Schizophrenia is a heterogeneous disorder defined by sustained periods of psychosis and functional deterioration in the major arenas of life, such as interpersonal relations, education, employment, and self-care (R1). The boundaries of the disorder have been a source of debate since its earliest delineation by Kraepelin (R2) and Bleuler (R3) a century ago. Formerly, especially in American psychiatry, all chronic and recurrent psychotic symptoms were grouped together under the category of schizophrenia. Since the advent of DSM-III in 1980, however, it has been recognized that episodic psychosis in the context of severe mood symptoms and psychosis attributable to identifiable medical causes are separate entities that must be ruled out before a diagnosis of schizophrenia can be made.

The presentation of schizophrenia varies enormously. The current definition of schizophrenia includes distinct symptom domains, each of which may vary independently, as well as diagnostic subtypes. The concept of symptom domains represents a dimensional approach to the disorder, in which positive, negative, cognitive, and affective symptoms, which are present to varying degrees in every patient with schizophrenia, are characterized as separate components of the illness. The subtypes of the disorder represent a categorical approach to symptom description, each defined by the presence or absence of specific symptoms. These two approaches to symptom description serve different, complementary functions and aid the clinician in conceptual understanding, prognostication, and treatment selection.

Symptom domains

Positive symptoms are synonymous with active psychosis; they are referred to as "positive" in the sense that they are added onto normal experience. They include delusions, hallucinations, and thought disorganization (Table 1T1). Delusions are the most common psychotic symptom in schizophrenia, occurring in 65% of patients (R4, R5). Hallucinations and thought disorganization are each described in about 50% of patients. Most patients experience multiple symptoms (R4). Positive symptoms are highly correlated with hospitalization but only minimally with other measures of dysfunction.

Delusions are false beliefs based on internally generated concepts of reality or on incorrect inferences about actual events. By definition they are not amenable to correction by contrary evidence or persuasive argument. Although they may occur within the context of specific cultural customs or belief systems, they are not shared by others who have a similar background, and they are readily acknowledged by peers to be aberrant. They occur in a wide range of categories, such as persecutory, religious, somatic, grandiose, and so forth. Ideas of reference, the belief that neutral events are uniquely directed to the individual, are of special note because of their frequency. Among such delusions are the belief that messages from the television or radio are directed specifically at the person, that newspaper articles secretly refer to the person, or that conversations among strangers are about the person. Also common is the belief that one’s thoughts are being read or controlled by others, a symptom so striking that it was once thought to be pathognomonic for schizophrenia (R6). In fact, none of these symptoms are unique to schizophrenia, nor do they carry special significance for treatment or prognosis (R5), but their diversity is mentioned here to alert the clinician to the range of possibilities that must be considered in the assessment of possible psychotic symptoms.

Hallucinations are false perceptions occurring in any sensory modality. Auditory hallucinations, especially voices, are the most common in patients with schizophrenia (R7). Hallucinatory voices may vary in frequency, intensity, number of speakers, and content. Command hallucinations, voices instructing the patient to do something, may place the patient at risk of acting on the commands. Visual hallucinations, although sometimes associated with substance intoxication or withdrawal, are also common in schizophrenia (R7). Similarly, tactile, olfactory, and gustatory hallucinations may occur in schizophrenia and are not uniquely attributable, as some older texts suggested, to seizures, substance abuse, or other medically identifiable causes.

Thought disorganization is most often noted as bizarre behavior or a disruption of the logical associations of speech. Disorganization is especially disruptive to the routine tasks of life and creates severe problems in social, occupational, and self-care functions.

Catatonia is an altered state of motor activity and attention. Lack of movement is common and may be accompanied by either catalepsy (waxy flexibility), in which the person maintains a posture into which he or she has been passively moved by the examiner, or negativism, in which the patient actively resists such attempts by the examiner. Echopraxia refers to the patient’s inappropriately mirroring movements of the examiner. Spontaneous movements may be stereotypic and bizarre, or patients may exhibit excessive motor activity that appears without direction or purpose. Speech may be absent (mutism) or may repeat the examiner’s speech in a rigid and stereotypic way (echolalia). The patient may appear stuporous or may be internally preoccupied with other psychotic symptoms.

Positive symptoms may occur episodically or continuously. Even in episodic cases, however, it is rare for positive symptoms to remit completely. Most patients experience some degree of active psychosis whether or not they receive treatment. However, treatment has a significant impact on symptom severity and on the long-term course of illness.

Negative symptoms of schizophrenia are characterized by the absence of certain functions (see Table 1T1) (R8). Blunted affect refers to the loss of an individual’s ability to communicate via facial expression, tone of voice, eye contact, posture, gestures, respiratory pattern, and so forth. Normal communication depends heavily on these affective cues, and their diminution leads to significant impairment in interpersonal relations. Alogia is a reduction in language production, including both the number of words spoken and the number of ideas communicated by those words. Avolition refers to a loss of the inherent motivation that leads most individuals to sustained efforts at some form of productive activity, such as schoolwork, employment, care of a family, or recreational pursuits. Anhedonia is a loss of enjoyment of and satisfaction in productive and recreational activities, even when they are pursued. Asociality describes a decline in interest in social interactions, such as the development and maintenance of acquaintanceships, friendships, and family relations. Inattention to interpersonal cues, work expectations, and the social environment is common. Negative symptoms do not usually lead to hospitalization, but their presence is highly correlated with other areas of dysfunction, such as the inability to work, maintain family relations, or live independently (R9, R10).

