Guideline Statements and Implementation

The importance of the psychiatric evaluation cannot be underestimated because it serves as the initial basis for a therapeutic relationship with the patient and provides information that is crucial to differential diagnosis, shared decision-making about treatment, and educating patients and family members about such factors as illness course and prognosis. APA’s Practice Guidelines for the Psychiatric Evaluation of Adults, 3rd edition (American Psychiatric Association 2016a) describe recommended and suggested elements of assessment for any individual who presents with psychiatric symptoms (Table 1). These elements are by no means comprehensive, and additional areas of inquiry will become apparent as the evaluation unfolds, depending on the responses to initial questions, the presenting concerns, the observations of the clinician during the assessment, the complexity and urgency of clinical decision-making, and other aspects of the clinical context. In many circumstances, aspects of the evaluation will extend across multiple visits (American Psychiatric Association 2016a).

The specific approach to the interview will depend on many factors, including the patient’s ability to communicate, degree of cooperation, level of insight, illness severity, and ability to recall historical details (American Psychiatric Association 2016a). Such factors as the patient’s health literacy (Clausen et al. 2016) and cultural background (Lewis-Fernández et al. 2016) can also influence the patient’s understanding or interpretation of questions. Typically, a psychiatric evaluation involves a direct interview between the patient and the clinician (American Psychiatric Association 2016a). The use of open-ended empathic questions about the patient’s current life circumstances and reasons for evaluation can provide an initial picture of the individual and serve as a way of establishing rapport. Such questions can be followed up with additional structured inquiry about history, symptoms, or observations made during the assessment.

Throughout the assessment process, it is important to gain an understanding of the patient’s goals, their view of the illness, and preferences for treatment. This information will serve as a starting point for person-centered care and shared decision-making with the patient, family, and other persons of support (Dixon et al. 2016; Hamann and Heres 2019). It will also provide a framework for recovery, which has been defined as “a process of change through which individuals improve their health and wellness, live self-directed lives, and strive to reach their full potential” (Substance Abuse and Mental Health Services Administration 2012a, p. 3). Consequently, discussions of goals should be focused beyond symptom relief and may include goals related to schooling, employment, living situation, relationships, leisure activities, and other aspects of functioning and quality of life. Questions about the patient’s views may help determine whether the patient is aware of having an illness and whether the patient has other explanations for symptoms that are helpful to them (Saks 2009). Patients may have specific views about such topics as medications, other treatment approaches, mechanical restraints, or involuntary treatment based on prior treatment experiences. They may also be able to delineate strategies that have been helpful for them in coping with or managing their symptoms in the past (Cohen et al. 2017). Some patients will have completed a psychiatric advance directive (Murray and Wortzel 2019), which is important to review with the patient if it exists.

In addition to direct interview, patients may be asked to complete electronic or paper-based forms that ask about psychiatric symptoms or key aspects of the history (American Psychiatric Association 2016a). When available, prior medical records, electronic prescription databases, and input from other treating clinicians can add further details to the history or corroborate information obtained in the interview (American Psychiatric Association 2016a).

Family members, friends, and other individuals involved in the patient’s support network can be an important part of the patient’s care team and valuable sources of collateral information about the reason for evaluation, the patient’s past history, and current symptoms and behavior (American Psychiatric Association 2016a). Outreach to family, friends, and others in the support network will typically occur with the patient’s permission. In situations in which the patient is given the opportunity and does not object, necessary information can be shared with family members or other persons involved in the patient’s care or payment for care (Office for Civil Rights 2017b). For example, if a relative or person of support is present with the patient at an appointment, the clinician may discuss information about medications or give education about warning signs of a developing emergency.

In some instances, however, patients may ask that family or others not be contacted. When this is the case, patients can usually identify someone whom they trust to provide additional information, and they are often willing to reconsider contact as treatment proceeds. It is also useful to discuss the reasons that the patient has concerns about contacts with family members or other important people in the patient’s life. For example, a patient may wish to avoid burdening a loved one, may have felt unsupported by a particular family member in the past, or may be experiencing delusional beliefs that involve a family member or friend. The patient may also want to limit the information that clinicians receive about past or recent treatment, symptoms, or behaviors. Even when a patient does not want a specific person to be contacted, the clinician may listen to information provided by that individual, as long as confidential information is not provided to the informant (American Psychiatric Association 2016a). Also, to prevent or lessen a serious and imminent threat to the health or safety of the patient or others, The Principles of Medical Ethics (American Psychiatric Association 2013f) and the Health Insurance Portability and Accountability Act of 1996 (HIPAA; Office for Civil Rights 2017b) permit clinicians to disclose necessary information about a patient to family members, caregivers, law enforcement, or other persons involved with the patient as well as to jails, prisons, and law enforcement officials having lawful custody of the patient. HIPAA also permits health care providers to disclose necessary information to the patient’s family, friends, or other persons involved in the patient’s care or payment for care when such disclosure is judged to be in the best interests of the patient and the patient either is not present or is unable to agree or object to a disclosure because of incapacity or emergency circumstances. Examples of such circumstances are not limited to unconsciousness and may also include such circumstances as temporary psychosis or intoxication with alcohol or other substances (Office for Civil Rights 2017b).

Although it is beyond the scope of this guideline to discuss the differential diagnosis and evaluation of psychotic disorders, many features and aspects of clinical course will enter into such a determination in addition to psychotic symptoms per se. Clinicians should also be mindful that biases can influence assessment and diagnosis, with disparities in diagnosis based on race being particularly common (Olbert et al. 2018; Schwartz and Blankenship 2014). The clinician should be alert to features of the history, including family, developmental, and academic history, that may suggest specific conditions or a need for additional physical or laboratory evaluation. Examples of conditions that can mimic schizophrenia in their initial presentation include neurosyphilis, Huntington’s disease, Wilson’s disease, and anti-N-methyl-d-aspartate (NMDA) receptor encephalitis (Lieberman and First 2018). Individuals with 22q11.2 deletion syndrome have a substantially increased risk of developing schizophrenia (Bassett et al. 2017; McDonald-McGinn et al. 2015; Van et al. 2017). In addition, the presence of a 22q11.2 deletion is associated with an increased likelihood of neurocognitive and physical health impairments (McDonald-McGinn et al. 2015; Moberg et al. 2018; Swillen and McDonald-McGinn 2015), which has implications for treatment (Fung et al. 2015; Mosheva et al. 2019). Psychotic symptoms can also occur in the context of other neurological and systemic illnesses, with or without delirium, and such acute states can at times be mistaken for an acute exacerbation of schizophrenia. Furthermore, because a significant fraction of individuals with psychosis will have a shift in diagnosis over time, the diagnosis may need to be reevaluated as new information about the patient’s illness course and symptoms becomes available (Bromet et al. 2011). Specialty consultation can be helpful in establishing and clarifying diagnosis (Coulter et al. 2019), particularly if the illness symptoms or course appear to be atypical or if the patient is not responding to treatment.

A thorough history is also important for identifying the presence of co-occurring psychiatric conditions or physical disorders that need to be addressed in treatment planning (American Psychiatric Association 2016a; Firth et al. 2019). For example, individuals with serious mental illness have higher rates of smoking, higher rates of heavy smoking, and lower rates of smoking cessation than do community samples (Cook et al. 2014; de Leon and Diaz 2005; Myles et al. 2012; Wium-Andersen et al. 2015). Furthermore, the use of cannabis may be more frequent in individuals with schizophrenia (Koskinen et al. 2010) and associated with greater symptom severity or earlier onset of psychosis (Carney et al. 2017; Large et al. 2011). Other substance use disorders, if present, can also produce or exacerbate symptoms of psychosis (American Psychiatric Association 2016a; Large et al. 2014). Thus, as part of the initial evaluation, it is important to determine whether the patient uses tobacco, cannabis, or other substances such as alcohol, caffeine, cocaine, opioids, sedative-hypnotic agents, stimulants, 3,4-methylenedioxymethamphetamine (MDMA), solvents, androgenic steroids, hallucinogens, or synthetic substances (e.g., “bath salts,” K2, Spice). The route by which substances are used (e.g., ingestion, smoking, vaping, intranasal, intravenous) is similarly important to document.

Mortality is increased in individuals with schizophrenia (Brown et al. 2000; Fazel et al. 2014; Olfson et al. 2015), and the average life span is shortened by a decade or more, with much of this decrease related to increased rates of co-occurring physical conditions (Laursen et al. 2013; Saha et al. 2007; Walker et al. 2015). Adverse health effects of smoking also contribute to an increased risk of mortality among individuals with schizophrenia (Lariscy et al. 2018; Reynolds et al. 2018; Tam et al. 2016). Many other conditions are more frequent in individuals with serious mental illness in general (Janssen et al. 2015; McGinty et al. 2016) and schizophrenia in particular (Henderson et al. 2015), including, but not limited to, poor oral health (Kisely et al. 2015), hepatitis C infection (Chasser et al. 2017; Hauser and Kern 2015; Hughes et al. 2016), HIV infection (Hobkirk et al. 2015; Hughes et al. 2016), cancer (Olfson et al. 2015), sleep apnea (Myles et al. 2016; Stubbs et al. 2016b), obesity (Janssen et al. 2015), diabetes mellitus (Vancampfort et al. 2016a), metabolic syndrome (Vancampfort et al. 2015), and cardiovascular disease (Correll et al. 2017c). These disorders, if present, can contribute to mortality or reduced quality of life, and some may be induced or exacerbated by psychiatric medications. Laboratory tests and physical examination as part of the initial evaluation can help to identify common co-occurring conditions and can serve as a baseline for subsequent monitoring during treatment (Table 2).

As part of the initial evaluation, it is also useful to inquire about the course and duration of symptoms prior to treatment (i.e., duration of untreated psychosis) (Penttilä et al. 2014; Register-Brown and Hong 2014; Santesteban-Echarri et al. 2017) and whether the patient has received any mental health treatment. If the patient has received treatment previously, it is important to ask about a broad range of treatments and other approaches to addressing the patient’s symptoms and functioning and to specifically ask about the full range of treatment settings (e.g., outpatient, partial hospitalization, inpatient) and approaches that the patient has found helpful or problematic (American Psychiatric Association 2016a). Although most patients will comment on prior medications, psychotherapy, or psychiatric hospitalizations if asked about treatment history, specific questions may be needed to gather details of such treatments. Prompting may be needed to learn information about the patient’s experiences with other interventions such as psychosocial rehabilitation, supported employment, assertive community treatment (ACT), court-ordered treatment, treatment while incarcerated, substance use treatments, neuromodulatory therapies (e.g., electroconvulsive therapy [ECT], transcranial magnetic stimulation [TMS]), 12-step programs, self-help groups, spiritual healers, and complementary or alternative treatment approaches. Pharmacy databases and patients’ lists of active medications are not likely to include long-acting injectable (LAI) medications (e.g., antipsychotics, naltrexone, buprenorphine) or implants (e.g., buprenorphine, contraceptive agents), over-the-counter medications, herbal products, or nutritional supplements. For each specific type of intervention that the patient has received, it is helpful to learn more about the duration, mode of delivery (e.g., formulation, route, and dose for medications; format, type, and frequency of treatment for psychotherapy), response (including tolerability, changes in quality of life, level of functioning, and symptom response/remission), and degree of adherence.

The psychosocial history reviews the stages of the patient’s life and may include attention to perinatal events, delays in developmental milestones, academic history and performance (including learning difficulties, special education interventions, or disciplinary actions), relationship and sexual history, interpersonal functioning (including in social and family roles, such as parenting), occupational history (including military history), legal history, and identification of major life events (e.g., parental loss, divorce, traumatic experiences, migration history) and psychosocial stressors (e.g., financial, housing, legal, school/occupational, or interpersonal/relationship problems; lack of social support; painful, disfiguring, or terminal medical illness) (American Psychiatric Association 2016a; Barnhill 2014; MacKinnon et al. 2016; Smith et al. 2019). Information about the patient’s family constellation and persons who provide support will serve as a foundation for working collaboratively with the patient and their support network. The cultural history also emphasizes relationships, both familial and nonfamilial, and the role of important cultural and religious influences on the patient’s life (Aggarwal and Lewis-Fernández 2015; American Psychiatric Association 2013a; Lewis-Fernández et al. 2016).

The mental status examination is an integral part of the initial assessment. A full delineation of the mental status examination is beyond the scope of this document, and detailed information on conducting the examination is available elsewhere (American Psychiatric Association 2016a; Barnhill 2014; MacKinnon et al. 2016; Smith et al. 2019; Strub and Black 2000). However, for individuals with possible schizophrenia, a detailed inquiry into hallucinations and delusions will often identify psychotic experiences in addition to the presenting concerns. Negative symptoms and cognitive impairment are common and influence outcomes (Bowie et al. 2006; Green 2016; Jordan et al. 2014; Rabinowitz et al. 2012; Santesteban-Echarri et al. 2017) but may go undetected without specific attention during the evaluation. Negative symptoms also can be difficult to differentiate from lack of interest or reduced motivation due to depression, medication side effects, substance use, or neurological conditions.

Insight is also impaired in a significant proportion of individuals with schizophrenia (Mohamed et al. 2009) and can manifest as a decreased awareness of having a disorder, symptoms, consequences of illness, or a need for treatment (Mintz et al. 2003). Consequently, inquiring about the patient’s degree of insight and judgment will provide information relevant to risk assessment, treatment outcomes, and adherence (Mintz et al. 2003; Mohamed et al. 2009).

Risk assessment is another essential part of the initial psychiatric evaluation (American Psychiatric Association 2004). It requires synthesizing information gathered in the history and mental status examination and identifying modifiable risk factors for suicidal or aggressive behaviors that can serve as targets of intervention in constructing a plan of treatment. Suicidal ideas are common in individuals who have had a psychotic experience (Bromet et al. 2017). Death due to suicide has been estimated to occur in about 4%–10% of individuals with schizophrenia (Drake et al. 1985; Heilä et al. 2005; Hor and Taylor 2010; Inskip et al. 1998; Laursen et al. 2014; Nordentoft et al. 2011; Palmer et al. 2005; Popovic et al. 2014; Tanskanen et al. 2018; Yates et al. 2019), yielding a greater than tenfold increase in standardized mortality ratios (Saha et al. 2007). Among individuals with schizophrenia, suicide attempts and suicide may be more common early in the course of the illness (Popovic et al. 2014) and can occur even before initial treatment for psychosis (Challis et al. 2013).

In individuals with schizophrenia, many of the risk factors that contribute to the risks of suicidal or aggressive behaviors are the same as factors increasing risk in other disorders. For example, in individuals with schizophrenia, an increased risk of suicidal or aggressive behaviors has been associated with male sex, expressed suicidal ideation, a history of attempted suicide or other suicide-related behaviors, and the presence of alcohol use disorder or other substance use disorder (Cassidy et al. 2018; Challis et al. 2013; Fazel et al. 2009a, 2014; Fleischhacker et al. 2014; Hawton et al. 2005; Hor and Taylor 2010; Østergaard et al. 2017; Pompili et al. 2007; Popovic et al. 2014; Roché et al. 2018; Sariaslan et al. 2016; Singh et al. 2012; Swanson et al. 2006; Witt et al. 2013, 2014). Firearm access is an additional contributor to suicide risk (Alban et al. 2018; Anestis and Houtsma 2018; Siegel and Rothman 2016). Additional factors that have been identified as increasing risk for suicide among individuals with schizophrenia include depressive symptoms, hopelessness, agitation or motor restlessness, fear of mental disintegration, recent loss, recency of diagnosis or hospitalization, repeated hospitalizations, high intelligence, young age, and poor adherence to treatment (Cassidy et al. 2018; Fleischhacker et al. 2014; Hawton et al. 2005; Lopez-Morinigo et al. 2014; Pompili et al. 2007; Popovic et al. 2014; Randall et al. 2014). It is not clear whether preserved insight is associated with an increase in suicide risk among individuals with schizophrenia (Hor and Taylor 2010) or whether this is an apparent increase that is mediated by other factors such as hopelessness (López-Moríñigo et al. 2012).

