Evidence-Based Reviews

Medication nonadherence is a common problem when treating patients with schizophrenia that can worsen prognosis and lead to sub-optimal treatment outcomes. In this article, we discuss common reasons for nonadherence and describe evidence-based treatments intended to increase adherence and improve outcomes (Box).^1-6

Common reasons for nonadherence

The primary predictor of future nonadherence is a history of nonadherence. It is important to understand patients’ reasons for nonadherence so that practical and evidence-based solutions can be implemented into the treatment plans of individual patients.

The 2009 Expert Consensus Guidelines on Adherence Problems in Patients with Serious and Persistent Mental Illness divided variables related to nonadherence into 3 categories:

• those that lie within the patient (intrinsic)
• those that are related to the patient’s relationship with healthcare providers, family, or caregivers (extrinsic)
• those that are related to the healthcare delivery system (extrinsic).⁷

Among intrinsic variables, studies have shown a correlation between nonadherence and education level, lower socioeconomic status, homelessness, and male sex.⁷ (The Expert Consensus Guidelines considered homelessness to be an intrinsic factor because it was used as a demographic variable in the studies.)

Cognitive and negative symptoms associated with schizophrenia are an intrinsic risk factor for nonadherence because patients might not remember when or how to take medication.⁷ In a study by Freudenreich and co-workers⁸ of 81 outpatients who had a diagnosis of schizophrenia, the presence of negative symptoms predicted a negative attitude toward psychotropic medications. Poor insight might be the result of cognitive dysfunction associated with schizophrenia, and often is due to a lack of awareness of the importance of taking medications.

Limited insight into the need for treatment can be problematic early in the course of the illness when it may be directly related to positive symptoms. Perkins and colleagues⁹ demonstrated that patients recovering from a first psychotic episode who had limited insight into their illness and lacked desire to seek treatment were less adherent with medication. In another study, 5% of psychiatrists surveyed thought that many of their patients with schizophrenia were nonadherent because those patients did not believe that medications were effective or useful.¹⁰

Comorbid substance abuse disorders can contribute to medication nonadherence. In an analysis of 6,731 patients with schizophrenia, Novick and co-workers reported that alcohol dependence and substance abuse in the previous month predicted medication nonadherence.¹¹ Hunt and colleagues demonstrated that, among 99 nonadherent patients with schizophrenia, time to first readmission was shorter for patients with comorbid substance abuse disorders compared with patients who had a diagnosis of schizophrenia only. Over the 4-year study period, the 28 patients who had a dual diagnosis (schizophrenia and substance abuse) accounted for 57% of all hospital readmissions.¹²

Several variables that affect medication adherence are related to the patient’s relationship with healthcare providers, family, caregivers, and the service delivery system.⁷ These include:

• the perceived stigma of being given a diagnosis of a serious mental illness
• adverse effects related to medications
• poor social and family support
• difficulty gaining access to mental health services.^7,10

Societal stigma associated with seeking treatment from a mental health professional may contribute to nonadherence in some patients. In 1 study,¹³ 36% of people surveyed would not want to work closely with a person who has a serious mental illness.

Adverse effects contribute significantly to nonadherence

Limited treatment options (which may be expensive) can make it difficult to manage the adverse effects of antipsychotics. In a cross-sectional survey of 876 patients, investigators reported that: 1) <50% of patients were adherent with medication, and 2) 80% experienced ≥1 side effect that was reported to be “somewhat bothersome” in self-ratings (Table 1).¹⁴ Extrapyramidal symptoms (EPS) and agitation were most strongly associated with nonadherence; weight gain, akathisia, and sexual dysfunction also were associated with nonadherence.¹⁴ This study did not distinguish adverse effects associated with first-generation antipsychotics (FGAs) from those associated with second-generation antipsychotics (SGAs), even though 71.7% of patients studied were taking an SGA.

A meta-analysis by Leucht and co-workers¹⁵ compared 15 antipsychotics (the FGAs haloperidol and chlorpromazine and 13 SGAs) for efficacy and tolerability in schizophrenia. Haloperidol had the highest rate of discontinuation for any

Antipsychotic binding affinities to dopamine 2 (D2), serotonin 2A (5-HT2A), histamine (H1), and other receptors have an impact on a medication’s side-effect profile. Because of individual patient characteristics, you might be faced with choosing a medication that has a lower risk of EPS but a higher risk of weight gain and metabolic complications—or the inverse. Understanding binding affinities, side-effect profiles, and how to minimize or utilize adverse effects (ie, giving a drug that is approved to treat schizophrenia and is associated with weight gain to a patient with schizophrenia who has lost weight) may lead to greater adherence (Table 2¹⁶ and Table 3¹⁷).

Adequate support is essential

The therapeutic alliance plays a key role in patients’ attitudes toward taking medication. Magura and colleagues¹⁸ found that one-third of psychiatric patients (13% of whom had a diagnosis of schizophrenia) reported that their psychiatrist did not spend enough time with them explaining side effects, and felt “rushed.”

Patients with schizophrenia often require access to social support systems provided by family members, friends, and community agencies that provide case management and attendant care services. Patients who are adherent to medication tend to have greater perceived family involvement in medication treatment, and tend to have been raised in a family that had more of a positive attitude toward medication.¹⁹

In our practice, we have observed that recent state and federal budget cuts have resulted in patients having greater difficulty gaining access to case management and attendant care services, which then leads to increased rates of medication nonadherence. Be aware that variables such as limited office hours, financial hardship, and cultural and language barriers can compromise a patient’s ability to seek and continue care.

In the following section, we lay out techniques for improving adherence in patients with schizophrenia.

Employ general and specific strategies to boost adherence

How can you raise medication adherence concerns with patients, keeping in mind that they often overestimate their adherence?

Ask. Some clinicians ask questions such as “Are you taking your medication?”, although a more effective approach might be to ask how the patient is taking his (her) medication. Asking questions such as “When do you take your medication?” and “In the past week, how many doses do you think you missed?” might be more effective ways to inquire about adherence.⁷

The Expert Consensus Guidelines recommend asking patients about medication adherence monthly for those who are stable, doing well, and believed to be adherent. For those who are new to a practice or who are not doing well, inquire about medication adherence at least weekly.⁷

In our practice, patients who are unstable but do not require inpatient hospitalization typically are seen more often in the clinic, or are referred to intensive outpatient or partial hospitalization programs. If an unstable patient is unable to come in for more frequent appointments, we arrange phone conferences between her and her provider. If a patient is not doing well and has a case manager, we often ask that case manager to visit the patient, in person, more often than he (she) would otherwise.

Take a nonjudgmental approach when raising these issues with patients. Questions such as “We all forget to take our medication sometimes; do you?” help to normalize nonadherence, and improve the therapeutic alliance, and might result in the patient being more honest with the clinician.⁷ Because patients may be apprehensive about discussing adverse events, clinicians must be proactive about improving the therapeutic alliance and making patients feel comfortable when discussing sensitive topics. Clinicians should try to convey the idea that, although adherence is a concern, so is quality of life. A clinicians’ willingness to take a flexible approach that is nonpunitive nor authoritarian can aid the therapeutic alliance and improve overall adherence.

Be sensitive to financial, cultural, and language variables that can affect access to care. The Expert Consensus Guidelines recommend asking patients if they can afford their medication. In our practice, we have seen patients with schizophrenia discharged from the hospital only to be readmitted 1 month later because they could not afford to fill their prescriptions.

It is important to have translation services available, in person or by phone, for patients who do not speak English. Furthermore, it is important to understand the limitations that your practice might place on access to care. Ask patients if they have ever had trouble making an appointment when they needed to be seen, or if they called the office with a question and did not receive an answer in a timely fashion; doing so allows you to assess the practice’s ability to meet patients’ needs and helps you build a therapeutic alliance.

Make objective assessments. It is important for practitioners to not base their assessment of medication adherence solely on subjective findings. Asking patients to bring in their medication bottles for pill counts and checking with the patients’ pharmacies for information about refill frequency can provide some objective data. Electronic monitoring systems use microprocessors inserted into bottle caps to record the occurrence and timing of each bottle opening. Studies show that these electronic monitoring systems are the gold standard for determining medication adherence and could be used in cases where it is unclear if the patient is taking his (her) medication.^7,20 Such systems have successfully monitored medication adherence in clinical trials, but their use in clinical practice is complicated by ethical and legal considerations and cost issues.

Simplify the regimen. Using medications with once-daily dosing, for example, can help improve adherence. Pfeiffer and co-workers²¹ found that patients whose medication regimens were changed from once daily to more than once daily experienced a decrease in medication adherence. Conversely, a decrease in dosing frequency was significantly associated with improved adherence. More than once-daily dosing was only weakly associated with poorer adherence among patients already on a stable regimen.

Discussing positive and negative aspects of past medication trials with a patient and inquiring if she prefers a specific medication can be an effective way to build the therapeutic relationship and help with adherence.

Direct patients to psychosocial interventions. These can be broadly classified as:

• educational approaches
• group therapy approaches
• family interventions
• cognitive treatments
• combination approaches.

Psychoeducational approaches have limited effect on improving adherence when delivered to individual patients. However, 1 study showed that psychoeducation was effective at improving adherence when extended to include the patient’s family.²²

Motivational interviewing techniques, behavioral approaches, and family interventions are effective at increasing medication adherence. One study looked at the value of training a patient-identified informant to supervise and administer medication. This person, usually a family member or close support, was responsible for obtaining medication from the pharmacy, administering the medication, and recording adherence. After 1 year, 67% of patients who used an informant were adherent, compared with only 45% in the group that did not have informant support.²² Case managers, attendant care workers, home health nurses, and assertive community treatment (ACT) teams also can participate in this manner; it is important, therefore, for you to be aware of the resources available in your community and to understand your role as patient advocate.

Substance abuse is a strong risk factor for nonadherence among patients with schizophrenia,¹⁸ which makes it important to assess patients for substance use and encourage those who do abuse to seek treatment. Although 1 study showed no correlation between Alcoholics Anonymous (AA) attendance and medication adherence,¹² many AA and Narcotics Anonymous groups do not discuss psychiatric medications during group meetings. Magura and colleagues encouraged the use of “dual focus” groups that involve mental health professionals and addiction treatment specialists discussing mental health and substance abuse issues at the same setting.¹⁸

Prescribe long-acting injectable antipsychotics. Typically, long-acting injectable antipsychotics (LAIs) are reserved for patients who have a history of nonadherence. In a small study (N = 97) comparing LAI risperidone and oral risperidone or oral haloperidol, patients treated with an LAI had significantly fewer all-cause discontinuations (26.0%, compared with 70.2%) at 24 months.²³ The Adherence to Treatment and Therapeutic Strategies in Schizophrenic Patients study examined 1,848 patients with schizophrenia and reached similar findings regarding LAI antipsychotics.²⁴ (Note: Aripiprazole, fluphenazine, haloperidol, olanzapine, and paliperidone also are available in an LAI formulation.)

Bottom Line

Antipsychotic nonadherence in schizophrenia is a major problem for patients, families, and society. Being able to identify patients at risk for nonadherence, understanding the reasons for their nonadherence, and seeking practical solutions to the problem are all the responsibility of the treating physician. Psychoeducation, addressing substance abuse, modifying dosing, and using long-acting injectable antipsychotics may help improve adherence.

Related Resources

• Velligan DI, Weiden PJ, Sajatovic M, et al; Expert Consensus Panel on Adherence Problems in Serious and Persistent Mental Illness. Assessment of adherence problems in patients with serious and persistent mental illness: recommendations from the Expert Consensus Guidelines. J Clin Psychiatry. 2009;70(suppl 4):1-46.
• National Alliance on Mental Illness. www.nami.org.
• Assertive Community Treatment (ACT) Organization. www.actassociation.org.

Drug Brand Names

Aripiprazole • Abilify        Chlorpromazine • Thorazine      Clozapine • Clozaril       Fluphenazine • Permitil    Haloperidol • Haldol                  Iloperidone • Fanapt       Olanzapine • Zyprexa      Paliperidone • Invega               Perphenazine • Trilafon    Quetiapine • Seroquel      Risperidone • Risperdal            Ziprasidone • Geodon

Disclosures

Dr. Macaluso has been the principal investigator for clinical trials for AbbVie, Eisai, Envivo, Janssen, Naurex, and Pfizer. All clinical trial and study contracts and payments were made through the Kansas University Medical Center Research Institute. Dr. McKnight reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Medication nonadherence is a common problem when treating patients with schizophrenia that can worsen prognosis and lead to sub-optimal treatment outcomes. In this article, we discuss common reasons for nonadherence and describe evidence-based treatments intended to increase adherence and improve outcomes (Box).^1-6

Common reasons for nonadherence

The primary predictor of future nonadherence is a history of nonadherence. It is important to understand patients’ reasons for nonadherence so that practical and evidence-based solutions can be implemented into the treatment plans of individual patients.

The 2009 Expert Consensus Guidelines on Adherence Problems in Patients with Serious and Persistent Mental Illness divided variables related to nonadherence into 3 categories:

• those that lie within the patient (intrinsic)
• those that are related to the patient’s relationship with healthcare providers, family, or caregivers (extrinsic)
• those that are related to the healthcare delivery system (extrinsic).⁷

Among intrinsic variables, studies have shown a correlation between nonadherence and education level, lower socioeconomic status, homelessness, and male sex.⁷ (The Expert Consensus Guidelines considered homelessness to be an intrinsic factor because it was used as a demographic variable in the studies.)

Cognitive and negative symptoms associated with schizophrenia are an intrinsic risk factor for nonadherence because patients might not remember when or how to take medication.⁷ In a study by Freudenreich and co-workers⁸ of 81 outpatients who had a diagnosis of schizophrenia, the presence of negative symptoms predicted a negative attitude toward psychotropic medications. Poor insight might be the result of cognitive dysfunction associated with schizophrenia, and often is due to a lack of awareness of the importance of taking medications.

Limited insight into the need for treatment can be problematic early in the course of the illness when it may be directly related to positive symptoms. Perkins and colleagues⁹ demonstrated that patients recovering from a first psychotic episode who had limited insight into their illness and lacked desire to seek treatment were less adherent with medication. In another study, 5% of psychiatrists surveyed thought that many of their patients with schizophrenia were nonadherent because those patients did not believe that medications were effective or useful.¹⁰

Comorbid substance abuse disorders can contribute to medication nonadherence. In an analysis of 6,731 patients with schizophrenia, Novick and co-workers reported that alcohol dependence and substance abuse in the previous month predicted medication nonadherence.¹¹ Hunt and colleagues demonstrated that, among 99 nonadherent patients with schizophrenia, time to first readmission was shorter for patients with comorbid substance abuse disorders compared with patients who had a diagnosis of schizophrenia only. Over the 4-year study period, the 28 patients who had a dual diagnosis (schizophrenia and substance abuse) accounted for 57% of all hospital readmissions.¹²

Several variables that affect medication adherence are related to the patient’s relationship with healthcare providers, family, caregivers, and the service delivery system.⁷ These include:

• the perceived stigma of being given a diagnosis of a serious mental illness
• adverse effects related to medications
• poor social and family support
• difficulty gaining access to mental health services.^7,10

Societal stigma associated with seeking treatment from a mental health professional may contribute to nonadherence in some patients. In 1 study,¹³ 36% of people surveyed would not want to work closely with a person who has a serious mental illness.

