Both before and after the reformation of the early 1800s we know that disturbed children and adolescents were at times placed in these facilities along with adults. Beyond a few scholarly reports whose aim was to document the occurrence of severe mental illness in the young as a source of curiosity in itself, little is known about the care and outcomes of children in asylums. The first dedicated child psychiatric inpatient units as such in the United States were created in the 1920s and ’30s, mostly as custodial services for children with postencephalitic brain disorders. The prevailing philosophies of these settings and their successors are discussed later.

By the mid-1980s, inpatient beds for children and adolescents proliferated markedly, chiefly in private sector general and specialty psychiatric hospitals. The U.S. Supreme Court’s 1985 decision in Massachusetts v. Metropolitan (3) supported mental health coverage by insurance plans, and earlier the Parham v. J.R. decision affirmed that parents’ could compel admission to psychiatric inpatient care for an unwilling minor, much as any other necessary medical treatment (4). Lengths of stay were extensive, standards for admission were liberal, and many inpatient settings adopted rather high, if subjective, criteria for judging wellness to warrant discharge. Direct-to-consumer advertising by these facilities, aimed at parents worried about their sullen or unruly teenagers, became commonplace. Inpatient care also served an evaluative purpose, with some referrals made for diagnostic clarification.

However, by the early 1990s this trend had rapidly reversed. Scandals plagued certain for-profit facilities. Managed care established increasingly strict criteria to justify inpatient admission. The Parham decision was partially blamed for the ease with which parents could have their adolescents psychiatrically hospitalized, often for rebellious or obnoxious behavior alone. The managed care revolution greatly reduced the length of stay, though actual rates of admission have remained mostly unaltered or increased, and perhaps readmissions have increased (5,6). In the public sector, policymakers also recognized that a disproportionate amount of the mental healthcare dollar was spent on very costly inpatient care to the detriment of less expensive community-based options that were far more appropriate for many children. This line of thought followed naturally from the somewhat earlier deinstitutionalization movement for adults with chronic psychiatric conditions.

Child advocates, who had earlier called attention to the scarcity of community-based resources that might help impaired children to remain at home (7,8), found that cost concerns were aligning policymakers’ interests toward development of a fuller range of supports in the community calibrated to the needs of individual children and families. The espousal of such continuum-of-care principles by federal, state, and county levels of mental health planning increased significantly the array of community-based programming. Such services included in-home and out-of-home respite services, supportive case management, therapeutic after-school programs, innovative programs based on “blended funding” from several agencies— all aimed at avoiding or reducing hospitalizations and optimizing a child’s opportunities for successful retention in the community. Localities, though, still vary widely in the availability and quality of these resources. Moreover, some evaluation projects raised the prospect that such enhanced services do not necessarily produce more favorable outcomes, although families do find them preferable to service systems that omit them (9).

At the present time, inpatient psychiatric treatment in the United States is regarded, properly, as an expensive resource to be used sparingly and as a last resort for the most ill of youngsters. Comparatively few children now depend on long-term psychiatric inpatient settings to receive care, and acute-care lengths of stay are shorter. Admissions for evaluative purposes often occur only upon court order. Nonetheless, inpatient treatment remains an important component of the system of care for very ill youngsters. Eligibility for many of the community wraparound services discussed earlier often depends on prior psychiatric hospitalization, or at least on the risk for admission. Moreover, despite the value placed on alternatives to restrictive placements, admissions to acute inpatient settings with a principal diagnosis of psychiatric disorder have increased strikingly between 1996 and 2004 among children (45% population-adjusted increase) and adolescents (25%), greatly outpacing the 11% increase for adults (10).

The first children’s units were essentially custodial in emphasis, due to the mostly organic impairments of the patient population (11). Most would be regarded today as fundamentally mentally retarded, whether by congenital or acquired (usually infectious) conditions.

By the mid-twentieth century, psychoanalytic thinking dominated child and adolescent psychiatry. Child psychoanalysis early on emphasized the primacy of interpersonal experience in development and emotional disturbances, and different strains of thought came to converge on a basic notion that early attachment and nurturance, and the promotion of autonomy, formed the template for a person’s manner of relating to the world (12). Some features of hospital treatment now considered at best “necessary evils” came to be seen as rather integral elements of the era’s psychoanalytically oriented model of inpatient care. In particular, separation of the child from his or her putatively pathogenic home environment was felt “to be the first requirement for successful treatment … since he is comparatively helpless to re-order his own surroundings or change them to better suit his needs” (13). Consequently, hospital settings did not seek to redress directly perceived deficiencies in the family. The premise of inpatient and residential settings for quite disturbed children was that in a more capable caregiving environment the child might have corrective experiences that would allay basic insecurities and foster ego development so that better modulated behavior and affect might blossom. Until such time, residential settings also were thought to provide an empathic surrogate “holding environment” (14) that would help the child manage destructive urges or disorganized behavior. There was no expectation that this would be a rapid or easy process and long hospital stays were common.

