Evaluating and Treating Violent Adolescents in the Managed Care Era

The evaluation and treatment of adolescents who threaten violence are great concerns to psychiatrists. The psychiatric literature identifies the comorbid disorders associated with violence as attention deficit disorder, residual type, or attention deficit disorder with hyperactivity, major depression (1), and posttraumatic stress disorder (2). The adult literature describes how the population of patients treated as inpatients has changed to being more dangerous, a primary effect of the application of the managed care paradigm’s stringent admission criteria of medical necessity (3). While this phenomenon of increased dangerousness has yet to be described for child and adolescent inpatient units, the adult literature notes that the demographics show that young male subjects are the more violent patients on adult units. Even without studies showing an increase in violent or potentially violent adolescents on inpatient units, the general statistics show that arrests for violence offenses increased by 61% between 1988 and 1994 and the juvenile arrest rate for homicide increased by 90% between 1987 and 1991 (4). Recent events, such as the youths who have committed violent murders at their schools in Paducah, Ky., Pearl, Miss., and Jonesboro, Ark., have drawn unprecedented attention to the problem of violent adolescents and preteens.

The problem, however, of predicting who is at greatest risk to actually commit a violent act has not been adequately addressed in the literature (5). Marohn, who has written extensively about the management of assaultive adolescents states, “The adolescent who misuses and abuses property or who threatens to harm another is dangerous and must be taken seriously” (6). However, the evaluation and treatment process that he recommends is not consistent with managed care cost containment. He describes a difficult evaluation process because healthy adolescent behavior is a function of complex biological, psychological, and sociocultural foundations. From the early work of Aichhorn, who tried to apply Freud’s ideas about perversions to the treatment of the “wayward youth” in Vienna (7, 8), to more contemporary writings (9), it is well recognized that it is the nature of adolescents to act, and to act instantly. A contemporary approach to examining adolescent defense mechanisms used the Defense Style Questionnaire and identified three groups: immature, mature, and prosocial (10). The authors suggest that the patients with a high level of immature defenses will need more structure and support and are likely to have a lengthier and more difficult course of treatment. While such findings are interesting and compelling, they have yet to be accepted as guidelines for care and certainly are not used by managed care companies to ascertain who gets what kind of and how much treatment.

Given the large number of adolescents who present with some sort of aggressivity or threat to commit an aggressive act, combined with the normal predilection to act, the dilemma of predicting who is at greatest risk to commit a violent act and warrants extended treatment can be a daunting task. Factors associated with adolescent violence include alcohol and substance use, depression and suicidality, overstimulation, sociocultural and family ambience, property damage, threats of violence, and alexythymia. While the ability to predict precisely who will commit a violent act may be difficult, it is well recognized that the child who behaves violently at a young age and who persists in behaving violently is likely to have a long and serious career of violence (11–14). Loeber, who has written extensively about the development of juvenile aggression and violence, succinctly states that desistance of violence is more probable in less serious than in more serious forms of aggression. He describes two distinct developmental trajectories: one subgroup of boys in which fighting decreases over time and another in which fighting escalates to violence. This kind of aggression, characterized as the life-course type, persists and worsens in severity from childhood to adulthood. It is postulated that many individuals with the life-course type have an onset of aggression in the preschool years and meet criteria for attention deficit hyperactivity disorder (15, 16). Satterfield and Schell’s recent prospective study shows that children with hyperactivity and conduct problems in childhood are at significant risk for later criminality (17).

With shortened inpatient stays focused on stabilization of acute symptoms, the problem of diagnostically assessing adolescents from the perspective of understanding their inner world, their coping skills, their external support systems, and their defense mechanisms and, from this mixture, their potential for violence has been lost to the more simplistic managed care paradigm of immediate dangerousness and medical necessity. This shift to acute stabilization as the model for most inpatient care has made comprehensive assessment and treatment planning for potentially violent patients increasingly challenging.