Adverse effects contribute significantly to nonadherence

Limited treatment options (which may be expensive) can make it difficult to manage the adverse effects of antipsychotics. In a cross-sectional survey of 876 patients, investigators reported that: 1) <50% of patients were adherent with medication, and 2) 80% experienced ≥1 side effect that was reported to be “somewhat bothersome” in self-ratings (Table 1).¹⁴ Extrapyramidal symptoms (EPS) and agitation were most strongly associated with nonadherence; weight gain, akathisia, and sexual dysfunction also were associated with nonadherence.¹⁴ This study did not distinguish adverse effects associated with first-generation antipsychotics (FGAs) from those associated with second-generation antipsychotics (SGAs), even though 71.7% of patients studied were taking an SGA.

A meta-analysis by Leucht and co-workers¹⁵ compared 15 antipsychotics (the FGAs haloperidol and chlorpromazine and 13 SGAs) for efficacy and tolerability in schizophrenia. Haloperidol had the highest rate of discontinuation for any

Antipsychotic binding affinities to dopamine 2 (D2), serotonin 2A (5-HT2A), histamine (H1), and other receptors have an impact on a medication’s side-effect profile. Because of individual patient characteristics, you might be faced with choosing a medication that has a lower risk of EPS but a higher risk of weight gain and metabolic complications—or the inverse. Understanding binding affinities, side-effect profiles, and how to minimize or utilize adverse effects (ie, giving a drug that is approved to treat schizophrenia and is associated with weight gain to a patient with schizophrenia who has lost weight) may lead to greater adherence (Table 2¹⁶ and Table 3¹⁷).

Adequate support is essential

The therapeutic alliance plays a key role in patients’ attitudes toward taking medication. Magura and colleagues¹⁸ found that one-third of psychiatric patients (13% of whom had a diagnosis of schizophrenia) reported that their psychiatrist did not spend enough time with them explaining side effects, and felt “rushed.”

Patients with schizophrenia often require access to social support systems provided by family members, friends, and community agencies that provide case management and attendant care services. Patients who are adherent to medication tend to have greater perceived family involvement in medication treatment, and tend to have been raised in a family that had more of a positive attitude toward medication.¹⁹

In our practice, we have observed that recent state and federal budget cuts have resulted in patients having greater difficulty gaining access to case management and attendant care services, which then leads to increased rates of medication nonadherence. Be aware that variables such as limited office hours, financial hardship, and cultural and language barriers can compromise a patient’s ability to seek and continue care.

In the following section, we lay out techniques for improving adherence in patients with schizophrenia.

Employ general and specific strategies to boost adherence

How can you raise medication adherence concerns with patients, keeping in mind that they often overestimate their adherence?

Ask. Some clinicians ask questions such as “Are you taking your medication?”, although a more effective approach might be to ask how the patient is taking his (her) medication. Asking questions such as “When do you take your medication?” and “In the past week, how many doses do you think you missed?” might be more effective ways to inquire about adherence.⁷

The Expert Consensus Guidelines recommend asking patients about medication adherence monthly for those who are stable, doing well, and believed to be adherent. For those who are new to a practice or who are not doing well, inquire about medication adherence at least weekly.⁷

In our practice, patients who are unstable but do not require inpatient hospitalization typically are seen more often in the clinic, or are referred to intensive outpatient or partial hospitalization programs. If an unstable patient is unable to come in for more frequent appointments, we arrange phone conferences between her and her provider. If a patient is not doing well and has a case manager, we often ask that case manager to visit the patient, in person, more often than he (she) would otherwise.

Take a nonjudgmental approach when raising these issues with patients. Questions such as “We all forget to take our medication sometimes; do you?” help to normalize nonadherence, and improve the therapeutic alliance, and might result in the patient being more honest with the clinician.⁷ Because patients may be apprehensive about discussing adverse events, clinicians must be proactive about improving the therapeutic alliance and making patients feel comfortable when discussing sensitive topics. Clinicians should try to convey the idea that, although adherence is a concern, so is quality of life. A clinicians’ willingness to take a flexible approach that is nonpunitive nor authoritarian can aid the therapeutic alliance and improve overall adherence.

Be sensitive to financial, cultural, and language variables that can affect access to care. The Expert Consensus Guidelines recommend asking patients if they can afford their medication. In our practice, we have seen patients with schizophrenia discharged from the hospital only to be readmitted 1 month later because they could not afford to fill their prescriptions.

It is important to have translation services available, in person or by phone, for patients who do not speak English. Furthermore, it is important to understand the limitations that your practice might place on access to care. Ask patients if they have ever had trouble making an appointment when they needed to be seen, or if they called the office with a question and did not receive an answer in a timely fashion; doing so allows you to assess the practice’s ability to meet patients’ needs and helps you build a therapeutic alliance.

Make objective assessments. It is important for practitioners to not base their assessment of medication adherence solely on subjective findings. Asking patients to bring in their medication bottles for pill counts and checking with the patients’ pharmacies for information about refill frequency can provide some objective data. Electronic monitoring systems use microprocessors inserted into bottle caps to record the occurrence and timing of each bottle opening. Studies show that these electronic monitoring systems are the gold standard for determining medication adherence and could be used in cases where it is unclear if the patient is taking his (her) medication.^7,20 Such systems have successfully monitored medication adherence in clinical trials, but their use in clinical practice is complicated by ethical and legal considerations and cost issues.

Simplify the regimen. Using medications with once-daily dosing, for example, can help improve adherence. Pfeiffer and co-workers²¹ found that patients whose medication regimens were changed from once daily to more than once daily experienced a decrease in medication adherence. Conversely, a decrease in dosing frequency was significantly associated with improved adherence. More than once-daily dosing was only weakly associated with poorer adherence among patients already on a stable regimen.

Discussing positive and negative aspects of past medication trials with a patient and inquiring if she prefers a specific medication can be an effective way to build the therapeutic relationship and help with adherence.

Direct patients to psychosocial interventions. These can be broadly classified as:

• educational approaches
• group therapy approaches
• family interventions
• cognitive treatments
• combination approaches.

Psychoeducational approaches have limited effect on improving adherence when delivered to individual patients. However, 1 study showed that psychoeducation was effective at improving adherence when extended to include the patient’s family.²²

Motivational interviewing techniques, behavioral approaches, and family interventions are effective at increasing medication adherence. One study looked at the value of training a patient-identified informant to supervise and administer medication. This person, usually a family member or close support, was responsible for obtaining medication from the pharmacy, administering the medication, and recording adherence. After 1 year, 67% of patients who used an informant were adherent, compared with only 45% in the group that did not have informant support.²² Case managers, attendant care workers, home health nurses, and assertive community treatment (ACT) teams also can participate in this manner; it is important, therefore, for you to be aware of the resources available in your community and to understand your role as patient advocate.

Substance abuse is a strong risk factor for nonadherence among patients with schizophrenia,¹⁸ which makes it important to assess patients for substance use and encourage those who do abuse to seek treatment. Although 1 study showed no correlation between Alcoholics Anonymous (AA) attendance and medication adherence,¹² many AA and Narcotics Anonymous groups do not discuss psychiatric medications during group meetings. Magura and colleagues encouraged the use of “dual focus” groups that involve mental health professionals and addiction treatment specialists discussing mental health and substance abuse issues at the same setting.¹⁸

Prescribe long-acting injectable antipsychotics. Typically, long-acting injectable antipsychotics (LAIs) are reserved for patients who have a history of nonadherence. In a small study (N = 97) comparing LAI risperidone and oral risperidone or oral haloperidol, patients treated with an LAI had significantly fewer all-cause discontinuations (26.0%, compared with 70.2%) at 24 months.²³ The Adherence to Treatment and Therapeutic Strategies in Schizophrenic Patients study examined 1,848 patients with schizophrenia and reached similar findings regarding LAI antipsychotics.²⁴ (Note: Aripiprazole, fluphenazine, haloperidol, olanzapine, and paliperidone also are available in an LAI formulation.)

Bottom Line

Antipsychotic nonadherence in schizophrenia is a major problem for patients, families, and society. Being able to identify patients at risk for nonadherence, understanding the reasons for their nonadherence, and seeking practical solutions to the problem are all the responsibility of the treating physician. Psychoeducation, addressing substance abuse, modifying dosing, and using long-acting injectable antipsychotics may help improve adherence.

Related Resources

• Velligan DI, Weiden PJ, Sajatovic M, et al; Expert Consensus Panel on Adherence Problems in Serious and Persistent Mental Illness. Assessment of adherence problems in patients with serious and persistent mental illness: recommendations from the Expert Consensus Guidelines. J Clin Psychiatry. 2009;70(suppl 4):1-46.
• National Alliance on Mental Illness. www.nami.org.
• Assertive Community Treatment (ACT) Organization. www.actassociation.org.

Drug Brand Names

Aripiprazole • Abilify        Chlorpromazine • Thorazine      Clozapine • Clozaril       Fluphenazine • Permitil    Haloperidol • Haldol                  Iloperidone • Fanapt       Olanzapine • Zyprexa      Paliperidone • Invega               Perphenazine • Trilafon    Quetiapine • Seroquel      Risperidone • Risperdal            Ziprasidone • Geodon

Disclosures

Dr. Macaluso has been the principal investigator for clinical trials for AbbVie, Eisai, Envivo, Janssen, Naurex, and Pfizer. All clinical trial and study contracts and payments were made through the Kansas University Medical Center Research Institute. Dr. McKnight reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Medication nonadherence is a common problem when treating patients with schizophrenia that can worsen prognosis and lead to sub-optimal treatment outcomes. In this article, we discuss common reasons for nonadherence and describe evidence-based treatments intended to increase adherence and improve outcomes (Box).^1-6

Common reasons for nonadherence

The primary predictor of future nonadherence is a history of nonadherence. It is important to understand patients’ reasons for nonadherence so that practical and evidence-based solutions can be implemented into the treatment plans of individual patients.

The 2009 Expert Consensus Guidelines on Adherence Problems in Patients with Serious and Persistent Mental Illness divided variables related to nonadherence into 3 categories:

• those that lie within the patient (intrinsic)
• those that are related to the patient’s relationship with healthcare providers, family, or caregivers (extrinsic)
• those that are related to the healthcare delivery system (extrinsic).⁷

Among intrinsic variables, studies have shown a correlation between nonadherence and education level, lower socioeconomic status, homelessness, and male sex.⁷ (The Expert Consensus Guidelines considered homelessness to be an intrinsic factor because it was used as a demographic variable in the studies.)

Cognitive and negative symptoms associated with schizophrenia are an intrinsic risk factor for nonadherence because patients might not remember when or how to take medication.⁷ In a study by Freudenreich and co-workers⁸ of 81 outpatients who had a diagnosis of schizophrenia, the presence of negative symptoms predicted a negative attitude toward psychotropic medications. Poor insight might be the result of cognitive dysfunction associated with schizophrenia, and often is due to a lack of awareness of the importance of taking medications.

Limited insight into the need for treatment can be problematic early in the course of the illness when it may be directly related to positive symptoms. Perkins and colleagues⁹ demonstrated that patients recovering from a first psychotic episode who had limited insight into their illness and lacked desire to seek treatment were less adherent with medication. In another study, 5% of psychiatrists surveyed thought that many of their patients with schizophrenia were nonadherent because those patients did not believe that medications were effective or useful.¹⁰

Comorbid substance abuse disorders can contribute to medication nonadherence. In an analysis of 6,731 patients with schizophrenia, Novick and co-workers reported that alcohol dependence and substance abuse in the previous month predicted medication nonadherence.¹¹ Hunt and colleagues demonstrated that, among 99 nonadherent patients with schizophrenia, time to first readmission was shorter for patients with comorbid substance abuse disorders compared with patients who had a diagnosis of schizophrenia only. Over the 4-year study period, the 28 patients who had a dual diagnosis (schizophrenia and substance abuse) accounted for 57% of all hospital readmissions.¹²

Several variables that affect medication adherence are related to the patient’s relationship with healthcare providers, family, caregivers, and the service delivery system.⁷ These include:

• the perceived stigma of being given a diagnosis of a serious mental illness
• adverse effects related to medications
• poor social and family support
• difficulty gaining access to mental health services.^7,10

Societal stigma associated with seeking treatment from a mental health professional may contribute to nonadherence in some patients. In 1 study,¹³ 36% of people surveyed would not want to work closely with a person who has a serious mental illness.

Adverse effects contribute significantly to nonadherence

Limited treatment options (which may be expensive) can make it difficult to manage the adverse effects of antipsychotics. In a cross-sectional survey of 876 patients, investigators reported that: 1) <50% of patients were adherent with medication, and 2) 80% experienced ≥1 side effect that was reported to be “somewhat bothersome” in self-ratings (Table 1).¹⁴ Extrapyramidal symptoms (EPS) and agitation were most strongly associated with nonadherence; weight gain, akathisia, and sexual dysfunction also were associated with nonadherence.¹⁴ This study did not distinguish adverse effects associated with first-generation antipsychotics (FGAs) from those associated with second-generation antipsychotics (SGAs), even though 71.7% of patients studied were taking an SGA.

A meta-analysis by Leucht and co-workers¹⁵ compared 15 antipsychotics (the FGAs haloperidol and chlorpromazine and 13 SGAs) for efficacy and tolerability in schizophrenia. Haloperidol had the highest rate of discontinuation for any

Antipsychotic binding affinities to dopamine 2 (D2), serotonin 2A (5-HT2A), histamine (H1), and other receptors have an impact on a medication’s side-effect profile. Because of individual patient characteristics, you might be faced with choosing a medication that has a lower risk of EPS but a higher risk of weight gain and metabolic complications—or the inverse. Understanding binding affinities, side-effect profiles, and how to minimize or utilize adverse effects (ie, giving a drug that is approved to treat schizophrenia and is associated with weight gain to a patient with schizophrenia who has lost weight) may lead to greater adherence (Table 2¹⁶ and Table 3¹⁷).

Adequate support is essential

The therapeutic alliance plays a key role in patients’ attitudes toward taking medication. Magura and colleagues¹⁸ found that one-third of psychiatric patients (13% of whom had a diagnosis of schizophrenia) reported that their psychiatrist did not spend enough time with them explaining side effects, and felt “rushed.”

Patients with schizophrenia often require access to social support systems provided by family members, friends, and community agencies that provide case management and attendant care services. Patients who are adherent to medication tend to have greater perceived family involvement in medication treatment, and tend to have been raised in a family that had more of a positive attitude toward medication.¹⁹

In our practice, we have observed that recent state and federal budget cuts have resulted in patients having greater difficulty gaining access to case management and attendant care services, which then leads to increased rates of medication nonadherence. Be aware that variables such as limited office hours, financial hardship, and cultural and language barriers can compromise a patient’s ability to seek and continue care.

In the following section, we lay out techniques for improving adherence in patients with schizophrenia.

Employ general and specific strategies to boost adherence

How can you raise medication adherence concerns with patients, keeping in mind that they often overestimate their adherence?

Ask. Some clinicians ask questions such as “Are you taking your medication?”, although a more effective approach might be to ask how the patient is taking his (her) medication. Asking questions such as “When do you take your medication?” and “In the past week, how many doses do you think you missed?” might be more effective ways to inquire about adherence.⁷

The Expert Consensus Guidelines recommend asking patients about medication adherence monthly for those who are stable, doing well, and believed to be adherent. For those who are new to a practice or who are not doing well, inquire about medication adherence at least weekly.⁷

In our practice, patients who are unstable but do not require inpatient hospitalization typically are seen more often in the clinic, or are referred to intensive outpatient or partial hospitalization programs. If an unstable patient is unable to come in for more frequent appointments, we arrange phone conferences between her and her provider. If a patient is not doing well and has a case manager, we often ask that case manager to visit the patient, in person, more often than he (she) would otherwise.