In the 1960s and 1970s, learning theories from experimental psychology, especially its neobehaviorist schools, acquired steadily greater traction in clinical psychology. Interventions based on environmental manipulation to modify pathological behavior, known broadly as behavior modification or behavior therapy, gained wider application in facilities for developmentally impaired and chronically mentally ill adults. Approaches based on operant conditioning principles prioritized the adaptive behaviors that the individual lacked, and sought to promote them by following their appearance
with rewards or reinforcers. Likewise, efforts to eliminate (or “extinguish”) the problematic behaviors involved withholding the reinforcer that it usually elicited (such as attention or avoidance) or by applying an aversive consequence. Approaches rooted in classical conditioning sought to unlink troubling exaggerated emotional responses, such as intense anxiety, from the relatively benign stimuli that had come to elicit them, or to reduce the attraction of a problematic stimulus, such as tobacco, by associating it with something unpleasant. Integrating these and other principles, applied behavioral analysis provided some elegant demonstrations of how systematic assessment of antecedents, behaviors, and consequences could lead to interventions that resulted in marked behavior change. Offering patients explicit training and practice in specific behavioral skills, such as assertiveness, anger control, and social interaction, were also undertaken in a variety of formats.

Particularly influential reports showed dramatic improvements in the social engagement and activities of daily living among chronically ill adults, the acquisition of some language by autistic children, and reductions in self-abusive behavior by those with mental retardation (15,16,17,18).

Settings that provided round-the-clock care were ideal for the implementation of treatments that required consistent monitoring of behavior and the systematic manipulation of the consequences for that behavior. Moreover, constant supervision by a professional staff facilitated recording that quantifies behavior, a methodological necessity of behaviorism and an appealing feature of behavior therapy. “Token economies,” in which patients earned chits toward various privileges for prespecified behaviors, became especially widespread, and influenced the rather ubiquitous point or level systems in the inpatient and residential settings of today. The appeal of these systems may derive partly from their implementation on a unit-wide basis, in that many patients will share similar behavioral objectives and thus offer a common template for the whole service. In contrast, classical conditioning and applied behavioral analytic approaches are highly idiographic and the staffing of most psychiatric settings seldom permits such intense staff training and individualized implementation efforts on a routine basis. The obvious availability of a peer group also enables on-the-spot opportunities to develop and practice social and other skills. Behavioral interventions are conceived to yield dividends in weeks, or at least that is the period for evaluating the usefulness of a particular treatment plan.

However, the very intensity of specialized out-of-home settings that facilitates ecological interventions of these types militates against the generalization of behavior changes to other settings once the individual no longer experiences the environmental contingencies that supported them. One potential remedy is to regard the patient’s family as a suitable locus for intervention that might enable maintenance of gains after the youngster’s return to the community. Outpatient “child guidance” clinics for youth with conduct problems had since the 1920s included parents seeing a social worker as ancillary to the child’s psychotherapy (19), and family therapy as an identifiable treatment modality for outpatients had been practiced since at least the 1940s. However, family-focused treatment did not assume a widespread central position in the treatment of hospitalized children’s psychiatric disorders until the 1980s. In 1980, the most common type of treatment received across all settings was individual therapy, received by 89%, while family therapy was provided to only 38% (20). At that time, some settings had programs where entire families were admitted and under constant observation.

Family therapy encompasses a range of approaches and theories, so its incorporation into psychiatric inpatient settings displays eclecticism. However, child psychiatry is perhaps unique in that the recognized standards of care for several of its most common disorders explicitly include interventions focused on parent–child interaction. These include parent management training (PMT) for conduct problems, the mainstay of this patient population. The obvious continuity between behavior therapy approaches in the inpatient setting and PMT has led to the latter’s becoming a significant component of family intervention in many settings. Moreover, parents can see methods modeled by staff and undertake them with guidance and support during hospitalization. Regardless of the specific disorder and treatment approach, psychoeducation and support to families coping with an ill child are now universally judged as essential to compassionate care by all members of the treatment team. This is obviously a significant philosophical shift from earlier times, when families were at best ancillary to the child’s treatment, when not the object of clinicians’ reproach as the source of the patient’s illness.

As behavioral therapy matured as an influential clinical specialty, a significant development was the view that a person’s thoughts and beliefs are behaviors in their own right that could be altered and thereby affect overt behavior or mood. Imparting systematic methods of thinking through problems to suppress disadvantageous “automatic” responses (problem solving), and reevaluating unhelpful beliefs that motivate maladaptive behavior (cognitive restructuring) were melded with learning approaches to yield the area we know today as cognitive-behavioral therapy (CBT). These treatments have had modest impact on the overall inpatient milieu, but are often incorporated into group therapy as befits clinician preference. A few settings have woven some of these approaches, such as dialectical behavior therapy for self-injurious adolescents, into their milieus, such that patients having trouble are prompted and coached in situ to utilize these skills. Such milieu-based adjuncts to CBT are probably underutilized, especially with adolescents, but multidisciplinary staff training in these methods requires a substantial commitment that may be the limiting factor in many settings with short stays.