This article describes a complex case as it presented to an acute adolescent inpatient unit. We will discuss how and why this adolescent’s evaluation made discharge to anything less than a residential treatment center an untenable and unacceptable plan. The logical consequence of this was that the patient’s length of stay was well beyond the current average lengths of stay for hospitalization. The difficult process of negotiating with the third-party payer, the interface with public health care funding, and the coordination ultimately with the local school system to finally effect residential treatment will also be described. The discussion will focus on the inherent policy issues raised by the need to provide comprehensive evaluation and treatment for potentially violent adolescents, while simultaneously protecting the community and operating within the confines of today’s health care reform, which emphasizes market-driven managed cost containment.

Case Presentation

CJ was a 15-year-old boy admitted after setting a small fire in his bedroom. In addition, the outpatient therapist was concerned about the patient appearing psychotic and expressing extremely sexual and violent thoughts. On admission, the patient openly expressed the desire to thrust a sword into a female’s vagina through to her intestines and up to her head. He reported experiencing “weird and freaky and frightening thoughts” all of his life. He noted that these thoughts were easily triggered, either by hearing a song on the radio or watching a television program. He stated that he felt increasingly unable to refrain from acting on these thoughts.

Past History

This was CJ’s fourth psychiatric hospitalization. His first hospitalization occurred at age 8 for firesetting, aggressive behaviors, and cruelty to animals. The second hospitalization occurred at age 10 for similar symptoms, with the addition of depressive symptoms including wrist cutting and a hanging attempt. Diagnoses included major depression and attention deficit disorder. The current admission was preceded 3 weeks earlier by CJ’s third hospitalization, which lasted only 9 days. Although at that time he presented with more self-destructive behaviors, including self-excoriation of acne wounds on his face and torso and self-induced scratches to his penis, he seemed to stabilize in the hospital and was discharged to his home. During this admission, he was switched from methylphenidate, sustained release, 40 mg every morning, to fluoxetine, 20 mg every morning.

CJ was the youngest of three boys by his mother’s first marriage. This marriage was characterized by physical abuse of the mother and physical and sexual abuse of CJ by his biological father. He was also allegedly abused by a cousin and a school bus driver. His mother’s current husband stated that he tried to be supportive and helpful with all three boys. However, CJ was described as quite different from the older brothers in terms of not being athletic or popular with peers. He was reported to have almost no friends and was never invited to parties or other outings. He was recently placed on telephone restriction at home secondary to his making inappropriate calls of a sexual nature to female peers. It was also noted that CJ watched unsupervised television, including cable channel movies late at night in his bedroom.

The family history of psychiatric illness, as reported by the mother, included unspecified psychiatric illnesses on the father’s side. CJ had no history of alcohol or substance abuse, except for having smoked marijuana on a few occasions.

Mental Status

CJ was a diminutive boy who looked younger than his stated age. There were multiple scabs on his face, neck, and arms, and he was generally disheveled. Eye contact was good, and he was cooperative. He had a looseness of associations with tangentiality and circumstantiality. Rate of speech was increased also. He reported having obsessive thoughts of a sexually violent nature and compulsive behaviors such as picking pimples on his face and torso; this extended to picking at his penis and causing lesions. His affect was labile and often inappropriate. He became easily frustrated and irritable and described feeling depressed, yet denied suicidal ideation. He reported homicidal ideation, but this was vague and not directed toward anyone specifically. He did report distress over his violent or homicidal thoughts. He was oriented to person, place, and time and appeared to be of average intelligence.

Hospital Course

The working diagnoses at admission included attention deficit hyperactivity disorder; depressive disorder not otherwise specified; rule out psychotic disorder not otherwise specified; and rule out obsessive-compulsive disorder. Fluoxetine was discontinued because it was thought that it might have been activating for CJ. He was started back on a regimen of methylphenidate with valproic acid, 250 mg b.i.d., added for mood stabilization.

Within the first few days of hospitalization, CJ became increasingly aggressive. He started several fights with his peers, requiring the need for quiet room, with occasional escalation to locked door seclusion. He continued to report being preoccupied with sexual and violent thoughts that were constantly present and not controllable. He showed no emotional discomfort regarding his thoughts or actions, consistent with their being ego syntonic, not dystonic.