Take a nonjudgmental approach when raising these issues with patients. Questions such as “We all forget to take our medication sometimes; do you?” help to normalize nonadherence, and improve the therapeutic alliance, and might result in the patient being more honest with the clinician.⁷ Because patients may be apprehensive about discussing adverse events, clinicians must be proactive about improving the therapeutic alliance and making patients feel comfortable when discussing sensitive topics. Clinicians should try to convey the idea that, although adherence is a concern, so is quality of life. A clinicians’ willingness to take a flexible approach that is nonpunitive nor authoritarian can aid the therapeutic alliance and improve overall adherence.

Be sensitive to financial, cultural, and language variables that can affect access to care. The Expert Consensus Guidelines recommend asking patients if they can afford their medication. In our practice, we have seen patients with schizophrenia discharged from the hospital only to be readmitted 1 month later because they could not afford to fill their prescriptions.

It is important to have translation services available, in person or by phone, for patients who do not speak English. Furthermore, it is important to understand the limitations that your practice might place on access to care. Ask patients if they have ever had trouble making an appointment when they needed to be seen, or if they called the office with a question and did not receive an answer in a timely fashion; doing so allows you to assess the practice’s ability to meet patients’ needs and helps you build a therapeutic alliance.

Make objective assessments. It is important for practitioners to not base their assessment of medication adherence solely on subjective findings. Asking patients to bring in their medication bottles for pill counts and checking with the patients’ pharmacies for information about refill frequency can provide some objective data. Electronic monitoring systems use microprocessors inserted into bottle caps to record the occurrence and timing of each bottle opening. Studies show that these electronic monitoring systems are the gold standard for determining medication adherence and could be used in cases where it is unclear if the patient is taking his (her) medication.^7,20 Such systems have successfully monitored medication adherence in clinical trials, but their use in clinical practice is complicated by ethical and legal considerations and cost issues.

Simplify the regimen. Using medications with once-daily dosing, for example, can help improve adherence. Pfeiffer and co-workers²¹ found that patients whose medication regimens were changed from once daily to more than once daily experienced a decrease in medication adherence. Conversely, a decrease in dosing frequency was significantly associated with improved adherence. More than once-daily dosing was only weakly associated with poorer adherence among patients already on a stable regimen.

Discussing positive and negative aspects of past medication trials with a patient and inquiring if she prefers a specific medication can be an effective way to build the therapeutic relationship and help with adherence.

Direct patients to psychosocial interventions. These can be broadly classified as:

• educational approaches
• group therapy approaches
• family interventions
• cognitive treatments
• combination approaches.

Psychoeducational approaches have limited effect on improving adherence when delivered to individual patients. However, 1 study showed that psychoeducation was effective at improving adherence when extended to include the patient’s family.²²

Motivational interviewing techniques, behavioral approaches, and family interventions are effective at increasing medication adherence. One study looked at the value of training a patient-identified informant to supervise and administer medication. This person, usually a family member or close support, was responsible for obtaining medication from the pharmacy, administering the medication, and recording adherence. After 1 year, 67% of patients who used an informant were adherent, compared with only 45% in the group that did not have informant support.²² Case managers, attendant care workers, home health nurses, and assertive community treatment (ACT) teams also can participate in this manner; it is important, therefore, for you to be aware of the resources available in your community and to understand your role as patient advocate.

Substance abuse is a strong risk factor for nonadherence among patients with schizophrenia,¹⁸ which makes it important to assess patients for substance use and encourage those who do abuse to seek treatment. Although 1 study showed no correlation between Alcoholics Anonymous (AA) attendance and medication adherence,¹² many AA and Narcotics Anonymous groups do not discuss psychiatric medications during group meetings. Magura and colleagues encouraged the use of “dual focus” groups that involve mental health professionals and addiction treatment specialists discussing mental health and substance abuse issues at the same setting.¹⁸

Prescribe long-acting injectable antipsychotics. Typically, long-acting injectable antipsychotics (LAIs) are reserved for patients who have a history of nonadherence. In a small study (N = 97) comparing LAI risperidone and oral risperidone or oral haloperidol, patients treated with an LAI had significantly fewer all-cause discontinuations (26.0%, compared with 70.2%) at 24 months.²³ The Adherence to Treatment and Therapeutic Strategies in Schizophrenic Patients study examined 1,848 patients with schizophrenia and reached similar findings regarding LAI antipsychotics.²⁴ (Note: Aripiprazole, fluphenazine, haloperidol, olanzapine, and paliperidone also are available in an LAI formulation.)

Bottom Line

Antipsychotic nonadherence in schizophrenia is a major problem for patients, families, and society. Being able to identify patients at risk for nonadherence, understanding the reasons for their nonadherence, and seeking practical solutions to the problem are all the responsibility of the treating physician. Psychoeducation, addressing substance abuse, modifying dosing, and using long-acting injectable antipsychotics may help improve adherence.

Related Resources

• Velligan DI, Weiden PJ, Sajatovic M, et al; Expert Consensus Panel on Adherence Problems in Serious and Persistent Mental Illness. Assessment of adherence problems in patients with serious and persistent mental illness: recommendations from the Expert Consensus Guidelines. J Clin Psychiatry. 2009;70(suppl 4):1-46.
• National Alliance on Mental Illness. www.nami.org.
• Assertive Community Treatment (ACT) Organization. www.actassociation.org.

Drug Brand Names

Aripiprazole • Abilify        Chlorpromazine • Thorazine      Clozapine • Clozaril       Fluphenazine • Permitil    Haloperidol • Haldol                  Iloperidone • Fanapt       Olanzapine • Zyprexa      Paliperidone • Invega               Perphenazine • Trilafon    Quetiapine • Seroquel      Risperidone • Risperdal            Ziprasidone • Geodon

Disclosures

Dr. Macaluso has been the principal investigator for clinical trials for AbbVie, Eisai, Envivo, Janssen, Naurex, and Pfizer. All clinical trial and study contracts and payments were made through the Kansas University Medical Center Research Institute. Dr. McKnight reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

When Kanner first described autism,¹ the disorder was believed to be an uncommon condition, occurring in 4 of every 10,000 children. Over the past few years, however, the rate of autism has increased substantially. Autism is now regarded as a childhood-onset spectrum disorder^a characterized by persistent deficits in social communication, with a restricted pattern of interests and activities, occurring in approximately 1% of children.³

In DSM-IV-TR, Asperger’s disorder (AD), first described as “autistic psychopathy,”⁴ is categorized as a subtype of ASD in which the patient, without a history of language delay or mental retardation, has autistic social deficits that do not meet full criteria for autism.

DSM-5 eliminated AD as an independent category, including it instead as part of ASD.⁵ The label “high-functioning autism” is sometimes used to refer to persons with autism who have normal intelligence (usually defined as full-scale IQ >70), whereas those who have severe intellectual and communication disability are referred to as “low-functioning.” I use “high-functioning autism” and “Asperger’s disorder” interchangeably.

Violent crime and ASD/AD

Reports in the past 2 decades have described violent behavior in persons with ASD/AD. Because of the sensational and unusual nature of these criminal incidents, there is a perception by the public that persons with these disorders, especially those with AD, are predisposed to violent behavior. (Incidents allegedly committed by persons with ASD include the 2007 Virginia Tech campus shooting and the 2012 Newtown, Connecticut, school massacre.⁶)

Yet neither the original descriptions by Kanner (of autism) and Asperger, nor follow-up studies based on the initial samples studied, showed an increased prevalence of violent crime among persons with ASD/AD.⁷

In this article, I examine the evidence behind the claim that people who have ASD/AD are predisposed to criminal violence. At the conclusion, you should, as a physician without special training in autism, have a better understanding of when to suspect ASD/AD in an adult who is involved in criminal behavior.

When should you suspect ASD/AD in an adult?

Although autism is a childhood-onset disorder, its symptoms persist across the life
span. If the diagnosis is missed in childhood, which is likely to happen if the person has normal intelligence and relatively good verbal skills, he (she) might come to medical attention for the first time as an adult.

Because most psychiatrists who treat adults do not receive adequate training in the assessment of childhood psychiatric disorders, ASD/AD might be misdiagnosed as schizophrenia or another psychotic disorder. What clues help identify underlying ASD/AD when a patient is referred to you for psychiatric evaluation after allegedly committing a violent crime?

Clue #1. He makes no attempt to deny or conceal the act. The behavior appears to be part of ritualistic behavior or excessive interest (Table).

Often, the alleged crime occurs when the patient’s excessive interests “get out of control,” perhaps because of an external event. For example, a teenager with AD who is fixated on video games might stumble upon pornographic web sites and begin making obscene telephone calls. Particular attention should be paid to a history of rigid, restricted interests beginning in early childhood.

These restricted interests change over time and correlate with intelligence level: The higher the level of intelligence, the more sophisticated the level of fixation.  Examples of fixations include computers, technology, and scientific experiments and pursuits. Repeated acts of arson have been reported to be part of an autistic person’s fixation with starting fires.⁸

Clue #2. He appears to lack sound and prudent judgment despite normal intelligence.

Although most patients with ASD score in the intellectually disabled or mentally retarded range, at least one-third have an IQ in the normal range.⁹ Examine school records and reports from other agencies when evaluating a patient. Pay attention to a history of difficulty relating to peers at an early age, combined with evidence of rigid, restricted fixations and interests.

It is important to obtain a reliable history going back to early childhood, and not rely just on the patient’s mental status; presenting symptoms might mask underlying traits of ASD, especially in higher-functioning adults. (I once cared for a young man with ASD who had been fired a few days after landing his first job selling used cars because he was “sexually harassing” his colleagues. When questioned, he said that he was only trying to be “friendly” and “practicing his social skills.”)

Clue #3. He has been given a diagnosis of schizophrenia without a clear history of hallucinations or delusions.

Differentiating chronic schizophrenia and autism in adults is not always easy, especially in those who have an intellectual disability. In patients whose cognitive and verbal skills are relatively well preserved (such as AD), the presence of intense, focused interests, a pedantic manner of speaking, and abnormalities of nonverbal communication can help clarify the diagnosis. In particular, a recorded history of “childhood schizophrenia” or “obsessive-compulsive behavior” going back to preschool years should alert you to possible ASD.

Scales and screens. Apart from obtaining an accurate developmental history from a variety of sources, you can use rating scales and screening instruments, such as the Social and Communication Questionnaire¹⁰—although their utility is limited in adults. It is important not to risk overdiagnosis on the basis of these instruments alone: The gold standard of diagnosis remains clinical. The critical point is that the combination of core symptoms of social communication deficits and restricted interests is more important than the presence of a single symptom. A touch of oddity does not mean that one has ASD/AD.

Is the prevalence of violent crime increased in ASD/AD?

It is important to distinguish violent crime from aggressive behavior. The latter, which can be verbal or nonverbal, is not always intentional or malevolent. In some persons who have an intellectual disability, a desire to communicate might lead to inappropriate touching or pushing. This distinction is particularly relevant to psychiatrists because many people who have ASD have an intellectual disability.

Violent crime is more deliberate, serious, and planned. It involves force or threat of force. According to the Federal Bureau of Investigation Uniform Crime Reporting Program, violent crime comprises four offenses: murder and non-negligent manslaughter, forcible rape, robbery, and aggravated assault.¹¹

Earlier descriptions of ASD/AD did not mention criminal violence as an important feature of these disorders. However, reports began to emerge about two decades ago suggesting that people who have ASD—particularly AD—are prone to violent crime. Some of the patients described in Wing’s original series¹² of AD showed violent tendencies, ranging from sudden outbursts of violence to injury to others because of fixation on hobbies such as chemistry experimentation.

Reports such as these were based on isolated case reports or select samples, such as residents of maximum-security hospitals. Scragg and Shah, for example, surveyed the male population of Broadmoor Hospital, a high-security facility in the United Kingdom, and found that the prevalence of AD was higher than expected in the general population.¹³

Recent reports have not been able to confirm that violent crime is increased in persons with ASD, however:

• In a clinical sample of 313 Danish adults with ASD (age 25 to 59) drawn from the Danish Register of Criminality, Mouridsen and colleagues found that persons with ASD had a lower rate of criminal conviction than matched controls (9%, compared with 18%).¹⁴
• In a small community study, Woodbury-Smith and colleagues examined the prevalence rates and types of offending behavior in persons with ASD. Based on official records, only two (18%) had a history of criminal conviction.¹⁵

The role of psychiatric comorbidity

Psychiatric disorders are common in persons who have ASD. In one study, 70% of a sample of 114 children with ASD (age 10 to 14) had a psychiatric disorder, based on a parent interview.¹⁶ Although people with mental illness are not inherently criminal or violent, having an additional psychiatric disorder independently increases the risk of offending behavior.¹⁷ For example, the association of attention-deficit/hyperactivity disorder with criminality is well established.¹⁶ Some patients with severe depression and psychotic disorders, including schizophrenia, also are at increased risk of committing a violent act.

To examine the contribution of mental health factors to the commission of crime by persons with ASD, Newman and Ghaziuddin¹⁸ used online databases to identify relevant articles, which were then cross-referenced with keyword searches for “violence,” “crime,” “murder,” “assault,” “rape,” and “sex offenses.” Thirty-seven cases were identified in the 17 publications that met inclusion criteria. Out of these, 30% had a definite psychiatric disorder and 54% had a probable psychiatric disorder at the time they committed the crime.¹⁸

Any patient with ASD/AD who is evaluated for criminal behavior should be screened for a comorbid psychiatric disorder. In adolescents, stressors such as bullying in school and problems surrounding dating might contribute to offending behavior.

What are management options in the face of violence?

Managing ASD/AD when an offending behavior has occurred first requires a correct diagnosis.¹⁹ Professionals working in the criminal justice system have little awareness of the variants of ASD; a defendant with an intellectual disability and a characteristic facial appearance (for example, someone with Down syndrome) can be easily identified, but a high-functioning person who has mild autistic features often is missed. This is more likely to occur in adults because the symptoms of ASD, including the type and severity of isolated interests, change over time.

Here is how I recommend that you proceed:

Step #1. Confirm the ASD diagnosis based on developmental history and the presence of persistent social and communication deficits plus restricted interests.

Step #2. Screen for comorbid psychiatric and medical disorders, including depression, psychosis, and seizure disorder.

Step #3. Treat any disorders you identify with a combination of medication and behavioral intervention.

Step #4. Carefully examine the circumstances surrounding the offending behavior. Involve forensic services on a case-by-case basis, depending on the type and seriousness of the offending behavior (see Related Resources for information on the role of forensic services). When the crime does not involve serious violence, lengthy incarceration might be unnecessary. Because psychopathy and ASD/AD are not mutually exclusive, persons who commit a heinous crime, such as rape or murder, should be dealt with in accordance with the law.

Need for greater awareness of the complexion of ASD

Patients who have ASD/AD form a heterogeneous group in which the levels of cognitive and communication skills are variable. Those who are low-functioning and who have severe behavioral and adaptive deficits occasionally commit aggressive acts against their caregivers.

Most patients with ASD/AD are neither violent nor criminal. Those who are at the higher end of the spectrum, with relatively preserved communication and intellectual skills, occasionally indulge in criminal behavior—behavior that is nonviolent and results from their inability to read social cues or excessive preoccupations.

Most reports that link criminal violence with ASD are based on isolated case reports or on biased samples that use unreliable diagnostic criteria. In higher-functioning persons with ASD, violent crime is almost always precipitated by a comorbid psychiatric disorder, such as severe depression and psychosis.

In short: There is a need to increase our awareness of the special challenges faced by persons with ASD/AD in the criminal justice system.

aGiven the term pervasive developmental disorders (PDD) in the DSM-IV-TR, the spectrum includes autistic disorder, Asperger’s disorder, and pervasive developmental disorder not otherwise specified.2

Bottom Line

Most people who have an autism spectrum disorder (ASD) do not commit violent crime. When violent crime occurs at the hands of a person with ASD, it is almost always precipitated by a comorbid psychiatric disorder, such as severe depression or psychosis. Treating a person with ASD who has committed a violent crime is multimodal, including forensic services when necessary.