Pharmacotherapy now plays a prominent role in the psychiatric treatment of youth. The 1980 NIMH report (20) indicated that 42% of child and adolescent inpatients were treated with standing psychotropic medication. While the pharmacologic evidence base still has extensive gaps, the use of medication in inpatient settings has increased dramatically. It is now the rare youngster whose inpatient or residential treatment does not include medication (21,22,23). Greater severity of illness to obtain approval for admission may account for some of this change, but it does mirror data showing a corresponding increase among outpatients. Consequently, youth are more likely to be receiving treatment with psychotropics, often two or more, at the time of admission. Medication trials tend to be a reason for continuing stay most acceptable to managed care reviewers. The combined effect is that the role of the child psychiatrist in these settings has increasingly focused, perhaps to the detriment of other areas, on which preadmission agents were doing any good, which were potentially making things worse, and what to try next, all in the context of constrained lengths of stay with a possibly more treatment-refractory patient population.

The current ethos prevailing in psychiatric care is that it should be provided in the least restrictive environment possible. Many innovations in psychiatric services for children have therefore aimed at reducing reliance on congregate care settings such as inpatient and residential facilities. These include family support services, such as emergency respite— both in and outside of the home— using trained professional or paraprofessionals, volunteers or other parents. This service may take the form of regular after-school or weekend specialized recreation or therapeutic care programs. Home care programs may include
child supervision, instruction in parenting skills, and case advocacy assistance. Home-based crisis intervention programs provide in-home services to families for 4–6 weeks with the goal of avoiding hospitalization. Family-based treatment uses surrogate families who are “professional parents” to care for and treat youth with serious emotional disturbance (24). Some well developed programs have shown benefits compared to “usual care” in the hands of the group that designed them and a major challenge concerns the exportability of such services to other facilities, especially when their adoption represents a marked shift from practitioners’ prior mode of functioning (25).

Although an important development, these alternatives remain unevenly available and demand exceeds supply in many localities. Priority for intensive community-based services often goes to youth deemed to be at risk for out-of-home placement. As a practical matter, such risk is quite often demonstrated by prior hospitalization. Consequently, a more prominent function of inpatient and residential settings has become liaison with community-based care providers and schools to recommend and arrange implementation of appropriate postdischarge services from within this continuum of care. Several states and service regions have also implemented a centralization mechanism whereby a common application for services is submitted to the single point of entry (SPOA) reviewing body for appropriate assignment to community-based intensive treatment options. As it happens, though, there is little empirical basis to support these recommendations, or the prognostic judgments that are implicit in them.

A form of extended psychiatric triage has also emerged that may divert some admissions. Such services usually constitute an enhanced psychiatric emergency room service by providing a small number of “holding beds” for up to 72 hours, a mobile crisis team that can be called to a home, school, or other community setting by a parent, concerned citizen, police officer, etc. Such an enhanced emergency service usually has 24-hour socialwork coverage to work rapidly on community-based disposition whenever possible. Referred to as CPEPs in some localities (comprehensive psychiatric emergency program), their general objective is to treat in the emergency room if possible, and thus avoid inpatient hospitalization. Only rarely, however, do these settings have a separate section for children.

In hospital settings, daily rounds nowadays typically begin with the question, “Why does this child need to be in the hospital?” All hospitalizations covered by managed care plans are constantly monitored by the insurance companies’ reviewers. Publicly funded care is also subject to retroactive denial of payments if inspection of the medical record is judged to lack sufficient justification for inpatient care. Although minimizing the time a child spends in a hospital is not a controversial goal, a widespread sentiment is that aggressive cost containment may have compromised care. For instance, payers often regard as inertia the observation of a child after withdrawing preadmission medications, which biases the system toward initiating new, possibly superfluous pharmacotherapy. This is another area deserving more systematic study.

As noted earlier, acute care is now only deemed appropriate when less restrictive alternatives have been considered, have failed, or are not available. The most common reason for admission is behavior felt to place the child or others in danger. This may translate to suicidal ideation, intent, or attempt, or may reflect sufficient threat of aggression or actual aggression such that the caretaking system, school or home, is concerned and unable to handle the youth. It is difficult nowadays to get authorization from payers to admit for a purely diagnostic assessment. Indeed, many components of such evaluations, say, psychological testing, MRI, lab tests, and the like can generally be secured on an outpatient basis, and payers seldom find the putative value of inpatient observation a cogent rationale for admission.

