CorrectCare

Violent and Agitated Inmates: A Review of Management
and a Call for Research

by Scott Savage, DO, CCHP

Since the 1998 Hartford Courant investigative reports on deaths associated with seclusion and restraint, the use of these methods has come under stricter control by regulatory agencies. However, there is a shocking lack of data on the safety or danger of using seclusion or restraint. This article reviews gaps in scientific knowledge about the effective use of seclusion and restraint in correctional health settings, points to areas where studies may prove fruitful and indicates a model for improving care.

Restraining people is dangerous, both for the instigator and the intervener. In recent years, there has been progress in understanding how to manage agitated and violent behavior. Unfortunately, it is inconsistently practiced. This article outlines basic principles and methods of managing agitated behavior of healthy-appearing adult males in correctional medical and mental health settings. It excludes groups such as medically frail people, those under pharma­cological treatment for chronic violent conditions, the elderly and those suspected to be chronically ill.

Currently, there are many ways of managing agitated and violent behavior. In order of generally accepted risk, they are de-escalation, seclusion, physical restraint, chemical restraint, nonballistic weapons and firearms. This article omits discussion of the latter two, and limits the term physical restraint to the use of standard four-point manual restraint systems used in medical settings.

Methods of Restraint and Seclusion

De-escalation
De-escalation is the use of talking to help calm an agitated person. Many methods exist, but unfortunately none have been empirically researched. Two are worth noting. Verbal Judo is used widely in law enforcement. This simple method has five steps: listening with empathy, giving a context, providing options, confirming noncooperation and acting. For situations where there is not likely to be an ongoing relationship with the agitated person, this method  has been reported to be useful. Another method is Crucial Confrontation. Although empirical data is not given, the authors state they studied and interviewed 25,000 people over 20 years to develop the method. Designed for business and family situations, it is useful for corrections settings where an ongoing relationship with the agitated person exists. During my time as a corrections medical director I used both methods successfully while performing medical treatment in volatile situations.

Seclusion
Confining an agitated person to a quiet room can be an effective way to reduce agitated behavior. In one study of 263 episodes, the authors found that 83% demonstrated agitated behavior at the initiation of seclusion, but only 23% remained agitated upon release.

Seclusion can be voluntary (a time-out), verbally ordered or physically mandated. Although protocols vary, the Joint Commission and the Centers for Medicare and Medicaid Services have issued guidelines on the use of seclusion and restraint. The Joint Commission standard states that seclusion and restraint are used only to protect the immediate physical safety of people and not as a means of coercion or discipline; they are used only when other means are less effective; the least restrictive method of effective seclusion or restraint is used; and it is discontinued at the earliest possible time. NCCHC standard I-01 Restraint and Seclusion makes a sharp distinction between medically necessary restraint and seclusion and restraints ordered by custody staff. Furthermore, it does not permit medical staff to participate in custody-ordered restraint except to monitor the health status of restrained people.

Physical Restraint
Physical restraint can be a dangerous practice, yet there is little literature concerning the rate of injury in correctional settings. Unfortunately, even death in restraint is not a separately reportable incident, and there is debate on how to classify it: accidental, homicidal, natural or undetermined.

In medical settings, physical restraint has been thought to cause death through asphyxiation, aspiration, cardiac arrest and other reasons. However, the data for injury or death in these settings is equally scant. But there is indirect evidence that an alarming problem exists. In 1994, the New York State Commission on Quality of Care reported 111 patient deaths in the preceding 10-year period in the medical setting that were related to restraint use. Also, the Joint Commission now collects data and has recorded 76 incidents of injury or death between 2004 and 2008.

Similar data for correctional settings could not be located but the potential magnitude of the problem is suggested by the following statistics. The Criminal Justice Statistics Center reported that in the 10-year period from 1994 through 2003, the State of California had 428 accidental deaths in custody. In the period between 1990 and 2004, the State of Maryland reported 45 accidental deaths in custody for which the cause of death remains unexplained. Finally, the U.S. Bureau of Justice Statistics reported that between 2001 and 2006, there were 377 deaths from either accidental or unknown causes. While most of these deaths may have been not related to restraint use, it is concerning that the data does not appear to be available.

Injuries related to the placement and use of restraints in correctional settings is likewise unknown. However, exertional rhabdomyolysis, pulmonary embolism and persistent emotional anxiety states have been attributed at least in part to the use of physical restraint and are consistent with the known mechanisms of these conditions. The safety or danger must be considered in the context of the rate of complications of alternate methods of management, which unfortunately, are also largely unknown.

Pharmacological Restraint
One alternative to physical restraint is pharmacological restraint. But it, too, can be dangerous. A variety of methods and medications can be used. Medications can be given orally (PO), intravenously (IV) or intramuscularly (IM).