CJ was started on a regimen of risperidone, titrated up to 3 mg/day, in an attempt to dampen his aggressive impulses and control his violent fantasies; risperidone was then titrated down to 1 mg because of extrapyramidal side effects. He was also started on a regimen of fluvoxamine, 40 mg/day, to help with his obsessive thoughts. Even with these psychopharmacologic interventions, CJ continued to have difficulty on the unit. There were several incidents of exhibitionism, provocative and lewd sexual comments, and inappropriate sexual behavior toward female peers.

Given this course, during the third week of hospitalization, a consultation was requested from a forensic child psychiatrist for an assessment of CJ’s potential for dangerousness. Although the consultant felt he was not acutely dangerous to himself or to others, the potential for violence if he remained untreated was assessed to be great. The consultant highlighted that CJ would grow physically stronger over the next years and would likely have access to disinhibitory drugs such as alcohol and potentially even weapons. The consultant strongly recommended that CJ be treated for a longer period of time in a residential treatment center. Other recommendations included discontinuing the methylphenidate and fluvoxamine because they might be disinhibiting and starting paroxetine treatment to address his underlying dysthymia and dysphoria, while potentially taking advantage of the side effect of decreased sexual drive.

While these changes were being made, the treatment team considered whether CJ could be safely discharged home until funding for residential placement could be secured. (The process for this in Maryland includes extensive documentation to the local school system about the patient’s needs, with the goal of securing funding for the educational component of a nonpublic residential treatment center under Public Law 92-142. This law mandates public schools to provide an education for all students regardless of their physical or emotional condition.) During the ensuing days, however, CJ’s behavior on the unit continued to be of great concern. He engaged in several covert sexually inappropriate behaviors with female peers, including comments and touching. He was also observed to be touching his penis when he thought staff was not looking and would then adamantly deny what he had done.

Tests done during hospitalization included an EEG, with normal findings; psychological testing, which revealed a full-scale IQ of 106 with some distortion of reality testing, looseness of associations, and a potential for dissociation; and standard laboratory tests including thyroid function, with all findings within normal limits.

CJ’s course in the hospital became complicated by a mild leukopenia that developed during treatment with val­proic acid. The valproic acid was stopped, and his WBC returned to normal. A rechallenge produced a drop again, so he was switched to lithium carbonate to assist with mood stability; a therapeutic blood level was reached at 300 mg every morning and 600 mg at bedtime. Psychopharmacology consultation to provide additional input suggested switching CJ gradually from paroxetine to clomipramine and discontinuing benztropine, which had been added to treat extrapyramidal symptoms secondary to the risperidone. An ECG was obtained before clomipramine treatment was started, and it was repeated periodically to ensure there were no conduction delays secondary to tricyclic effect.

CJ’s hospital course became quite extended secondary to multiple medication trials, continued inappropriate behaviors on the unit even with close staff supervision, and difficulty with securing funding for his transfer to a residential treatment center. Although attempts were made to have him go off the unit with his parents for brief periods of time in order to continually assess his ability to be safely maintained at home while awaiting residential treatment, these were thwarted by his behaviors on the unit, which were dangerous to other patients. These behaviors included grabbing at a female subject’s breasts or saying to another patient that he wanted to have sex with her in a vulgar, violent way and then yelling, storming down the hall, and going into his room and slamming the door shut when staff confronted CJ about his behavior. CJ was ultimately discharged to a residential treatment center after 4 months in the hospital. At the time of discharge, he was mildly improved and there were fewer episodes of escalation requiring quiet room and locked door seclusion.

DISCUSSION

This discussion will focus on five areas: 1) the process of securing funding, 2) the clinical criteria to decide which patients pose the greatest risk, 3) treatment considerations, 4) the societal and policy implications of the current state of predicting violence in adolescence, and 5) ethical concerns.