Although reduced risk of suicide was associated with hallucinations in one meta-analysis (Hawton et al. 2005), the presence of auditory command hallucinations may confer increased risk (Harkavy-Friedman et al. 2003; Wong et al. 2013). Command hallucinations can also be relevant when assessing individuals for a risk of aggressive behaviors (McNiel et al. 2000; Swanson et al. 2006), although the relationship between experiencing commands and acting on them is complex (Braham et al. 2004). Persecutory delusions may also contribute to risk of aggression, particularly in the absence of treatment or in association with significant anger (Coid et al. 2013; Keers et al. 2014; Swanson et al. 2006).

Among individuals with psychotic illnesses, prior suicidal threats, angry affect, impulsivity, hostility, recent violent victimization, childhood sexual abuse, medication nonadherence, and a history of involuntary treatment were also associated with an increased risk of aggressive behavior (Buchanan et al. 2019; Large and Nielssen 2011; Reagu et al. 2013; Swanson et al. 2006; Witt et al. 2013, 2014). Other factors associated with a risk of aggression are similar to findings in individuals without psychosis and include male sex, young age, access to firearms, the presence of substance use, traumatic brain injury, a history of attempted suicide or other suicide-related behaviors, and prior aggressive behavior, including that associated with legal consequences (Buchanan et al. 2019; Cassidy et al. 2018; Fazel et al. 2009a, 2009b, 2014; Fleischhacker et al. 2014; Large and Nielssen 2011; Monuteaux et al. 2015; Østergaard et al. 2017; Popovic et al. 2014; Roché et al. 2018; Sariaslan et al. 2016; Short et al. 2013; Singh et al. 2012; Swanson et al. 2006; Witt et al. 2013, 2015).

Relevant guidelines from the following organizations were reviewed: British Association for Psychopharmacology (BAP), Canadian Psychiatric Association and Schizophrenia Society of Canada, National Institute for Health and Care Excellence (NICE), Royal Australian and New Zealand College of Psychiatry (RANZCP), Scottish Intercollegiate Guidelines Network (SIGN), World Federation of Societies of Biological Psychiatry (WFSBP), and Schizophrenia Patient Outcomes Research Team (PORT). Information from them is generally consistent with this guideline statement. Other guidelines on the treatment of schizophrenia incorporate recommendations related to the need for a comprehensive initial assessment (Addington et al. 2017a; National Institute for Health and Care Excellence 2014), including identification of prior and current psychiatric symptoms and diagnoses (Addington et al. 2017a; Hasan et al. 2015; National Institute for Health and Care Excellence 2014), assessment of tobacco use (Addington et al. 2017a; National Institute for Health and Care Excellence 2014), assessment of substance use (Addington et al. 2017a; Barnes et al. 2011; Crockford and Addington 2017; Galletly et al. 2016; National Institute for Health and Care Excellence 2014), physical health history and examination (Addington et al. 2017a; National Institute for Health and Care Excellence 2014), assessment of psychosocial factors (Addington et al. 2017a; Galletly et al. 2016; National Institute for Health and Care Excellence 2014), and mental status examination (Addington et al. 2017a), including assessment of the risk of harm to self or others (Addington et al. 2017a; Hasan et al. 2015; National Institute for Health and Care Excellence 2014). Several other guidelines also provide information on the circumstances in which an electrocardiogram is suggested (Barnes et al. 2011; National Institute for Health and Care Excellence 2014; Pringsheim et al. 2017).

For patients with psychotic disorders, including schizophrenia, several components of the initial psychiatric evaluation have potential relevance for quality measure development, although such quality measures do not exist at present. A first step toward development of scientifically sound quality measures is identification of discrete indicators that signal the delivery of high-quality care. This step may be challenging to accomplish given the breadth of content within the initial psychiatric assessment and the difficulty in ascertaining evaluation details from chart or administrative data. However, it may still be possible to use available evidence and expert-recommended consensus to develop and specify electronic and clinical data registry quality measures. Additionally, as discussed in the APA Practice Guidelines for the Psychiatric Evaluation of Adults, 3rd edition (American Psychiatric Association 2016a), quality improvement efforts at the local level could assess whether specific aspects of the evaluation such as a risk assessment were completed while still allowing flexibility in the documentation of findings.

APA’s Practice Guidelines for the Psychiatric Evaluation of Adults, 3rd edition (American Psychiatric Association 2016a) provides a general description of the use of quantitative measures as part of the initial psychiatric evaluation. In the assessment of a patient with a possible psychotic disorder, quantitative measures can also be used to help detect and determine the severity of psychosis and associated symptoms. The intent of using a quantitative measure is not to establish a diagnosis but rather to complement other aspects of the screening and assessment process. Depending on the measure, it can aid in treatment planning by providing a structured replicable way to document the patient’s baseline symptoms. It can also help to determine which symptoms should be the target of intervention on the basis of such factors as frequency of occurrence, magnitude, potential for associated harm to the patient or others, and associated distress to the patient.

As treatment proceeds, use of quantitative measures allows more precise tracking of whether nonpharmacological and pharmacological treatments are having their intended effect or whether a shift in the treatment plan is needed. This record of a patient’s response to treatment is of particular value when the treatment is nonstandard (e.g., combination of antipsychotics) or expensive. It can also provide helpful information about the actual effects of prior treatments. In addition, patients’ ratings can be compared with family members’ impressions of treatment effects to clarify the longitudinal course of the patient’s illness.

Much of the treatment-related research in psychiatry has used clinician-rated scales to determine patient outcomes; however, patient-rated scales are typically less time-consuming to administer than clinician-rated scales. In addition, they provide important insights into the patient’s experience that support person-centered care. The use of anchored, self-rated scales with criteria to assess the severity and frequency of symptoms can also help patients become more informed self-observers. However, correlations between patient- and clinician-rated scales are often modest (Harvey 2011; Spitz et al. 2017), suggesting that both types of quantitative measures provide useful information. If a mismatch is noted in the self-assessment of patients as compared with assessments of other observers, this may provide information relevant to outcomes. The accuracy of self-assessments of ability, skills, performance, or decisions (also termed introspective accuracy) is a better predictor of everyday functional deficits than objective measures of neurocognitive or social cognitive performance (Silberstein and Harvey 2019).

The exact frequency at which measures are warranted will depend on clinical circumstances. Use of quantitative measures over time will help assure that key elements of information are collected to guide treatment (Lewis et al. 2019). Consequently, it is preferable to use a consistent approach to quantitative measurement for a given patient because each rating scale defines and measures psychosis and other symptoms differently.

Although recommending a particular scale, patient- or clinician-rated, is outside the scope of this practice guideline, a number of objective, quantitative rating scales to monitor clinical status in schizophrenia are available (American Psychiatric Association 2013a; Rush et al. 2008). The Clinician-Rated Dimensions of Psychosis Symptom Severity (American Psychiatric Association 2013b), which is included in DSM-5 for further research and clinical evaluation, contains eight domains that are rated on a scale from 0 (not present) to 4 (present and severe) on the basis of symptoms in the prior 7 days. A 6-item version of the Positive and Negative Syndrome Scale (PANSS-6; Bech et al. 2018; Østergaard et al. 2018b) consists of the PANSS items for delusions, conceptual disorganization, hallucinations, blunted affect, social withdrawal, and lack of spontaneity and flow of conversation (Kay et al. 1987). Data on the PANSS-6 suggest that it correlates highly with scores on the 30-item version of the Positive and Negative Syndrome Scale (PANSS-30; Østergaard et al. 2018a, 2018b). Furthermore, the PANSS-6 is sensitive to changes with treatment and able to identify symptom remission with a high degree of accuracy (Østergaard et al. 2018a, 2018b) if individuals who are performing ratings are appropriately trained (Opler et al. 2017). Clinician-rated scales may also be selected to assess specific clinical presentations, such as using the Bush-Francis Catatonia Rating Scale for individuals with catatonic features (Bush et al. 1996a). Other clinician-rated scales are commonly used for monitoring psychopathology in research but are likely to be too lengthy for routine clinical use. These include the PANSS-30 (Kay et al. 1987), the Scale for the Assessment of Negative Symptoms (SANS; Andreasen 1984a), the Scale for the Assessment of Positive Symptoms (SAPS; Andreasen 1984b), and the Brief Psychiatric Rating Scale (BPRS; https://mha.ohio.gov/Portals/0/assets/HealthProfessionals/About MH and Addiction Treatment/First Episode Psychosis/Brief-Psychiatric-Rating-Scale.pdf) (Leucht et al. 2005; Overall and Gorham 1962; Ventura et al. 1993).

In terms of patient-rated scales, the DSM-5 Self-Rated Level 1 Cross-Cutting Symptom Measure—Adult (American Psychiatric Association 2013c) includes a total of 23 items in 13 domains, with only 2 items related to psychosis. Nevertheless, it may be useful for identifying and tracking symptoms other than psychosis, including those related to co-occurring disorders. DSM-5 also includes 36-item self- and proxy-administered versions of the World Health Organization Disability Schedule 2.0 (WHODAS 2.0) for assessing functioning difficulties due to health and mental health conditions (American Psychiatric Association 2013d; Üstün et al. 2010). Other options for assessing functioning include the Social and Occupational Functioning Assessment Scale (SOFAS; American Psychiatric Association 2000) and the Personal and Social Performance scale (Morosini et al. 2000). Several versions of Patient-Reported Outcomes Measurement Information System (PROMIS) scales, which address social roles and functioning, are also available (www.healthmeasures.net/explore-measurement-systems/promis). The use of ratings from other informants is particularly helpful in assessing the patient’s level of functioning because individuals with schizophrenia often have a different view of their functioning than do family members or others involved in their lives (Harvey 2011).

For a nonspecific measure of quality of life, patients can be asked to rate their overall (physical and mental) quality of life in the past month on a scale from 0 (“about as bad as dying”) to 10 (“life is perfect”) (Unützer et al. 2002). In individuals with chronic mental illness, the Satisfaction With Life Scale (Diener et al. 1985) has been developed and used to assess life satisfaction and quality of life. Quality of life can also be measured using a scale developed by the World Health Organization, the WHOQOL-BREF (Skevington et al. 2004; WHOQOL Group 1998) (http://depts.washington.edu/seaqol/WHOQOL-BREF). The Centers for Disease Control and Prevention Healthy Days Measure (HRQOL-14) and Healthy Days Core Module (HRQOL-4) (www.cdc.gov/hrqol/hrqol14_measure.htm) have also been used in general population samples to assess physical and emotional symptoms as related to an individual’s perceived sense of well-being (Moriarty et al. 2003).

Rating scales should always be implemented in a way that supports developing and maintaining the therapeutic relationship with the patient. Reviewing scale results with the patient can help foster a collaborative dialogue about progress toward symptom improvement, functioning gains, and recovery goals. Such review may help clinicians, patients, families, and other support persons recognize that improvement is taking place or, conversely, identify issues that need further attention.

If more than one quantitative measure is being used, it is important to minimize duplication of questions and avoid overwhelming the patient with an excessive number of scales to complete. In addition, when choosing among available quantitative measures, objectives of scale use (e.g., screening, documenting baseline symptoms, ongoing monitoring) should be considered. Optimal scale properties (e.g., sensitivity, specificity) will differ depending on the desired purpose, but assessments of scale validity and reliability are typically conducted cross-sectionally in research contexts.

Because many scales ask the patient to rate symptoms over several weeks, they may not be sensitive to change. This can be problematic in acute care settings, where treatment adjustments and symptom improvement can occur fairly quickly. Some symptom-based quantitative measures focus either on symptom frequency over the observation period or on symptom severity. Although these features often increase or decrease in parallel, that is not invariably the case. Other quantitative measures ask the patient to consider both symptom frequency and severity, which can also make the findings difficult to interpret.

Other factors that can affect the statistical reliability and validity of rating scale measures include comorbid illnesses and patient age, language, race, ethnicity, cultural background, literacy, and health literacy. These factors and others can lead patients to misinterpret questions or bias the ratings that they record, either unintentionally (e.g., to please the clinician with their progress) or intentionally (e.g., to obtain controlled substances, to support claims of disability). Thus, the answers to questions and the summative scores on quantitative measures need to be interpreted in the context of the clinical presentation.

The type and extent of quantitative measures used will also be mediated by the clinical setting, the time available for evaluation, and the urgency of the situation. In some clinical contexts, such as a planned outpatient assessment, patients may be asked to complete electronic- or paper-based quantitative measures, either prior to the visit or on arrival at the office (Allen et al. 2009; Harding et al. 2011). Between or prior to visits, electronic approaches (e.g., mobile phone applications, clinical registries, patient portal sites in electronic health records) may also facilitate obtaining quantitative measurements (Lewis et al. 2019; Palmier-Claus et al. 2012; K. Wang et al. 2018). In other clinical contexts, such as acute inpatient settings, electronic modes of data capture may be more cumbersome, and patients may need more assistance in completion of scales. As an alternative, printed versions of scales may be completed by the patient or a proxy or administered by the clinician. In other clinical circumstances, however, printed or electronic versions of quantitative scales may not be readily available or information may not be available to complete all scale items. In emergency settings, use of a quantitative rating scale may need to be postponed until the acute crisis has subsided or until the patient’s clinical status permits a detailed examination.

Although available information suggests that ambulatory patients are generally cooperative, some individuals may be unwilling to complete quantitative measures (Narrow et al. 2013). Severe symptoms, co-occurring psychiatric conditions, low health literacy, reading difficulties, or cognitive impairment may limit some patients’ ability to complete self-report instruments (Harding et al. 2011; Valenstein et al. 2009; Zimmerman et al. 2011). In these circumstances, it may be necessary to place greater reliance on collateral sources of information such as family members, other treating health professionals, or staff members of community residence programs, if applicable. If collateral sources of information are not immediately available, treatment may also need to proceed, with adjustments in the plan, if indicated, as additional knowledge is gained. If time constraints are present, the clinician may wish to focus on rating of relevant target symptoms (e.g., on a Likert scale). In emergent circumstances, safety of the patient and others must take precedence; the initial assessment may need to be brief, with a more detailed assessment and incorporation of quantitative measures once the acute clinical situation has been stabilized.

Multiple guidelines from other organizations were reviewed (Addington et al. 2017a, 2017b; Barnes et al. 2011; Buchanan et al. 2010; Crockford and Addington 2017; Galletly et al. 2016; Hasan et al. 2012, 2013, 2015; National Institute for Health and Care Excellence 2014; Norman et al. 2017; Pringsheim et al. 2017; Scottish Intercollegiate Guidelines Network 2013). None of these guidelines specifically recommend using quantitative measures as part of the initial assessment in individuals with schizophrenia, but several guidelines (Barnes et al. 2011; Galletly et al. 2016) do recommend use of rating scales under some circumstances at baseline or as part of ongoing monitoring.