Oral medications are painless to administer, but require cooperation of the person receiving them and take much longer to have therapeutic effect. For example, one commonly used agent, haloperidol, is effective 20 to 40 minutes after IM administration but takes three to six hours when given orally. However, oral medications have the advantage of not bringing a sharp needle into a volatile situation. IV administration has the most rapid onset, but gaining intravenous access can be difficult in uncooperative patients.

IM injections have an advantage over IV techniques in that they are easier to give and can be administered without the patient’s permission. However, they are painful and may cause complications such as cellulitis, abscess, phlebitis and hematoma (which, fortunately, are uncommon). The combination of rapid onset of action, ease of administration and the ability to give the medication to an uncooperative person make IM injection the route of choice in most patients with agitated behavior.

Another important set of risks is the side-effect profile of the drug given. Medications currently used are antipsychotics and benzodiazepines. Haloperidol is a typical antipsychotic drug. Its cardiovascular side effects include tachycardia, alteration in blood pressure and potentially fatal cardiac arrhythmia. Its central nervous system side effects include extrapyramidal symptoms, exacerbation of psychosis and neuroleptic malignant syndromes. Lorazepam is a typical benzodiazepine drug. Its side effect profile includes amnesia, weakness, unsteadiness, paradoxical disinhibition and potentially fatal apnea.

Newer antipsychotics such as olanzapine and ziprasidone have been used to alleviate acute agitation states. Both have few or no extrapyramidal side effects. Olanzipine does not carry a significant risk of cardiac arrhythmia in short-term use, but can cause diabetes with chronic use. Ziprasidone reaches peak therapeutic effect in as little as 30 minutes, but is associated with prolonged QTc syndrome, which may cause fatal cardiac arrhythmia. Both medications are much more expensive than haloperidol.

Few medications other than those named above have been studied in agitated behavior. This could be a useful area of study. Several agents used for procedural sedation in adults and children have rapid onset, short duration of action and limited side effects. For example, methohexital is a short-acting barbiturate that can be given intramuscularly with a rapid onset of action, typically less than 10 minutes. Its duration of action is less than 30 minutes when given to children for CT scans. Its use for IM injection in adults has yet to be studied. Midazolam has been shown to be much faster in onset than haloperidol or lorazepam (18 minutes vs. 28 to 32), and the time to arousal was also significantly faster, 82 minutes vs. 127 and 217. Various beta-blockers have been studied in the use of long-term therapy for violence disorders. Given the ability of these drugs to reduce tachycardia, blood pressure and other symptoms of arousal and agitated states, they may be useful in reducing agitated behavior acutely. This has yet to be studied.

Unintended Consequences?
The 1998 investigative series on restraint and seclusion raised public awareness of the problem and, in the intervening years, U.S. regulatory agencies have placed tighter control on the use of these methods. These regulations center on documenting the justification for restraint and intense monitoring of the restrained person in medical and psychiatric settings. This has prompted some institutions to develop policies of zero seclusion and restraint. This appears to be a laudable goal, and efforts to reduce the unnecessary use of restraints should be encouraged.

Unfortunately, resulting staff injury rates are rarely reported. In a letter to the editor, one program participant noted that while his directors reduced the use of restraints by 60%, the staff injury rate increase by 30%. Another problem is that agitated behavior can be the presenting symptom for a variety of potentially serious illnesses, including infection, metabolic disorders, endocrine disorders, trauma, pain, toxicologic disorders, structural brain abnormalities and psychiatric disorders. Even de-escalation may not be benign. I have seen cases where patients appeared to respond to de-escalation techniques, only to discover that they were actually getting too ill to mount further physical response.

Incarceration settings have a complex and volatile mix of detainees with a variety of medical, emotional and social illnesses. Most commonly accepted methods of managing violence in this setting have never been empirically tested. The publicity of death and poor outcomes associated with physical restraints has led to a trend toward limiting their use. Without adequate data, it is impossible to know if we have solved a problem or simply hidden it better.

There is a lot to learn. How do custody officers rapidly identify patients who are agitated because they require medical attention? Which de-escalation techniques are most effective? Are nonballistic weapons safer than five-person tackle techniques? When do the risks of physical restraint outweigh the risks of chemical restraint? These are just a few of the important questions that need to be answered.

And once we learn, we need to act. There are effective models for managing complex medical problems. Advanced cardiac life support was so successful that it was replicated by the American College of Surgeons as advanced trauma life support and by the American Academy of Pediatrics as pediatric advanced life support. It is time to develop a system of care that involves people and organizations from law enforcement, psychology, social work and medicine to work together to develop effective, safe and humane treatment of people with agitated and violent behavior.

[This article first appeared in the Summer 2010 issue of CorrectCare.]