Securing Funding

In this era of managed care, third-party payers use medical necessity as the criterion that guides the payment of hospital care. This has created acute stabilization inpatient care. Patients whose needs are persistent and chronic fall to the public sector. The guiding principles of acute stabilization, immediate risk to others, and medical necessity, however, are not so easy to apply when we are faced with a patient such as CJ. The clinician is required to carefully translate the patient’s behavior and risk to the criteria used by the managed care paradigm. CJ’s initial insurance coverage was provided by a local managed care company that required frequent review of his need for continued inpatient stay. While it was clear to the evaluating clinician that CJ posed a significant risk if discharged, the reviewers were not easily in agreement; they wanted precise language using examples from CJ’s current behavior on the unit. Risk factors from the past were not regarded as relevant by the reviewers. Over the course of a few weeks, the review process was moved up the review ladder to the medical director of the managed care company, who ultimately stated that CJ could be converted to medical assistance, i.e., public dollars, after residing in the hospital for 30 days. (In Maryland, before the recent implementation of managed mental health care for the chronic mentally ill, after 30 days out of home for a child or adolescent, the patient was eligible for medical assistance and thus public health dollars in the yet-to-be-managed system.) The medical director’s willingness to continue to authorize payment of the first 30 days of inpatient care was reinforced by the forensic consultation process. This provided clear documentation in the medical chart of the risk for violence should CJ be discharged to live in a less restrictive setting where his behavior could not be carefully monitored.

The process of providing documentation to the local school system to ensure funding for the educational component of residential treatment was parallel to the process of dealing with the third-party payer. Fortunately, CJ’s mother was a responsible, concerned parent who was able to work with the treatment team and shared their concerns for CJ’s future. In fact, at one point his mother stated that she feared that he would eventually rape someone. Although she attempted to provide structure at home, CJ was adept at eluding limits and, given his age, could not be monitored while living in a home setting. His deceptiveness on the unit, coupled with denial and escalation to aggression when he was confronted with his behaviors, also made placement in a group home untenable. The child psychiatrist in charge of CJ’s care had to carefully write a report to the school that documented his need for treatment and why this could not occur anyplace other than a residential treatment center. Given the relative lack of confidentiality of school reports, the report had to be faxed to CJ’s mother before the school meeting in order to seek her approval that the information be publicly shared. This resulted in editing on the telephone but kept the mother and treatment team allied in the pursuit of adequate treatment for CJ. The diagnostic and prescriptive teacher from the inpatient service also attended school meetings to provide support and help translate the clinical needs of Maryland’s current special educational terminology to the mother. Ultimately, it was necessary for the treatment team to guide her to contact a lawyer to assist her and also attend meetings when CJ’s treatment would be reviewed. This dramatically speeded the approval process for the funding of CJ’s educational needs.

Clinical Criteria

The process of clinically determining which potentially violent patients warrant more extended care and treatment and which do not has yet to be clarified by psychiatric research. Without a research knowledge base from which to draw, the treatment team used a combination of risk factors, current cultural themes, and CJ’s unit behavior and response to confrontation as the basis of decision making. This was coupled with parental concern, which was consistent with clinical thinking.

CJ’s risk factors were many and included history of early abuse; history of early age at onset of symptoms, before age 5 years, and first psychiatric hospitalization at age 8; symptoms that included attentional problems coupled with aggression and cruelty to animals; failed outpatient treatment in spite of active parental involvement and psychopharmacologic interventions; poor identity formation; long-standing, extremely poor peer relationships; and attraction to current violent television and music now ubiquitous in our society.

It is perhaps the interaction of CJ’s early risk factors with today’s cultural themes that posed one of the greatest problems for his parents to monitor in a home setting. CJ’s access to television and music violence, specifically “shock rock,” coupled with his poor identity formation and poor peer relationships, placed him at great risk to coalesce his identity with someone who commits heinous acts. He was particularly enamored with the character that Woody Harrelson played in “Natural Born Killers,” a film he had seen on cable television many times. His attentional problems, lack of impulse control, obsessional thinking, and increased libidinal drive served to further increase his propensity to act out a violent fantasy. These concerns were confounded by the clinician’s experiences of meeting with the patient, during which time he had little insight into his behaviors and frequently denied them. When pressed further to discuss his inner world, CJ continued to be preoccupied with violent fantasies, often involving women. These were discussed without any evidence of emotional connectedness, leaving the clinician feeling uncomfortable and sickened. This kind of probing and countertransference was tolerable only within a highly structured therapeutic environment in which everyone’s safety could be ensured.