There is insufficient consensus on the most appropriate quantitative measures (i.e., rating scales) to use in assessing individuals with psychotic disorders, including schizophrenia. Nevertheless, data do support the use of clinician-reported, symptom-based ratings to guide treatment. Patient-rated scales may be clinically useful in identifying patient concerns or subjective experiences (e.g., medication side effects). On the basis of these potential benefits, a process-focused internal or health system–based quality improvement measure could determine rates of quantitative measure use, and quality improvement initiatives could be implemented to increase the frequency with which such measures are used in individuals with schizophrenia.

When treating individuals with schizophrenia, a person-centered treatment plan should be developed, documented in the medical record, and updated with the patient at appropriate intervals. A person-centered treatment plan can be recorded as part of an evaluation note or progress note and does not need to adhere to a defined development process (e.g., face-to-face multidisciplinary team meeting) or format (e.g., time-specified goals and objectives). Depending on the urgency of the initial clinical presentation, the availability of laboratory results, and other sources of information, the initial treatment plan may need to be augmented over several visits as more details of history and treatment response are obtained.

As treatment proceeds, the treatment plan will require iterative reevaluation and adjustment prompted by such factors as inadequate treatment response, difficulties with tolerability or adherence, impairments in insight, changes in presenting issues or symptoms, or revisions in diagnosis. In adapting treatment to the needs of the individual patient, tailoring of the treatment plan may also be needed on the basis of sociocultural or demographic factors, with an aim of enhancing quality of life or aspects of functioning (e.g., social, academic, occupational). Factors that influence medication metabolism (e.g., age, sex, body weight, renal or hepatic function, smoking status, use of multiple concurrent medications) may also require adjustments to the treatment plan in terms of either typical medication doses or frequency of monitoring. For most individuals with schizophrenia, it is challenging to piece together a coherent picture of the patient’s longitudinal course from medical records. Thus, it is important to note the rationale for any changes in the treatment plan as well as the specific changes that are being made because an accurate history of past and current treatments and responses to them is a key part of future treatment planning.

Information from other guidelines (Addington et al. 2017a, 2017b; Barnes et al. 2011; Buchanan et al. 2010; Crockford and Addington 2017; Galletly et al. 2016; Hasan et al. 2012, 2013, 2015; National Institute for Health and Care Excellence 2014; Norman et al. 2017; Pringsheim et al. 2017; Scottish Intercollegiate Guidelines Network 2013) is generally consistent with this guideline statement in either explicitly or implicitly recommending development of a person-centered treatment plan that includes evidence-based nonpharmacological and pharmacological treatments.

It is not known whether psychiatrists and other mental health professionals typically document a comprehensive and person-centered treatment plan that includes evidence-based nonpharmacological and pharmacological treatments, and there is likely to be variability. Although a well-defined and scientifically sound quality measure could be developed to assess for the implementation of an evidence-based treatment plan that meets consensus-based features of person-centered care, clinical judgment would still be needed to determine whether a documented treatment plan is comprehensive and adapted to individual needs and preferences. Manual review of charts to evaluate for the presence of such a person-centered treatment plan would be burdensome and time-consuming to implement.

A quality measure could assess the presence or absence of text in the medical record that would reflect treatment planning. When considering the development of such quality measures, there should be a thorough examination of the potential for unintended negative consequences, such as increased documentation burden or overuse of standardized language that meets the quality measure criteria but would inaccurately reflect what occurred in practice.

Information from other guidelines is consistent with this guideline statement. Other guidelines on the treatment of schizophrenia (BAP, Canadian Schizophrenia Guidelines [CSG], NICE, PORT, RANZCP, SIGN, WFSBP) all recommend use of an antipsychotic medication in the treatment of schizophrenia, with the selection of a specific medication on an individualized basis with consideration of medication characteristics, patient characteristics, and patient preferences (Addington et al. 2017a, 2017b; Barnes et al. 2011; Buchanan et al. 2010; Crockford and Addington 2017; Galletly et al. 2016; Hasan et al. 2012; National Institute for Health and Care Excellence 2014; Pringsheim et al. 2017; Remington et al. 2017; Scottish Intercollegiate Guidelines Network 2013). Each guideline also recommends monitoring during the course of treatment to assess therapeutic response and treatment-related side effects.

In clinical practice, almost all individuals with schizophrenia are offered an antipsychotic medication. Thus, a quality measure is unlikely to enhance outcomes if it examines only whether an individual with schizophrenia is offered or receives an initial prescription for antipsychotic treatment. An existing National Quality Forum–endorsed measure, “Adherence to Antipsychotic Medications for Individuals with Schizophrenia” (NQF #1879, www.qualityforum.org/QPS/1879), is aimed at assessing whether an antipsychotic medication is continued once it is begun. For individuals who are at least 18 years old and who have a diagnosis of schizophrenia or schizoaffective disorder, this measure assesses the percentage who have been dispensed an antipsychotic medication (as reflected by at least two such prescriptions being filled) and who had a proportion of covered days of at least 0.8 during a 12 consecutive month measurement period. By requiring ongoing prescribing of antipsychotic medication, this measure is more likely to be associated with improvements in outcomes for patients. Nevertheless, this measure does have several limitations. It uses pharmacy claims data or electronic prescription orders to examine whether a medication has been prescribed, but such measures do not guarantee treatment adherence. For instance, a prescriber could submit an antipsychotic medication prescription and the patient could fill the prescription at the pharmacy, but the patient might not actually take the medication. This measure also does not determine the adequacy of the medication or the medication dose and could be met through continuous prescriptions of a subtherapeutic dose or clinically ineffective antipsychotic. Another limitation is the difficulty in determining the proportion of covered days in a 12 consecutive month period, particularly when patients have transitions in care between settings or treating clinicians.

Quality measures, quality improvement initiatives, or electronic decision supports may be appropriate for monitoring side effects of antipsychotic treatment. Evidence suggests that rates of guideline concordant monitoring are low for metabolic risk factors, including lipids, diabetes, and weight (Mitchell et al. 2012; Morrato et al. 2009). Several measures endorsed by NQF address such monitoring. NQF #1932, “Diabetes Screening for People With Schizophrenia or Bipolar Disorder Who Are Using Antipsychotic Medications” (www.qualityforum.org/QPS/1932) measures the percentage of patients ages 18–64 years with schizophrenia or bipolar disorder “who were dispensed an antipsychotic medication and had a diabetes screening test during the measurement year.” Because this measure is focused on screening, it excludes monitoring of individuals who had diabetes in the measurement year or in the preceding year.

NQF #1927, “Cardiovascular Health Screening for People With Schizophrenia or Bipolar Disorder Who Are Prescribed Antipsychotic Medications” (www.qualityforum.org/QPS/1927) measures the percentage of individuals ages 25–64 years with schizophrenia or bipolar disorder “who were prescribed any antipsychotic medication and who received a cardiovascular health screening during the measurement year,” where cardiovascular health screening consists of “one or more LDL-C screenings performed during the measurement year.” Individuals are excluded from this screening measure for having evidence of preexisting cardiovascular disease as defined by precise criteria in the measure text. Presently, the specific elements of these criteria can often be challenging to determine from unstructured electronic health records.

Two additional NQF-approved measures, NQF #1933 (www.qualityforum.org/QPS/1933) and NQF #1934 (www.qualityforum.org/QPS/1934) address cardiovascular and diabetes monitoring, respectively, for individuals with preexisting cardiovascular disease or diabetes. Each of these measures is limited to individuals who are ages 18–64 years. The cardiovascular monitoring measure requires that individuals receive “an LDL-C test performed during the measurement year,” whereas the diabetes monitoring measure requires “[o]ne or more HbA1c tests and one or more LDL-C tests performed during the measurement year.” These measures have been tested for feasibility, usability, reliability, and validity at the health plan, integrated delivery system, and population levels; however, before holding individual clinicians or facilities accountable for the delivered quality of care, the measures would need additional testing at these levels.

For individuals with a diagnosis of schizophrenia whose symptoms have improved with an antipsychotic medication, there are a number of benefits to maintenance treatment, including reduced risks of relapse (Bowtell et al. 2018; Goff et al. 2017; Hui et al. 2018; Kishi et al. 2019; Leucht et al. 2012; Thompson et al. 2018), rehospitalization (Tiihonen et al. 2018), and death (Tiihonen et al. 2018; Vermeulen et al. 2017). When administered on a long-term basis, however, antipsychotic medications are also associated with a greater incidence of weight gain, sedation, and movement disorders (Leucht et al. 2012). In addition, some studies have raised questions about whether long-term antipsychotic treatment might be associated with other adverse effects on functioning or health, including loss of brain volume (Davidson 2018; Goff et al. 2017). These data are heterogeneous, and when compared with withholding treatment, there was minimal evidence to suggest negative effects of maintenance treatments on outcomes (Goff et al. 2017; Huhtaniska et al. 2017). In addition, it may be possible to mitigate these risks by preventive interventions (e.g., early intervention for weight gain, screening for lipid and glucose abnormalities) and careful monitoring for side effects of medication. Nevertheless, as treatment proceeds, the pluses and minuses of continuing treatment with an antipsychotic medication should be reviewed with the patient in the context of shared decision-making. It will typically be beneficial to include family members or other persons of support in such discussions (Hamann and Heres 2019).

Despite the benefits of continued antipsychotic treatment for the majority of patients, maintaining adherence to an antipsychotic medication can be difficult (Acosta et al. 2012; Shafrin et al. 2016; Valenstein et al. 2006). Barriers to, facilitators of, and motivators of treatment adherence will differ for each patient. Engaging family members or other persons of support can be helpful in fostering adherence (Mueser et al. 2015). Other approaches to assessing and enhancing adherence are described in detail in Statement 3.

For some patients, the formulation of the antipsychotic medication may influence adherence (see Statement 4, Table 3). For example, rapidly dissolving tablets or oral concentrates may be preferable for patients who have difficulty swallowing pills or who are ambivalent about medications and inconsistent in swallowing them. LAI formulations may be preferred by some patients (Heres et al. 2007; Patel et al. 2009; Walburn et al. 2001) and may be particularly useful for patients with a history of poor or uncertain adherence (see Statement 10).

It is also important to assess the ongoing benefits and side effects of treatment that may indicate a need for adjustments to medication doses or changes in medications. The use of quantitative measures can be helpful in systematically assessing each of these realms (see Statement 2). The optimal dose of medication is one that provides the best medication benefits yet is tolerable in terms of medication side effects. For some patients, adjustments in dose will be required during the course of treatment to maintain this balance (Essock et al. 2006). Such factors as addition or discontinuation of interacting medications, changes in smoking status, changes in patient body mass, changes in renal or hepatic status, or changes in drug absorption (e.g., with bariatric surgery) may influence medication pharmacokinetics and require increases or decreases in medication dose. In order to minimize the risk of extended side effects, when medications with a long half-life are used and particularly when LAI antipsychotic medications are used, considerations about changes in dose need to consider the extended actions of these medications.

Evidence on the rationale and approach to planned reductions of medication doses is limited. Unless a medication requires emergent discontinuation, gradual reductions in doses are preferable, with close monitoring for recurrent symptoms.

Shared decision-making discussions with the patient should consider the patient’s recovery goals, the potential benefits of medication changes or dose reductions in terms of changes or diminution of side effects, and the potential harms of medication changes or dose reductions (Davidson 2018). The longitudinal course of the patient’s episode and the certainty of the diagnosis should also be considered. There may be some individuals with a brief episode of psychosis or uncertain psychotic diagnosis (e.g., possible substance-induced psychosis or mood-related psychosis) who may not require continuing antipsychotic treatment. On the other hand, individuals with chronic symptoms, repeated relapses, and clear diagnostic features of schizophrenia will likely have poorer outcomes if medications are stopped. In addition to symptom recurrence and relapse, medication cessation may be associated with hospitalization, legal difficulties, reduced likelihood of response with reinstatement of treatment, or poorer psychosocial outcomes (Correll et al. 2018; Hui et al. 2018; Takeuchi et al. 2019; Wilper et al. 2009). It will typically be beneficial to include family members or other persons of support in discussions of medication changes or dose reductions.

Information from other guidelines is consistent with this guideline statement. Other guidelines on the treatment of schizophrenia (BAP, NICE, PORT, SIGN, WFSBP) recommend continued use of an antipsychotic medication in the treatment of schizophrenia once symptom response has been achieved (Barnes et al. 2011; Buchanan et al. 2010; Hasan et al. 2013; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013). Other guidelines (RANZCP, BAP, SIGN, NICE) also suggest the use of LAI antipsychotic medications on the basis of patient preference or when adherence has been poor or uncertain (Barnes et al. 2011; Galletly et al. 2016; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013). Use of a gradual reduction in dose, including a gradual cross-taper when changing medications, is noted by several guidelines (BAP, NICE, SIGN) along with an emphasis on close monitoring for signs of relapse (Barnes et al. 2011; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013).

See Statement 4 for a discussion of quality measures related to initiation and ongoing use of an antipsychotic medication.

As noted in Statement 5, it is important for treatment with an antipsychotic medication to be maintained once symptoms have improved. Specifically, for individuals with a diagnosis of schizophrenia, there are a number of benefits to continued treatment with an antipsychotic medication, including reduced risks of relapse (Bowtell et al. 2018; Goff et al. 2017; Hui et al. 2018; Kishi et al. 2019; Leucht et al. 2012; Thompson et al. 2018), rehospitalization (Tiihonen et al. 2018), and death (Tiihonen et al. 2018; Vermeulen et al. 2017). Implicitly, continued treatment with an effective and tolerable medication would be preferable to potential destabilization or treatment discontinuation. This inference is also consistent with clinical observations that individualizing choice of an antipsychotic medication is important. In clinical trials, a change to a different medication has been associated with earlier discontinuation of treatment as compared with continuation of the same antipsychotic medication (Essock et al. 2006; Stroup et al. 2011).

For these reasons, it will be optimal to continue on the same medication for most patients. Nevertheless, under some circumstances, it may be necessary to consider a change from one antipsychotic medication to another one. For example, a patient may have experienced some degree of response to initial treatment but may still have significant symptoms or difficulties in functioning that would warrant a trial of a different medication. Another reason to change medications would be to initiate treatment with an LAI antipsychotic if the current oral medication is unavailable in an LAI formulation (see Statement 10). A medication change may also be considered because of patient preferences, medication availability, or side effects. Given the long-term health risks of metabolic syndrome and obesity, weight gain and development of diabetes or metabolic syndrome are common reasons that a change to a different medication may be discussed.

In a randomized study that examined the effects of switching from olanzapine, quetiapine, or risperidone to aripiprazole to reduce metabolic risk, a change to aripiprazole was associated with improvements in non-HDL cholesterol, serum triglycerides, and weight as well as a small reduction in 10-year risk of coronary heart disease but no difference in the odds of having metabolic syndrome (Stroup et al. 2011, 2013). Individuals who switched to aripiprazole, as compared with those who remained on their initial medication, had a higher rate of discontinuing treatment but showed no significant increases in symptoms or hospitalizations. In addition, the Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) showed that individuals who experienced issues with medication efficacy or tolerability in an early phase of the trial could still go on to do well with a different medication in a subsequent phase of the trial (McEvoy et al. 2006; Rosenheck et al. 2009; Stroup et al. 2007, 2009).

These findings suggest that a change in medication can be of benefit to patients under some circumstances but also suggest that the possible benefits and risks of a medication change should be reviewed with the patient in the context of shared decision-making. Such discussions with the patient should consider the patient’s recovery goals, the potential benefits of medication changes or dose reductions in terms of changes in or diminution of side effects, and the potential harms of medication changes or dose reductions (Davidson 2018). It will typically be beneficial to include family members or other persons of support in such discussions.