Cognitive dysfunction has long been described in schizophrenia but has only recently become the focus of treatment interventions. The basic cognitive domains of memory, attention, executive function, and language (see Table 1T1) may all be affected (R11). Impairments in these areas are often noted in patients with schizophrenia long before the onset of psychotic symptoms (R12), and retrospective examination of IQ testing, school records, and formal neuropsychological testing typically shows difficulties in this area early in life. The mean IQ of individuals who go on to develop schizophrenia is 80—85 (R13), in contrast to the established norm of 100 for the population at large. Among the four symptom domains of schizophrenia, cognitive dysfunction has the highest correlation with functional impairment, and improvements in cognition are highly correlated with better quality of life (R14).

Affective symptoms include both mood and affective disturbance (see Table 1T1). The most common affective symptoms of schizophrenia are blunted, inappropriate, or bizarre affect and the inability to perceive affective signals in other people (R15). Inappropriate affect refers to a decoupling of the content and style of communication; for example, a person may describe with little emotion a tragic event or respond angrily to a trivial issue. Bizarre affect includes expressions or gestures that are not commonly observed in the general population in any context.

Disturbance of mood is also common in schizophrenia. Dysphoria in association with demoralization has been reported in about 60% of patients with schizophrenia (R16). Although lower mood has most often been described in the period immediately following an acute psychotic episode, dysphoric symptoms are common throughout such episodes but may not come to the attention of the clinician until after the more florid psychotic symptoms have subsided (R17). For many patients these symptoms are relieved with antipsychotic medication treatment alone (R18), particularly with the atypical agents (R19). Demoralization arising from an awareness of the impact of the illness on the patient’s life, in contrast, requires psychosocial interventions.

Diagnostic subtypes

The subtypes of the disorder include paranoid, disorganized, undifferentiated, residual, and catatonic schizophrenia. These subtypes have their historical precedents, unique clinical features, and characteristic courses, but it is not clear that they represent unitary pathological processes (R20). However, they have significant prognostic implications.

The paranoid subtype is characterized by prominent delusions and hallucinations and less prominent thought disorganization, affective disturbance, volitional impairment, and social dysfunction. The term "paranoid" is somewhat misleading, as the content of the delusions need not be persecutory but may run the gamut of religious, grandiose, somatic, jealous, and other delusions. The paranoid subtype generally has a more favorable prognosis than others, with a greater likelihood that some level of work, family interaction, and social life will be maintained (R20).

The disorganized subtype, in contrast, has the least favorable outcome (R21). This subtype is defined by prominent thought disorganization and gross disturbance of affect, neither of which is compatible with employment or social relations. Even independent living is less likely for patients in this group. This disorganized subtype is similar to the hebephrenic schizophrenia described in the older literature.

Catatonic schizophrenia is the least commonly encountered. Periods of catatonia are the defining feature of this subtype, although other positive and negative symptoms may be present in conjunction with and between catatonic episodes. It is important for clinicians to bear in mind that catatonic symptoms are more commonly associated with mood disorders than with schizophrenia (R22, R23).

Residual schizophrenia refers to an absence of active psychotic symptoms but continued impairments in function due to mild oddities of thought, affect, and social interaction. The episodic nature of psychotic symptoms makes it somewhat difficult to distinguish between this subtype and the others between acute episodes of illness.

Undifferentiated schizophrenia is the most common subtype (R24), with elements of hallucinations and delusions, disorganization, and affective disturbance all contributing to the patient’s symptoms. Outcomes for patients with this subtype are intermediate between the paranoid and disorganized subtypes. The fact that patients’ symptoms are more likely to cross subtype boundaries than to fall into a single subtype raises questions about the validity and the value of these categories.

The prevalence of schizophrenia has been estimated at 0.5%—1%, with most estimates approaching 1%. The annual incidence of new cases of schizophrenia is 11 per 100,000 (R25).

The onset of schizophrenia, when onset is defined as the first episode of active psychosis, most commonly occurs in young adulthood. However, psychotic symptoms typically begin 1—2 years before a diagnosis of schizophrenia is made (R26), but it is difficult to pinpoint the age at which the first prodromal symptoms developed. Numerous studies have identified premorbid cognitive (R12), affective (R16), and interpersonal problems (R27) from an early age in patients with schizophrenia. None of these symptoms is uniquely associated with schizophrenia, however, and a large percentage of children who have such difficulties early in life do not develop the disorder.

Negative symptoms also may develop well before the onset of active psychosis. Among patients studied who had prominent negative symptoms, the onset of these symptoms preceded that of positive symptoms by an average of 4.5 years and preceded diagnosis of the illness by 6.5 years (R28). DSM-IV-TR does not provide for the prospective diagnosis of patients identified between the onset of negative symptoms and that of positive symptoms; until the patient experiences overt psychosis, the diagnosis of schizophrenia cannot be made. This raises troubling issues for the diagnosis and treatment of patients in this prodromal phase of illness, and studies are under way to address this problem (R29).

The course of illness is highly variable, with 10%—20% of patients maintaining remission for five years after a first psychotic episode but most patients experiencing a pattern of repeated relapses (R30). Failure to continue taking antipsychotic medications is associated with a fivefold increase in the risk of relapse. Somewhat better outcomes are associated with later onset, more acute initial presentation, and concurrent mood symptoms (R31). Once the illness begins, the recurrence rate of active psychosis is high, and about 75% of untreated patients relapse within one year (R32). Between psychotic episodes, most patient continue to experience some positive symptoms, residual negative symptoms, and significant impairment in functioning. In general there is a worsening of negative and cognitive symptoms early in the course of illness as well as a trend toward less differentiation of symptom subtypes (R33, R34). A modest lessening of active pathology has been reported to occur later in life (R35). Despite the greater accessibility of treatment in industrialized nations, the long-term course of the illness appears to be less malignant in developing countries (R31).