Unfortunately, there is an entire generation of youth being raised by the repetitive watching of violent media. With the national average of more than 4 hours of television watched per day (18), and with half of network programs and 85% of cable movie channels showing violence (19), the sheer input of violent images to youth is staggering. The concern about the impact of media violence as a public health issue has been steadfastly growing over the last several years, as evidenced by the American Medical Association’s development of a lengthy educational pamphlet, “The Physician’s Guide to Media Violence” (20), which was distributed to more than 66,000 physicians nationwide in September 1997.

While the studies to date have looked at the impact of media violence on “normal viewers,” there has been no research on its effects on children and adolescents with preexisting emotional problems. This may well create a subset of viewers, like CJ, who are, indeed, more vulnerable. These vulnerable youth have low self-esteem and a poorly differentiated sense of self. They lack the capacity to see themselves as separate from what they are watching and yearn for the glamorized, sensationalized life on the screen. Seeking to define themselves, and without the necessary physical and mental skills to identify with more positive role models, they become easily drawn to coalesce their identity around violent figures, either real or imagined.

The degree to which an adolescent is preoccupied with violent themes and negative heroes can best be assessed by taking a detailed “media history.” A media history informs the clinician about the amount and kind of television and movies watched, whether a television is in the child’s bedroom and used for self-soothing at night, and whether the youth is permitted to watch R- or X-rated material or both. With well over 50% of youth having television in their bedrooms, young children and adolescents are watching more television unsupervised, and many parents are unaware of what they are seeing. The media history should include questions about favorite movies that may have been repetitively watched and therefore may have been even more of an influence on the developing youth’s identity. It is not uncommon for youth to watch a favorite movie multiple times because of easy access to rental videos and the ability to videotape from television. Many adolescents describe an almost obsessional fascination with “Natural Born Killers,” a movie that is full of graphic violence.

Musical preferences should also be noted, with particular attention to shock rock and violent rap, which are degrading to women. For example, the shock rock group that CJ was drawn to, Marilyn Manson, has each of its band members given a stage name that combines the name of a famous, beautiful woman with a serial killer. This kind of creative fusion of sex and aggression glamorizes the killers to pop icon status. The victims of the crimes and their pain and suffering, as well as the suffering of their families, are callously cast aside by the musicians, agents, and marketers. The thrust is solely to produce an image that will captivate the youth market and sell the musical product, with no attention paid to the subliminal message of condoning and encouraging violence.

While adolescent development has at its core the challenging of parental values as the child seeks to differentiate and establish his or her own separate identity, the fascination with negative, dangerous images should not be ignored. It needs to be assessed in the context of understanding the adolescent’s sense of self and his or her particular vulnerabilities, such as history of abuse, presence of attention deficit disorder, depressive disorder, or schizoid personality, before the preference is cast aside as a part of developmentally normal adolescent rebellion. Comorbid substance abuse and the presence of guns in the home only further increase the risk of untoward outcome.