Only a limited amount of research has explored the optimal approach for changing antipsychotic medications when warranted. The typical approach is a gradual cross-taper in which the second antipsychotic medication is begun and gradually increased in dose as the initial antipsychotic medication is gradually tapered. However, the few studies that are available do not suggest differences between gradual discontinuation as compared with immediate discontinuation of the first medication (Takeuchi et al. 2017a). In addition, no differences have been seen between starting the second antipsychotic and discontinuing the first antipsychotic at the same time as compared with starting the second antipsychotic and waiting before discontinuing the first antipsychotic agent (Takeuchi et al. 2017b). Regardless of the approach that is taken, careful monitoring is essential to avoid the risks of reduced adherence and clinical destabilization if a change in medications is undertaken. Depending on the pharmacological properties of the medications, including pharmacokinetic and receptor binding profiles (see Statement 4, Tables 4 and 5), side effects of medications may also emerge (e.g., insomnia with a shift to a less sedating medication, withdrawal dyskinesia with a shift to a medication with less prominent dopamine D₂ receptor blockade) (Cerovecki et al. 2013).

Information from other guidelines is consistent with this guideline statement. Other guidelines on the treatment of schizophrenia (SIGN, WFSBP) note that treatment should usually continue with the same antipsychotic medication that led to the best response and had the best individual side-effect profile, given the risk of destabilization with switching an antipsychotic regimen (Hasan et al. 2013; Scottish Intercollegiate Guidelines Network 2013).

As a suggestion, this guideline statement is not appropriate for use as a quality measure or for electronic decision support. However, health plans may wish to implement internal process measures to assess and reduce rates at which changes to stable medication regimens are made on the basis of nonclinical factors such as pre-authorization requirements or formulary changes.

Practice guidelines (BAP, CSG, NICE, RANZCP, SIGN, WFSBP, PORT) are consistent in recommending clozapine for individuals with treatment-resistant schizophrenia (Barnes et al. 2011; Buchanan et al. 2010; Galletly et al. 2016; Hasan et al. 2012; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013). In terms of other therapies for treatment-resistant schizophrenia, several guidelines (SIGN, WFSBP, BAP) recommend augmentation treatment with an antidepressant for treatment-resistant illness associated with negative symptoms (Barnes et al. 2011; Hasan et al. 2012; Scottish Intercollegiate Guidelines Network 2013). For individuals with catatonia, the WFSBP recommends benzodiazepines and ECT (Hasan et al. 2012). In addition, ECT is mentioned in several guidelines (e.g., SIGN, RANZCP, WFSBP) as being appropriate in individuals with treatment-resistant schizophrenia.

Studies suggest that clozapine is underused and that a significant proportion of individuals with treatment-resistant schizophrenia do not receive treatment with clozapine, although there is significant variation between and within countries (Addington et al. 2012; Bachmann et al. 2017; Carruthers et al. 2016; Keller et al. 2014; Olfson et al. 2016; Stroup et al. 2014; Tang et al. 2017). Thus, internal quality improvement programs may wish to focus on ways to increase use of clozapine in individuals with treatment-resistant schizophrenia and track rates of clozapine use in this patient population. Internal quality improvement programs could also focus on increasing use of quantitative measures (i.e., rating scales) to improve identification of individuals with treatment-resistant schizophrenia and facilitate systematic longitudinal tracking of functioning, symptoms, and side-effect burdens. If quality measures are considered for development at the provider, facility, health plan, integrated delivery system, or population level, testing of feasibility, usability, reliability, and validity would be essential prior to use for purposes of accountability.

Electronic decision support would be challenging to implement and would depend on accurate and consistent entry of structured information. Nevertheless, in combination with rating scale data and prior prescribing histories, electronic decision support could help identify individuals with treatment-resistant illness who would benefit from a trial of clozapine.

Treatment with clozapine can be effective in reducing suicidal behavior if risk remains substantial despite other treatments. In addition, treatment with clozapine can be effective in reducing rates of suicide attempts and suicide in individuals with schizophrenia, regardless of whether formal criteria for treatment resistance have been met. Risk factors for suicidal behavior in individuals with schizophrenia are described in Statement 1, subsection “Implementation.” Although demographic and historical risk factors are static, a number of other risk factors are potentially modifiable and can serve as targets of intervention in constructing a plan of treatment (for additional details, see Statement 3). As in other circumstances in which patients do not appear to be responding fully to treatment, attention to adherence is crucial (for additional details, see Statement 3). For details of initiating and monitoring clozapine treatment, see Statement 7, subsections “Initiation of Treatment With Clozapine,” “Use of Clozapine Levels During Treatment With Clozapine,” and “Monitoring for Side Effects During Treatment With Clozapine.”

Other guidelines do not specifically mention the use of clozapine for individuals with schizophrenia who are at substantial risk for suicide attempts or suicide despite other treatment. Guidelines (BAP, CSG, NICE, RANZCP, SIGN, WFSBP, PORT) are consistent, however, in recommending clozapine for individuals with treatment-resistant schizophrenia (Barnes et al. 2011; Buchanan et al. 2010; Galletly et al. 2016; Hasan et al. 2012; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013).

Studies suggest that clozapine is underused and that a significant proportion of individuals with treatment-resistant schizophrenia do not receive treatment with clozapine, although there is significant variation between and within countries (Addington et al. 2012; Bachmann et al. 2017; Carruthers et al. 2016; Keller et al. 2014; Olfson et al. 2016; Stroup et al. 2014; Tang et al. 2017). Given low utilization of clozapine in general (Addington et al. 2012; Bachmann et al. 2017; Carruthers et al. 2016; Keller et al. 2014; Olfson et al. 2016; Stroup et al. 2014; Tang et al. 2017) and the high rates of suicidal ideas among individuals with treatment-resistant schizophrenia (Kennedy et al. 2014), it is likely that many individuals at significant suicide risk are not receiving treatment with clozapine. Thus, internal quality improvement programs may wish to focus on ways to increase and track use of clozapine in individuals with schizophrenia who have significant suicide risk that persists despite other treatments. Internal quality improvement programs could also focus on increasing the use of quantitative measures to improve identification and monitoring of individuals with risk factors for suicide.

If quality measures are considered for development at the provider, facility, health plan, integrated delivery system, or population level, testing of feasibility, usability, reliability, and validity would be essential prior to use for purposes of accountability. In particular, it is currently not possible to identify from most administrative data people at increased risk for suicide. Clinical assessment or patient self-report data are likely to be required.

Electronic decision support using passive alerts may be able to prompt clinicians to consider clozapine; however, such prompts would be challenging to implement because they would depend on accurate and consistent entry of structured information about diagnosis, suicidal ideation, and suicide attempts. Nevertheless, in combination with rating scale data, electronic decision support could help identify individuals with schizophrenia and significant suicide risk who would benefit from a trial of clozapine.

Treatment with clozapine can be effective in reducing aggressive behavior if risk remains substantial despite other treatments. As in other circumstances in which patients do not appear to be responding fully to treatment, attention to adherence is crucial (for additional details, see Statement 3). Risk factors for aggressive behavior in individuals with schizophrenia are described in Statement 1, subsection “Implementation.” Although demographic and historical risk factors are static, a number of other risk factors are potentially modifiable and can serve as targets of intervention in constructing a plan of treatment (for additional details, see Statement 3). For details of initiating and monitoring clozapine treatment, see Statement 7, subsections “Initiation of Treatment With Clozapine,” “Use of Clozapine Levels During Treatment With Clozapine,” and “Monitoring for Side Effects During Treatment With Clozapine.”

Information from other guidelines is consistent with this guideline statement. The SIGN, RANZCP, BAP, and PORT guidelines all suggest consideration of clozapine for individuals with hostility or aggressive behaviors that do not respond to other interventions (Barnes et al. 2011; Buchanan et al. 2010; Galletly et al. 2016; Hasan et al. 2012; Scottish Intercollegiate Guidelines Network 2013). In addition, guidelines (BAP, CSG, NICE, RANZCP, SIGN, WFSBP, PORT) are consistent in recommending clozapine for individuals with treatment-resistant schizophrenia (Barnes et al. 2011; Buchanan et al. 2010; Galletly et al. 2016; Hasan et al. 2012; National Institute for Health and Care Excellence 2014; Scottish Intercollegiate Guidelines Network 2013).

As a suggestion, this guideline statement is not appropriate for use as a quality measure for purposes of accountability. Electronic decision support using passive alerts may be able to prompt clinicians to consider clozapine; however, such prompts would be challenging to implement because they would depend on accurate and consistent entry of structured information about diagnosis and risk factors for aggression. Nevertheless, in combination with rating scale data, electronic decision support could help identify individuals with schizophrenia and significant aggression risk who may benefit from a trial of clozapine.

LAI formulations of antipsychotic medications can provide a number of benefits for patients, families, and clinicians, yet they are often underutilized (Brown et al. 2014; Correll et al. 2016; Lawson et al. 2015; Sultana et al. 2019). Racial differences also exist in the proportion of individuals who are treated with LAI antipsychotic medications, with greater use of these formulations in Black patients than in white patients (Brown et al. 2014; Lawson et al. 2015).

With LAI antipsychotic medications, there is greater assurance that a patient will receive medication continuously because there are fewer opportunities to miss a medication dose and clinicians will be immediately aware of a missed visit or injection, yielding greater time for intervention before symptoms recur (Correll et al. 2016; Velligan et al. 2010; West et al. 2008). Presumably due to improved adherence, advantages of LAI antipsychotics include the potential for a decreased risk of mortality; reduced risk of hospitalization; and decreased rates of treatment discontinuation, including discontinuation due to inefficacy (see Appendix C, Statement 10). Other benefits for patients include a subjective sense of better symptom control, greater convenience as a result of needing to take fewer medications daily, and reduced conflict with family members or other persons of support related to medication-related reminders (Caroli et al. 2011; Correll et al. 2016; Iyer et al. 2013; Yeo et al. 2019). Although some patients may not wish to experience the discomfort associated with receiving injections of medications, this is not a major barrier for most patients. In addition, discomfort can often be minimized by using SGA LAIs rather than FGA LAIs, which have sesame oil–based vehicles, or by using an LAI with a small injection volume or lower administration frequency (Correll et al. 2016).

Consistent with principles of patient-centered care, it may be preferable to educate patients about the availability of LAI antipsychotic medications when discussing other aspects of antipsychotic treatment. Indeed, many patients will accept and may prefer LAIs if provided with information about the pluses and minuses of LAIs in the context of shared decision-making (Caroli et al. 2011; Heres et al. 2007; Kane et al. 2019; Waddell and Taylor 2009; Weiden et al. 2015; Yeo et al. 2019). Preference rates for LAIs are even higher among individuals who have personal experience in receiving an LAI formulation of an antipsychotic medication (Heres et al. 2007; Patel et al. 2009; Waddell and Taylor 2009).

Discussions about LAI antipsychotic medications often occur with patients who have had difficulty in adhering to oral medications. However, such discussions can also take place at other junctures. For example, if an individual has not responded to treatment with an oral antipsychotic medication, a trial of an LAI may be warranted (Howes et al. 2017; see Statement 4, subsection “Strategies to Address Initial Nonresponse or Partial Response to Antipsychotic Treatment”) because breaks in the continuity of oral medication therapy can be unrecognized (Lopez et al. 2017). An LAI formulation of an antipsychotic may also be considered when patients are transitioning between settings (e.g., at inpatient discharge, on release from a correctional facility), when future adherence is uncertain and the risk of reduced adherence may be increased. Although LAI antipsychotic medications have typically been used in individuals with multiple episodes of schizophrenia, some studies have used an LAI antipsychotic formulation earlier in the course of illness (Kane et al. 2019; Subotnik et al. 2015), when rates of poor adherence may be greater. Earlier discussion of an LAI may also be considered in individuals who are at increased risk of poor adherence due to a limited awareness of needing treatment or a co-occurring substance use disorder (García et al. 2016; Velligan et al. 2009).

If a decision is made to initiate treatment with an LAI, aspects of medication selection are similar to those for selection of an oral medication in terms of considering prior response, prior tolerability, pharmacological considerations, and side-effect profiles (see Statement 4). Patients may also have specific preferences and values related to the frequency of injection and the type and location of the injection (e.g., IM deltoid or gluteal sites, subcutaneous abdominal site; Heres et al. 2012). Because the oral and LAI formulations of a specific antipsychotic medication are comparable, a trial of the same oral antipsychotic will typically occur first to assure efficacy and tolerability (SMI Adviser 2019). Patients will experience medication side effects for a longer period of time after drug discontinuation with LAIs than with oral medications because of pharmacokinetic differences in the formulations, but this is not usually a problem if the oral formulation has first been well tolerated. Nevertheless, caution is warranted in a patient who has experienced NMS previously (Correll et al. 2016).

The conversion from an oral dose of medication to a corresponding dose of an LAI antipsychotic depends on the specific medication (see Statement 4, Tables 7, 8, and 9; Meyer 2013, 2017); product labeling for each medication describes approximate conversion ratios and whether a period of concomitant oral and LAI medication is needed. If the patient is taking an oral medication that lacks a corresponding LAI formulation, a change in antipsychotic medication may be needed if an LAI formulation is clinically indicated or preferred (see Statement 6).

When administering LAI antipsychotic injections, it is important to follow recommendations for safe injection practices (Centers for Disease Control and Prevention 2019a, 2019b) and infection control precautions (Centers for Disease Control and Prevention 2016, 2019b) as well as instructions from product labeling. Product labeling also contains important information on storage and reconstitution of LAI antipsychotic formulations as well as information on how to handle missed or late doses of medications (see Statement 4, Tables 7–9).

There are several barriers related to the use of LAI formulations of antipsychotic medications. For patients, there may be logistical barriers (e.g., access to transportation, childcare, school or work schedules) that depend on the frequency of appointments needed to receive an LAI antipsychotic medication. Other logistical barriers for patients may relate to such factors as cost or insurance authorizations for LAI antipsychotic agents. For clinicians, lack of knowledge and limited experience in using LAI antipsychotic medications contribute to underuse (Correll et al. 2016). Skill and experience in administering injections may be lacking, and nursing staff may not be available to give injections.

Additional barriers relate to the decision to suggest an LAI antipsychotic medication. Clinicians often do not consider LAI antipsychotic medications as a treatment option (Hamann et al. 2014; Heres et al. 2006; Kirschner et al. 2013; Weiden et al. 2015), even when such use is appropriate. Furthermore, clinicians may overestimate patients’ adherence with oral medications when considering relative benefits of LAIs (Correll et al. 2016; Lopez et al. 2017) or underestimate the acceptability of LAIs to patients (Hamann et al. 2010; Iyer et al. 2013; Patel et al. 2010a, 2010b; Weiden et al. 2015). At an organizational level, there may be a lack of resources, space, or trained personnel to administer injections (Correll et al. 2016; Velligan et al. 2011). Thus, workflows may need to be adjusted, partnerships may need to be developed with primary care clinicians to administer LAI antipsychotic injections, or other concerted efforts may be needed to address logistical and clinical barriers to LAI antipsychotic use.

Information from other guidelines is consistent with this guideline statement (Barnes et al. 2011; Buchanan et al. 2010; Galletly et al. 2016; Hasan et al. 2012; National Institute for Health and Care Excellence 2014; Remington et al. 2017; Scottish Intercollegiate Guidelines Network 2013). Other guidelines on the treatment of schizophrenia suggest use of LAIs based on patient preference (BAP, CSG, NICE, PORT, RANZCP, SIGN) or in the context of poor or uncertain adherence (BAP, NICE, RANZCP, SIGN). The RANZCP and WFSBP guidelines also note that LAIs should be considered if there has been a poor response to oral medication, and the RANZCP guideline notes that LAIs should be offered to patients early in the clinical course of schizophrenia.

