Although liability issues in the treatment of children and adolescents are similar in many ways to those in the assessment and treatment of adults, some characteristics of children and adolescents raise special issues. This article focuses on the differences in work with children and adolescents that pertain to liability in the assessment and treatment management of youths at risk of harming themselves or others.
Consent
Children and adolescents under the age of 18 are, for most purposes, deemed legally incompetent. However, the age of consent for medical treatment varies by jurisdiction and by the treatment involved. The most common exceptions to the rule of needing parental consent include emergencies and emancipated minors (minors who are married; in the military; or found to be emancipated by a court, generally because they are living apart from their parents and are self-supporting). Certain procedures, such as research that will not directly benefit the minor, require the minor’s assent as well as the consent of a parent or guardian (
1). Minors can also generally provide consent for certain medical procedures, such as issues related to pregnancy, sexually transmitted diseases, and, if permitted by a judge, abortion. About half the states allow some minors to make some other treatment decisions, most commonly either by specifying an age below 18 for consent or by allowing “mature minors” (generally not clearly defined) to consent (
2). Some states also allow certain minors who are admitted to a psychiatric hospital over their objections to have their admission reviewed by a judge at a hearing after admission. In jurisdictions in which a minor can consent in certain situations, that consent authority is typically in addition to the ability of parents to consent. Clinicians who work with minors need to know the consent law in their state, because not having appropriate consent can be a cause of liability.
For minors who cannot consent, a parent, guardian, or—if the youth is in state custody—the state or other person legally empowered to act as a parent must generally provide consent for treatment. For brevity’s sake, the term parent will be used in this article to refer to any adult legally empowered to make decisions regarding a child, including guardians or, for a child in state custody, a state-designated official. A natural parent may not always be a parent for purposes of legal decision making, such as when only one parent is awarded legal decision-making authority after a divorce or when parental rights are terminated.
Notwithstanding legal rules that may require a parent or guardian to provide consent, extensive research on children’s capacity to make appropriate medical decisions supports the finding that most children 12 years or older of normal intelligence have the capacity to make medical decisions similar to those of adults (
3,
4).
A requirement for parental consent has important implications for assessment and treatment: If parental consent is required or sufficient, the hospitalization of a minor will be considered voluntary if the parent agrees, even if the minor objects. Thus, as will be discussed further, in cases of risk of harm to self or others, minors can often be hospitalized over their objection without meeting civil commitment criteria.
Suicide Risk Assessment
Liability for an adolescent patient suicide is the second most common claim for malpractice against psychiatrists who treat this population, after “incorrect treatment” (“The Psychiatrists’ Program Cause of Loss [2009–2018],” personal communication, Professional Risk Management Services, 2019). Although the general principles of assessing risk for suicide are similar to those of assessing such risk in adults—see the Pinals article in this issue (
5)—there are a number of significant differences, summarized in
Table 1.
Suicidal ideation and attempts are surprisingly common in adolescents. The Youth Risk Behavior Survey of high school students conducted annually by federal agencies contains questions about suicidal thinking and behavior. The 2017 survey found that, in the previous year, over 17% of students had seriously considered suicide, over 13% had made a plan, 7.4% had made an attempt, and 2.4% had made an attempt that resulted in an injury sufficiently severe to require treatment by a doctor or nurse (
6). Girls had higher rates in all four categories than boys, although adolescent boys die by suicide at a higher rate than girls (15.9 per 100,000 for boys 15–18 years old in 2017, and 5.4 per 100,000 for girls) (
7). With the high rates of suicidal thinking and planning, given that the rate of suicide deaths for 15- to 18-year-olds in 2017 was 10.76 per 100,000 (0.0176%), it’s clear that the ratio of seriously considering suicide and having a plan, generally considered to be a high-risk situation, to completed suicide is over 700:1. Thus, the specificity for predicting suicide by virtue of having suicidal ideation and a plan is quite low.
A previous suicide attempt is often considered the most significant risk factor for suicide. However, a recent study found that approximately three quarters of youth suicides occurred on the first attempt (
8), a higher rate of completed suicide at the initial attempt than was found for other age groups. Thus, most youth suicides are not predicted by a previous attempt. About 80% of these suicides were by boys, and 85% involved the use of a firearm. A study of Finnish adolescents found a precipitating stressor in 70% of completed suicides (
9), and half the precipitants occurred within the 24 hours preceding the suicide, which allows only a very narrow time window for intervention.