Treatment Considerations

While there is abundant literature that examines how to rehabilitate convicted violent offenders with the clear goal of preventing future criminal behavior, the literature on prevention is scant. The problems with this are multiple and complex, not the least of which is the heterogeneous nature of adolescents who present with risk of violence and the variety of settings in which they may arise. In this discussion of CJ, we are examining a youth who is considered at risk by virtue of his early childhood identifiable antecedents and his current preoccupations. Yet even with this group of at-risk youth, the studies evaluating treatment of violence and aggression are often flawed methodologically. Few studies adequately describe the same population, and fewer yet have no-treatment control groups. Even the use of “aggression” and “violence” is without clarity, with both being used interchangeably throughout the research literature. The diversity of the population notwithstanding, several types of interventions are widely acknowledged as forming the basis of intervention and are therefore looked to for their preventive potential. These include individual, group, and family treatment; biological interventions; cognitive behavioral approaches; social skills training; problem-solving skills training; and multisystemic therapy. Interesting differences occur in the areas of acceptance of risk and the measures of outcome when one shifts to reviewing the literature that is part of the juvenile justice system, primarily concerned with rehabilitative issues. For example, the multisystemic therapy approach is, to date, the only treatment program to demonstrate short- and long-term efficacy with chronic, serious, and violent juvenile offenders (21–23). This kind of program, however, can be accessed only through the juvenile justice system for an adolescent who has already committed and been convicted of a violent offense. It is not an available treatment for an adolescent like CJ, who enters through the mental health system and has a long history of previous psychiatric treatment. As a modality, multisystemic therapy is offered outside the medical model, is accepted as experimental, and therefore is not within the realm of malpractice law. Paradoxically, when one looks at the adolescents who enter through the mental health care system before an offense, the professionals involved in evaluation and treatment are held to a different standard, which includes possible culpability regarding malpractice should an untoward event occur.

An underlying principle for prevention and intervention is the individual comprehensive assessment of the adolescent, family, school, and community support system and then the crafting of the individualized treatment plan. This is a cornerstone of multisystemic therapy, which designs individualized interventions that are child focused, family centered, and directed toward solving multiple problems across the many contexts in which adolescents live—family, peers, school, and neighborhoods. The prescribed treatment for CJ, referral to a residential treatment center, had many of these same elements, yet simultaneously would prevent a violent act in the community because treatment would occur within the safety of a highly supervised therapeutic milieu. Residential treatment centers have comprehensive services including: individual treatment, family therapy, child psychiatric care with attention to psychopharmacology, special educational services, group therapy, and recreational and vocational services. While these modalities are available in the community, one crucial aspect of CJ’s pathology was his pervasive denial and callousness. In order to treat CJ, the individual therapist needs to be able to safely confront him with information about his behavior from the milieu. Similarly, group therapy, which promotes confrontation by peers and community responsibility, can safely occur within a residential treatment center and is not possible in less structured settings.

The role of psychopharmacology for CJ is an additional aspect of his treatment that warrants careful attention. While his history of attention deficit disorder is important and likely warrants ongoing stimulant treatment to facilitate learning in school, it is only one medication of many that may have a role in his treatment. Medications such as valproic acid, carbamazepam, and lithium carbonate have been effective in helping to decrease mood lability and dampen aggressive drives. In addition, recent research has looked at the serotonin system as possibly related to violence, with the assertion that high-dose selective serotonin reuptake inhibitors are effective as treatment. The new antipsychotic medications, especially risperidone, are also being increasingly used for the prevention of impulsive, aggressive outbursts. Often more than one medication is used with targeting of specific symptoms for each medication tried. Finding the optimal combination that minimizes aggressive thoughts and actions, yet permits and even facilitates learning, without significant side effects, can be a laborious process. Medications need to be added one at a time with adequate trial periods of 4 to 6 weeks, during which accurate observational data are essential. Conducting such medication trials outside the structured, observing environment of a residential treatment center, or a hospital, as occurred in the past, does not provide the psychiatrist with sufficient reliable data to know what is effective and what is not. In addition, if CJ’s sexual preoccupations continue and are not contained by optimal standard psychiatric interventions, he may well be a candidate for hormone treatment with agents like medroxyprogesterone. Although such agents are not often given to male subjects as young as CJ, the persistence and lack of response to other treatments could make this a treatment of last resort. Again, the need for such an extraordinary treatment could be justified only if all else failed and CJ were carefully monitored within a residential setting.