As a suggestion, this guideline statement is not appropriate for use as a quality measure. However, only a small proportion of individuals with schizophrenia receive an LAI antipsychotic medication in clinical practice (Brown et al. 2014; Correll et al. 2016; Lawson et al. 2015; Sultana et al. 2019). Because adherence is poor or uncertain in many individuals with schizophrenia, patients may benefit from or prefer to receive an LAI antipsychotic medication if one is offered. Consequently, electronic decision support could suggest that clinicians consider an LAI antipsychotic medication if poor adherence is documented or with repeated hospitalizations or emergency visits. In addition, health care organizations may wish to ensure that they will provide treatment with an LAI antipsychotic medication to patients when appropriate. Health care organizations and health plans may also wish to implement internal process measures to assess and increase rates at which LAI antipsychotics are used.

Medication-induced acute dystonia is defined in DSM-5 as

[a]bnormal and prolonged contraction of the muscles of the eyes (oculogyric crisis), head, neck (torticollis or retrocollis), limbs, or trunk developing within a few days of starting or raising the dosage of a medication (such as a neuroleptic) or after reducing the dosage of a medication used to treat extrapyramidal symptoms. (American Psychiatric Association 2013a, p. 711)

A dystonic spasm of the axial muscles along the spinal cord can result in opisthotonos, in which the head, neck, and spinal column are hyperextended in an arched position. Rarely, acute dystonia can also present as life-threatening laryngospasm, which results in an inability to breathe (Ganesh et al. 2015; Koek and Pi 1989). Acute dystonia is sudden in onset and painful and can cause patients great distress. Because of the dramatic appearance of acute dystonia, health professionals who are unfamiliar with the condition may incorrectly attribute these reactions to catatonic signs or unusual behavior on the part of patients, and oculogyric crises can sometimes be misinterpreted as indicative of seizure activity. In individuals treated with FGAs, it is estimated that up to 10% of patients may experience an acute dystonic episode, and with SGAs, rates of acute dystonia may be less than 2% (Martino et al. 2018; Miller et al. 2008; Satterthwaite et al. 2008). Additional factors that increase the risk of acute dystonia with antipsychotic medication include young age, male sex, ethnicity, recent cocaine use, high medication dose, and intramuscular route of medication administration (Gray and Pi 1998; Spina et al. 1993; van Harten et al. 1999).

There are a limited number of clinical studies of anticholinergic medications in acute dystonia associated with antipsychotic therapy. Nevertheless, a large amount of clinical experience suggests that acute dystonia can be reversed by administration of diphenhydramine, a histamine receptor antagonist with anticholinergic properties. Typically, it is administered intramuscularly to treat acute dystonia, but it can also be administered intravenously in emergent situations, as with acute dystonia associated with laryngospasm. Alternatively, benztropine can also be administered intramuscularly. Once the acute dystonia has resolved, it may be necessary to continue an oral anticholinergic medication to prevent recurrence, at least until other changes in medications can take place such as reducing the dose of medication or changing to an antipsychotic medication that is less likely to be associated with acute dystonia. Typically, a medication such as benztropine or trihexyphenidyl is used for this purpose because of the shorter half-life of oral diphenhydramine and a need for more frequent dosing. For additional details of dosing and use of these medications, see Statement 12, Table 10. Regardless of the anticholinergic medication that is chosen, it is important to use the lowest dose that is able to treat acute dystonia and continue the anticholinergic medication for the shortest time needed to prevent dystonia from recurring. After several weeks to months, anticholinergic medications can sometimes be reduced or withdrawn without recurrence of dystonia or worsening of other antipsychotic-induced neurological symptoms (Desmarais et al. 2012). Medications with anticholinergic effects can result in multiple difficulties for patients, including impaired quality of life, impaired cognition, and significant health complications (Salahudeen et al. 2015). Dry mouth due to anticholinergic effects is associated with an increased risk for multiple dental complications (Singh and Papas 2014), and drinking high-calorie fluids in response to dry mouth can contribute to weight gain. Medications with anticholinergic effects can also precipitate acute angle-closure glaucoma (Lachkar and Bouassida 2007), although patients with treated glaucoma seem to be able to tolerate these medications with careful monitoring (Bower et al. 2018). Other peripheral side effects of anticholinergic medications can include blurred vision, constipation, tachycardia, urinary retention, and effects on thermoregulation (e.g., hyperthermia in hot weather) (Nasrallah and Tandon 2017; Ozbilen and Adams 2009), and central anticholinergic effects can include impaired learning and memory and slowed cognition (Ang et al. 2017; Vinogradov et al. 2009). Older individuals can be particularly sensitive to these anticholinergic effects and can develop problems such as urinary retention, confusion, fecal impaction, and anticholinergic toxicity (with delirium, somnolence, and hallucinations) (Nasrallah and Tandon 2017). In addition, it is important to consider the anticholinergic side effects associated with other medications that a patient is taking, such as antipsychotic medications, some antidepressant medications, urological medications (e.g., oxybutynin), and nonselective antihistamines (e.g., hydroxyzine, diphenhydramine).

The guidelines of the WFSBP are in agreement with this recommendation, noting that acute dystonia responds dramatically to administration of anticholinergic or antihistaminic medication (Hasan et al. 2013). The guideline of the BAP notes that use for acute dystonia “should be determined on an individual basis, taking account of factors such as the patient’s history of extrapyramidal side effects and the risk of anticholinergic side effects” (Barnes et al. 2011, p. 7).

This guideline statement is not appropriate for use as a quality measure or as part of electronic clinical decision support. Even with short-term treatment, some patients could have the potential to develop significant anticholinergic side effects, which would need to be incorporated into exclusion and exemption criteria. With reductions in the use of high doses of high-potency FGAs, the frequency of acute dystonia is significantly reduced. Any measure would apply only to a small number of individuals, which would complicate testing for feasibility, usability, reliability, and validity.

Medication-induced parkinsonism, which is termed neuroleptic-induced parkinsonism and other medication-induced parkinsonism in DSM-5, is defined as

[p]arkinsonian tremor, muscular rigidity, akinesia (i.e., loss of movement or difficulty initiating movement), or bradykinesia (i.e., slowing movement) developing within a few weeks of starting or raising the dosage of a medication (e.g., a neuroleptic) or after reducing the dosage of a medication used to treat extrapyramidal symptoms. (American Psychiatric Association 2013a, p. 709)

These symptoms of medication-induced parkinsonism are dose-dependent and generally resolve with discontinuation of antipsychotic medication. It is important to appreciate that medication-induced parkinsonism can affect emotional and cognitive function, at times in the absence of detectable motor symptoms. As a result, it can be difficult to distinguish the negative symptoms of schizophrenia or concomitant depression from medication-induced parkinsonism. In addition, emotional and cognitive features of medication-induced parkinsonism can be subjectively unpleasant and can contribute to poor medication adherence (Acosta et al. 2012; Ascher-Svanum et al. 2006).

A number of approaches can be taken when a patient is experiencing medication-induced parkinsonism. A reduction in the dose of the antipsychotic medication, if feasible, is often helpful in reducing parkinsonism. In some individuals, it may be appropriate to change the antipsychotic medication to one with a lower likelihood of parkinsonism (see Statement 4, Table 6). For individuals who are highly sensitive to medication-induced parkinsonism, clozapine may be considered. However, before reducing the dose of medication or changing to another antipsychotic medication, the benefits of reduced parkinsonism should be weighed against the potential for an increase in psychotic symptoms. Careful monitoring for symptom recurrence is always important when making changes or reducing doses of antipsychotic medications, and use of quantitative measures can be helpful in this regard, as described in Statement 2.

The use of an anticholinergic medication is another option, either on a short-term basis, until a change in dose or a change in medication can occur, or on a longer-term basis, if a change in dose or change in medication is not feasible. In most circumstances, an anticholinergic medication will be started only after parkinsonian symptoms are apparent. However, some individuals may be at increased risk of developing parkinsonism (e.g., those with significant parkinsonism with prior treatment), and prophylactic use of an anticholinergic medication may occasionally be warranted.

Typically, a medication such as benztropine or trihexyphenidyl is used to treat medication-induced parkinsonism because diphenhydramine has a shorter half-life and greater likelihood of sedation. However, oral or intramuscular diphenhydramine can also be used on an acute basis. Additional details on these medications are provided in Table 10. It should also be noted that different symptoms of parkinsonism (e.g., rigidity, tremors, akinesia) may have a differential response to anticholinergic medications, and different treatment approaches may be needed to address each of these symptoms. Parkinsonism should also be distinguished from tardive dyskinesia, which can be worsened by use of anticholinergic medications (Bergman and Soares-Weiser 2018; Cogentin 2013).

If an anticholinergic medication is used, it is important to adjust the medication to the lowest dose that is able to treat the parkinsonian symptoms. In addition, it is also important to use the medication for the shortest time necessary. After several weeks to months, anticholinergic medications can sometimes be reduced or withdrawn without recurrence of parkinsonism or worsening of other antipsychotic-induced neurological symptoms (Desmarais et al. 2012). Medications with anticholinergic effects can result in multiple difficulties for patients, including impaired quality of life, impaired cognition, and significant health complications (Salahudeen et al. 2015). Dry mouth due to anticholinergic effects is associated with an increased risk for multiple dental complications (Singh and Papas 2014), and drinking high-calorie fluids in response to dry mouth can contribute to weight gain. Medications with anticholinergic effects can also precipitate acute angle-closure glaucoma (Lachkar and Bouassida 2007), although patients with treated glaucoma seem to be able to tolerate these medications with careful monitoring (Bower et al. 2018).

Other peripheral side effects of anticholinergic medications can include blurred vision, constipation, tachycardia, urinary retention, and effects on thermoregulation (e.g., hyperthermia in hot weather) (Nasrallah and Tandon 2017; Ozbilen and Adams 2009), and central anticholinergic effects can include impaired learning and memory and slowed cognition (Ang et al. 2017; Vinogradov et al. 2009). Older individuals can be particularly sensitive to these anticholinergic effects and can develop problems such as urinary retention, confusion, fecal impaction, and anticholinergic toxicity (with delirium, somnolence, and hallucinations) (Nasrallah and Tandon 2017). In addition, it is important to consider the anticholinergic side effects associated with other medications that a patient is taking, such as antipsychotic medications, some antidepressant medications, urological medications (e.g., oxybutynin), and nonselective antihistamines (e.g., hydroxyzine, diphenhydramine).

Amantadine is an alternative to using an anticholinergic medication to treat medication-induced parkinsonism. Studies of amantadine have had small samples, but the available evidence and clinical experience suggest that amantadine may have comparable or somewhat less benefit than anticholinergic agents in treating medication-induced parkinsonism (Ananth et al. 1975; Borison 1983; DiMascio et al. 1976; Fann and Lake 1976; Greenblatt et al. 1977; Kelly et al. 1974; König et al. 1996; McEvoy 1987; McEvoy et al. 1987; Mindham et al. 1972; Silver et al. 1995). With the absence of anticholinergic properties, side effects, including cognitive impairment, are less prominent with amantadine than with anticholinergic agents. Common adverse effects with amantadine include nausea, dizziness, insomnia, nervousness, impaired concentration, fatigue, and livedo reticularis. Hallucinations and suicidal thoughts have also been reported, as has an increased seizure frequency in individuals with preexisting seizure disorder (Micromedex 2019).

Statements from other guidelines vary in their approach to medication-induced parkinsonism. The WFSBP guideline notes that use of SGAs or reductions in medication doses should be the primary treatment for medication-induced parkinsonism (Hasan et al. 2013). The BAP guideline notes that decisions about the use of anticholinergic medications for medication-induced parkinsonism should be made on an individual basis, but these medications should not be given prophylactically (Barnes et al. 2011). The PORT guideline notes that prophylactic use of antiparkinsonian agents is not warranted in patients treated with SGAs but may be indicated on an individual basis in patients treated with FGAs (Buchanan et al. 2010).

As a suggestion, this statement is not appropriate for use as a quality measure. It is also not appropriate for incorporation into electronic decision support.

Medication-induced acute akathisia is defined in DSM-5 as

[s]ubjective complaints of restlessness, often accompanied by observed excessive movements (e.g., fidgety movements of the legs, rocking from foot to foot, pacing, inability to sit or stand still), developing within a few weeks of starting or raising the dosage of a medication (such as a neuroleptic) or after reducing the dosage of a medication used to treat extrapyramidal symptoms. (American Psychiatric Association 2013a, p. 711)

Akathisia is sometimes difficult to distinguish from psychomotor agitation associated with psychosis, leading to a cycle of increasing doses of antipsychotic medication that lead to further increases in akathisia. Even in mild forms in which the patient is able to control most movements, akathisia is often extremely distressing to patients and is a frequent cause of nonadherence with antipsychotic treatment. If allowed to persist, akathisia can contribute to feelings of dysphoria and, in some instances, suicidal behaviors. The reported rates of akathisia vary from 10%–15% to as many as one-third of patients treated with antipsychotic medication, even when SGAs are used (Juncal-Ruiz et al. 2017; Martino et al. 2018; Mentzel et al. 2017; Miller et al. 2008).

A number of approaches can be taken when a patient is experiencing antipsychotic-induced akathisia. A reduction in the dose of the antipsychotic medication, if feasible, is often helpful in reducing akathisia. In some individuals, it may be appropriate to change the antipsychotic medication to one with a lower likelihood of akathisia (see Statement 4, Table 6). However, before reducing the dose of medication or changing to another antipsychotic medication, the benefits of reduced akathisia should be weighed against the potential for an increase in psychotic symptoms. Careful monitoring for symptom recurrence is always important when making changes or reducing doses of antipsychotic medications, and use of quantitative measures can be helpful in this regard, as described in Statement 2.

Benzodiazepine medications, including lorazepam and clonazepam, can also be helpful in the treatment of akathisia. Among other side effects, somnolence and cognitive difficulties can be associated with benzodiazepine use (Lexicomp 2019; Micromedex 2019). In addition, problems with coordination as a result of benzodiazepines can contribute to falls, particularly in older individuals (Donnelly et al. 2017). Although benzodiazepines are much safer than older sedative agents, respiratory depression can be seen with high doses of a benzodiazepine, particularly in combination with alcohol, other sedating medications, or opioids (Hirschtritt et al. 2017). Caution may also be indicated in prescribing benzodiazepines to individuals with sleep apnea, although few studies are available (Mason et al. 2015). Individuals who are treated with a benzodiazepine may also take them in higher amounts or more frequently than intended. In some patients, a sedative, hypnotic, or anxiolytic use disorder may develop, particularly in individuals with a past or current diagnosis of alcohol use disorder or another substance use disorder.

Another option for treatment of akathisia is the β-adrenergic blocking agent propranolol (Pringsheim et al. 2018), which is typically administered in divided doses, with a total daily dose of 30–120 mg. When using propranolol, it is important to monitor blood pressure with increases in dose and recognize that taking propranolol with protein-rich foods can increase bioavailability by 50%. In addition, propranolol is metabolized by CYP1A2, CYP2D6, CYP2C19, and CYP3A4, which can contribute to drug-drug interactions. Some literature also suggests that mirtazapine may reduce akathisia in some patients (Perry et al. 2018; Poyurovsky and Weizman 2018; Praharaj et al. 2015). In contrast, akathisia tends not to respond to anticholinergic agents (Pringsheim et al. 2018; Rathbone and Soares-Weiser 2006).