Related Resources

• Autism Speaks. No link between autism and violence. www.autismspeaks.org/science/science-news/no-link-between-autism-and-violence.
• Haskins BG, Silva JA. Asperger’s disorder and criminal behavior: Forensic-psychiatric considerations. J Am Acad Psychiatry Law. 2006;34(3):374-384.
• Newman SS, Ghaziuddin M. Violent crime and Asperger syndrome: the role of psychiatric comorbidity. J Autism Dev Disord. 2008;38:1848-1852.
• Wing L. Asperger’s syndrome: a clinical account. Psychol Med. 1981;11(1):115-129.

Disclosure

Dr. Ghaziuddin reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

When Kanner first described autism,¹ the disorder was believed to be an uncommon condition, occurring in 4 of every 10,000 children. Over the past few years, however, the rate of autism has increased substantially. Autism is now regarded as a childhood-onset spectrum disorder^a characterized by persistent deficits in social communication, with a restricted pattern of interests and activities, occurring in approximately 1% of children.³

In DSM-IV-TR, Asperger’s disorder (AD), first described as “autistic psychopathy,”⁴ is categorized as a subtype of ASD in which the patient, without a history of language delay or mental retardation, has autistic social deficits that do not meet full criteria for autism.

DSM-5 eliminated AD as an independent category, including it instead as part of ASD.⁵ The label “high-functioning autism” is sometimes used to refer to persons with autism who have normal intelligence (usually defined as full-scale IQ >70), whereas those who have severe intellectual and communication disability are referred to as “low-functioning.” I use “high-functioning autism” and “Asperger’s disorder” interchangeably.

Violent crime and ASD/AD

Reports in the past 2 decades have described violent behavior in persons with ASD/AD. Because of the sensational and unusual nature of these criminal incidents, there is a perception by the public that persons with these disorders, especially those with AD, are predisposed to violent behavior. (Incidents allegedly committed by persons with ASD include the 2007 Virginia Tech campus shooting and the 2012 Newtown, Connecticut, school massacre.⁶)

Yet neither the original descriptions by Kanner (of autism) and Asperger, nor follow-up studies based on the initial samples studied, showed an increased prevalence of violent crime among persons with ASD/AD.⁷

In this article, I examine the evidence behind the claim that people who have ASD/AD are predisposed to criminal violence. At the conclusion, you should, as a physician without special training in autism, have a better understanding of when to suspect ASD/AD in an adult who is involved in criminal behavior.

When should you suspect ASD/AD in an adult?

Although autism is a childhood-onset disorder, its symptoms persist across the life
span. If the diagnosis is missed in childhood, which is likely to happen if the person has normal intelligence and relatively good verbal skills, he (she) might come to medical attention for the first time as an adult.

Because most psychiatrists who treat adults do not receive adequate training in the assessment of childhood psychiatric disorders, ASD/AD might be misdiagnosed as schizophrenia or another psychotic disorder. What clues help identify underlying ASD/AD when a patient is referred to you for psychiatric evaluation after allegedly committing a violent crime?

Clue #1. He makes no attempt to deny or conceal the act. The behavior appears to be part of ritualistic behavior or excessive interest (Table).

Often, the alleged crime occurs when the patient’s excessive interests “get out of control,” perhaps because of an external event. For example, a teenager with AD who is fixated on video games might stumble upon pornographic web sites and begin making obscene telephone calls. Particular attention should be paid to a history of rigid, restricted interests beginning in early childhood.

These restricted interests change over time and correlate with intelligence level: The higher the level of intelligence, the more sophisticated the level of fixation.  Examples of fixations include computers, technology, and scientific experiments and pursuits. Repeated acts of arson have been reported to be part of an autistic person’s fixation with starting fires.⁸

Clue #2. He appears to lack sound and prudent judgment despite normal intelligence.

Although most patients with ASD score in the intellectually disabled or mentally retarded range, at least one-third have an IQ in the normal range.⁹ Examine school records and reports from other agencies when evaluating a patient. Pay attention to a history of difficulty relating to peers at an early age, combined with evidence of rigid, restricted fixations and interests.

It is important to obtain a reliable history going back to early childhood, and not rely just on the patient’s mental status; presenting symptoms might mask underlying traits of ASD, especially in higher-functioning adults. (I once cared for a young man with ASD who had been fired a few days after landing his first job selling used cars because he was “sexually harassing” his colleagues. When questioned, he said that he was only trying to be “friendly” and “practicing his social skills.”)

Clue #3. He has been given a diagnosis of schizophrenia without a clear history of hallucinations or delusions.

Differentiating chronic schizophrenia and autism in adults is not always easy, especially in those who have an intellectual disability. In patients whose cognitive and verbal skills are relatively well preserved (such as AD), the presence of intense, focused interests, a pedantic manner of speaking, and abnormalities of nonverbal communication can help clarify the diagnosis. In particular, a recorded history of “childhood schizophrenia” or “obsessive-compulsive behavior” going back to preschool years should alert you to possible ASD.

Scales and screens. Apart from obtaining an accurate developmental history from a variety of sources, you can use rating scales and screening instruments, such as the Social and Communication Questionnaire¹⁰—although their utility is limited in adults. It is important not to risk overdiagnosis on the basis of these instruments alone: The gold standard of diagnosis remains clinical. The critical point is that the combination of core symptoms of social communication deficits and restricted interests is more important than the presence of a single symptom. A touch of oddity does not mean that one has ASD/AD.

Is the prevalence of violent crime increased in ASD/AD?

It is important to distinguish violent crime from aggressive behavior. The latter, which can be verbal or nonverbal, is not always intentional or malevolent. In some persons who have an intellectual disability, a desire to communicate might lead to inappropriate touching or pushing. This distinction is particularly relevant to psychiatrists because many people who have ASD have an intellectual disability.

Violent crime is more deliberate, serious, and planned. It involves force or threat of force. According to the Federal Bureau of Investigation Uniform Crime Reporting Program, violent crime comprises four offenses: murder and non-negligent manslaughter, forcible rape, robbery, and aggravated assault.¹¹

Earlier descriptions of ASD/AD did not mention criminal violence as an important feature of these disorders. However, reports began to emerge about two decades ago suggesting that people who have ASD—particularly AD—are prone to violent crime. Some of the patients described in Wing’s original series¹² of AD showed violent tendencies, ranging from sudden outbursts of violence to injury to others because of fixation on hobbies such as chemistry experimentation.

Reports such as these were based on isolated case reports or select samples, such as residents of maximum-security hospitals. Scragg and Shah, for example, surveyed the male population of Broadmoor Hospital, a high-security facility in the United Kingdom, and found that the prevalence of AD was higher than expected in the general population.¹³

Recent reports have not been able to confirm that violent crime is increased in persons with ASD, however:

• In a clinical sample of 313 Danish adults with ASD (age 25 to 59) drawn from the Danish Register of Criminality, Mouridsen and colleagues found that persons with ASD had a lower rate of criminal conviction than matched controls (9%, compared with 18%).¹⁴
• In a small community study, Woodbury-Smith and colleagues examined the prevalence rates and types of offending behavior in persons with ASD. Based on official records, only two (18%) had a history of criminal conviction.¹⁵

The role of psychiatric comorbidity

Psychiatric disorders are common in persons who have ASD. In one study, 70% of a sample of 114 children with ASD (age 10 to 14) had a psychiatric disorder, based on a parent interview.¹⁶ Although people with mental illness are not inherently criminal or violent, having an additional psychiatric disorder independently increases the risk of offending behavior.¹⁷ For example, the association of attention-deficit/hyperactivity disorder with criminality is well established.¹⁶ Some patients with severe depression and psychotic disorders, including schizophrenia, also are at increased risk of committing a violent act.

To examine the contribution of mental health factors to the commission of crime by persons with ASD, Newman and Ghaziuddin¹⁸ used online databases to identify relevant articles, which were then cross-referenced with keyword searches for “violence,” “crime,” “murder,” “assault,” “rape,” and “sex offenses.” Thirty-seven cases were identified in the 17 publications that met inclusion criteria. Out of these, 30% had a definite psychiatric disorder and 54% had a probable psychiatric disorder at the time they committed the crime.¹⁸

Any patient with ASD/AD who is evaluated for criminal behavior should be screened for a comorbid psychiatric disorder. In adolescents, stressors such as bullying in school and problems surrounding dating might contribute to offending behavior.

What are management options in the face of violence?

Managing ASD/AD when an offending behavior has occurred first requires a correct diagnosis.¹⁹ Professionals working in the criminal justice system have little awareness of the variants of ASD; a defendant with an intellectual disability and a characteristic facial appearance (for example, someone with Down syndrome) can be easily identified, but a high-functioning person who has mild autistic features often is missed. This is more likely to occur in adults because the symptoms of ASD, including the type and severity of isolated interests, change over time.

Here is how I recommend that you proceed:

Step #1. Confirm the ASD diagnosis based on developmental history and the presence of persistent social and communication deficits plus restricted interests.

Step #2. Screen for comorbid psychiatric and medical disorders, including depression, psychosis, and seizure disorder.

Step #3. Treat any disorders you identify with a combination of medication and behavioral intervention.

Step #4. Carefully examine the circumstances surrounding the offending behavior. Involve forensic services on a case-by-case basis, depending on the type and seriousness of the offending behavior (see Related Resources for information on the role of forensic services). When the crime does not involve serious violence, lengthy incarceration might be unnecessary. Because psychopathy and ASD/AD are not mutually exclusive, persons who commit a heinous crime, such as rape or murder, should be dealt with in accordance with the law.

Need for greater awareness of the complexion of ASD

Patients who have ASD/AD form a heterogeneous group in which the levels of cognitive and communication skills are variable. Those who are low-functioning and who have severe behavioral and adaptive deficits occasionally commit aggressive acts against their caregivers.

Most patients with ASD/AD are neither violent nor criminal. Those who are at the higher end of the spectrum, with relatively preserved communication and intellectual skills, occasionally indulge in criminal behavior—behavior that is nonviolent and results from their inability to read social cues or excessive preoccupations.

Most reports that link criminal violence with ASD are based on isolated case reports or on biased samples that use unreliable diagnostic criteria. In higher-functioning persons with ASD, violent crime is almost always precipitated by a comorbid psychiatric disorder, such as severe depression and psychosis.

In short: There is a need to increase our awareness of the special challenges faced by persons with ASD/AD in the criminal justice system.

aGiven the term pervasive developmental disorders (PDD) in the DSM-IV-TR, the spectrum includes autistic disorder, Asperger’s disorder, and pervasive developmental disorder not otherwise specified.2

Bottom Line

Most people who have an autism spectrum disorder (ASD) do not commit violent crime. When violent crime occurs at the hands of a person with ASD, it is almost always precipitated by a comorbid psychiatric disorder, such as severe depression or psychosis. Treating a person with ASD who has committed a violent crime is multimodal, including forensic services when necessary.

Related Resources

• Autism Speaks. No link between autism and violence. www.autismspeaks.org/science/science-news/no-link-between-autism-and-violence.
• Haskins BG, Silva JA. Asperger’s disorder and criminal behavior: Forensic-psychiatric considerations. J Am Acad Psychiatry Law. 2006;34(3):374-384.
• Newman SS, Ghaziuddin M. Violent crime and Asperger syndrome: the role of psychiatric comorbidity. J Autism Dev Disord. 2008;38:1848-1852.
• Wing L. Asperger’s syndrome: a clinical account. Psychol Med. 1981;11(1):115-129.

Disclosure

Dr. Ghaziuddin reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

When Kanner first described autism,¹ the disorder was believed to be an uncommon condition, occurring in 4 of every 10,000 children. Over the past few years, however, the rate of autism has increased substantially. Autism is now regarded as a childhood-onset spectrum disorder^a characterized by persistent deficits in social communication, with a restricted pattern of interests and activities, occurring in approximately 1% of children.³

In DSM-IV-TR, Asperger’s disorder (AD), first described as “autistic psychopathy,”⁴ is categorized as a subtype of ASD in which the patient, without a history of language delay or mental retardation, has autistic social deficits that do not meet full criteria for autism.

DSM-5 eliminated AD as an independent category, including it instead as part of ASD.⁵ The label “high-functioning autism” is sometimes used to refer to persons with autism who have normal intelligence (usually defined as full-scale IQ >70), whereas those who have severe intellectual and communication disability are referred to as “low-functioning.” I use “high-functioning autism” and “Asperger’s disorder” interchangeably.

Violent crime and ASD/AD

Reports in the past 2 decades have described violent behavior in persons with ASD/AD. Because of the sensational and unusual nature of these criminal incidents, there is a perception by the public that persons with these disorders, especially those with AD, are predisposed to violent behavior. (Incidents allegedly committed by persons with ASD include the 2007 Virginia Tech campus shooting and the 2012 Newtown, Connecticut, school massacre.⁶)

Yet neither the original descriptions by Kanner (of autism) and Asperger, nor follow-up studies based on the initial samples studied, showed an increased prevalence of violent crime among persons with ASD/AD.⁷

In this article, I examine the evidence behind the claim that people who have ASD/AD are predisposed to criminal violence. At the conclusion, you should, as a physician without special training in autism, have a better understanding of when to suspect ASD/AD in an adult who is involved in criminal behavior.

When should you suspect ASD/AD in an adult?

Although autism is a childhood-onset disorder, its symptoms persist across the life
span. If the diagnosis is missed in childhood, which is likely to happen if the person has normal intelligence and relatively good verbal skills, he (she) might come to medical attention for the first time as an adult.

Because most psychiatrists who treat adults do not receive adequate training in the assessment of childhood psychiatric disorders, ASD/AD might be misdiagnosed as schizophrenia or another psychotic disorder. What clues help identify underlying ASD/AD when a patient is referred to you for psychiatric evaluation after allegedly committing a violent crime?

Clue #1. He makes no attempt to deny or conceal the act. The behavior appears to be part of ritualistic behavior or excessive interest (Table).

Often, the alleged crime occurs when the patient’s excessive interests “get out of control,” perhaps because of an external event. For example, a teenager with AD who is fixated on video games might stumble upon pornographic web sites and begin making obscene telephone calls. Particular attention should be paid to a history of rigid, restricted interests beginning in early childhood.

These restricted interests change over time and correlate with intelligence level: The higher the level of intelligence, the more sophisticated the level of fixation.  Examples of fixations include computers, technology, and scientific experiments and pursuits. Repeated acts of arson have been reported to be part of an autistic person’s fixation with starting fires.⁸

Clue #2. He appears to lack sound and prudent judgment despite normal intelligence.

Although most patients with ASD score in the intellectually disabled or mentally retarded range, at least one-third have an IQ in the normal range.⁹ Examine school records and reports from other agencies when evaluating a patient. Pay attention to a history of difficulty relating to peers at an early age, combined with evidence of rigid, restricted fixations and interests.

It is important to obtain a reliable history going back to early childhood, and not rely just on the patient’s mental status; presenting symptoms might mask underlying traits of ASD, especially in higher-functioning adults. (I once cared for a young man with ASD who had been fired a few days after landing his first job selling used cars because he was “sexually harassing” his colleagues. When questioned, he said that he was only trying to be “friendly” and “practicing his social skills.”)

Clue #3. He has been given a diagnosis of schizophrenia without a clear history of hallucinations or delusions.

Differentiating chronic schizophrenia and autism in adults is not always easy, especially in those who have an intellectual disability. In patients whose cognitive and verbal skills are relatively well preserved (such as AD), the presence of intense, focused interests, a pedantic manner of speaking, and abnormalities of nonverbal communication can help clarify the diagnosis. In particular, a recorded history of “childhood schizophrenia” or “obsessive-compulsive behavior” going back to preschool years should alert you to possible ASD.