Societal and Policy Implications

Child and adolescent inpatient services have undergone a profound transformation as a result of managed care. During the late 1980s, the cost of psychiatric and substance abuse treatment rose at a substantially higher rate than the cost for all of health care—20% for psychiatric and 32% for substance abuse, as compared to an overall health cost increase of 13%. Inpatient care especially increased more rapidly, and most of this increase (almost three-quarters) between the years 1986 and 1988 was due to an increase in inpatient utilization by children and adolescents. Limits were placed on such utilization in 1989 and began over the next several years a dramatic curtailment in inpatient use by children and adolescents. Most of the growth during the 1980s was due to the rapid increase in for-profit psychiatric hospitals and units devoted to care for children and adolescents. The for-profit managed care industry owes its inception to the consensus among payers that this use was overuse and in need of reform (24).

In addition to the growth of inpatient beds, lengths of stay during the late 1980s were rather long and costly. At Sheppard Pratt the average length of stay for children and adolescents in 1989 was 80 days. The average length of stay in 1996 was 13 days, with a median of 6 days. Again, payer expectations and managed care protocols have led to the dramatically shortened inpatient stay and the development of less restrictive and less costly levels of care: a continuum of hospitalization, residential treatment (including group homes and specialized foster care), day treatment (including therapeutic school and therapeutic vocational placements), evening treatment, outpatient treatment, and in-home crisis stabilization (25).

Although these shortened stays have led to increased treatment opportunities for many children and adolescents, the case discussed here raises important policy implications on the role of psychiatry and psychiatric hospitals in predicting and preventing violence among the most disturbed of our patients. While there has been great fear that applying risks factors in clinical decision making would increase length of stay and once again lead to excessive treatment and possible violation of civil liberties, there is currently an equal or greater concern of putting community members at risk by using only the presence of acute symptoms as the guiding principle for who receives treatment in a secure setting. Is the current paradigm of short stays a game of Russian roulette with the lives of patients, family, and community members at stake under the guise of immediate dangerousness?

Clinicians who can be sued for their decisions related to safety are understandably uncomfortable with this set of risks, but the third-party payers are not. This liability issue is not trivial. Many Americans are covered by a health insurance plan protected by an Employee Retirement Income Security Act (ERISA) preemption that precludes the plan from being sued for managed care decisions. Increasingly, managed care companies, through the utilization review process, deny payment by using medical necessity criteria that focus on acute symptoms. Recent court decisions have been rather explicit in their concern about ERISA. The U.S. Court in Massachusetts noted that when ERISA was passed, fee-for-service medicine, not today’s managed care model, was the norm. The court stated that it was troubled that “in the health insurance context, ERISA has evolved into a shield of immunity that thwarts the legitimate claims of the very people that it was designed to protect.” The district court further asserted that the case in question, Andrews-Clarke v. Travelers Insurance Co.(26), was “yet another illustration of the glaring need for Congress to amend ERISA to account for the changing realities of the modern health care system.” The court further noted that it was up to Congress to amend ERISA and that it was not the court’s purview to reconstruct it.

Discussions with managed care reviewers about the risks to societal members and family should a patient be discharged often are exercises in frustration for clinicians and are viewed as irrelevant in the decision process as to whether to extend hospital stay. In fact, it is our experience that when the risk of dangerousness is clear and suggested, the managed care reviewer’s response is often that “the justice system will have to take over.” We have already witnessed the cost shift from mental health to the penal system in the context of deinstitutionalization (27). The difference here is the shift in the locus of responsibility for adolescent behavior from the mental health system to the justice system, with increasing numbers of adolescents tried as adults. Young children will live long lives in peril and with inadequate treatment in jail if the intervention they receive at critical times is not adequate. The short-term orientation of managed care will produce long-term costs for society, which it will continue to pay for many years.

From the perspective of public policy, our criteria for who gets treatment, when, and how much has changed. Risk to individuals and families is no longer an acceptable criterion to justify the expenditure of increased health dollars. We need to reevaluate this decision and ask whether people would rather pay for increased mental health benefits that allow for early identification and comprehensive treatment of youths at risk or whether they are willing to accept the added risk in their communities and schools that results from the fewer dollars expended for treatment and support. Satterfield et al. (28) reported in 1987 a multimodal treatment model for hyperactive boys that resulted in less antisocial behavior, improved academic performance, and better social adjustment at home and at school after 3 years of treatment. When followed into adolescence, these boys were found to have a juvenile arrest rate that was 50% lower than that of a group treated only with drugs.