For men, most cases of schizophrenia are diagnosed in young adulthood, although childhood and late-life cases also occur. For women, the onset of illness tends to be somewhat later, although usually before age 40. The illness course tends to be somewhat more malignant for men than for women, although individual variation is large (R36, R37).

Both antipsychotic medications and psychosocial treatments have been shown to prevent recurrence of psychotic symptoms, particularly when they are used in conjunction with one another (R38, R39). Noncompliance with treatment is highly correlated with relapse (R40), as are psychosocial stressors (R41, R42) and substance abuse (R43). Even in a perfectly compliant patient with good social supports, however, periodic recrudescence of psychosis is the rule. The reason for this pattern is unknown.

Acute episodes of illness are correlated with deterioration in function and worsening of negative and cognitive symptoms. Thus, prevention of relapse is not only desirable to avoid the distress of active psychotic symptoms and disruption of hospitalization but also to favorably affect the long-term course of illness (R45—R47).

The secondary consequences of schizophrenia are enormous. Patients with schizophrenia drift downward on all socioeconomic measures (R48). Unemployment is common, and underemployment is nearly universal among these patients (R49). Schizophrenia is overrepresented in the homeless population (R50). Although these patients are more often the victims than the perpetrators of violent crime, they are at a higher risk of assaultive acts than the general population (R51). Substance abuse is common, with prevalence estimates as high as 80% (R52). Not surprisingly, all quality of life measures are adversely affected by the disorder (R10). The lifetime risk of suicide among patients with schizophrenia is 10% (R53).

The economic impact of the disorder is also high, estimated at $15—$20 billion in annual direct treatment costs and an additional $15 billion annually in lost productivity (R54). This translates to $7,000—$8,000 per year per patient, irrespective of treatment setting, and additional costs to the family of $11,000—$12,000 per year (R55). Both short-term and long-term institutionalization are common. A high percentage of hospital beds are used for this population, and treatment of the disorder accounts for an estimated 2.5% of all health care expenditures (R56).

Schizophrenia very likely represents the final pathway of a complex interaction of a variety of genetic, developmental, and environmental factors (R57). The absence of a conclusive biological model for the development of schizophrenia may be attributed in part to the wide range of findings that must be accounted for in such a model. Genetic factors are clearly prominent, as indicated by the 50% concurrence rate for the illness among identical twins (R58). Second-trimester viral infection has been associated with an elevated risk of illness (R59). Perinatal anoxia, birth trauma, and toxic exposures have also been shown to predispose affected individuals to the development of the disorder (R60). A number of biological and psychological correlates of illness early in life have also been identified and must be accounted for in any neuropathologic model of the illness (R61). These considerations led to the development of the "two-hit" hypothesis, which suggests that both genetic predisposition and environmental or developmental factors must be present for the development of the disorder (R62). Despite the challenge of integrating these broad-ranging factors, an early delineation of a model linking them to the symptomatic manifestations of the disorder has been sketched.

Dopamine hypothesis

The involvement of dopaminergic pathways in the pathogenesis of psychosis was originally suggested by the correlation of the efficacy of antipsychotic medications with dopamine D₂ receptor blockade (R63). This correlation, is as strong with the atypical antipsychotics (R64) as with the conventional neuroleptics, where it was originally observed. There is also an increasing recognition of the importance of the unique anatomical distributions of the five dopamine receptor subtypes and of the preferential effects of atypical agents at these receptor subtypes (R65).

The dopamine receptor subtypes are grouped into two families on the basis of their similarity in structure and function to either dopamine D₁ or D₂ receptors. The D₁ family consists of the D₁ and D₅ receptor subtypes. These subtypes are widespread in the human CNS, are found only postsynaptically, and have not been directly implicated in the pathogenesis or treatment of psychotic symptoms. It has been noted, however, that they may play a modulatory role in pathways involving the D₂ receptor.

The D₂ family comprises dopamine D₂, D₃, and D₄ receptors, each of which has a unique anatomic distribution and binding affinity for antipsychotic drugs. D₂ and D₃ receptors are found both postsynaptically and as presynaptic autoreceptors; D₄ receptors are only found postsynaptically and are localized to mesolimbic and mesocortical structures. Attention was drawn to the D₄ receptor when researchers noted that clozapine, an antipsychotic with unique efficacy and no association with extrapyramidal side effects, has a higher affinity for D₄ receptors than for other subtypes (R66). Subsequent studies involving the therapeutic efficacy of selective dopamine D₄ antagonists and a neuropathologic role for D₄ receptors in the schizophrenia brain have been more equivocal (R67). The precise role of the D₄ receptor subtype in the pathogenesis and treatment of schizophrenia thus remains unclear.

Aberrations in dopaminergic pathways also differ by location. In the prefrontal cortex, hypoactivity of dopaminergic systems appears to be correlated with cognitive and negative symptoms of illness (R61). In contrast, active psychosis appears to be related to an excess of dopaminergic activity in mesocortical and mesolimbic pathways. Motor symptoms may arise in response to dopamine blockade of the nigrostriatal pathway. Finally, elevation of prolactin levels with antipsychotic medication treatment is the result of dopamine blockade of the tuberoinfundibular pathway (R65). The therapeutic need to increase dopamine in some brain regions, to decrease dopaminergic transmission in others, and to avoid alteration of this activity in still others poses unique problems in the pharmacotherapy of schizophrenia.