Scales and screens. Apart from obtaining an accurate developmental history from a variety of sources, you can use rating scales and screening instruments, such as the Social and Communication Questionnaire¹⁰—although their utility is limited in adults. It is important not to risk overdiagnosis on the basis of these instruments alone: The gold standard of diagnosis remains clinical. The critical point is that the combination of core symptoms of social communication deficits and restricted interests is more important than the presence of a single symptom. A touch of oddity does not mean that one has ASD/AD.

Is the prevalence of violent crime increased in ASD/AD?

It is important to distinguish violent crime from aggressive behavior. The latter, which can be verbal or nonverbal, is not always intentional or malevolent. In some persons who have an intellectual disability, a desire to communicate might lead to inappropriate touching or pushing. This distinction is particularly relevant to psychiatrists because many people who have ASD have an intellectual disability.

Violent crime is more deliberate, serious, and planned. It involves force or threat of force. According to the Federal Bureau of Investigation Uniform Crime Reporting Program, violent crime comprises four offenses: murder and non-negligent manslaughter, forcible rape, robbery, and aggravated assault.¹¹

Earlier descriptions of ASD/AD did not mention criminal violence as an important feature of these disorders. However, reports began to emerge about two decades ago suggesting that people who have ASD—particularly AD—are prone to violent crime. Some of the patients described in Wing’s original series¹² of AD showed violent tendencies, ranging from sudden outbursts of violence to injury to others because of fixation on hobbies such as chemistry experimentation.

Reports such as these were based on isolated case reports or select samples, such as residents of maximum-security hospitals. Scragg and Shah, for example, surveyed the male population of Broadmoor Hospital, a high-security facility in the United Kingdom, and found that the prevalence of AD was higher than expected in the general population.¹³

Recent reports have not been able to confirm that violent crime is increased in persons with ASD, however:

• In a clinical sample of 313 Danish adults with ASD (age 25 to 59) drawn from the Danish Register of Criminality, Mouridsen and colleagues found that persons with ASD had a lower rate of criminal conviction than matched controls (9%, compared with 18%).¹⁴
• In a small community study, Woodbury-Smith and colleagues examined the prevalence rates and types of offending behavior in persons with ASD. Based on official records, only two (18%) had a history of criminal conviction.¹⁵

The role of psychiatric comorbidity

Psychiatric disorders are common in persons who have ASD. In one study, 70% of a sample of 114 children with ASD (age 10 to 14) had a psychiatric disorder, based on a parent interview.¹⁶ Although people with mental illness are not inherently criminal or violent, having an additional psychiatric disorder independently increases the risk of offending behavior.¹⁷ For example, the association of attention-deficit/hyperactivity disorder with criminality is well established.¹⁶ Some patients with severe depression and psychotic disorders, including schizophrenia, also are at increased risk of committing a violent act.

To examine the contribution of mental health factors to the commission of crime by persons with ASD, Newman and Ghaziuddin¹⁸ used online databases to identify relevant articles, which were then cross-referenced with keyword searches for “violence,” “crime,” “murder,” “assault,” “rape,” and “sex offenses.” Thirty-seven cases were identified in the 17 publications that met inclusion criteria. Out of these, 30% had a definite psychiatric disorder and 54% had a probable psychiatric disorder at the time they committed the crime.¹⁸

Any patient with ASD/AD who is evaluated for criminal behavior should be screened for a comorbid psychiatric disorder. In adolescents, stressors such as bullying in school and problems surrounding dating might contribute to offending behavior.

What are management options in the face of violence?

Managing ASD/AD when an offending behavior has occurred first requires a correct diagnosis.¹⁹ Professionals working in the criminal justice system have little awareness of the variants of ASD; a defendant with an intellectual disability and a characteristic facial appearance (for example, someone with Down syndrome) can be easily identified, but a high-functioning person who has mild autistic features often is missed. This is more likely to occur in adults because the symptoms of ASD, including the type and severity of isolated interests, change over time.

Here is how I recommend that you proceed:

Step #1. Confirm the ASD diagnosis based on developmental history and the presence of persistent social and communication deficits plus restricted interests.

Step #2. Screen for comorbid psychiatric and medical disorders, including depression, psychosis, and seizure disorder.

Step #3. Treat any disorders you identify with a combination of medication and behavioral intervention.

Step #4. Carefully examine the circumstances surrounding the offending behavior. Involve forensic services on a case-by-case basis, depending on the type and seriousness of the offending behavior (see Related Resources for information on the role of forensic services). When the crime does not involve serious violence, lengthy incarceration might be unnecessary. Because psychopathy and ASD/AD are not mutually exclusive, persons who commit a heinous crime, such as rape or murder, should be dealt with in accordance with the law.

Need for greater awareness of the complexion of ASD

Patients who have ASD/AD form a heterogeneous group in which the levels of cognitive and communication skills are variable. Those who are low-functioning and who have severe behavioral and adaptive deficits occasionally commit aggressive acts against their caregivers.

Most patients with ASD/AD are neither violent nor criminal. Those who are at the higher end of the spectrum, with relatively preserved communication and intellectual skills, occasionally indulge in criminal behavior—behavior that is nonviolent and results from their inability to read social cues or excessive preoccupations.

Most reports that link criminal violence with ASD are based on isolated case reports or on biased samples that use unreliable diagnostic criteria. In higher-functioning persons with ASD, violent crime is almost always precipitated by a comorbid psychiatric disorder, such as severe depression and psychosis.

In short: There is a need to increase our awareness of the special challenges faced by persons with ASD/AD in the criminal justice system.

aGiven the term pervasive developmental disorders (PDD) in the DSM-IV-TR, the spectrum includes autistic disorder, Asperger’s disorder, and pervasive developmental disorder not otherwise specified.2

Bottom Line

Most people who have an autism spectrum disorder (ASD) do not commit violent crime. When violent crime occurs at the hands of a person with ASD, it is almost always precipitated by a comorbid psychiatric disorder, such as severe depression or psychosis. Treating a person with ASD who has committed a violent crime is multimodal, including forensic services when necessary.

Related Resources

• Autism Speaks. No link between autism and violence. www.autismspeaks.org/science/science-news/no-link-between-autism-and-violence.
• Haskins BG, Silva JA. Asperger’s disorder and criminal behavior: Forensic-psychiatric considerations. J Am Acad Psychiatry Law. 2006;34(3):374-384.
• Newman SS, Ghaziuddin M. Violent crime and Asperger syndrome: the role of psychiatric comorbidity. J Autism Dev Disord. 2008;38:1848-1852.
• Wing L. Asperger’s syndrome: a clinical account. Psychol Med. 1981;11(1):115-129.

Disclosure

Dr. Ghaziuddin reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

The economic downturn in the United States has prompted numerous state and county budget cuts, in turn forcing many patients to receive their mental health care in the emergency room (ER). Most patients evaluated in the ER for mental health-related reasons have a legitimate psychiatric crisis—but that isn’t always the case. And as the number of people seeking care in the ER has increased, it appears that so too has the number of those who feign symptoms for secondary gain—that is, who are malingering.

This article highlights several red flags for malingered behavior; emphasizes typical (compared with atypical) symptoms of psychosis; and provides an overview of four instruments that you can use to help assess for malingering in the ED.

A difficult diagnosis

No single factor is indicative of malingering, and no objective tests exist to diagnose malingering definitively. Rather, the tests we discuss provide additional information that can help formulate a clinical impression. 

According to DSM-5, malingering is “…the intentional production of false or grossly exaggerated physical or psychological symptoms, motivated by external incentives…”¹ Despite a relatively straightforward definition, the diagnosis is difficult to make because it is a diagnosis of exclusion.

Even with sufficient evidence, many clinicians are reluctant to diagnose malingering because they fear retaliation and diagnostic uncertainty. Psychiatrists also might be reluctant to diagnose malingering because the negative connotation that the label carries risks stigmatizing a patient who might, in fact, be suffering. This is true especially when there is suspicion of partial malingering, the conscious exaggeration of existing symptoms.

Despite physicians’ reluctance to diagnose malingering, it is a real problem, especially in the ER. Research suggests that as many as 13% of patients in the ER feign illness, and that their secondary gain most often includes food, shelter, prescription drugs, financial gain, and avoidance of jail, work, or family responsibilities.²

CASE REPORT ‘The voices are telling me to kill myself’

Mr. K, a 36-year-old white man, walks into the ER on a late December day. He tells the triage nurse that he suicidal; she escorts him to the psychiatric pod of the ER. Nursing staff provide line-of-sight care, monitor his vital signs, and draw blood for testing.

Within hours, Mr. K is deemed “medically cleared” and ready for assessment by the psychiatric social worker.

Interview and assessment. During the interview with the social worker, Mr. K reports that he has been depressed, adamantly maintaining that he is suicidal, with a plan to “walk in traffic” or “eat the end of a gun.” The social worker places him on a 72-hour involuntary psychiatric hold. ER physicians order psychiatric consultation.

Mr. K is well-known to the psychiatrist on call, from prior ER visits and psychiatric hospital admissions. In fact, two days earlier, he put a psychiatric nurse in a headlock while being escorted from the psychiatric inpatient unit under protest.

On assessment by the psychiatrist, Mr. K continues to endorse feeling suicidal; he adds: “If I don’t get some help, I’m gonna kill somebody else!”

Without prompting, the patient states that “the voices are telling me to kill myself.” He says that those voices have been relentless since he left the hospital two days earlier. According to Mr. K, nothing he did helped quiet the voices, although previous prescriptions for quetiapine have been helpful.

Mr. K says that he is unable to recall the clinic or name of his prior psychiatrist. He claims that he was hospitalized four months ago, (despite the psychiatrist’s knowledge that he had been discharged two days ago) and estimates that his psychotic symptoms began one year ago. He explains that he is homeless and does not have social support. He is unable to provide a telephone number or a name to contact family for collateral information.

Mental status exam. The mental status examination reveals a tall, thin, disheveled man who has poor dentition. He is now calm and cooperative despite his reported level of distress. His speech is unremarkable and his eye contact is appropriate. His thought process is linear, organized, and coherent.

Mr. K does not endorse additional symptoms, but is quick to agree with the psychiatrist’s follow-up questions about hallucinations: “Yeah! I’ve been seeing all kinds of crazy stuff.” When prompted for details, he says, “I just saw Big Bird… He was 100 feet tall!”

Lab testing. Mr. K’s blood work is remarkable for positive urine toxicology for amphetamines.

Nursing notes indicate that Mr. K slept overnight and ate 100% of the food on his dinner and breakfast trays.

Red flags flying

Mr. K’s case highlights several red flags that should raise suspicion of malingering (Table 1)^3,4:

• A conditional statement by which a patient threatens to harm himself or others, contingent upon a demand—for example, “If I don’t get A, I’ll do B.”
• An overly dramatic presentation, in which the patient is quick to endorse
  distressing symptoms. Consider Mr. K: He was quick to report that he saw Big Bird, and that this Sesame Street character “was 100 feet tall.” Patients who have been experiencing true psychotic symptoms might be reluctant to speak of their distressing symptoms, especially if they have not experienced such symptoms in the past (the first psychotic break). Mr. K, however, volunteered and called attention to particularly dramatic psychotic symptoms.
• A subjective report of distress that is inconsistent with the objective presentation. Mr. K’s report of depression—a diagnosis that typically includes insomnia and poor appetite—was inconsistent with his behavior: He slept and he ate all of his meals.

Atypical (vs typical) psychosis

Malingering can occur in various arenas and take many different forms. In forensic settings, such as prison, malingered conditions more often present as posttraumatic stress disorder or cognitive impairment.⁵ In non-forensic settings, such as the ER, the most commonly malingered conditions include suicidality and psychosis.

To detect malingered psychosis, one must first understand how true psychotic symptoms manifest. The following discussion describes and compares typical and atypical symptoms of psychosis; examples are given in Table 2.^6,7No single atypical psychotic symptom is indicative of malingering. Rather, a collection of atypical symptoms, when considered in clinical context, should raise suspicion of malingering and prompt you to seek additional collateral information or perform appropriate testing for malingering.

Hallucinations

Typically, hallucinations take three forms: auditory, visual, and tactile. In primary psychiatric conditions, auditory hallucinations are the most common of those three.

Tactile hallucinations can be present during episodes of substance intoxication or withdrawal (eg, so-called coke bugs). 

Auditory hallucinations. Patients who malinger psychosis are often unaware of the nuances of hallucinations. For example, they might report the atypical symptom of continuous voices; in fact, most patients who have schizophrenia hear voices intermittently. Keep in mind, too, that 75% of patients who have schizophrenia hear male and female voices, and that 70% have some type of coping strategy to minimize their internal stimuli (eg, listening to music).^6,7

Visual hallucinations are most often associated with neurologic disease, but also occur often in primary psychotic disorders, such as schizophrenia.

Patients who malinger psychotic symptoms often are open to suggestion, and are quick to endorse visual hallucinations. When asked to describe their hallucinations, however, they often respond without details (“I don’t know”). Other times, they overcompensate with wild exaggeration of atypical visions—recall Mr. K’s description of a towering Big Bird. Asked if the visions are in black and white, they might eagerly agree. Research suggests, however, that patients who have schizophrenia more often experience life-sized hallucinations of vivid scenes with family members, religious figures, or animals.⁸ Furthermore, genuine visual hallucinations typically are in color.

Putting malingering in the differential

Regardless of the number of atypical symptoms a patient exhibits, malingering will be missed if you do not include it in the differential diagnosis. This fact was made evident in a 1973 study.⁹

In that study, Rosenhan and seven of his colleagues—a psychology graduate student, three psychologists, a pediatrician, a psychiatrist, a painter, and a housewife—presented to various ERs and intake units, and, as they had been instructed, endorsed vague auditory hallucinations of “empty,” “hollow,” or “thud” sounds—but nothing more. All were admitted to psychiatric hospitals. Once admitted, they refrained (again, as instructed) from endorsing or exhibiting any psychotic symptoms.

Despite the vague nature of the reported auditory hallucinations and how rapidly symptoms resolved on admission, seven of these pseudo-patients were given a diagnosis of schizophrenia, and one was given a diagnosis of manic-depressive psychosis. Duration of admission ranged from 7 to 52 days (average, 19 days). None of the study participants were suspected of feigning symptoms.

It’s fortunate that, since then, mental health professionals have developed more structured techniques of assessment to detect malingering in inpatient and triage settings.

Testing to identify and assess malingering

The ER is a fast-paced environment, in which treatment teams are challenged to make rapid clinical assessments. With the overwhelming number of patients seeking mental health care in the ER, however, overall wait times are increasing; in some regions, it is common to write, then to rewrite, involuntary psychiatric holds for patients awaiting transfer to a psychiatric hospital. This extended duration presents an opportunity to serially evaluate patients suspected of malingering.

Even in environments that allow for a more comprehensive evaluation (eg, jail or inpatient psychiatric wards), few psychometric tests have been validated to detect malingering. The most validated tests include the Structured Interview of Reported Symptoms (SIRS), distributed now as the Structured Interview of Reported Symptoms, 2nd edition (SIRS-2), and the Minnesota Multiphasic Personality Inventory Revised (MMPI-2). These tests typically require ≥30 minutes to administer and generally are not feasible in the fast-paced ER.

Despite the high prevalence of malingered behaviors in the ER, no single test has been validated in such a setting. Furthermore, there is no test designed to specifically assess for malingered suicidality or homicidality. The results of one test do not, in isolation, represent a comprehensive neuropsychological examination; rather, those results provide additional data to formulate a clinical impression. The instruments discussed below are administered and scored in a defined, objective manner.

