quality control should include debriefing of the intervention team after the utilization of physical restraint, tracking and reporting of events and circumstances.

Cardiopulmonary Resuscitation (CPR)

1.    Discussion

Psychiatric patients, especially pediatric psychiatric patients, are generally considered a low-risk population for cardiac arrest; nevertheless, risk exists. Psychiatric patients are frequently placed on medications which may have as a side effect an alteration in conduction patterns of the heart, causing a small but increased risk of serious rhythm disturbance. Such complications are seen with both anti-psychotic and anti-depressant medications. Psychiatric patients may also suffer from other medical conditions which may lead to cardio-respiratory arrest. Sufficient personnel should be trained in CPR to ensure its proper and timely utilization if an emergency arises. One-person CPR is rapidly exhausting and is less likely to be effective than two-person CPR.

The need for cardiopulmonary resuscitation is rare in the pediatric population.   Few medical personnel participate in such resuscitations frequently unless they practice in a pediatric intensive care unit or a large pediatric emergency department. Cardiac arrest in any age group is rare in psychiatric in-patients. In general, a deteriorating physical condition is recognized and the child is transferred to a medical institution or medical unit for treatment of the condition. The ability to maintain emergency skills such as resuscitation is difficult when those skills are not maintained and refreshed in a reasonably frequent manner.

CPR is most effective when it is initiated promptly after cardiac arrest due to a serious arrhythmia. In such a case, neither the heart nor any other vital organ has suffered oxygen deprivation at the moment when the cardiac arrest has occurred, and prompt intervention can ensure that the brain and heart are preserved from further harm. If the cardiac arrest has been precipitated by ventricular fibrillation, prompt defibrillation (within one to two minutes) is the most effective technique for resuscitation; delay in defibrillation decreases the likelihood of successful intervention. For this reason, a working defibrillator should be present on every in-patient unit or floor.

Where the cardiac arrest has been precipitated by respiratory failure, as in Andrew's case, such as traumatic asphyxia, the heart has stopped because it has suffered from oxygen deprivation; all of the other vital organs, including the brain, have been deprived of oxygen for the same interval of time. This reduces the likelihood of resuscitation even with properly performed CPR or advanced life support measures. Recognition and treatment of respiratory distress prior to cardiac arrest is far more likely to have a desirable outcome than prompt institution of CPR after cardiac arrest has occurred.

Unfortunately, the staff who attended Andrew during this critical incident failed to recognized the signs of respiratory distress and air hunger that he displayed (increasing agitation, yelling "Get off me!" and, finally, continuing to struggle after he stopped screaming). The record does not reflect that anyone checked his airway or his lips or nailbeds (to check the level of oxygenation in a dark-skinned person). All of the staff attributed his struggles and verbalizations as reflecting anger and a desire to escape from restraint. Even when he urinated on himself, it was interpreted as a hostile act, not as a sign of distress. In reality, Andrew’s urination most likely reflected a release of all his sphincters after cardiac arrest had occurred.

It is noteworthy that the Facility did not train its mental health workers in cardiopulmonary resuscitation. A secondary benefit of training in cardiopulmonary resuscitation is that it teaches people to appreciate the importance of maintaining an airway and the need for children to have oxygenated blood.

Standard resuscitation protocol calls for the emergency medical services (EMS) to be notified as soon as resuscitation has been started. In Connecticut, the most prudent way to do this is to call 911. By doing so, the caller can ensure the speediest response to the emergency, with all necessary responders being notified. This also permits automatic identification of the caller’s address in those regions with extended 911 service; such identification will further speed the emergency response.

The Facility’s policy requires 911 notification in the event of a cardiac arrest. It is unclear what went wrong in the process on March 22. Nurse 2, incorrectly believed she was speaking to the 911 operator rather than to a local ambulance company, leading to a short delay in paramedic response time.

The mental health workers who restrained Andrew were not trained in cardiopulmonary resuscitation, and resuscitation of Andrew began only after the child was unresponsive for three to five minutes, and after the nurse checked for a pulse, asked for a stethoscope and checked for a heartbeat. While it is likely that, given the cause of Andrew's death, CPR would not have made a discernible difference in the outcome of this case, a children’s unit within a psychiatric hospital should be equipped to respond immediately in both behavioral and medical capacities.

At the time of Andrew's admission, the policy of the Facility was to train only nurses and doctors in CPR, although all direct care providers were trained in their off-campus programs. They have recently instituted a new policy of training all direct care providers in their institutions as well.

2.   Recommendations

·         At least two individuals trained in CPR (basic life support) should be present on a psychiatric inpatient unit at all times. This will require at least three to four trained individuals each shift to account for meal breaks.  This may require that some personnel other than nurses receive basic life support training.

·         In a residential facility, at least one individual trained in CPR should be present in each housing unit on each shift. Appropriate measures should be in place to ensure prompt response by other personnel trained in CPR and advanced life support techniques in the event of an emergency.

·         All personnel who provide direct care to children should receive some training in recognition of medical emergencies likely to produce cardiac arrest, and should be instructed in the appropriate measures to take to ensure a prompt response by trained personnel.

·         At least one person trained in defibrillation should be present at all times on a psychiatric unit unless other trained personnel can be available in one to two minutes. A working defibrillator should be immediately available and should be brought to the site of the resuscitation immediately, even before the arrival of the person trained in defibrillation.\

·         At least one person trained in advanced cardiac life support techniques should be available within five minutes to any psychiatric unit. Any physician or nurse in charge of medical emergencies should be certified in advanced cardiac life support and recertified every two years.

·         Any physician who will be responsible for responding to a "doctor stat" or cardiac arrest should have advanced cardiac life support training and be recertified every two years. "Mock doctor stats" should be run intermittently throughout the year, which may encourage the retention of these infrequently used skills.

·         In the event of cardio-respiratory arrest or any life-threatening emergency in a psychiatric hospit