The WFSBP guideline notes that there is some evidence that benzodiazepines are effective in the treatment of akathisia and that there is very limited evidence to support the use of centrally active β-adrenergic blocking agents in the treatment of akathisia (Hasan et al. 2013).

As a suggestion, this statement is not appropriate for use as a quality measure. It is also not appropriate for incorporation into electronic decision support.

Tardive syndromes are persistent abnormal involuntary movement disorders caused by sustained exposure to antipsychotic medication, the most common of which are tardive dyskinesia, tardive dystonia, and tardive akathisia (Frei et al. 2018). They begin later in treatment than acute dystonia, akathisia, or medication-induced parkinsonism, and they persist and may even increase, despite reduction in dose or discontinuation of the antipsychotic medication. Typically, tardive dyskinesia presents as “[i]nvoluntary athetoid or choreiform movements (lasting at least a few weeks) generally of the tongue, lower face and jaw, and extremities (but sometimes involving the pharyngeal, diaphragmatic, or trunk muscles)” (American Psychiatric Association 2013a, p. 712), whereas tardive dystonia and tardive akathisia resemble their acute counterparts in phenomenology.

Tardive dyskinesia has been reported after exposure to any of the available antipsychotic medications (Carbon et al. 2017, 2018). It occurs at a rate of approximately 4%–8% per year in adult patients treated with FGAs (Carbon et al. 2018; Woods et al. 2010), a risk that appears to be at least three times that observed with SGAs (Carbon et al. 2018; O’Brien 2016; Woods et al. 2010). Various factors are associated with greater vulnerability to tardive dyskinesia, including age greater than 55 years; female sex; white or African race/ethnicity; presence of a mood disorder, intellectual disability, or central nervous system injury; and past or current akathisia, clinically significant parkinsonism, or acute dystonic reactions (Patterson-Lomba et al. 2019; Solmi et al. 2018a).

Evaluation for the presence of tardive syndromes is important in order to identify them, minimize worsening, and institute clinically indicated treatment. However, evaluation of the risk of tardive dyskinesia is complicated by the fact that dyskinetic movements may be observed with a reduction in antipsychotic medication dose, which is termed a withdrawal-emergent dyskinesia (American Psychiatric Association 2013a). Fluctuations in symptoms are also common and may be influenced by such factors as psychosocial stressors. Furthermore, spontaneous dyskinesias, which are clinically indistinguishable from tardive dyskinesia, have been described in elderly patients before the advent of antipsychotic medications and in up to 20% of never-medicated patients with chronic schizophrenia (Blanchet et al. 2004; Crow et al. 1982; Fenton et al. 1997; Saltz et al. 1991).

Regular assessment of patients for tardive syndromes through clinical examination or through the use of a structured evaluative tool can aid in identifying tardive syndromes, clarifying their likely etiology, monitoring their longitudinal course, and determining the effects of medication changes or treatments for tardive dyskinesia (see Statement 1, Table 2). The AIMS and the DISCUS are examples of such tools (Guy 1976; Kalachnik and Sprague 1993; Munetz and Benjamin 1988). When using scales such as the AIMS or the DISCUS, it should be noted that there is no specific score threshold that suggests a need for intervention, although ranges of scores are noted to correspond with mild, moderate, and severe symptoms. In addition, the same total score can be associated with significantly different clinical manifestations and varying impacts on the patient. Patients, family members, and other persons of support may be able to provide information about the onset of movements; their longitudinal course in relation to treatment or other precipitants; and their impact on functioning, health status (including dentition), and quality of life.

Although the majority of patients who develop tardive dyskinesia have mild symptoms, a small proportion will develop symptoms of moderate or severe degree. In such circumstances, assessment for other contributors to a movement disorder is also warranted (Jinnah and Factor 2015; Mehta et al. 2015; Poewe and Djamshidian-Tehrani 2015; Preskorn et al. 2015; Waln and Jankovic 2015). In addition to a neurological examination and complete history of motor symptoms and past and current medications, history and laboratory testing may include liver function tests, thyroid function tests, serum calcium, complete blood count, and antiphospholipid antibodies. Depending on the results of the history and evaluation, additional studies may be indicated (e.g., ceruloplasmin for Wilson’s disease; brain MRI for basal ganglia changes with Huntington’s disease, stroke, or other lesions; lumbar puncture for anti-NMDA receptor encephalitis). If dyskinetic movements have begun or have increased in the context of antipsychotic dose reduction, it is important to assess the longitudinal course of symptoms for up to several months because spontaneous reductions or resolution of the dyskinesia may occur.

If no contributing etiology is identified and moderate to severe or disabling tardive dyskinesia persists, treatment is recommended with a reversible inhibitor of the VMAT2. Treatment with a VMAT2 inhibitor can also be considered for patients with mild tardive dyskinesia on the basis of such factors as patient preference, associated impairment, or effect on psychosocial functioning. Table 11 shows the characteristics of VMAT2 inhibitors that are currently available in the United States.

In general, deutetrabenazine or valbenazine is preferred over tetrabenazine because of the greater evidence base supporting their use. In addition, tetrabenazine has a shorter half-life and greater rates of associated depression when used in the treatment of patients with Huntington’s disease. Other factors that may influence choice of a VMAT2 inhibitor relate to hepatic or renal function; tetrabenazine and deutetrabenazine are contraindicated in individuals with hepatic impairment, whereas valbenazine is not recommended for use in individuals with severe renal impairment. The metabolism of these medications is also somewhat different. Although all of these medications are substrates for CYP2D6 and CYP3A4, tetrabenazine and deutetrabenazine are major substrates for CYP2D6, whereas valbenazine is a major substrate for CYP3A4. Consequently, the patient’s CYP2D6 metabolizer status or use of concomitant medications that influence these metabolic enzymes may affect the choice of a VMAT2 inhibitor. In terms of side effects, these medications are generally well tolerated, with sedation being most common. In initial studies of tetrabenazine in patients with Huntington’s disease, significant rates of depression were noted as well as concerns about suicidal ideas and behaviors (Shen et al. 2013). However, in studies of deutetrabenazine and valbenazine in patients with tardive dyskinesia, there were no apparent increases in depression or suicidal ideas either in the randomized portions of the clinical trials or in longer open-label extension periods (Solmi et al. 2018b). Nevertheless, depression or suicidal ideas could occur during treatment for tardive dyskinesia, and clinicians will want to be alert to this possibility.

Small clinical trials and case series have examined other treatments for tardive dyskinesia. A lower dose of antipsychotic medication can be considered, although evidence for this approach is minimal (Bergman et al. 2017), and the potential for benefit needs to be weighed against the possibility of recurrent symptoms or relapse. Some benefits have been noted with benzodiazepines (Bergman et al. 2018a), although the potential for benefits must be weighed against the potential side effects of these medications, including somnolence, cognitive difficulties, problems with coordination, and risk of misuse or development of a sedative use disorder. In high doses and particularly in combination with alcohol, other sedating medications, or opioids, respiratory depression may occur. A change in antipsychotic therapy to a lower-potency medication (particularly clozapine) may also be associated with a reduction in tardive dyskinesia, particularly for individuals with moderate to severe symptoms (Mentzel et al. 2018). Again, however, the potential benefits of changing medication should be considered in light of the possibility of symptom recurrence.

In general, giving a higher dose of an antipsychotic may suppress movements of tardive dyskinesia in the short term but would be expected to escalate further development of tardive dyskinesia in the long term. Nevertheless, there may be life-threatening circumstances (e.g., patients with constant movement, gagging, or choking) in which rapid suppression of dyskinesia is needed, and judicious use of an antipsychotic may be appropriate. Anticholinergic medications do not improve and may even worsen tardive dyskinesia (Bergman and Soares-Weiser 2018; Cogentin 2013) in addition to producing significant side effects.

For individuals with other tardive syndromes, other approaches may be helpful on an individual basis. For example, depending on the muscle group that is affected, injections of botulinum toxin have been used to treat tardive dystonia (Brashear et al. 1998; Jinnah and Factor 2015; Kiriakakis et al. 1998). In addition, tardive dystonia may respond to β-adrenergic blocking agents (Hatcher-Martin et al. 2016), and in rare cases of severe intractable tardive dystonia, deep brain stimulation might be considered (Paschen and Deuschl 2018). High doses of anticholinergic agents have also been used to treat severe tardive dystonia (Burke et al. 1982; Kang et al. 1986; Wojcik et al. 1991), although these medications are not useful in treating tardive dyskinesia (Bergman and Soares-Weiser 2018; Cogentin 2013). Reserpine, which also depletes monoamines, should not be used to treat tardive syndromes because it has high rates of associated depression and suicidal ideas and lowers blood pressure (Micromedex 2019). Other treatments, such as vitamin B₆ or vitamin E, are less likely to be associated with harms but do not appear to be associated with benefits in treating tardive dyskinesia (Adelufosi et al. 2015; Soares-Weiser et al. 2018a).

The WFSBP guideline notes that tetrabenazine might have positive effects on tardive dyskinesia (Hasan et al. 2013). It also notes that the risk of tardive dyskinesia is less with SGAs than with FGAs and that there is limited evidence of benefit with clozapine for tardive dyskinesia. Information on a range of other treatments is noted to be even less conclusive. The practice guideline of the American Academy of Neurology, which was published before the availability of deutetrabenazine and valbenazine, notes that tetrabenazine might be considered as a treatment for tardive syndromes (Bhidayasiri et al. 2013).

If a quality measure on VMAT2 inhibitor treatment of tardive syndromes is considered at the provider, facility, health plan, integrated delivery system, or population level, testing of feasibility, usability, reliability, and validity would be essential prior to use for purposes of accountability. However, it may be possible and preferable to incorporate this recommendation into internal facility or health plan initiatives focused on enhanced identification and treatment of tardive syndromes.

Electronic decision support using passive alerts may be able to prompt clinicians to consider a VMAT2 inhibitor, although such prompts would depend on accurate and consistent entry of structured information about the presence of a tardive syndrome, its severity, and its associated degree of disability. Nevertheless, in combination with rating scale data (e.g., AIMS), electronic decision support could help identify individuals with a tardive syndrome who may benefit from a trial of a VMAT2 inhibitor. Information from laboratory data, diagnoses, or problem lists would also be helpful to incorporate in terms of potential contraindications to VMAT2 inhibitor treatment (e.g., tetrabenazine and deutetrabenazine are contraindicated in the presence of hepatic impairment; valbenazine is not recommended for use in individuals with severe renal impairment).

For individuals with a first episode of psychosis, coordinated specialty care (CSC) programs have been developed that integrate a number of evidence-based interventions into a comprehensive treatment package. For example, the NAVIGATE program, which was developed for the Recovery After an Initial Schizophrenia Episode (RAISE) Early Treatment research initiative, uses a collaborative, shared decision-making approach that incorporates family involvement and education, individual resiliency training, supported employment and education, and individualized medication treatment (Mueser et al. 2015; National Institute of Mental Health 2019a). Similar CSC programs, which are sometimes referred to as team-based, multicomponent interventions, have also been used in other countries for treatment of early psychosis (Anderson et al. 2018; Craig et al. 2004; Secher et al. 2015). These treatment programs often include individuals with diagnoses other than schizophrenia but have been associated with a number of benefits, including lower mortality (Anderson et al. 2018), lower rates of relapse, better quality of life, better global function, and greater likelihood of working or being in school after receiving up to 2 years of treatment (McDonagh et al. 2017). Patients in such programs may also experience a greater sense of empowerment and support for their autonomy (Browne et al. 2017). Programs are also available that are aimed at early identification and treatment of attenuated psychosis syndrome or related syndromes of high psychosis risk (J. Addington et al. 2017; Cotton et al. 2016); however, these programs are not within the scope of this guideline recommendation because they include individuals who do not have a psychiatric diagnosis or who have diagnoses other than schizophrenia at later follow-up times (Fusar-Poli et al. 2016; Iorfino et al. 2019).

The main barriers to implementing this recommendation in practice relate to the limited availability of first-episode, multicomponent treatment programs. For state health agencies, health systems, or organizations that are implementing these programs, barriers include such issues as funding, training, and implementation support. However, consultation and implementation materials are available to help guide the establishment of new programs with evidence-based approaches (National Institute of Mental Health 2019b; NAVIGATE 2019; OnTrackNY 2019). Tools are also available to assess fidelity of CSC programs to intended implementation principles (Addington et al. 2016, 2018; Durbin et al. 2019; Essock and Addington 2018).

Other guidelines did not specifically address the use of CSC programs, but they do endorse many of the individual elements of such programs (e.g., family engagement, psychoeducation, supported employment, medication treatment).

More patients may benefit from CSC programs in the United States than currently receive it. Consequently, state mental health agencies, health plans, and health organizations may wish to implement initiatives to increase the use of a CSC program among individuals with a first episode of psychosis.

This guideline statement would not be appropriate for a performance-based quality measure unless it were tested for feasibility, usability, reliability, and validity. Factors such as geographic variations in treatment availability would need to be considered in testing any quality measures related to CSC program use. It may also be difficult to determine whether a patient is receiving appropriate CSC services.

Electronic decision support using passive alerts may be able to prompt clinicians to consider referral to a CSC program if the presence of a first episode of schizophrenia and the patient’s history of prior treatment were accurately and consistently entered into the electronic record as structured data. The electronic record could also incorporate reference information on the location and referral processes for CSC treatment programs in the local area.

The use of cognitive-behavioral therapy (CBT) for individuals with schizophrenia has a number of potential benefits, including improvements in quality of life and global, social, and occupational function and reductions in core symptoms of illness, such as positive symptoms. However, it is important to appreciate that these benefits have been found in studies of CBT that is adapted to use for individuals with psychosis (CBTp), which has some differences from CBT that is focused on other indications. More specifically, CBTp focuses on guiding patients to develop their own alternative explanations for maladaptive cognitive assumptions that are healthier and realistic and do not perpetuate the patient’s convictions regarding the veracity of delusional beliefs or hallucinatory experiences. Thus, the overall approach with CBTp includes developing a collaborative and nonjudgmental therapeutic relationship in which patients can learn to monitor relationships between thoughts, feelings, behaviors, and symptoms and to evaluate the perceptions, beliefs, and thought processes that contribute to symptoms (Beck and Rector 2005; Beck et al. 2009; Beck Institute 2019; Hardy 2019; Kingdon and Turkington 2019; Landa 2019; Lecomte et al. 2016; Morrison 2017; Turkington et al. 2006; Wright et al. 2009). (Videos that demonstrate some of the approaches to CBTp are available at I Can Feel Better, www.icanfeelbetter.org/cbtpskills.) Through this dual focus on monitoring and evaluation, patients can develop beneficial coping strategies and improve functioning, with behavioral self-monitoring serving as a basis for graded task assignments or activity scheduling. In addition, symptoms can be discussed as being within a range of normal experiences (e.g., hearing a loved one’s voice in the context of grief), and alternative explanations for symptoms can be developed that help reduce associated stress (Turkington et al. 2006).

CBTp can be started in any treatment setting, including inpatient settings, and during any phase of illness (Turkington et al. 2006), although some initial reduction in symptoms may be needed for optimal participation (Burns et al. 2014; Valmaggia et al. 2008). It can also be conducted in group as well as in individual formats, either in person or via Web-based delivery platforms. CBTp can also be made available to family members or other persons of support (Turkington and Spencer 2019). Although patient preferences and treatment availability may influence choice of a delivery method, there do not appear to be clear-cut differences in the treatment benefits of group as compared with individual CBTp (McDonagh et al. 2017; Wykes et al. 2008).