The atypical antipsychotics introduced in the 1990s began to address some of these problems. The appellation "atypical" refers to the lower propensity of these medications to cause parkinsonism, or extrapyramidal side effects, as well as long-term choreoathetotic movement disorders such as tardive dyskinesia. Compared with the conventional agents, atypical antipsychotics cause significantly fewer parkinsonian symptoms at standard doses. In addition, in contrast to the 5%—7% cumulative annual risk of tardive dyskinesia with conventional agents, atypical antipsychotics are associated with these symptoms in only about 0.5% of cases per year of medication exposure (R68, R69). In addition to the dramatic reduction in the risk of some side effects, modest reductions in the severity of negative and cognitive symptoms have been noted with atypical agents (R70, R71). The atypical antipsychotics are all effective antagonists of the dopamine D₂ receptor, and at clinically relevant doses their blockade of this receptor is comparable to that of conventional agents, but with the advantage of a greater degree of separation between clinically effective doses and doses required to produce extrapyramidal side effects (R71).

Two major hypotheses have been proposed to account for the favorable side effect profile—the "atypicality"—of the novel agents. The first is based on the observation that these agents all have greater activity at serotonin 5-HT₂ receptors than at dopamine D₂ sites. It has been proposed that blockade of the serotonin system has a modulatory effect on motor pathways (R72). A second, more recent, observation is that the atypical agents occupy the dopamine D₂ receptor only briefly, in contrast with the longer occupancy observed for conventional agents (R73). It has been suggested that this looser binding precludes these drugs from causing motor side effects. A comparison of these qualities among various conventional and atypical antipsychotics provides some support for this model (R64).

A second approach to the problem of simultaneous pharmacology of multiple dopaminergic systems is the use of partial dopamine agonists (R74). Unlike antagonists, which bind to receptors and prevent their activation by neurotransmitters, partial agonists bind and activate target receptors but elicit a lower level of receptor activation than the intrinsic neurotransmitter. The consequences of this partial activation depend on the condition of the receptor before initiation of treatment. When the receptor is hyperstimulated, competitive partial activation has the effect of reducing activity. When the untreated receptor is at a low level of stimulation, partial activation has the effect of increasing its activity.

The potential benefits of partial agonists in the treatment of schizophrenia are twofold (R75). First, by avoiding high levels of dopamine blockade in motor systems, the extrapyramidal side effects and tardive dyskinesia associated with conventional antipsychotic agents may be avoided. Second, partial agonists have the potential to simultaneously address the hypodopaminergic activity of the prefrontal cortex and the excessive dopamine of the mesolimbic and mesocortical pathways, thus treating both positive and negative symptoms. A number of clinical trials have been conducted with partial agonists (R74), and one agent, aripiprazole, has been found effective and approved for clinical use in the United States. However, the optimal role of partial agonists in schizophrenia treatment remains to be determined.

Another intriguing development in the dopamine hypothesis of schizophrenia involves catechol O-methyltransferase (COMT), an enzyme in the catabolic pathway of dopamine to its major metabolite, homovanillic acid. In animal models, greater COMT activity is correlated with more rapid metabolism of dopamine and a corresponding decrease in memory functions that are dependent on dopaminergic neurons. In humans, polymorphism of the COMT gene is associated with significant changes in prefrontal dopaminergic activity and predicts performance on some cognitive tests. Interestingly, the more active form of COMT, which has valine at position 158 of the membrane-bound enzyme (position 108 of the solubilized enzyme), in contrast to the less active variant with serine at that position, is correlated—although weakly—with schizophrenia in some family-based studies (R61, R76). This finding represents the first time a family linkage study has been correlated with a specific genetic defect that is plausibly connected to the pathogenesis of schizophrenia.

Structural hypothesis

The most consistently demonstrated biological finding in schizophrenia is enlargement of the lateral ventricles, often accompanied by dilatation of the cortex, but localization of a specific structural lesion has proved elusive. Neuroimaging and postmortem studies suggest that this reduction of structural material is due to widespread diminution of neuritic processes rather than the loss of cell bodies from a specific brain structure (R77). This finding provides support for a neurodevelopmental model for the illness, in which genetic and environmental factors interact to adversely affect the normal developmental processes of synaptic pruning, programmed cell death, experience-modulated neuronal proliferation, and axonal myelination (R47) that typically occur during adolescence and young adulthood. This model corresponds well with the development of active psychotic symptoms in early adulthood, but less well with studies of the progression of anatomic abnormalities. Although some studies have shown evidence of progression of ventricular enlargement with continued or recurrent illness, other studies suggest more static lesions (R47). This issue remains under active investigation.

There are no pathognomonic features or specific diagnostic tests for schizophrenia. The diagnosis is based on the presence of psychotic symptoms in the context of functional deterioration, after mood disorders and medical causes have been excluded. Hence, the diagnosis may be characterized by significant uncertainty early in the course of illness. Another obstacle to early diagnosis is resistance on the part of physicians, even when symptoms are obviously consistent with the disorder, possibly because of the stigma associated with schizophrenia.