These factors imply that looking at risk factors alone has little predictive power in an individual case. Therefore, it is important in a clinical assessment to look at details that are specific to the case. Utilizing the Beck Suicide Intent Scale, Kingsbury (
10) identified four factors that are useful to consider in assessing intent in a recent attempt, summarized in
Table 2. Adolescents may minimize talking about their intent after an attempt, so it is important to obtain corroborative data about what occurred.
There is an extensive literature regarding the details of the assessment of risk and treatment of suicidal adolescents (
11–
13) that will not be summarized here. In malpractice cases involving the suicide of an adolescent outpatient, the most common issue is whether the patient should have been hospitalized, a decision that turns on whether the suicide was reasonably foreseeable. In most cases, parents need to be brought into the decision as to whether to hospitalize. In jurisdictions where only a parent’s consent is required for voluntary admission, the bar for hospitalization is lower than for an unwilling adult at risk. On the other hand, in many cases, close observations by parents can provide enough support, structure, and surveillance that a patient can be reasonably treated as an outpatient. However, parental cooperation needs to be carefully assessed. Brent et al. (
14) made the surprising finding that less than a third of parents of depressed adolescents were compliant with a recommendation to remove firearms from the home.
The low predictive value of risk factors for completed suicide means that considerable clinical judgment is involved in weighing the risk factors and protective factors in an individual case. From a liability standpoint, the key is to conduct appropriate initial and ongoing assessments. Almost as important is the documentation of not only the risk and protective factors but also the clinician’s reasoning as to the weight to be given to those factors. This is especially important in those cases in which some risk is identified but the decision is not to hospitalize the patient. In particularly difficult cases, consider Jonas Rappeport’s suggestion: “When in doubt, shout!” (
15). Have another clinician assess the patient and document his or her judgment. In case of an adverse outcome, the question will be whether the clinician acted reasonably. The opinion of a second clinician formulated at the time can go a long way in combating the later opinion of a plaintiff expert who wasn’t there looking back through the “retrospectoscope.”
Danger to Others
Violence is surprisingly common in adolescence. One review reported that by age 17, 30%–40% of boys and 16%–32% of girls had committed a serious violent offense, defined as an aggravated assault, robbery, gang fight, or rape (
16). Most of these did not result in arrest. Serious violence toward others generally starts in adolescence. The peak age for onset of violent offending is 16, and it is rare for a person to first become violent after age 21 (
17). Homicide, for many years the second leading cause of death in adolescents, has now fallen to third place, after suicide (
18). The good news is that for a majority of violent adolescents, violent offending is an adolescent phenomenon, and most violent adolescents do not progress to adult criminal careers (
16).
Mental health clinicians who work with children and adolescents see aggression at all ages, from the aggressive toddler to the latency-aged bully to the angry adolescent. Although all of these present clinical challenges, liability concerns in dealing with aggression (apart from those arising from medication use) are most common with adolescents who are at significant risk of seriously harming others.
In this issue, Resnick and Saxton (
19) discuss the principles in assessing violence risk in adults. Many of those principles also apply to work with adolescents, but because of developmental and legal differences, there are some major differences, summarized in
Table 3.
Adolescent patients at risk of violence are a heterogeneous population but tend to fall into two major groups. The first group includes those whose violence is closely tied to mental illness: the psychotic and paranoid youth, the adolescent with an impulse control disorder, and the patient with a developmental disorder and accompanying serious deficits in social judgment. The key to reducing violence risk in this group is treating the underlying mental disorder. The second group contains those adolescents who have a history of conduct disorder, multiple prior diverse offenses, repeated violent behavior, and association with delinquent peers and who are often referred because of their involvement with the juvenile justice system. Although this group has a high rate of diagnosable mental disorder, their violence risk is not easily explained by a mental disorder (
20). Treatment of their mental disorder is useful in ameliorating suffering, but successful interventions to reduce violence generally focus on strategies that address criminogenic factors.