The duration of treatment with CBTp has varied in research and clinical practice, with a range from 8 weeks to 5 years of treatment reported in the literature (McDonagh et al. 2017). However, guidelines from other countries recommend a minimum treatment duration of 16 sessions of CBTp (National Institute for Health and Care Excellence 2014; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013). Although the available research suggests that treatment benefits are no longer significant when assessed more than 6 months after the end of a CBTp course (McDonagh et al. 2017), it is unclear whether longer durations of treatment with CBTp will result in greater benefits or will help in maintaining treatment-related improvements.

Issues with implementation of CBTp have also been examined. Although the methodological rigor of most studies has been low (Ince et al. 2016), common barriers to CBTp have been identified. For example, some individuals with schizophrenia may be too symptomatic or are experiencing too many side effects (e.g., sedation) to allow effective participation, particularly in inpatient settings. From a patient-centered perspective, CBTp was sometimes viewed as more emotionally challenging and requiring more effort (e.g., homework) than other psychological therapies (Wood et al. 2015). However, engagement strategies, such as motivational interviewing, can be useful in helping patients explore whether CBTp might have potential benefits for them that would offset such concerns.

Attitudinal barriers of staff and organizational management were also found to be common and included a lack of understanding of CBTp and negative expectancies about its value. In addition to inadequate availability of trained staff, staff reported difficulty in identifying patients who were most likely to benefit from CBTp as well as a lack of dedicated time to provide CBTp. Insufficient initial training and insufficient reinforcement of training were also common. Thus, for CBTp to be effective, individuals who are providing CBTp should have appropriate training using established approaches, supervision in CBTp techniques, and experience in treating individuals with schizophrenia. In addition, concerted efforts may be needed to foster positive attitudes and assure adequate time to deliver CBTp. At organizational or health system levels, attention to enhancing the availability of CBTp is also important given the limited availability of CBTp in the United States. Stepped-care approaches, learning collaborative models, and other approaches to best-practice consultations show promise as ways to enhance delivery of CBTp in community settings (Creed et al. 2014; Kopelovich et al. 2019a, 2019b; Stirman et al. 2010).

Statements from other practice guidelines are consistent with this recommendation. The RANZCP and NICE guidelines recommend the use of CBTp for all individuals with schizophrenia (Galletly et al. 2016; National Institute for Health and Care Excellence 2014), whereas the SIGN, CSG, and PORT guidelines recommend CBTp for individuals who have persistent symptoms despite treatment with an antipsychotic medication (Dixon et al. 2010; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013). The NICE, SIGN, and CSG guidelines note that CBTp should include at least 16 planned sessions (National Institute for Health and Care Excellence 2014; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013).

This guideline statement may not be appropriate for a performance-based quality measure because of the impact of logistical barriers to CBTp, including geographic variations in availability of CBTp, and the difficulty in identifying whether delivered psychotherapy is CBTp. Reminders about CBTp are also not well suited to incorporation into electronic health record clinical decision support. Anecdotal observations suggest that use of CBTp is infrequent in the United States (Kopelovich et al. 2019b). Consequently, health organizations and health plans may wish to implement programs to increase the use of CBTp among individuals with schizophrenia.

Elements of psychoeducation are an integral part of good clinical practice. For example, APA’s Practice Guidelines for the Psychiatric Evaluation of Adults emphasizes the importance of involving patients in treatment-related decision-making and recommends providing the patient with education about the differential diagnosis, risks of untreated illness, treatment options, and benefits and risks of treatment (American Psychiatric Association 2016a). In addition, these informal approaches to psychoeducation have been expanded into formal, systematically delivered programs of psychoeducation that have been evaluated through clinical trials (Pekkala and Merinder 2002; Xia et al. 2011).

The psychoeducational programs that have been studied have varied in their format, duration, and scope. Some psychoeducational programs are delivered on an individual basis, whereas others are delivered in a group format, often in conjunction with family members or other individuals who are involved in the patient’s life. In clinical trials, a 12-session program of psychoeducation is the norm; however, briefer psychoeducation programs of 10 sessions or fewer have also been studied (Pekkala and Merinder 2002; Xia et al. 2011). Typically, psychoeducation is conducted on an outpatient basis, but elements of formal psychoeducation programs can also be incorporated into care in inpatient settings.

Information that is commonly conveyed in a psychoeducation program includes key information about diagnosis, symptoms, psychosocial interventions, medications, and side effects as well as information about stress and coping, crisis plans, early warning signs, and suicide and relapse prevention (Bäuml et al. 2006). Teaching of illness management or self-management strategies (Substance Abuse and Mental Health Services Administration 2010a), as discussed in Statement 21, is often incorporated into psychoeducation. In addition to conveying empathy and respect for the individual, psychoeducation is delivered in a manner that aims to stimulate hope, reassurance, resilience, and empowerment. Typically, psychoeducation incorporates multiple educational modalities, such as workbooks (McCrary et al. 2019), pamphlets, videos, and individual or group discussions in achieving the goals of psychoeducation. Information that may be useful to patients and families as a part of psychoeducation is available through SMI Adviser (https://smiadviser.org). Barriers to providing psychoeducation as a part of the treatment plan relate primarily to program availability. Online delivery of psychoeducation may be one approach to enhancing availability.

Guidelines from other organizations, including CSG, RANZCP, and SIGN, note the value of psychoeducation for individuals with schizophrenia, including information about diagnosis (Galletly et al. 2016; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013).

This guideline statement may not be appropriate for a performance-based quality measure because of the diversity of psychoeducational approaches and services and uncertainty regarding linking specific patient needs for psychoeducation with markers of delivery of psychoeducation. Reminders about psychoeducation are also not well suited to incorporation into electronic health record clinical decision support. However, health organizations and health plans may wish to implement quality improvement efforts to increase the use of formal psychoeducational programs among individuals with schizophrenia.

Supported employment differs from other vocational rehabilitation services in providing assistance in searching for and maintaining competitive employment concurrently with job training, embedded job support, and mental health treatment (Becker and Drake 2003; Frederick and VanderWeele 2019; Substance Abuse and Mental Health Services Administration 2010b). In contrast, other vocational rehabilitation approaches focus on training before placement and put greater emphasis on placement in sheltered and transitional employment rather than in a competitive employment setting (Marino and Dixon 2014). For individuals whose goals are related to educational advancement prior to pursuit of employment, supported educational services may also be pursued (Substance Abuse and Mental Health Services Administration 2012b).

Of approaches to supported employment, the bulk of studies involve individual placement and support (IPS; Becker and Drake 2003; Frederick and VanderWeele 2019; McDonagh et al. 2017; Substance Abuse and Mental Health Services Administration 2010b). In addition to a focus on rapid attainment of competitive employment, IPS emphasizes patient preferences in the types of jobs sought, the nature of the services that are delivered, and the outreach that occurs with potential employers (Marino and Dixon 2014). Patient preferences also guide whether to disclose the presence of a psychiatric illness to the employer, with more than half of individuals choosing to disclose this information (DeTore et al. 2019). Services are offered to anyone who is interested, with no exclusion criteria for participation. Additional principles of IPS include individualized long-term job support and integration of employment specialists with the clinical team. Employment specialists also develop relationships with community employers and provide personalized benefits counseling to participants.

Evidence consistently shows that supported employment is associated with greater rates of competitive employment than transitional employment or prevocational training, although prevocational training is superior to no vocational intervention at all (Marshall et al. 2014; McDonagh et al. 2017; Metcalfe et al. 2018; Modini et al. 2016; Richter and Hoffmann 2019; Suijkerbuijk et al. 2017). Augmenting supported employment with symptom-related skills training, training in workplace fundamentals, or cognitive training may assist in gaining and maintaining competitive employment (Dewa et al. 2018; Suijkerbuijk et al. 2017). Other benefits of supported employment include greater number of hours worked per week, a longer duration of each job, a longer duration of total employment, and an increase in earnings (McDonagh et al. 2017). Individuals receiving supported employment are also more likely to obtain job-related accommodations than individuals with mental illness who are not receiving supported employment (McDowell and Fossey 2015). Such accommodations typically relate to assistance from the supported employment coach but may also include flexible scheduling, reduced hours, modified job duties, and modified training and supervision. In addition to job-related benefits of supported employment, there is some evidence of modest reductions in symptoms and a reduced risk of hospitalization associated with obtaining a job (Bouwmans et al. 2015; Burns et al. 2007; Charzyńska et al. 2015; Hoffmann et al. 2014; Luciano et al. 2014).

Among individuals who receive supported employment, factors that may be associated with a greater likelihood of success include lower levels of symptoms, higher levels of cognitive functioning (e.g., attention, memory, executive functioning, psychomotor speed), greater work success in the past, higher levels of educational attainment, and greater interest in obtaining employment (Kirsh 2016). Peer support and support from families and others in the patient’s social network may also be associated with better outcomes, although these factors have been less well studied (Kirsh 2016). Even when individuals do not experience initial success with supported employment, addition of cognitive remediation may improve vocational outcomes (McGurk et al. 2015, 2016).

There are a number of barriers to supported employment, including economic and regulatory factors (Kirsh 2016; Metcalfe et al. 2018; Modini et al. 2016) and the limited number of available programs (Marshall et al. 2014; Sherman et al. 2017). Although data are limited, employers may be reluctant to participate in supported employment out of concern about the impact of providing work-related accommodations and because of discrimination and bias toward individuals with serious mental illness (Kirsh 2016). Treating clinicians may also serve as a barrier by having inappropriately limited expectations (Kirsh 2016) and being unaware that some individuals with schizophrenia are able to function at high levels of occupational achievement (Cohen et al. 2017). In addition, concerns about losing disability benefits or health insurance may lead some individuals to forgo supported employment opportunities (Kirsh 2016). Such concerns are not entirely unrealistic because many of the competitive jobs that individuals do obtain are entry-level and/or part-time positions without health insurance benefits (Kirsh 2016). Within supported employment programs, organizational barriers to success have included poor fidelity to supported employment principles (Marshall et al. 2014); insufficient time devoted to leading and management of the programs; and insufficient training, skills, and business and public relations knowledge of program staff (Kirsh 2016; Swanson et al. 2013). Each of these barriers is important to address at individual, systems, and policy levels so that more patients can benefit from supported employment interventions.

For clinicians and organizations wishing to learn more about supported employment or develop supported employment programs, additional information is available through SMI Adviser (https://smiadviser.org), NAVIGATE (https://navigateconsultants.org/manuals), and the Boston University Center for Psychiatric Rehabilitation (https://cpr.bu.edu).

Guidelines from other organizations are generally consistent with this recommendation. NICE and PORT recommend that supported employment be offered to individuals with schizophrenia who wish to find or return to work (Dixon et al. 2010; National Institute for Health and Care Excellence 2014), and RANZCP recommends IPS services for individuals with first-episode psychosis (Galletly et al. 2016). RANZCP, NICE, and CSG also emphasize the appropriateness of other occupational or educational activities for individuals with schizophrenia (Galletly et al. 2016; National Institute for Health and Care Excellence 2014; Norman et al. 2017).

This guideline statement may not be appropriate for a performance-based quality measure because of the barriers to supported employment, including variations in availability, and difficulty identifying when patients desire competitive employment. Reminders about supported employment are also not well suited to incorporation into electronic health record clinical decision support. However, given the infrequent availability of supported employment in the United States, health organizations and health plans may wish to implement programs to increase the use of supported employment among individuals with schizophrenia.

ACT, sometimes referred to as programs of assertive community treatment, is a multidisciplinary, team-based approach in which patients receive individualized care outside a formal clinical setting. Thus, individuals may be engaged in their homes, workplaces, or other community locations. Continuity of care is enhanced because individuals work with an assigned team, which has 24/7 availability, rather than being assigned to a designated clinician for care. Team members typically include a psychiatrist, nurse, and social worker or case manager. Peer specialists, vocational specialists, and clinicians with expertise in substance use treatment are often part of the team as well.

Other features of ACT include its provision of personalized and flexible care that addresses the patient’s needs and preferences without time limits or other constraints on services. Particularly in rural areas, some ACT teams are augmenting face-to-face visits with telepsychiatry visits, although research will be needed to determine whether such an approach alters the benefits of ACT (Swanson and Trestman 2018). ACT teams also work with a smaller number of individuals than traditional outpatient clinicians or case managers do, which contributes to the ability to provide frequent visits and a more personalized and comprehensive approach to care. For these reasons, ACT is often used in individuals who have a history of poor engagement with services that leads to frequent relapse or social disruption (e.g., homelessness; legal difficulties, including imprisonment), although it can also be used for individuals who are better engaged but still have high rates of service utilization. Many individuals who are referred for ACT are also at risk for poor adherence and may benefit from consideration of an LAI antipsychotic medication (see Statement 10).

Studies of ACT suggest that it is associated with symptom improvement comparable to other treatment delivery approaches and that individuals who receive ACT are more likely to be domiciled, living independently, and working and less likely to be hospitalized as compared with individuals who receive treatment as usual (McDonagh et al. 2017). Although ACT has multiple strengths that would make it an attractive approach in individuals with co-occurring disorders and schizophrenia, the impact of ACT on physical health has not been well studied (Vanderlip et al. 2017). Also, in individuals with a concomitant substance use disorder, research to date has not shown associated improvements in functioning, mortality, or substance use as compared with usual care (McDonagh et al. 2017).

In terms of implementation barriers, there is often limited availability of ACT programs. Funding these programs can be challenging because the comprehensive and multidisciplinary nature of ACT services is not well aligned with payment models in the U.S. health care delivery system (Monroe-DeVita et al. 2012). Effective delivery of ACT services is also dependent on having a high level of fidelity to ACT program standards (Monroe-DeVita et al. 2012; Thorning et al. 2016), and this requires considerable training as well as ongoing mentoring, collaboration, and consultation with individuals who are skilled in ACT implementation. Attention to outcomes and organizational culture is also important in providing a team-based approach that is warm, flexible, pragmatic, collaborative, and supportive of patients’ recovery (Monroe-DeVita et al. 2012). For organizations or state mental health systems that are implementing ACT programs, a number of resources are available (Center for Evidence-Based Practices 2019; Substance Abuse and Mental Health Services Administration 2008; Thorning et al. 2016).

This guideline recommendation is consistent with the SIGN recommendation to offer assertive outreach to individuals with schizophrenia who “make high use of inpatient services, who show residual psychotic symptoms and who have a history of poor engagement with services leading to frequent relapse and/or social breakdown (for example homelessness)” (Scottish Intercollegiate Guidelines Network 2013, p. 8). The PORT guideline also notes that ACT should be included in systems of care that serve individuals with schizophrenia and that it “should be provided to individuals who are at risk for repeated hospitalizations or have recent homelessness” (Dixon et al. 2010, p. 49). In addition, RANZCP recommends the use of ACT “after initial contact, during crises and after discharge from hospital” in individuals with schizophrenia (Galletly et al. 2016, p. 29).

This guideline statement may not be appropriate for a performance-based quality measure because of the impact of logistical barriers to ACT, including geographic variations in availability. Reminders about ACT are also not well suited to incorporation into electronic health record clinical decision support because of the multiple patient-specific factors that may contribute to a decision to recommend ACT. However, anecdotal observations suggest that more patients may benefit from ACT in the United States than currently receive it. Consequently, state mental health agencies, health plans, and health organizations may wish to implement programs to increase the use of ACT among individuals with schizophrenia who have had a history of poor engagement with services, leading to frequent relapse or social disruption (e.g., homelessness; legal difficulties, including imprisonment).