Acute psychotic episodes arising in the context of schizophrenia, bipolar disorder, substance abuse, medical illness, and other disorders are remarkably similar, and they cannot be reliably distinguished from one another on the basis of their immediate presentation. The presence or absence of grandiose or persecutory delusions, visual hallucinations, agitation, or cognitive impairments is not uniquely related to any one cause of illness and should not be taken as evidence of a specific diagnosis.

Assessment of a patient begins with an appraisal of the presenting symptoms to determine whether they represent psychosis or some form of nonpsychotic reality distortion. Examples of nonpsychotic symptoms that might be mistaken for psychosis include dissociative episodes, flashbacks to traumatic events, extreme social phobia, body image distortions associated with eating disorders, obsessive ruminations, and cognitive disorganization associated with stress or anxiety. Projective psychological testing may sometimes aid the clinician in making these distinctions.

Medical causes of psychosis are sought through a medical history, a review of systems, a standard screening physical examination, and selected laboratory tests. Although no clear consensus exists on which tests are essential, a reasonable start includes a complete blood count, serum chemistry studies (electrolytes, liver enzymes, blood urea nitrogen, creatinine, calcium, etc.), a urine toxicology screen for substances of abuse, and urinalysis. Brain imaging may be useful for patients experiencing initial episodes of psychotic symptoms, but it rarely shows pathology in the absence of other neurological symptoms. Similarly, tests for rare but treatable conditions such as Wilson’s disease, porphyria, encephalitis, neurosyphilis, and so forth are best reserved for situations in which suggestive symptoms are present.

Substance-induced psychosis, related either to intoxication or to withdrawal from alcohol or depressant drugs, is commonly seen in emergency services settings. The diagnosis is made by history, toxicology screen results, and findings from the physical examination, with particular attention to vital signs and level of consciousness.

The chronological history of symptoms is a key element in distinguishing among the primary psychiatric disorders. Significant periods of intense and sustained mood disturbance, with or without psychotic symptoms, alert the clinician to the possibility of bipolar disorder or depression with psychotic features. Family history, although not conclusive, may also yield clues as to the presence or absence of a mood disorder. In contrast, a history of poor school and work performance, aberrant social function, and other operational impairments suggests a likely diagnosis of schizophrenia.

The diagnosis of schizoaffective disorder has been somewhat problematic over the years, given the lack of clarity as to diagnostic and prognostic features. The disorder is currently defined by periods of psychosis with clear mood features, interspersed with psychosis independent of affective symptoms. The prevalence, response to treatment, and expected outcome of the disorder remain ambiguous.

The two basic treatment domains, psychosocial and pharmacological, are essential elements of the treatment of schizophrenia. In general, patients are more likely to benefit from participation in community-based programs specializing in the treatment of chronically and severely mentally ill persons than from treatment with individual practitioners. Because of the severe global and wide-ranging psychopathology of the disorder, only specialty programs are likely to have the resources and expertise to address the range of treatment issues pertinent to this population. Elements of such programs include social support, vocational rehabilitation, social skills training, case management, family education and support, and intensive medication management.

Psychosocial treatments

A variety of psychosocial interventions for schizophrenia have been shown to be effective in decreasing relapse rates and improving function and quality of life. These treatments are based on the recognition of the critical role that social interactions and support play in most areas of function and in quality of life. Psychosocial treatments also assist the patient in maintaining essential services, including housing, income, and personal support. Perhaps most important, psychosocial interventions encourage and support compliance with medication treatment and maintenance of an ongoing relationship with a psychiatrist.

Social skills training focuses on the interpersonal deficits that patients with schizophrenia have. It addresses specifically the problems in communication and maintenance of social support systems that accompany the illness. This treatment begins with an assessment of the patient’s deficits in social performance in settings such as family interactions, contacts with roommates and other peers, and vocational contacts as the patient seeks or performs work. Through a progressive series of classroom and role-playing experiences, the patient is encouraged to strengthen social skills, first within the limited circle of the family and then within a broader circle that includes other supportive contacts, and finally to develop skills for use in novel situations such as a job interview. Patients have been shown to improve their skills in these areas with the training (R78), to generalize their improvements to novel settings, and to retain these improvements over time (R79). What has not been shown is improvement in the underlying deficits that the patients suffer. Rather, it appears that social skills training helps them compensate for these deficits.

Although there is no evidence to suggest that family interactions contribute to the development of schizophrenia, it is clear that some families respond more constructively than others to a family member with this disorder. Such variations are due in part to the diversity of relationships that exist between a patient and the family as well as the complexity of meanings associated with the illness for both the family and the patient. Few households are unaffected by the emergence of schizophrenic symptoms in a family member. Guilt, blame, anger, despair, and embarrassment are but a sample of the emotions experienced by all concerned. In addition, the family faces numerous difficult decisions about the patient’s care and treatment. Is the patient to remain at home? What level of care will the family provide? What degree of independence will be afforded the patient? Will treatment noncompliance be tolerated? It is unlikely that any two families will respond identically to these issues. It is equally unlikely that any family will be entirely satisfied with the outcome of its choices.

Some type of interaction between the family and mental health professionals is an essential element of schizophrenia treatment. Whether this interaction is in the form of a referral to a peer support group, psychoeducation, involvement in treatment planning, or formal family therapy—all of which we strongly support—attention to family issues is critical. Among the earliest studies of family interactions involving patients with schizophrenia were those that examined "expressed emotion" in the family system (R80). Expressed emotion refers to the number of critical and positive comments, hostility, warmth, and emotional overinvolvement of the family. Patients in families that had high levels of expressed emotion tended to have worse outcomes than those in other families. Also, when expressed emotion was decreased through family interventions, outcomes improved (R81). More recently developed family interventions have focused on education about the illness, basic communication skills, and practical suggestions for problem solving (R82).