When evaluating a patient whom you suspect of malingering, gathering collateral information—from family members, friends, nurses, social workers, emergency medicine physicians, and others—becomes important. You might discover pertinent information in ambulance and police reports and a review of the patient’s prior ER visits.

During the initial interview, ask open-ended questions; do not lead the patient by listing clusters of symptoms associated with a particular diagnosis. Because it is often difficult for a patient to malinger symptoms for a prolonged period, serial observations of a patient’s behavior and interview responses over time can provide additional information to make a clinical diagnosis of malingering.⁴

What testing is feasible in the ER?

Miller Forensic Assessment of Symptoms Test. The M-FAST measures rare symptom combinations, excessive reporting, and atypical symptoms of psychosis, using the same principles as the SIRS-2.

The 25-item screen begins by advising the examinee that he (she) will be asked questions about his psychological symptoms and that the questions that follow might or might not apply to his specific symptoms.

After that brief introduction, the examinee is asked if he hears ringing in his ears. Based on his response, the examiner reads one of two responses—both of which suggest the false notion that patients with true mental illness will suffer from ringing in their ears.

The examinee is then asked a series of Yes or No questions. Some pertain to legitimate symptoms a person with a psychotic illness might suffer (such as, “Do voices tell you to do things? Yes or No?”). Conversely, other questions screen for improbable symptoms that are atypical of patients who have a true psychotic disorder (such as “On many days I feel so bad that I can’t even remember my full name: Yes or No?”).

The exam concludes with a question about a ringing in the examinee’s ear. Affirmative responses are tallied; a score of ≥6 in a clinical setting is 83% specific and 93% sensitive for malingering.¹⁰

Visual Memory Test. Rey’s 15-Item Visual Memory Test capitalizes on the false belief that intellectual deficits, in addition to psychotic symptoms, make a claim of mental illness more believable.

In this simple test, the provider tells the examinee, “I am going to show you a card with 15 things on it that I want you to remember. When I take the card away, I want you to write down as many of the 15 things as you can remember.”³ The examinee is shown 15 common symbols (eg, 1, 2, 3; A, B, C; I, II, III, a, b, c; and the geometrics ●, ■, ▲).

At 5 seconds, the examinee is prompted, “Be sure to remember all of them.” After 10 seconds, the stimulus is removed, and the examinee is asked to recreate the figure.

Normative data indicate that even a patient who has a severe traumatic brain injury is able to recreate at least eight of the symbols. Although controversial, research indicates that a score of <9 symbols is predictive of malingering with 40% sensitivity and 100% specificity.¹¹

Critics argued that confounding variables (IQ, memory disorder, age) might skew the quantitative score. For that reason, the same group developed the Rey’s II Test, which includes a supplementary qualitative scoring system that emphasizes embellishment errors (eg, the wrong symbol) and ordering errors (eg, wrong row). The Rey’s II Test proved to be more sensitive (accurate classification of malingers): A cut-off score of ≥2 qualitative errors is predictive of malingering with 86% sensitivity and 100% specificity.¹²

Coin-in-the-Hand Test. Perhaps the simplest test to administer is the Coin-in-the-Hand, designed to seem—superficially—to be a challenging memory test.

The patient must guess in which hand the examiner is holding a coin. The patient is shown the coin for two seconds, and then asked to close his eyes and count back from 10. The patient then points to one of the two clenched hands.

This task is repeated 10 times; each time, the provider gives verbal feedback about the accuracy or inaccuracy of that attempt. Studies indicate that a patient who has a severe traumatic brain injury is able to score 85% correct. A score <85%, however, suggests feigning of symptoms (sensitivity, 92.5%; specificity 87.5%).¹³ Hanley and co-workers demonstrated that people who are simulating cognitive impairment had a mean accurate response of 4.1, whereas people who had true amnesia had a mean accurate response of 9.65.¹⁴

Persons who feign psychosis or mood symptoms often inaccurately believe that people with mental illness also have cognitive impairment. Both Rey’s test and the Coin-in-the-Hand Test capitalize on this misconception.

Mini-Mental State Examination. Research also has shown that the Folstein Mini-Mental State Examination (MMSE) can screen for malingered cognitive impairment. Powell compared 40 mental health clinicians who were instructed to feign psychosis and 40 patients with schizophrenia. Using the MMSE, the researchers found that the malingers more often gave approximate answers.¹⁵ Moreover, Myers argued that, when compared with Rey’s Test, the MMSE is superior for assessing malingered cognitive impairment because it has a higher positive predictive value (67%, compared with 43% for Rey’s Test) and a higher negative predictive value (93% and 89%).¹⁶

What can you do for these patients after diagnosis?

Malingering is not considered a psychiatric diagnosis; there are no indicated therapies with which to manage it—only guidelines. When you suspect a patient of malingering, you should avoid accusing him (her) of faking symptoms. Rather, when feasible, gently confront the person and provide the opportunity for him to explain his current behaviors. For example, you might say: “I’ve treated many patients with the symptoms that you’re reporting, but the details you provide are different, and don’t ring completely true. Is there anything else that could explain this?”¹⁷

Regardless of a patient’s challenging behaviors, it is important to remember that people who feign illness—whether partial malingering or pure malingering—often do need help. The assistance they require, however, might be best obtained from a housing agency, a chemical dependency program, or another social service—not from the ER. Identifying malingered behaviors saves time and money and shifts limited resources to people who have a legitimate mental health condition.

Last, despite an empathetic approach, some malingering patients continue to feign symptoms—as Mr. K did.

CASE CONTINUED

Although the psychiatrist on call considered forsaking the police to escort Mr. K out of the ER, he eventually agreed to leave the hospital on his own, stating, “My death is going to be on your hands.”

Eight days later, Mr. K visited the ER at a different hospital, endorsing chronic pain and demanding narcotics.

Bottom Line

As the number of people seeking care in the emergency room (ER) has increased, so has the number of those who feign symptoms for secondary gain. No single factor is indicative of malingering, and no objective tests exist to diagnose it definitively. Furthermore, there are no indicated therapies with which to manage malingering—only guidelines. Keep in mind that people who feign illness, whether partial or pure malingering, often do need help—although not the services of an ER.

Related Resources

• Miller Forensic Assessment of Symptoms Test (M-FAST). Psychological Assessment Resources, Inc. www4.parinc.com (enter “M-FAST” in search field).
• Duffy S. Malingering psychological symptoms: An empirical review. Illinois State University, Department of Psychology. http://psychology.illinoisstate.edu/cc/Comps/Duffy%20-%20Malingering.pdf. Accessed September 10, 2013.

Drug Brand Names

Quetiapine • Seroquel

Disclosure

The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers  of competing products.

Featured Audio
M. Cait Brady, MD, shares strategies for assessing malingering. Dr. Brady is a Third-Year Resident in General Psychiatry, University of California, Davis Medical Center - Sacramento, Sacramento, California.

The economic downturn in the United States has prompted numerous state and county budget cuts, in turn forcing many patients to receive their mental health care in the emergency room (ER). Most patients evaluated in the ER for mental health-related reasons have a legitimate psychiatric crisis—but that isn’t always the case. And as the number of people seeking care in the ER has increased, it appears that so too has the number of those who feign symptoms for secondary gain—that is, who are malingering.

This article highlights several red flags for malingered behavior; emphasizes typical (compared with atypical) symptoms of psychosis; and provides an overview of four instruments that you can use to help assess for malingering in the ED.

A difficult diagnosis

No single factor is indicative of malingering, and no objective tests exist to diagnose malingering definitively. Rather, the tests we discuss provide additional information that can help formulate a clinical impression. 

According to DSM-5, malingering is “…the intentional production of false or grossly exaggerated physical or psychological symptoms, motivated by external incentives…”¹ Despite a relatively straightforward definition, the diagnosis is difficult to make because it is a diagnosis of exclusion.

Even with sufficient evidence, many clinicians are reluctant to diagnose malingering because they fear retaliation and diagnostic uncertainty. Psychiatrists also might be reluctant to diagnose malingering because the negative connotation that the label carries risks stigmatizing a patient who might, in fact, be suffering. This is true especially when there is suspicion of partial malingering, the conscious exaggeration of existing symptoms.

Despite physicians’ reluctance to diagnose malingering, it is a real problem, especially in the ER. Research suggests that as many as 13% of patients in the ER feign illness, and that their secondary gain most often includes food, shelter, prescription drugs, financial gain, and avoidance of jail, work, or family responsibilities.²

CASE REPORT ‘The voices are telling me to kill myself’

Mr. K, a 36-year-old white man, walks into the ER on a late December day. He tells the triage nurse that he suicidal; she escorts him to the psychiatric pod of the ER. Nursing staff provide line-of-sight care, monitor his vital signs, and draw blood for testing.

Within hours, Mr. K is deemed “medically cleared” and ready for assessment by the psychiatric social worker.

Interview and assessment. During the interview with the social worker, Mr. K reports that he has been depressed, adamantly maintaining that he is suicidal, with a plan to “walk in traffic” or “eat the end of a gun.” The social worker places him on a 72-hour involuntary psychiatric hold. ER physicians order psychiatric consultation.

Mr. K is well-known to the psychiatrist on call, from prior ER visits and psychiatric hospital admissions. In fact, two days earlier, he put a psychiatric nurse in a headlock while being escorted from the psychiatric inpatient unit under protest.

On assessment by the psychiatrist, Mr. K continues to endorse feeling suicidal; he adds: “If I don’t get some help, I’m gonna kill somebody else!”

Without prompting, the patient states that “the voices are telling me to kill myself.” He says that those voices have been relentless since he left the hospital two days earlier. According to Mr. K, nothing he did helped quiet the voices, although previous prescriptions for quetiapine have been helpful.

Mr. K says that he is unable to recall the clinic or name of his prior psychiatrist. He claims that he was hospitalized four months ago, (despite the psychiatrist’s knowledge that he had been discharged two days ago) and estimates that his psychotic symptoms began one year ago. He explains that he is homeless and does not have social support. He is unable to provide a telephone number or a name to contact family for collateral information.

Mental status exam. The mental status examination reveals a tall, thin, disheveled man who has poor dentition. He is now calm and cooperative despite his reported level of distress. His speech is unremarkable and his eye contact is appropriate. His thought process is linear, organized, and coherent.

Mr. K does not endorse additional symptoms, but is quick to agree with the psychiatrist’s follow-up questions about hallucinations: “Yeah! I’ve been seeing all kinds of crazy stuff.” When prompted for details, he says, “I just saw Big Bird… He was 100 feet tall!”

Lab testing. Mr. K’s blood work is remarkable for positive urine toxicology for amphetamines.

Nursing notes indicate that Mr. K slept overnight and ate 100% of the food on his dinner and breakfast trays.

Red flags flying

Mr. K’s case highlights several red flags that should raise suspicion of malingering (Table 1)^3,4:

• A conditional statement by which a patient threatens to harm himself or others, contingent upon a demand—for example, “If I don’t get A, I’ll do B.”
• An overly dramatic presentation, in which the patient is quick to endorse
  distressing symptoms. Consider Mr. K: He was quick to report that he saw Big Bird, and that this Sesame Street character “was 100 feet tall.” Patients who have been experiencing true psychotic symptoms might be reluctant to speak of their distressing symptoms, especially if they have not experienced such symptoms in the past (the first psychotic break). Mr. K, however, volunteered and called attention to particularly dramatic psychotic symptoms.
• A subjective report of distress that is inconsistent with the objective presentation. Mr. K’s report of depression—a diagnosis that typically includes insomnia and poor appetite—was inconsistent with his behavior: He slept and he ate all of his meals.

Atypical (vs typical) psychosis

Malingering can occur in various arenas and take many different forms. In forensic settings, such as prison, malingered conditions more often present as posttraumatic stress disorder or cognitive impairment.⁵ In non-forensic settings, such as the ER, the most commonly malingered conditions include suicidality and psychosis.

To detect malingered psychosis, one must first understand how true psychotic symptoms manifest. The following discussion describes and compares typical and atypical symptoms of psychosis; examples are given in Table 2.^6,7No single atypical psychotic symptom is indicative of malingering. Rather, a collection of atypical symptoms, when considered in clinical context, should raise suspicion of malingering and prompt you to seek additional collateral information or perform appropriate testing for malingering.

Hallucinations

Typically, hallucinations take three forms: auditory, visual, and tactile. In primary psychiatric conditions, auditory hallucinations are the most common of those three.

Tactile hallucinations can be present during episodes of substance intoxication or withdrawal (eg, so-called coke bugs). 

Auditory hallucinations. Patients who malinger psychosis are often unaware of the nuances of hallucinations. For example, they might report the atypical symptom of continuous voices; in fact, most patients who have schizophrenia hear voices intermittently. Keep in mind, too, that 75% of patients who have schizophrenia hear male and female voices, and that 70% have some type of coping strategy to minimize their internal stimuli (eg, listening to music).^6,7

Visual hallucinations are most often associated with neurologic disease, but also occur often in primary psychotic disorders, such as schizophrenia.

Patients who malinger psychotic symptoms often are open to suggestion, and are quick to endorse visual hallucinations. When asked to describe their hallucinations, however, they often respond without details (“I don’t know”). Other times, they overcompensate with wild exaggeration of atypical visions—recall Mr. K’s description of a towering Big Bird. Asked if the visions are in black and white, they might eagerly agree. Research suggests, however, that patients who have schizophrenia more often experience life-sized hallucinations of vivid scenes with family members, religious figures, or animals.⁸ Furthermore, genuine visual hallucinations typically are in color.

Putting malingering in the differential

Regardless of the number of atypical symptoms a patient exhibits, malingering will be missed if you do not include it in the differential diagnosis. This fact was made evident in a 1973 study.⁹

In that study, Rosenhan and seven of his colleagues—a psychology graduate student, three psychologists, a pediatrician, a psychiatrist, a painter, and a housewife—presented to various ERs and intake units, and, as they had been instructed, endorsed vague auditory hallucinations of “empty,” “hollow,” or “thud” sounds—but nothing more. All were admitted to psychiatric hospitals. Once admitted, they refrained (again, as instructed) from endorsing or exhibiting any psychotic symptoms.

Despite the vague nature of the reported auditory hallucinations and how rapidly symptoms resolved on admission, seven of these pseudo-patients were given a diagnosis of schizophrenia, and one was given a diagnosis of manic-depressive psychosis. Duration of admission ranged from 7 to 52 days (average, 19 days). None of the study participants were suspected of feigning symptoms.

It’s fortunate that, since then, mental health professionals have developed more structured techniques of assessment to detect malingering in inpatient and triage settings.

Testing to identify and assess malingering

The ER is a fast-paced environment, in which treatment teams are challenged to make rapid clinical assessments. With the overwhelming number of patients seeking mental health care in the ER, however, overall wait times are increasing; in some regions, it is common to write, then to rewrite, involuntary psychiatric holds for patients awaiting transfer to a psychiatric hospital. This extended duration presents an opportunity to serially evaluate patients suspected of malingering.

Even in environments that allow for a more comprehensive evaluation (eg, jail or inpatient psychiatric wards), few psychometric tests have been validated to detect malingering. The most validated tests include the Structured Interview of Reported Symptoms (SIRS), distributed now as the Structured Interview of Reported Symptoms, 2nd edition (SIRS-2), and the Minnesota Multiphasic Personality Inventory Revised (MMPI-2). These tests typically require ≥30 minutes to administer and generally are not feasible in the fast-paced ER.