As with adults, the clinician working with youths will be held to the standard of conducting an appropriate risk assessment. However, risk assessments for the two groups are somewhat different. For the first group, whose violence stems from mental illness, risk assessment addresses dynamic factors such as particular stressors or triggers and details of parent–child interactions. For the delinquent group, a number of risk assessment measures, such as the Structured Assessment of Violence Risk for Youth; the Psychopathy Checklist, Youth Version; and the Level of Service/Case Management Inventory, have received extensive study. Although the measures are useful, the correlations with future violence (rs=.25–0.30) (
21) are too low to be predictive of short-term violence in an individual case. While a rating scale classification of a youth as high risk has implications for need for treatment and case management, it is not generally sufficient to justify hospitalization. For the clinician treating an individual patient, the decision as to whether the risk is sufficiently high to warrant hospitalization generally turns on whether the danger is reasonably foreseeable, which typically requires evidence of a recent act or specific threat rather than a more general judgment about the likelihood of dangerousness at some future time. Short of any significant evidence of intent to harm, clinicians are unlikely to be held liable for failing to prevent violence.
School-Shooting Threats
In the wake of multiple mass school shootings, school officials are highly sensitized to potential threats, and statements that students may make in jest or to be provocative often trigger an immediate suspension and referral for evaluation. Evaluations of such threats raise different issues from the typical assessment of a potentially violent adult or adolescent. For example, a history of prior violence, a potent risk factor in most dangerousness assessments, is seldom present in the history of K–12 students who have committed mass violence (
22).
Mass shootings are a form of predatory violence, rather than being impulsive, and it is easy for an adolescent being interviewed simply to deny intent. The one commonality found in one study of K–12 shooters is that almost all had communicated their intent to peers, but their threats had not been taken seriously (
22). These two facts imply that an interview of the suspected minor, alone, is very likely to be insufficient: collateral information, including talking to the friends and parents of the evaluee, as well as to school personnel, is necessary. When a parent brings a suspended student to a psychiatric emergency service for a “clearance letter” to return to school, the student should almost always be referred for a more detailed evaluation that includes obtaining collateral data. Although, in many cases, a mental health evaluator can conclude from interview and collateral sources that a threat actually was made in jest or otherwise was not serious, in more complex cases, a comprehensive evaluation may require a team approach, involving school personnel and law enforcement, and include such tasks as forensic assessment of a youth’s computer to determine whether there has been a preoccupation with internet sites related to school shootings. Details of the techniques for assessing a potential school shooter are beyond the scope of this article but are available in the literature (
23–
25). From a liability perspective, it is important that the clinician recognize that these are specialized evaluations; that an interview of the individual suspect is rarely a sufficient evaluation; and that, depending on the seriousness of the threat, involvement of school personnel and law enforcement may be required for a comprehensive assessment.
Malpractice Litigation
Although the general principles of malpractice law apply to malpractice cases involving minor patients, differences in working with patients who are minors lead to somewhat different considerations. A 2019 analysis of the previous 10 years of claims arising out of the treatment of minors covered by Professional Risk Management Services, Inc., shows that the most common claim is for the catch-all “incorrect treatment” category (25%), followed by claims involving suicide or attempted suicide (19%) (“The Psychiatrists’ Program Cause of Loss [2009–2018],” personal communication, Professional Risk Management Services, 2019). Suits involving failure to protect from violence were much less common (3%). Overall, child psychiatrists are at lower risk of being sued than general psychiatrists, and this is reflected in child psychiatrists’ often receiving premium discounts for working with a younger population. However, clinicians treating children have a longer exposure time for suit. Because a child cannot file suit on his or her own, the statute of limitations (the amount of time a plaintiff has to file suit after the injury or discovery of the injury)—typically 2 years for adults—is considerably longer and, in many jurisdictions, does not begin to run until the child patient turns 18.
The immaturity of minors affects the course of litigation in child malpractice cases. Because minors are limited legally in making decisions for themselves, parents have more involvement, and communications to parents are scrutinized. Because parents often receive advice from the clinician about how to interact with their children, the parents are often involved in the case as quasi-patients. Children and adolescents are less responsible for their actions than are adults, which tends to lead juries to shift more responsibility to the clinician in cases that involve noncompliance with treatment or suicide. Finally, juries may be more sympathetic to injured children and so be more inclined to grant them compensation.