Another recent development is the increasingly assertive involvement of families in patients’ care via advocacy groups and legal action. These groups have clearly enunciated families’ frustrations at being marginalized by care providers and have demanded greater involvement. As a result, through policy changes, and occasionally legislative action, families have won the right to be recognized as participants in care—to be informed and consulted, and sometimes even to participate in decision making with professional treatment teams.

Vocational rehabilitation seeks to address the marked deficiencies in constructive activity observed among patients with schizophrenia. The communication deficits, the cognitive impairments, the lack of intrinsic motivation, and the intrusive psychotic symptoms typical of schizophrenia converge to create major impairments in the ability to obtain and keep a job. To address these problems, a variety of training programs and sheltered work experiences have been suggested. Superior outcomes appear to be associated with more intensive programs, employment in a sheltered workshop, and continued involvement with the work program (R82).

For patients with greater disability, intensive outpatient programs such as assertive community treatment are appropriate (R83). In the assertive community treatment model of care, case managers, nurses, social workers, and physicians are intensely focused on a small group of patients in the community. High levels of contact are maintained between the various mental health professionals and the patients, often involving daily visits, and sometimes even more frequent visits. Specific areas of attention for the individual patient are identified, and interventions focus on those problems. A common service is "eyes-on medications," in which each dose of medication is administered to the patient by a member of the assertive community treatment team, thereby ensuring compliance and permitting frequent evaluation of the patient’s condition. Assertive community treatment has been shown to be highly effective in minimizing hospital readmission as long as the patient maintains involvement with the program.

Pharmacological treatments

Since the advent of chlorpromazine more than a half century ago, antipsychotic medications have been a central element of the treatment of schizophrenia. These medications have been shown to reduce or eliminate active psychotic symptoms in 60%—70% of patients, with corresponding improvements in psychotic agitation and mood disturbance. They are also highly effective in preventing psychotic relapse and associated hospital admissions. Thus, it is essential that patients with schizophrenia be maintained on an appropriate antipsychotic medication regimen as part of a comprehensive treatment plan.

Conventional antipsychotics ("neuroleptics") were the mainstay of antipsychotic treatment for more than 40 years. These medications had the advantages of proven short-term and long-term efficacy, multiple routes of administration, and the availability of depot formulations. Their primary disadvantages were short-term and long-term side effects, including extrapyramidal symptoms, tardive dyskinesia, sedation, sexual dysfunction, cardiac effects, elevation of prolactin levels, anticholinergic effects, and weight gain. It is little wonder that acceptance of these medications was limited among physicians and patients alike, and noncompliance was a major factor in symptom relapse. In recent years the Food and Drug Administration (FDA) has required that two of these medications, thioridazine and mesoridazine, include "black box" warnings regarding their potential for QT prolongation and potentially life-threatening cardiac arrhythmia, although reported cases of such adverse outcomes are rare.

A second generation of antipsychotics, the atypical agents, began in the late 1980s (Table 2T2). Clozapine, the first of the atypical antipsychotics, received FDA approval in 1990. Uniquely efficacious in the treatment of schizophrenia, this medication brings about a reduction in active psychotic symptoms in one-third to one-half of patients who do not respond to other agents (R84). Clozapine also has significant disadvantages, however. Common side effects include orthostatic hypotension, sedation, sialorrhea (excessive salivation), constipation, and weight gain. A small proportion of patients, 1%—2%, develop a potentially life-threatening agranulocytosis, which prompted the FDA to require weekly white blood cell counts during the first 6 months of treatment and every 2 weeks thereafter. With this intensive monitoring, the annual risk of death is reduced to 1:10,000. Not surprisingly, patient acceptance of clozapine is limited, despite its excellent efficacy. In addition, the frequent blood tests and associated administrative expenses keep the cost of clozapine treatment high, approaching $10,000 a year, even though the drug is available in generic form.

Risperidone, introduced in 1994, was the first "atypical" antipsychotic medication that was appropriate as first-line treatment of schizophrenia and other psychotic disorders (R85, R86). Risperidone is typically used at doses of 2—6 mg/day, with an average dose in the community of about 4 mg/day. With a 20-hour elimination half-time, the medication is appropriate for once-daily dosing. Few drug-drug interactions have been reported. The medication is available in tablets, liquid concentrate, and in a rapid-disintegrating formulation. Although the rapid-disintegrating tablet is not absorbed more rapidly than other forms, it offers the advantages of ease of administration and assurance of patient compliance. A depot formulation of the medication is being tested and may be available for clinical use in 2004. Common side effects of risperidone include mild sedation, mild hypotension, moderate akathisia, elevation of prolactin levels, and moderate weight gain. At doses above 6—8 mg/day, extrapyramidal side effects are more common than with other atypical antipsychotics. For this reason, higher doses of the medication should be avoided. Risperidone is relatively inexpensive compared with other atypical agents, costing about $2,800 per year at an average dose of 4 mg/day (R45).