Despite the high prevalence of malingered behaviors in the ER, no single test has been validated in such a setting. Furthermore, there is no test designed to specifically assess for malingered suicidality or homicidality. The results of one test do not, in isolation, represent a comprehensive neuropsychological examination; rather, those results provide additional data to formulate a clinical impression. The instruments discussed below are administered and scored in a defined, objective manner.

When evaluating a patient whom you suspect of malingering, gathering collateral information—from family members, friends, nurses, social workers, emergency medicine physicians, and others—becomes important. You might discover pertinent information in ambulance and police reports and a review of the patient’s prior ER visits.

During the initial interview, ask open-ended questions; do not lead the patient by listing clusters of symptoms associated with a particular diagnosis. Because it is often difficult for a patient to malinger symptoms for a prolonged period, serial observations of a patient’s behavior and interview responses over time can provide additional information to make a clinical diagnosis of malingering.⁴

What testing is feasible in the ER?

Miller Forensic Assessment of Symptoms Test. The M-FAST measures rare symptom combinations, excessive reporting, and atypical symptoms of psychosis, using the same principles as the SIRS-2.

The 25-item screen begins by advising the examinee that he (she) will be asked questions about his psychological symptoms and that the questions that follow might or might not apply to his specific symptoms.

After that brief introduction, the examinee is asked if he hears ringing in his ears. Based on his response, the examiner reads one of two responses—both of which suggest the false notion that patients with true mental illness will suffer from ringing in their ears.

The examinee is then asked a series of Yes or No questions. Some pertain to legitimate symptoms a person with a psychotic illness might suffer (such as, “Do voices tell you to do things? Yes or No?”). Conversely, other questions screen for improbable symptoms that are atypical of patients who have a true psychotic disorder (such as “On many days I feel so bad that I can’t even remember my full name: Yes or No?”).

The exam concludes with a question about a ringing in the examinee’s ear. Affirmative responses are tallied; a score of ≥6 in a clinical setting is 83% specific and 93% sensitive for malingering.¹⁰

Visual Memory Test. Rey’s 15-Item Visual Memory Test capitalizes on the false belief that intellectual deficits, in addition to psychotic symptoms, make a claim of mental illness more believable.

In this simple test, the provider tells the examinee, “I am going to show you a card with 15 things on it that I want you to remember. When I take the card away, I want you to write down as many of the 15 things as you can remember.”³ The examinee is shown 15 common symbols (eg, 1, 2, 3; A, B, C; I, II, III, a, b, c; and the geometrics ●, ■, ▲).

At 5 seconds, the examinee is prompted, “Be sure to remember all of them.” After 10 seconds, the stimulus is removed, and the examinee is asked to recreate the figure.

Normative data indicate that even a patient who has a severe traumatic brain injury is able to recreate at least eight of the symbols. Although controversial, research indicates that a score of <9 symbols is predictive of malingering with 40% sensitivity and 100% specificity.¹¹

Critics argued that confounding variables (IQ, memory disorder, age) might skew the quantitative score. For that reason, the same group developed the Rey’s II Test, which includes a supplementary qualitative scoring system that emphasizes embellishment errors (eg, the wrong symbol) and ordering errors (eg, wrong row). The Rey’s II Test proved to be more sensitive (accurate classification of malingers): A cut-off score of ≥2 qualitative errors is predictive of malingering with 86% sensitivity and 100% specificity.¹²

Coin-in-the-Hand Test. Perhaps the simplest test to administer is the Coin-in-the-Hand, designed to seem—superficially—to be a challenging memory test.

The patient must guess in which hand the examiner is holding a coin. The patient is shown the coin for two seconds, and then asked to close his eyes and count back from 10. The patient then points to one of the two clenched hands.

This task is repeated 10 times; each time, the provider gives verbal feedback about the accuracy or inaccuracy of that attempt. Studies indicate that a patient who has a severe traumatic brain injury is able to score 85% correct. A score <85%, however, suggests feigning of symptoms (sensitivity, 92.5%; specificity 87.5%).¹³ Hanley and co-workers demonstrated that people who are simulating cognitive impairment had a mean accurate response of 4.1, whereas people who had true amnesia had a mean accurate response of 9.65.¹⁴

Persons who feign psychosis or mood symptoms often inaccurately believe that people with mental illness also have cognitive impairment. Both Rey’s test and the Coin-in-the-Hand Test capitalize on this misconception.

Mini-Mental State Examination. Research also has shown that the Folstein Mini-Mental State Examination (MMSE) can screen for malingered cognitive impairment. Powell compared 40 mental health clinicians who were instructed to feign psychosis and 40 patients with schizophrenia. Using the MMSE, the researchers found that the malingers more often gave approximate answers.¹⁵ Moreover, Myers argued that, when compared with Rey’s Test, the MMSE is superior for assessing malingered cognitive impairment because it has a higher positive predictive value (67%, compared with 43% for Rey’s Test) and a higher negative predictive value (93% and 89%).¹⁶

What can you do for these patients after diagnosis?

Malingering is not considered a psychiatric diagnosis; there are no indicated therapies with which to manage it—only guidelines. When you suspect a patient of malingering, you should avoid accusing him (her) of faking symptoms. Rather, when feasible, gently confront the person and provide the opportunity for him to explain his current behaviors. For example, you might say: “I’ve treated many patients with the symptoms that you’re reporting, but the details you provide are different, and don’t ring completely true. Is there anything else that could explain this?”¹⁷

Regardless of a patient’s challenging behaviors, it is important to remember that people who feign illness—whether partial malingering or pure malingering—often do need help. The assistance they require, however, might be best obtained from a housing agency, a chemical dependency program, or another social service—not from the ER. Identifying malingered behaviors saves time and money and shifts limited resources to people who have a legitimate mental health condition.

Last, despite an empathetic approach, some malingering patients continue to feign symptoms—as Mr. K did.

CASE CONTINUED

Although the psychiatrist on call considered forsaking the police to escort Mr. K out of the ER, he eventually agreed to leave the hospital on his own, stating, “My death is going to be on your hands.”

Eight days later, Mr. K visited the ER at a different hospital, endorsing chronic pain and demanding narcotics.

Bottom Line

As the number of people seeking care in the emergency room (ER) has increased, so has the number of those who feign symptoms for secondary gain. No single factor is indicative of malingering, and no objective tests exist to diagnose it definitively. Furthermore, there are no indicated therapies with which to manage malingering—only guidelines. Keep in mind that people who feign illness, whether partial or pure malingering, often do need help—although not the services of an ER.

Related Resources

• Miller Forensic Assessment of Symptoms Test (M-FAST). Psychological Assessment Resources, Inc. www4.parinc.com (enter “M-FAST” in search field).
• Duffy S. Malingering psychological symptoms: An empirical review. Illinois State University, Department of Psychology. http://psychology.illinoisstate.edu/cc/Comps/Duffy%20-%20Malingering.pdf. Accessed September 10, 2013.

Drug Brand Names

Quetiapine • Seroquel

Disclosure

The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers  of competing products.

Featured Audio
M. Cait Brady, MD, shares strategies for assessing malingering. Dr. Brady is a Third-Year Resident in General Psychiatry, University of California, Davis Medical Center - Sacramento, Sacramento, California.

The economic downturn in the United States has prompted numerous state and county budget cuts, in turn forcing many patients to receive their mental health care in the emergency room (ER). Most patients evaluated in the ER for mental health-related reasons have a legitimate psychiatric crisis—but that isn’t always the case. And as the number of people seeking care in the ER has increased, it appears that so too has the number of those who feign symptoms for secondary gain—that is, who are malingering.

This article highlights several red flags for malingered behavior; emphasizes typical (compared with atypical) symptoms of psychosis; and provides an overview of four instruments that you can use to help assess for malingering in the ED.

A difficult diagnosis

No single factor is indicative of malingering, and no objective tests exist to diagnose malingering definitively. Rather, the tests we discuss provide additional information that can help formulate a clinical impression. 

According to DSM-5, malingering is “…the intentional production of false or grossly exaggerated physical or psychological symptoms, motivated by external incentives…”¹ Despite a relatively straightforward definition, the diagnosis is difficult to make because it is a diagnosis of exclusion.

Even with sufficient evidence, many clinicians are reluctant to diagnose malingering because they fear retaliation and diagnostic uncertainty. Psychiatrists also might be reluctant to diagnose malingering because the negative connotation that the label carries risks stigmatizing a patient who might, in fact, be suffering. This is true especially when there is suspicion of partial malingering, the conscious exaggeration of existing symptoms.

Despite physicians’ reluctance to diagnose malingering, it is a real problem, especially in the ER. Research suggests that as many as 13% of patients in the ER feign illness, and that their secondary gain most often includes food, shelter, prescription drugs, financial gain, and avoidance of jail, work, or family responsibilities.²

CASE REPORT ‘The voices are telling me to kill myself’

Mr. K, a 36-year-old white man, walks into the ER on a late December day. He tells the triage nurse that he suicidal; she escorts him to the psychiatric pod of the ER. Nursing staff provide line-of-sight care, monitor his vital signs, and draw blood for testing.

Within hours, Mr. K is deemed “medically cleared” and ready for assessment by the psychiatric social worker.

Interview and assessment. During the interview with the social worker, Mr. K reports that he has been depressed, adamantly maintaining that he is suicidal, with a plan to “walk in traffic” or “eat the end of a gun.” The social worker places him on a 72-hour involuntary psychiatric hold. ER physicians order psychiatric consultation.

Mr. K is well-known to the psychiatrist on call, from prior ER visits and psychiatric hospital admissions. In fact, two days earlier, he put a psychiatric nurse in a headlock while being escorted from the psychiatric inpatient unit under protest.

On assessment by the psychiatrist, Mr. K continues to endorse feeling suicidal; he adds: “If I don’t get some help, I’m gonna kill somebody else!”

Without prompting, the patient states that “the voices are telling me to kill myself.” He says that those voices have been relentless since he left the hospital two days earlier. According to Mr. K, nothing he did helped quiet the voices, although previous prescriptions for quetiapine have been helpful.

Mr. K says that he is unable to recall the clinic or name of his prior psychiatrist. He claims that he was hospitalized four months ago, (despite the psychiatrist’s knowledge that he had been discharged two days ago) and estimates that his psychotic symptoms began one year ago. He explains that he is homeless and does not have social support. He is unable to provide a telephone number or a name to contact family for collateral information.

Mental status exam. The mental status examination reveals a tall, thin, disheveled man who has poor dentition. He is now calm and cooperative despite his reported level of distress. His speech is unremarkable and his eye contact is appropriate. His thought process is linear, organized, and coherent.

Mr. K does not endorse additional symptoms, but is quick to agree with the psychiatrist’s follow-up questions about hallucinations: “Yeah! I’ve been seeing all kinds of crazy stuff.” When prompted for details, he says, “I just saw Big Bird… He was 100 feet tall!”

Lab testing. Mr. K’s blood work is remarkable for positive urine toxicology for amphetamines.

Nursing notes indicate that Mr. K slept overnight and ate 100% of the food on his dinner and breakfast trays.

Red flags flying

Mr. K’s case highlights several red flags that should raise suspicion of malingering (Table 1)^3,4:

• A conditional statement by which a patient threatens to harm himself or others, contingent upon a demand—for example, “If I don’t get A, I’ll do B.”
• An overly dramatic presentation, in which the patient is quick to endorse
  distressing symptoms. Consider Mr. K: He was quick to report that he saw Big Bird, and that this Sesame Street character “was 100 feet tall.” Patients who have been experiencing true psychotic symptoms might be reluctant to speak of their distressing symptoms, especially if they have not experienced such symptoms in the past (the first psychotic break). Mr. K, however, volunteered and called attention to particularly dramatic psychotic symptoms.
• A subjective report of distress that is inconsistent with the objective presentation. Mr. K’s report of depression—a diagnosis that typically includes insomnia and poor appetite—was inconsistent with his behavior: He slept and he ate all of his meals.

Atypical (vs typical) psychosis

Malingering can occur in various arenas and take many different forms. In forensic settings, such as prison, malingered conditions more often present as posttraumatic stress disorder or cognitive impairment.⁵ In non-forensic settings, such as the ER, the most commonly malingered conditions include suicidality and psychosis.

To detect malingered psychosis, one must first understand how true psychotic symptoms manifest. The following discussion describes and compares typical and atypical symptoms of psychosis; examples are given in Table 2.^6,7No single atypical psychotic symptom is indicative of malingering. Rather, a collection of atypical symptoms, when considered in clinical context, should raise suspicion of malingering and prompt you to seek additional collateral information or perform appropriate testing for malingering.

Hallucinations

Typically, hallucinations take three forms: auditory, visual, and tactile. In primary psychiatric conditions, auditory hallucinations are the most common of those three.

Tactile hallucinations can be present during episodes of substance intoxication or withdrawal (eg, so-called coke bugs). 

Auditory hallucinations. Patients who malinger psychosis are often unaware of the nuances of hallucinations. For example, they might report the atypical symptom of continuous voices; in fact, most patients who have schizophrenia hear voices intermittently. Keep in mind, too, that 75% of patients who have schizophrenia hear male and female voices, and that 70% have some type of coping strategy to minimize their internal stimuli (eg, listening to music).^6,7

Visual hallucinations are most often associated with neurologic disease, but also occur often in primary psychotic disorders, such as schizophrenia.

Patients who malinger psychotic symptoms often are open to suggestion, and are quick to endorse visual hallucinations. When asked to describe their hallucinations, however, they often respond without details (“I don’t know”). Other times, they overcompensate with wild exaggeration of atypical visions—recall Mr. K’s description of a towering Big Bird. Asked if the visions are in black and white, they might eagerly agree. Research suggests, however, that patients who have schizophrenia more often experience life-sized hallucinations of vivid scenes with family members, religious figures, or animals.⁸ Furthermore, genuine visual hallucinations typically are in color.

Putting malingering in the differential

Regardless of the number of atypical symptoms a patient exhibits, malingering will be missed if you do not include it in the differential diagnosis. This fact was made evident in a 1973 study.⁹

In that study, Rosenhan and seven of his colleagues—a psychology graduate student, three psychologists, a pediatrician, a psychiatrist, a painter, and a housewife—presented to various ERs and intake units, and, as they had been instructed, endorsed vague auditory hallucinations of “empty,” “hollow,” or “thud” sounds—but nothing more. All were admitted to psychiatric hospitals. Once admitted, they refrained (again, as instructed) from endorsing or exhibiting any psychotic symptoms.

Despite the vague nature of the reported auditory hallucinations and how rapidly symptoms resolved on admission, seven of these pseudo-patients were given a diagnosis of schizophrenia, and one was given a diagnosis of manic-depressive psychosis. Duration of admission ranged from 7 to 52 days (average, 19 days). None of the study participants were suspected of feigning symptoms.

It’s fortunate that, since then, mental health professionals have developed more structured techniques of assessment to detect malingering in inpatient and triage settings.

Testing to identify and assess malingering

The ER is a fast-paced environment, in which treatment teams are challenged to make rapid clinical assessments. With the overwhelming number of patients seeking mental health care in the ER, however, overall wait times are increasing; in some regions, it is common to write, then to rewrite, involuntary psychiatric holds for patients awaiting transfer to a psychiatric hospital. This extended duration presents an opportunity to serially evaluate patients suspected of malingering.

Even in environments that allow for a more comprehensive evaluation (eg, jail or inpatient psychiatric wards), few psychometric tests have been validated to detect malingering. The most validated tests include the Structured Interview of Reported Symptoms (SIRS), distributed now as the Structured Interview of Reported Symptoms, 2nd edition (SIRS-2), and the Minnesota Multiphasic Personality Inventory Revised (MMPI-2). These tests typically require ≥30 minutes to administer and generally are not feasible in the fast-paced ER.