At present, these questions are not articulated in the boardrooms of major employers and other payers in the private sector. Instead, these children are being lost to the chasms that exist among mental health care, juvenile justice, and social policy. There is no agreement about whether these children are “sick” or “bad.” Child mental health professionals and child advocates, however, are increasingly concerned about this as they see more and more money being given to juvenile justice to prosecute and incarcerate teens as adults. This seems particularly ironic to child psychiatrists who understand the history of the field. The first child guidance clinic, the Judge Baker Guidance Center in Boston, was founded by Judge Baker with the intention of helping the Boston Juvenile Court to better understand and treat troubled youth. While the data since have shown multiple causal relationships between early childhood behavioral disorders and later court-related problems, we have yet to develop a cohesive, rational policy that addresses these complex factors.

In the future, these questions will become unavoidable as states move to managing public sector health care dollars in the form of managed Medi­caid. The attempts to control health care dollars have, with children and adolescents, given rise to increased spending for juvenile justice systems and social services, much like deinstitutionalization gave rise to the increase in nursing home residents and homeless mentally ill. An optimistic view projects that it may well be the merger of the business world of medicine with the public sector that could ultimately bring the boardrooms together with the social policy makers. This will only happen, however, if the contractual arrangements stipulate long-term risk and responsibility with commensurate changes made in the ERISA statute that now protects the self-insured.

Ethical Concerns

The treatment of potentially violent patients poses complicated ethical questions for the responsible physician. While there are financial pressures present in most psychiatric facilities to remain “managed care friendly,” the primary ethical responsibility has to be to the patient. The elusive proprietary term “medically necessary,” used to guide payment, cannot be the sole criterion for the treating physician who is ultimately responsible for the patient. Simon (29), in his recent article about the responsibility of psychiatrists in discharging sicker and potentially violent inpatients, outlines the duty to treat, citing several legal references that clearly outline the physician’s responsibility to the patient regardless of whether or not the insurance company has authorized payment for the services.

Unfortunately, it takes considerable inner strength to withstand the short-term cost pressures from payers. This also raises questions of ethics. It has evolved that treating physicians and reviewing physicians may be operating under different ethical standards. The treating physician has an obligation to treat the patient to the best of his or her ability. This may include providing additional information about alternative treatments, some of which may not be covered by the third-party payer. The physician reviewer is in a contractual relationship with the managed care company to provide expertise about what is reasonable treatment within a covered benefit. The value system that guides the practices of reviewers is driven by cost-containment pressures.

This is in sharp contrast to the ethics and values of treating physicians, who bear direct responsibility to the patient and his or her family. While the treating physician in this case was continuously concerned about her responsibility to the patient and his mother, as well as to the community, the matter was rather simple for the reviewer. The reviewer, consistent with his contractual agreement with his employer, stated that the patient’s treatment was no longer reasonable within the covered benefit. He had no duty to disclose other treatments that might be available, such as residential treatment, and no duty to treat until another treatment setting could be found. Geraty et al. (30) and others argue that the ethics of managed care are based in society’s wish to draw limits on what can be paid for. However, this ethical view is in conflict with the Hippocratic oath, which states, in a version approved by the American Medical Association, that as a physician, “into whatsoever house you shall enter, it shall be for the good of the sick to the utmost of your power, your holding yourselves far aloof from wrong, from corruption, from the tempting of others to vice . . . that you will exercise your art solely for the cure of your patients.” While some might argue that physicians have always had difficulty upholding this oath, it has nevertheless remained the standard value system taught to physicians for 2,000 years. Managed care is a great challenge to the Hippocratic ethical tradition, which guides us to, above all, “do no harm.”

Received April 27, 1998; revision received Aug. 19, 1998; accepted Sept. 1, 1998. From the Sheppard Pratt Health System. Address reprint requests to Dr. Sharfstein, Sheppard Pratt, 6501 North Charles St., Baltimore, MD 21285