Olanzapine (R87, R88) was introduced in 1996, shortly after risperidone. Olanzapine is typically administered at doses of 15—20 mg/day but is routinely and safely given at doses up to 40 mg/day, well above the FDA-approved maximum of 20 mg. Given the drug’s 30-hour elimination half-time, blood levels remain steady, and once-daily dosing can be prescribed. Although drug-drug interactions involving olanzapine are rare, smoking is known to decrease its blood levels by as much as 30%. This can be a significant issue for patients who are stabilized on the medication in a smoke-free hospital environment and then resume smoking at home. The medication is available in coated tablets and a rapid-disintegrating formulation. The tablet coating protects the medication from oxidation; if the tablet is cut, it must be taken immediately. The rapid-disintegrating formulation is easily administered but is absorbed with the same 5-hour time to peak concentration as the coated tablet. Weight gain is a frequent and significant side effect of olanzapine and has the associated risks of hyperglycemia and diabetes. Other common side effects of olanzapine are moderate sedation, mild akathisia, mild hypotension, dry mouth, and constipation. Extrapyramidal side effects have been reported at higher doses, although they are rare and tend to be mild. Olanzapine is relatively expensive, at an average cost of $5,500 a year for 15—20 mg/day (R45).

Quetiapine (R89, R90), approved in 1997, is most effective at doses of 400—800 mg/day. The drug’s elimination half-time is relatively short at 6—7 hours, and its dopamine D₂ receptor occupancy time is also brief. Accordingly, quetiapine is approved for twice-daily dosing, although anecdotal accounts of stable patients being successfully maintained on once-daily regimens abound. The significance of quetiapine’s brief D₂ receptor occupancy is not clear, but it has been suggested as an explanation for the exceptionally low risk of extrapyramidal side effects with this medication (R73), a quality it has in common with clozapine. Quetiapine is available in tablets. This agent is somewhat more susceptible to drug-drug interactions than other antipsychotics, particularly inducers and inhibitors of the cytochrome P450 system, such as carbamazepine and fluvoxamine. Sedation and hypotension in the early stages of treatment are the most troublesome side effects of quetiapine. Both problems improve significantly with continued treatment (R91). Other side effects include mild akathisia, mild dry mouth, and mild weight gain. The primary advantage of this drug is its low risk of extrapyramidal side effects. The cost for treatment is in the intermediate range at $3,400 per year for 400 mg/day (R45).

Ziprasidone is most effectively used at doses of 120—160 mg/day (R92). The drug’s 7-hour elimination half-time is consistent with its recommended twice-daily dosing. Absorption of the medication is strongly affected by food, and oral doses should be administered with meals. Ziprasidone is available in tablets as well as in sterile solution for intramuscular injection (R93)—the only atypical agent available by this route. The use of injectable doses of 10—20 mg, up to a total of 40 mg/day, has been approved for the treatment of acute psychotic agitation. Blood levels of ziprasidone are modestly lower when carbamazepine is administered concurrently. Common side effects include mild sedation early in treatment, mild nausea, weakness, nasal congestion, and mild QTc prolongation; some patients find the medication mildly stimulating. Some anecdotal evidence suggests that side effects may be greater during transitions from this medication to others; it may be advantageous to stay with treatment rather than switch prematurely. Ziprasidone appears to cause only minimal weight gain, in contrast with other agents. The clinical risk associated with the QTc prolongation appears to be low. Clinical trials have shown no evidence of an increased risk of arrhythmias or sudden death (R94), and the FDA has required only an intermediate level of warning on the package insert. The drug should be used with caution, however, in the presence of other agents that prolong the QT interval. Ziprasidone is priced slightly lower than other atypical agents, at an average cost of $2,700 per year (R45).

The most recent development in antipsychotic therapy is the introduction of the dopamine partial agonist aripiprazole. The medication was studied at doses up to 30 mg/day (R75, R95) and is recommended for use at 10—15 mg in once-daily doses for most patients. The manufacturer estimates that 80% of patients who respond to the medication will do so in the middle dose range. The drug’s elimination half-time is 75 hours, which provides good stability between doses but makes dose adjustments rather slow. Aripiprazole is metabolized through the cytochrome P450 system, and serum levels of the drug are affected by inducers and inhibitors of those pathways. Common side effects include headache, nausea and vomiting, insomnia, and, in a few patients, weight gain in excess of 7% of base weight, although the drug is weight neutral in most patients. At higher doses, some patients also experience somnolence. Aripiprazole shows no evidence of extrapyramidal side effects at any dose tested. As with ziprasidone, some side effects may be more prominent during transitions to other antipsychotic medications. Preliminary estimates suggest that the drug will cost $2,700—$3,000 a year, making it competitive with other atypical antipsychotics.

The availability of several first-line atypical antipsychotics, including the partial agonist aripiprazole, raises the issue of how best to select medication treatment, for both patients with new-onset illness and patients with recurring illness. Several general principles can serve as useful guidelines in this regard. First, the first-line atypical antipsychotic medications appear to be equally efficacious in the treatment of psychotic disorders (R96). It is likely that aripiprazole has similar efficacy. Second, the first-line medications have been found in large controlled studies to be equally well tolerated. Third, although the number of patients who drop out of treatment because of side effects is the same for each medication (R96), each one has a unique set of side effects (Table 3T3). Fourth, there is some evidence that these agents have differential effects among patients. That is, although the same percentage of patients respond to each drug, an individual patient may respond preferentially to one rather than another. Finally, clozapine has clearly documented superior efficacy compared with either conventional or other atypical agents, and it should be considered for any patient who has not responded to other medications.