Specialty acute inpatient care is a locked and therefore secure setting, which includes round-the-clock staffing, and the capacity to restrain or seclude an out-of-control patient. Despite the emphasis on the therapeutic value of the “structure of the unit,” structure is a generic attribute of several other less restrictive settings, though it must be conceded that the locked setting confers a special level of environmental control.

The majority of short-term psychiatric treatment is provided in units located in freestanding psychiatric hospitals or in units located in general medical/surgical hospitals. Acute care was arbitrarily defined as being for 30 days or less. This was driven by the typical 30-day insurance policy rather than having any established relationship to diagnosis, progress, or prognosis. Nevertheless, this time frame certainly became embedded in the format of inpatient assessments and in the mindset driving disposition planning and implementation. The time frame, in turn, influenced many aspects of the inpatient therapeutic structure, such as how long it took to earn privileges, obtain an off-unit pass, etc. Beyond 30 days, applications for intermediate care (30–180 days) will be entertained, usually in the regional state facility.

Acute and intermediate care facilities generally serve children age 4 to 18 years. Very few programs serve a preschool population in an inpatient setting. The developmental disorders that in the past had occasioned the need for hospital care in this age group are now, thankfully, more successfully addressed by the universally available early intervention system and the significant growth of highly specialized education settings for very young children. It is customary to have separate inpatient settings for children up to age 12 or 13 and for adolescents up to age 18. Some inpatient units treat children and adolescents together, or those over the age of 16 may be admitted to adult settings, but these practices more often derive from necessity than philosophy.

Municipal and county facilities often have a public mandate to serve the local court system. Judges have the authority to mandate assessments in such units for defined timeframes (21 days is common). A complete assessment of the child’s mental condition, including psychological testing, and a psychosocial assessment of the child’s family, school, and community culminate in an advisory report to the court. In 2003, 2.2 million youth in the United States were arrested for delinquency or status offenses, of whom 1.8 million appeared in juvenile courts; 329,000 of these were detained in a residential setting for assessment. In addition, the census of youth in residential correctional facilities following adjudication was just under 100,000 (26). This population is considered to be massively underserved with respect to psychiatric illness, the prevalence of which is now known to be quite high (27). To the extent that mental health services variously exist within the juvenile justice system’s residential programs, they constitute another version of a psychiatric inpatient provider system for incarcerated youth.

Inpatient units can be subspecialized for the care of unique psychiatrically impaired populations. Eating disorder services are one example which enable the more intensive medical management these youth require initially with the specialized psychiatric care that does most of the heavy lifting toward
recovery adequate for the resumption of outpatient treatment. Special psychiatric units for the deaf, the blind, and for youth with cooccurring developmental disabilities also exist. Such facilities represent a type of nursing home care for the medically fragile who also have significant psychiatric problems, and whose families are unable to care for these children with the available community support services.

Of the varieties of noninpatient programs, partial hospitalization is the most intensive. The clinical challenge for this level of care is the provision of short-term, crisis stabilization as an alternative to inpatient care or as a step down from inpatient care.

Partial hospital programming may be provided on an inpatient unit. Some refer to this as unit-based aftercare. It allows the patient to continue working with the same treatment team and the same peer group rather than forcing a change for a short period of time. Partial day hospital licenses require treatment to be no longer than 6 weeks, as well as daily chart documentation of progress, much like in an inpatient setting. However, most inpatient units have a high inpatient census and staffing is not necessarily easily expanded to cope with “day patients,” so the model has obvious practical limitations.

More typically, partial hospital programs have their own staff, space, and school. Despite commonly being licensed for 6-week lengths of stay, managed care review generally constrains the actual duration of the patient’s involvement to only days or a couple of weeks at most. Since the setting is generally open, and regulation does not allow restraint or seclusion but does permit therapeutic hold and use of a quiet room (no locked door), there are practical limitations on the degree of psychopathology for which these settings are suitable. Programmatically, the range of therapeutic services are similar to inpatient settings, and include individual, group, and family therapies, recreation and rehabilitation therapies, medical care, and psychopharmacology.

Day treatment, sometimes referred to as continuing day treatment, differs from partial hospitals in several ways. Length of stay is much longer, often driven by the school year’s calendar. Children attending day treatment must usually be certified as being in need of that level of care by their local school district’s committee on special education, since these authorities generally assume the cost of the program’s educational component. The district is often intending that a child’s stay in day treatment will last for at least a full school term or even for the entire academic year. Day treatment settings are as likely to be, fundamentally, schools with high psychiatric involvement as they are to be psychiatric settings as such. Even when a psychiatric facility houses a day treatment program, it is often partnered with a local education systems’ special education division. The school service may therefore have some independence from the rest of the facility that the psychiatric staff should be aware of and respect to maintain good rapport with all those involved in the children’s care.