An important aspect of good psychiatric treatment is involvement of family members, person(s) of support, and other individuals who play a key role in the patient’s life. In addition to spouses, parents, children, or other biological or nonbiological relatives, such individuals may include people who reside with the patient, intimate partners, and close friends who are an integral part of the patient’s support network. Such individuals benefit from discussion of such topics as diagnosis and management of schizophrenia, types of support that are available, and ways to plan for and access help in a crisis. Other goals include helping individuals repair or strengthen their connections with family members and other members of their support system.

Family interventions are systematically delivered, extend beyond conveying of information, and focus on the future rather than on past events (Mueser et al. 2013). The family interventions that are suggested in this guideline statement go beyond the basics of family involvement and illness education that are important for good clinical care. They may include structured approaches to problem-solving, training in how to cope with illness symptoms, assistance with improving family communication, provision of emotional support, and strategies for reducing stress and enhancing social support networks (McDonagh et al. 2017; McFarlane 2016; Mueser et al. 2013). Family interventions can be particularly important early in the course of schizophrenia (McFarlane 2016) but can also be helpful during any phase of treatment.

Most patients want family to be involved in their treatment (Cohen et al. 2013). Nevertheless, even when a patient does not wish for a specific person to be involved in his or her care, the clinician may listen to information provided by that individual, as long as confidential information is not provided to the informant (American Psychiatric Association 2016a). Although some health professionals may be unsure about legal or regulatory aspects of sharing information, general information that is not specific to the patient can be provided (e.g., common approaches to treatment, general information about medications and their side effects, available support and emergency assistance). Also, to prevent or lessen a serious and imminent threat to the health or safety of the patient or others, the “Principles of Medical Ethics” (American Psychiatric Association 2013f) and HIPAA (Office for Civil Rights 2017a, 2017b) permit clinicians to disclose necessary information about a patient to family members, caregivers, law enforcement, or other persons involved with the patient. HIPAA also permits health care providers to disclose necessary information to the patient’s family, friends, or other persons involved in the patient’s care or payment for care when such disclosure is judged to be in the best interests of the patient and the patient is not present or is unable to agree or object to a disclosure because of incapacity or emergency circumstances (Office for Civil Rights 2017b).

The family interventions that have been studied include a variety of formats and approaches (McDonagh et al. 2017; McFarlane 2016). Interventions may or may not include the patient and can be conducted with a single family or a multifamily group. In many early studies, family interventions included the patient and were led by a member of the patient’s clinical care team. This approach allowed a liaison to develop among the care team, the patient, members of the family, and other person(s) of support. Other studies have been conducted independent of the patient’s care team. In terms of approach, some family interventions focus on psychoeducation, whereas other interventions incorporate other treatment elements (e.g., motivational interviewing, goal setting, cognitive-behavioral intervention, behavioral family therapy, support groups, social network development, communication training, role-playing, stress management, relaxation training). Given the diversity of options for family interventions, the selection of a specific approach should consider the preferences of the patient and family in collaboration with the clinician.

Benefits of family interventions include reductions in core symptoms of illness and reductions in relapses, including rehospitalization (McDonagh et al. 2017). Some studies have also shown benefits for family members such as reductions in levels of burden and distress or improvements in relationships among family members (McFarlane 2016; Sin et al. 2017a). Evidence suggests that benefits of family interventions are greatest when more than 10 treatment sessions are delivered over a period of at least 7 months (McDonagh et al. 2017). However, the Family-to-Family intervention available through the National Alliance on Mental Illness has shown significant benefits using a 12-week program consisting of weekly sessions of 2–3 hours each (Dixon et al. 2011; Lucksted et al. 2013; Marcus et al. 2013; Toohey et al. 2016).

A common barrier to implementing family interventions relates to program availability. However, guidance is available on developing family intervention programs focused on psychoeducation (Glynn et al. 2014; Substance Abuse and Mental Health Services Administration 2009). In addition, the National Alliance on Mental Illness has reduced this barrier through its Family-to-Family program, which has led to a significant expansion in the availability of family interventions (National Alliance on Mental Illness 2019b). Additional barriers include constraints of family members (e.g., work schedules, access to transportation, childcare, health issues) that may limit their ability to be involved in frequent family sessions. Similar logistical barriers can exist for patients when family interventions incorporate patient participation. Other implementation barriers include organizational and clinician-focused barriers such as time and cost constraints and insufficient understanding of the potential benefits of family intervention (Ince et al. 2016).

This guideline statement is consistent with guidelines from other organizations. CSG, NICE, RANZCP, SIGN, and PORT guidelines all recommend offering family interventions when an individual with schizophrenia reside with or are in close contact with family (Dixon et al. 2010; Galletly et al. 2016; National Institute for Health and Care Excellence 2014; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013). These guidelines also emphasize the importance of providing information to family and others involved in the patient’s care on such topics as diagnosis and management of schizophrenia, types of support that are available, and ways to access help in a crisis.

As a suggestion, this guideline statement is not appropriate for use as a performance-based quality measure or for incorporation into electronic decision support. Nevertheless, given the potential benefits of this approach, health care organizations and health plans may wish to track the availability and utilization of family interventions.

Illness self-management training programs have been applied to help address many chronic conditions and are designed to improve knowledge about one’s illness and management of symptoms (Grady and Gough 2014). Goals include reducing the risk of relapse, recognizing signs of relapse, developing a relapse prevention plan, and enhancing coping skills to address persistent symptoms, with the aim of improving quality of life and social and occupational functioning. In the studies included in the AHRQ review (McDonagh et al. 2017), self-management training was generally delivered in a group setting with sessions of 45–90 minutes each, and the number of intervention sessions ranged from 7 to 48 sessions. However, the evidence suggested better outcomes in patients who participated in at least 10 self-management intervention sessions. Self-management sessions were typically facilitated by clinicians, although peer-facilitated sessions have also been used. In addition, some studies have used individually targeted interventions, either face-to-face or via computer-based formats (Lean et al. 2019). Self-management approaches have also been used to address co-occurring medical conditions in individuals with serious mental illness, including schizophrenia, with benefits that included increased patient activation and improved health-related quality of life (Druss et al. 2010; Goldberg et al. 2013; Muralidharan et al. 2019).

Recovery-focused interventions have also been developed that focus on fostering self-determination in relation to a patient’s personal goals, needs, and strengths. Such approaches may include elements of self-management skill development, psychoeducation, and peer-based interventions but also include components and activities that allow participants to share experiences and receive support, learn and practice strategies for success, and identify and take steps toward reaching personal goals. Studies of peer-based interventions are limited (Castelein et al. 2015; Chien et al. 2019), and studies of recovery-focused interventions have also been small in number. Nevertheless, the available information suggests that these interventions may promote increased recovery, hope, and empowerment among individuals with serious mental illnesses (Le Boutillier et al. 2011; Mueser et al. 2013; Thomas et al. 2018).

The most common barrier to implementing this guideline statement is the availability of programs for developing self-management skills and enhancing person-oriented recovery. However, a toolkit for developing illness management and recovery-based programs in mental health is available through the Substance Abuse and Mental Health Services Administration (https://store.samhsa.gov/product/Illness-Management-and-Recovery-Evidence-Based-Practices-EBP-KIT/sma09-4463). Other resources are available through the Boston University Center for Psychiatric Rehabilitation (https://cpr.bu.edu), the Center on Integrated Health Care and Self-Directed Recovery (www.center4healthandsdc.org), Digital Opportunities for Outcomes in Recovery Services (https://skills.digitalpsych.org), Mental Health America (www.mhanational.org/self-help-tools), the National Alliance on Mental Illness (www.nami.org), NAVIGATE (https://navigateconsultants.org/manuals), SMI Adviser (https://smiadviser.org/individuals-families), and the Temple University Collaborative on Community Inclusion (www.tucollaborative.org).

This guideline statement is consistent with recommendations of other guidelines (RANZCP, NICE) that support the use of self-management and peer-support programs in the treatment of individuals with schizophrenia (Galletly et al. 2016; National Institute for Health and Care Excellence 2014).

As a suggestion, this guideline statement is not appropriate for use as a performance-based quality measure or for incorporation into electronic decision support. Nevertheless, given the potential benefits of such interventions, health care organizations and health plans may wish to track the availability and utilization of programs to develop self-management skills and enhance person-oriented recovery.

Cognitive remediation approaches are intended to address cognitive difficulties that can accompany schizophrenia, with the aim of enhancing function and quality of life. A number of different cognitive remediation approaches have been used, typically in group or computer-based formats, in an effort to enhance cognitive processes such as attention, memory, executive function, social cognition, or meta-cognition (Delahunty and Morice 1996; Medalia et al. 2018; Reeder et al. 2016; Wykes et al. 2011). Some programs have focused on improving cognitive flexibility (e.g., shifting cognitive sets), working memory (e.g., sequencing, multitasking, delayed recall), and planning (e.g., sequencing and chunking, active coding), whereas meta-cognitive approaches have attempted to teach patients how and when particular strategies that bypass specific cognitive limitations can be used. Some programs add aspects of social and communication skills to neurocognitive elements of remediation (Pentaraki et al. 2017).

Although this variability in program format and content confounds interpretation of the evidence, cognitive remediation does seem to result in improvements in cognition, symptoms, and function in individuals with schizophrenia, at least on a short-term basis (Harvey et al. 2018; McDonagh et al. 2017; Revell et al. 2015). Although long-term follow-up studies of cognitive remediation are not available in individuals with schizophrenia, data from healthy older individuals show long-term improvements as a result of cognitive training (Rebok et al. 2014). Beneficial effects on psychosocial outcomes seem particularly robust when cognitive remediation is used as a component of or adjunct to other forms of psychiatric rehabilitation rather than being delivered as a stand-alone intervention (McGurk et al. 2007; Revell et al. 2015; van Duin et al. 2019; Wykes et al. 2011). However, some apparent improvements in cognitive performance may result from practicing specific tasks and may not produce generalizable changes in other contexts. Furthermore, the specific elements of a particular cognitive remediation program may influence the benefits that are observed (Cella and Wykes 2019).

The primary barriers to use of cognitive remediation are related to program availability. Online delivery of cognitive remediation may be one way to overcome these barriers. Information and training on developing cognitive remediation programs are available (Medalia 2019; Medalia et al. 2018). In addition, Web-based programs have been used in clinical trials and may provide options for patients without access to in-person cognitive remediation programs (Jahshan et al. 2019; Kukla et al. 2018).

The RANZCP guideline recommends that cognitive remediation be available to individuals with schizophrenia if cognitive impairment is present and should be specifically “offered when cognitive deficits are affecting recovery and function” (Galletly et al. 2016, p. 20). The SIGN and CSG guidelines note that cognitive remediation “may be considered for individuals diagnosed with schizophrenia who have persisting problems associated with cognitive difficulties” (Norman et al. 2017, p. 621; Scottish Intercollegiate Guidelines Network 2013, p. 28).

As a suggestion, this guideline statement is not appropriate for use as a performance-based quality measure or for incorporation into electronic decision support.

Social skills training in the treatment of schizophrenia can improve social function, core illness symptoms, and negative symptoms more than usual care can (McDonagh et al. 2017). Reductions in relapse rates, including rehospitalization rates, have also been noted in some studies (McDonagh et al. 2017).

Social skills training has an overarching goal of improving interpersonal and social skills but can be delivered using a number of approaches (Almerie et al. 2015; Kopelowicz et al. 2006; Turner et al. 2018). These include cognitive-behavioral, social-cognitive, interpersonal, and functional adaptive skills training. Social skills training is delivered in a group format and includes homework assignments to facilitate skill acquisition. Other specific elements of the intervention will vary with the theoretical emphasis of the training. However, examples of techniques that can be used in social skills training include role-playing, modeling, and feedback approaches to enhance interpersonal interactions; behaviorally oriented exercises in assertiveness, appropriate contextual responses, and verbal and nonverbal communication; and instruction and practice with social and emotional perceptions (Almerie et al. 2015; Kopelowicz et al. 2006; Turner et al. 2018). These techniques are aimed at generating improvements in typical social behaviors such as making eye contact, smiling at appropriate times, actively listening to others, and sustaining conversations. In some social skills training programs, group sessions are augmented with video or technologically based interventions, in vivo community trips to practice social skills (Glynn et al. 2002; Mueser et al. 2010), and involvement of support people who are accessible, pleasant, and knowledgeable about the local environments’ resources and limitations (Mueser et al. 2010; Tauber et al. 2000).

As with other psychosocial interventions, availability of social skills training is a common barrier to its incorporation into treatment. However, information about social skills training is available for organizations that wish to develop such programs (Bellack and Goldberg 2019; Bellack et al. 2004; Granholm et al. 2016).

Other guidelines are generally consistent with this guideline statement. The PORT, RANZCP, CSG, and SIGN guidelines suggest offering social skills training to individuals with schizophrenia who have deficits in social skills (Buchanan et al. 2010; Galletly et al. 2016; Norman et al. 2017; Scottish Intercollegiate Guidelines Network 2013). However, the NICE guideline notes that social skills training should not be routinely offered to individuals with schizophrenia (National Institute for Health and Care Excellence 2014).

As a suggestion, this guideline statement is not appropriate for use as a performance-based quality measure or for incorporation into electronic decision support. Nevertheless, given the potential benefits of such an approach for some patients, health care organizations and health plans may wish to track the availability and utilization of social skills training for individuals with schizophrenia.

Supportive psychotherapy is commonly a part of the treatment plan in individuals with schizophrenia who are not receiving other modes of psychotherapy (e.g., CBTp). Because the evidence related to its benefits is limited, supportive psychotherapy should not take precedence over other evidence-based psychosocial treatments (e.g., CSC, CBTp, psychoeducation). When compared with treatment as usual, no advantage was seen for supportive psychotherapy in terms of global or social function (L. A. Buckley et al. 2015; McDonagh et al. 2017); however, these findings are difficult to interpret given the frequent use of supportive psychotherapy techniques as part of usual care. When compared with insight-oriented psychotherapies, a small number of early studies suggested that supportive psychotherapy might be associated with better outcomes in coping skills, adherence, and relapse (Fenton 2000; Hogarty et al. 1997; Stanton et al. 1984).

The focus of supportive psychotherapy is reality based and present centered (Kates and Rockland 1994; Novalis et al. 1993; Winston 2014; Winston et al. 2012). It commonly aims to help patients cope with symptoms, improve adaptive skills, and enhance self-esteem, although descriptions of the goals of supportive psychotherapy have varied. Examples of techniques used to foster these goals include reassurance; praise; encouragement; explanation; clarification; reframing; guidance; suggestion; and use of a conversational, nonconfrontational style of communication. Many of the common elements that have been identified in effective psychotherapies, including a positive therapeutic alliance, are also integral to supportive psychotherapy (Frank and Frank 1991; Wampold 2015). Typically, supportive psychotherapy is conducted in conjunction with medication management at a frequency that can vary from weekly to every few months depending on the needs of the individual patient. Other psychosocial treatments can also be used as part of the treatment plan in conjunction with these modalities.

The NICE guideline notes that supportive psychotherapy should not be offered routinely to individuals with schizophrenia if other psychosocial treatments that have greater efficacy are available (National Institute for Health and Care Excellence 2014). However, the NICE guideline also notes that patient preferences should be taken into account, particularly if other psychosocial interventions are not available locally.

As a suggestion, this guideline statement is not appropriate for use as a performance-based quality measure or for incorporation into electronic decision support.