The use of adjunctive medications in the treatment of schizophrenia is common but not well studied in controlled clinical trials. The use of adjunctive medications is often justified on the basis of a patient’s experiencing some level of psychotic symptoms even with optimal doses of an antipsychotic medication. In some instances, however, adjunctive medications clearly have a place in schizophrenia treatment. Benzodiazepines are effective for short-term use for rapid control of psychotic agitation, especially when used in conjunction with antipsychotics (R97). They are also well tolerated in long-term use for persistent agitation, anxiety, or akathisia, and they have modest benefits in some patients for control of psychosis (R98). Antidepressants should be reserved for cases of severe depression and anhedonia that have not responded to antipsychotic monotherapy. Patients with persistent demoralization and patients with negative symptoms such as avolition and social isolation, although presenting symptoms similar to the vegetative signs of depression, do not respond to antidepressant medications, and these symptoms are not appropriate targets of treatment. Mood stabilizers provide only transient benefit in the treatment of psychosis (R99) and should be reserved for cases of persistent and troublesome mood instability in patients who have been treated for an adequate duration with an appropriate dose of antipsychotics.

The simultaneous use of more than one antipsychotic agent is common in clinical practice, despite a dearth of studies to demonstrate the efficacy or safety of such combinations and a consistent bias against them in expert consensus guidelines (R100—R102). There is also evidence that untested antipsychotic combinations are sometimes used in preference to clozapine with patients who have treatment-refractory illness, even though clozapine is known to be effective in many such cases (R103). The pharmacological justification for the practice is weak, since antipsychotic medications all appear to work by the same mechanism—that is, blockade of the dopamine D₂ receptor. Differences in activities at other receptors and in receptor occupancy times, however, could serve as a rational basis for antipsychotic polypharmacy. Clozapine is sometimes combined with another agent in nonresponsive patients. Depot and oral medications are occasionally combined (see below). At present the appropriate role of antipsychotic polypharmacy in the treatment of schizophrenia is unresolved.

Depot antipsychotic agents also have a role in the treatment of schizophrenia. Although the conventional antipsychotics haloperidol and fluphenazine decanoate are the only options currently available, they should not be overlooked for responsive but noncompliant patients who remain at risk of symptom relapse. Depot medications offer the advantages of guaranteed compliance and steady serum drug levels. The absence of daily peak serum levels with these medications may also be associated with lower total side effects than equivalent doses of the oral formulation (R104). Depot antipsychotics may occasionally be used at low doses in combination with atypical agents to provide the patient with a measure of protection against relapse in the event of noncompliance with the oral medication. The major disadvantages of depot medications are poor patient acceptance and lengthy times for dose adjustments. Although atypical agents are not available in depot formulations in the United States, a long-acting formulation of risperidone has been developed and is currently available outside the United States. Depot risperidone has been shown to be effective and well tolerated in clinical trials (R105), and it now awaits FDA review.

Cost-effectiveness

Given the costs associated with schizophrenia and its treatment, it is not surprising that numerous attempts have been made to study and improve the cost-effectiveness of therapy. Despite the purported greater acceptability to patients of atypical antipsychotics, their effect on the economics of schizophrenia treatment has generally been found to be neutral because of their significantly higher acquisition cost (R106). It is more difficult to quantify the relationship between costs and quality of life measures (R107).

One widely used approach to cost containment in the treatment of severe and persistent mental disorders is the use of treatment guidelines and algorithms. Based on a combination of empirical studies and expert consensus, the more general guidelines and more specific algorithms attempt to define a rational approach to treatment involving both clinical and economic issues. Among the best known of these are the American Psychiatric Association’s practice guidelines and the Texas Medication Algorithm Project (TMAP). The APA practice guidelines focus on the most rigorously established empirical findings, and therefore tend to be general in their recommendations for treatment. Thus, the APA Practice Guideline for the Treatment of Patients With Schizophrenia specifies only that a combination of pharmacological and psychosocial interventions should be used (R108). TMAP, in contrast, freely incorporates expert consensus, but attempts to maintain a high level of flexibility for the physician. For treatment of schizophrenia, TMAP recommends atypical antipsychotic monotherapy for at least 3 weeks at doses up to the manufacturers’ recommended maximums. For nonresponding patients, it encourages trials of at least three first-line atypical antipsychotics and one conventional antipsychotic, followed by a trial of clozapine (R109). Additional algorithms for the handling of side effects and the use of adjunctive agents are also given. TMAP was designed not only as a treatment aid for clinicians but also to study the effectiveness of treatment algorithms in improving care and managing costs (R110). Each of these approaches has certain strengths and limitations, but it is not clear that they have had the desired impact on quality of care and cost-effectiveness (R111).

Despite more than a century of intensive research on the neurobiology of schizophrenia, many aspects of the disorder remain elusive. Nevertheless, the past few years have seen remarkable progress in this area, and it can reasonably be expected that progress will continue at a brisk pace. A number of other research questions are also currently under investigation, such as whether it is possible to predict and prevent the development of schizophrenia in predisposed individuals, whether there are biological markers for response to specific medications, and whether additional genetic family linkages can be identified. Laboratory studies provide ever more detailed maps of neurotransmitter and receptor localization. Further investigations into receptor subtypes may yet yield important insights into the function of antipsychotic medications.

Meanwhile, treatment continues in the real world. The trend of deinstitutionalization and community care places increasing numbers of patients in community mental health and similar settings. Such systems are currently at the limits of their capacity, and it is not clear that they are equal to the task of providing state-of-the-art care for each individual in their charge as the numbers of patients increase and funding becomes more scarce. The dilemma of advancing research and improved treatments offset by financially strained systems of care, the irony of American medicine, is particularly acute in psychiatry. The current challenge is to translate recent advances in research into available treatment improvements for patients.