Despite the high prevalence of malingered behaviors in the ER, no single test has been validated in such a setting. Furthermore, there is no test designed to specifically assess for malingered suicidality or homicidality. The results of one test do not, in isolation, represent a comprehensive neuropsychological examination; rather, those results provide additional data to formulate a clinical impression. The instruments discussed below are administered and scored in a defined, objective manner.

When evaluating a patient whom you suspect of malingering, gathering collateral information—from family members, friends, nurses, social workers, emergency medicine physicians, and others—becomes important. You might discover pertinent information in ambulance and police reports and a review of the patient’s prior ER visits.

During the initial interview, ask open-ended questions; do not lead the patient by listing clusters of symptoms associated with a particular diagnosis. Because it is often difficult for a patient to malinger symptoms for a prolonged period, serial observations of a patient’s behavior and interview responses over time can provide additional information to make a clinical diagnosis of malingering.⁴

What testing is feasible in the ER?

Miller Forensic Assessment of Symptoms Test. The M-FAST measures rare symptom combinations, excessive reporting, and atypical symptoms of psychosis, using the same principles as the SIRS-2.

The 25-item screen begins by advising the examinee that he (she) will be asked questions about his psychological symptoms and that the questions that follow might or might not apply to his specific symptoms.

After that brief introduction, the examinee is asked if he hears ringing in his ears. Based on his response, the examiner reads one of two responses—both of which suggest the false notion that patients with true mental illness will suffer from ringing in their ears.

The examinee is then asked a series of Yes or No questions. Some pertain to legitimate symptoms a person with a psychotic illness might suffer (such as, “Do voices tell you to do things? Yes or No?”). Conversely, other questions screen for improbable symptoms that are atypical of patients who have a true psychotic disorder (such as “On many days I feel so bad that I can’t even remember my full name: Yes or No?”).

The exam concludes with a question about a ringing in the examinee’s ear. Affirmative responses are tallied; a score of ≥6 in a clinical setting is 83% specific and 93% sensitive for malingering.¹⁰

Visual Memory Test. Rey’s 15-Item Visual Memory Test capitalizes on the false belief that intellectual deficits, in addition to psychotic symptoms, make a claim of mental illness more believable.

In this simple test, the provider tells the examinee, “I am going to show you a card with 15 things on it that I want you to remember. When I take the card away, I want you to write down as many of the 15 things as you can remember.”³ The examinee is shown 15 common symbols (eg, 1, 2, 3; A, B, C; I, II, III, a, b, c; and the geometrics ●, ■, ▲).

At 5 seconds, the examinee is prompted, “Be sure to remember all of them.” After 10 seconds, the stimulus is removed, and the examinee is asked to recreate the figure.

Normative data indicate that even a patient who has a severe traumatic brain injury is able to recreate at least eight of the symbols. Although controversial, research indicates that a score of <9 symbols is predictive of malingering with 40% sensitivity and 100% specificity.¹¹

Critics argued that confounding variables (IQ, memory disorder, age) might skew the quantitative score. For that reason, the same group developed the Rey’s II Test, which includes a supplementary qualitative scoring system that emphasizes embellishment errors (eg, the wrong symbol) and ordering errors (eg, wrong row). The Rey’s II Test proved to be more sensitive (accurate classification of malingers): A cut-off score of ≥2 qualitative errors is predictive of malingering with 86% sensitivity and 100% specificity.¹²

Coin-in-the-Hand Test. Perhaps the simplest test to administer is the Coin-in-the-Hand, designed to seem—superficially—to be a challenging memory test.

The patient must guess in which hand the examiner is holding a coin. The patient is shown the coin for two seconds, and then asked to close his eyes and count back from 10. The patient then points to one of the two clenched hands.

This task is repeated 10 times; each time, the provider gives verbal feedback about the accuracy or inaccuracy of that attempt. Studies indicate that a patient who has a severe traumatic brain injury is able to score 85% correct. A score <85%, however, suggests feigning of symptoms (sensitivity, 92.5%; specificity 87.5%).¹³ Hanley and co-workers demonstrated that people who are simulating cognitive impairment had a mean accurate response of 4.1, whereas people who had true amnesia had a mean accurate response of 9.65.¹⁴

Persons who feign psychosis or mood symptoms often inaccurately believe that people with mental illness also have cognitive impairment. Both Rey’s test and the Coin-in-the-Hand Test capitalize on this misconception.

Mini-Mental State Examination. Research also has shown that the Folstein Mini-Mental State Examination (MMSE) can screen for malingered cognitive impairment. Powell compared 40 mental health clinicians who were instructed to feign psychosis and 40 patients with schizophrenia. Using the MMSE, the researchers found that the malingers more often gave approximate answers.¹⁵ Moreover, Myers argued that, when compared with Rey’s Test, the MMSE is superior for assessing malingered cognitive impairment because it has a higher positive predictive value (67%, compared with 43% for Rey’s Test) and a higher negative predictive value (93% and 89%).¹⁶

What can you do for these patients after diagnosis?

Malingering is not considered a psychiatric diagnosis; there are no indicated therapies with which to manage it—only guidelines. When you suspect a patient of malingering, you should avoid accusing him (her) of faking symptoms. Rather, when feasible, gently confront the person and provide the opportunity for him to explain his current behaviors. For example, you might say: “I’ve treated many patients with the symptoms that you’re reporting, but the details you provide are different, and don’t ring completely true. Is there anything else that could explain this?”¹⁷

Regardless of a patient’s challenging behaviors, it is important to remember that people who feign illness—whether partial malingering or pure malingering—often do need help. The assistance they require, however, might be best obtained from a housing agency, a chemical dependency program, or another social service—not from the ER. Identifying malingered behaviors saves time and money and shifts limited resources to people who have a legitimate mental health condition.

Last, despite an empathetic approach, some malingering patients continue to feign symptoms—as Mr. K did.

CASE CONTINUED

Although the psychiatrist on call considered forsaking the police to escort Mr. K out of the ER, he eventually agreed to leave the hospital on his own, stating, “My death is going to be on your hands.”

Eight days later, Mr. K visited the ER at a different hospital, endorsing chronic pain and demanding narcotics.

Bottom Line

As the number of people seeking care in the emergency room (ER) has increased, so has the number of those who feign symptoms for secondary gain. No single factor is indicative of malingering, and no objective tests exist to diagnose it definitively. Furthermore, there are no indicated therapies with which to manage malingering—only guidelines. Keep in mind that people who feign illness, whether partial or pure malingering, often do need help—although not the services of an ER.

Related Resources

• Miller Forensic Assessment of Symptoms Test (M-FAST). Psychological Assessment Resources, Inc. www4.parinc.com (enter “M-FAST” in search field).
• Duffy S. Malingering psychological symptoms: An empirical review. Illinois State University, Department of Psychology. http://psychology.illinoisstate.edu/cc/Comps/Duffy%20-%20Malingering.pdf. Accessed September 10, 2013.

Drug Brand Names

Quetiapine • Seroquel

Disclosure

The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers  of competing products.

Featured Audio
M. Cait Brady, MD, shares strategies for assessing malingering. Dr. Brady is a Third-Year Resident in General Psychiatry, University of California, Davis Medical Center - Sacramento, Sacramento, California.

Providing patients and their families with information and education about their psychiatric illness is a central tenet of mental health care. Discussions about diagnostic impressions, treatment options, and the risks and benefits of interventions are customary. Additionally, patients and families often receive written material or referral to other information sources, including self-help books and a growing number of online resources. Although patient education remains a useful and expected element of good care, there is evidence that, alone, it is insufficient to change health behaviors.¹

A growing body of literature and clinical experience suggests that self-management strategies complement patient education and improve treatment outcomes for patients with chronic illnesses, including psychiatric conditions.^2,3 The Cochrane Collaboration describes patient education as “teaching or training of patients concerning their own health needs,” and self-management as “the individual’s ability to manage the symptoms, treatment, physical and psychosocial consequences and lifestyle changes inherent in living with a long-term disorder.”⁴

In this article, we review:

Principles of self-management

Chronic care models must account for conditions in which the clinical course can be variable, that are not amenable to a cure, and that demand long-term treatment. Optimal health outcomes rely on patients accurately monitoring, reporting, and responding to their symptoms, while engaging in critical health-related behaviors. In addition, clinicians must teach, partner with, and motivate patients to engage in crucial disease-management activities. Although typically not considered in this light, we believe most psychiatric disorders are best approached through a long-term care model, and benefit from self-management principles.

Basics of self-management include:

Corbin and Strauss believe effective “self-managers” achieve competency in three areas:

Proficiency in these three “self-manager” domains is enhanced by mastering five key self-management skills outlined in Table 1.⁷

Successful intervention programs vary widely with regard to individual vs group formats, communication interface, and involved health professionals. However, evidence indicates that problem solving, decision making, and action planning are key components.⁷ Successful planning includes:

Table 2 illustrates elements of successful action plans in our Action Planning Worksheet. Adapted from the work of Scharzer,⁸ Prochaska,⁹ and Clark,¹⁰ we developed this self-management tool for individual and group treatment settings. It serves as a vehicle for collaborative patient-provider discussion and planning. The nuts-and-bolts nature of the discussions inevitably leads to learning new, important, and often unexpected information about our patients’ daily lives and the challenges they face as they share their dominant priorities, fears, and insecurities.

Patients provide consistently positive feedback about action planning, and clinicians often find that the process reveals fruitful areas for further psychotherapeutic intervention. Examples include identifying a range of negative automatic thoughts or catastrophic thinking impeding initiation of important activation, and exposure activities for depressed or anxious patients.

Knowing what to do is different than actually doing it. Changing behavior is difficult in the best circumstances, let alone with the strain of a chronic illness. It is critical to recognize that the presence of depressive symptoms significantly reduces the likelihood that patients will employ self-management practices.¹¹ When combined with anxiety and the impairment of motivation and executive functioning that is common in psychiatric conditions, it is not surprising that patients with a mental health condition struggle to embrace ownership of their illness and engage in critical health behaviors—which may include adhering to medication regimens; maintaining a healthy sleep cycle, nutrition, and exercise routines; vigilant symptom surveillance; and carrying out an agreed-upon action plan.

Interventions

Professionally-guided “light-touch” interventions and technology-assisted self-management interventions also can improve patient engagement in activities through individual encounters, group forums, and technology-mediated exchanges, including telephone, email, text message, tele-health, and web-based interventions.¹³ The DE-STRESS (Delivery of Self Training and Education for Stressful Situations) model illustrates these principles. This 8-week program combines elements of face-to-face, email, telephone, and web-based assignments and exchanges, and demonstrates a decline in posttraumatic stress disorder, depression, and anxiety scores.¹⁴ Our Michigan Depression Outreach and Collaborative Care program is another example of a self-management intervention (Box 1).

Helping patients understand when common symptoms such as impaired concentration or dizziness should be “watched”—rather than responded to aggressively—is crucial for effective long-term management.

Challenges

Health care delivery and medical educational models have been slow to embrace this change to long-term care models because doing so involves what might be uncomfortable shifts in roles and responsibilities. Effective care for long-term illness necessitates that the patient become an expert on his (her) illness, and be an active participant and partner in their treatment. Preparing health professionals for this new role as teacher, mentor, and collaborator presents a challenge to health care systems and educational programs across disciplines.

When trying to facilitate effective patient-provider partnerships, it is important to recognize the variability in patient preference for what and how information is shared, how decisions are made, and the role patients are asked to play in their care. Patients differ in their desire for an active or collaborative shared-decision model; some prefer more directive provider communication and a passive role.¹⁸ Preferences are influenced by variables such as age, sex, race, anxiety level, and education.¹⁹ Open discussion of these matters between caregivers and patients is important; studies have shown that failure to address these issues of “fit” can
impede communication, healthy behavior, and positive outcomes.

Bottom Line

Emerging care models demand that health care providers become teachers and motivators to help patients develop and implement patterns of health surveillance and intervention that will optimize their well-being and functionality. As active collaborators in their care, patients form a partnership with their care teams, allowing for regular, reciprocal exchange of information and shared decision-making. This shift to a partnership creates new, exciting roles and responsibilities for all parties.

Related Resources

• Improving Chronic Illness Care. www.improvingchroniccare.org.
• Chronic disease self-management program (Better Choices, Better Health workshop). Stanford School of Medicine. http://patienteducation.stanford.edu/programs/cdsmp.html.

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Featured Audio
Thomas E. Fluent, MD, talks about addressing patient resistance to a self-management model of care. Dr. Fluent is Clinical Assistant Professor, Department of Psychiatry, University of Michigan, Ann Arbor, Michigan.

Providing patients and their families with information and education about their psychiatric illness is a central tenet of mental health care. Discussions about diagnostic impressions, treatment options, and the risks and benefits of interventions are customary. Additionally, patients and families often receive written material or referral to other information sources, including self-help books and a growing number of online resources. Although patient education remains a useful and expected element of good care, there is evidence that, alone, it is insufficient to change health behaviors.¹

A growing body of literature and clinical experience suggests that self-management strategies complement patient education and improve treatment outcomes for patients with chronic illnesses, including psychiatric conditions.^2,3 The Cochrane Collaboration describes patient education as “teaching or training of patients concerning their own health needs,” and self-management as “the individual’s ability to manage the symptoms, treatment, physical and psychosocial consequences and lifestyle changes inherent in living with a long-term disorder.”⁴

In this article, we review:

Principles of self-management

Chronic care models must account for conditions in which the clinical course can be variable, that are not amenable to a cure, and that demand long-term treatment. Optimal health outcomes rely on patients accurately monitoring, reporting, and responding to their symptoms, while engaging in critical health-related behaviors. In addition, clinicians must teach, partner with, and motivate patients to engage in crucial disease-management activities. Although typically not considered in this light, we believe most psychiatric disorders are best approached through a long-term care model, and benefit from self-management principles.

Basics of self-management include:

Corbin and Strauss believe effective “self-managers” achieve competency in three areas:

Proficiency in these three “self-manager” domains is enhanced by mastering five key self-management skills outlined in Table 1.⁷

Successful intervention programs vary widely with regard to individual vs group formats, communication interface, and involved health professionals. However, evidence indicates that problem solving, decision making, and action planning are key components.⁷ Successful planning includes:

Table 2 illustrates elements of successful action plans in our Action Planning Worksheet. Adapted from the work of Scharzer,⁸ Prochaska,⁹ and Clark,¹⁰ we developed this self-management tool for individual and group treatment settings. It serves as a vehicle for collaborative patient-provider discussion and planning. The nuts-and-bolts nature of the discussions inevitably leads to learning new, important, and often unexpected information about our patients’ daily lives and the challenges they face as they share their dominant priorities, fears, and insecurities.

Patients provide consistently positive feedback about action planning, and clinicians often find that the process reveals fruitful areas for further psychotherapeutic intervention. Examples include identifying a range of negative automatic thoughts or catastrophic thinking impeding initiation of important activation, and exposure activities for depressed or anxious patients.

Knowing what to do is different than actually doing it. Changing behavior is difficult in the best circumstances, let alone with the strain of a chronic illness. It is critical to recognize that the presence of depressive symptoms significantly reduces the likelihood that patients will employ self-management practices.¹¹ When combined with anxiety and the impairment of motivation and executive functioning that is common in psychiatric conditions, it is not surprising that patients with a mental health condition struggle to embrace ownership of their illness and engage in critical health behaviors—which may include adhering to medication regimens; maintaining a healthy sleep cycle, nutrition, and exercise routines; vigilant symptom surveillance; and carrying out an agreed-upon action plan.

Interventions

Professionally-guided “light-touch” interventions and technology-assisted self-management interventions also can improve patient engagement in activities through individual encounters, group forums, and technology-mediated exchanges, including telephone, email, text message, tele-health, and web-based interventions.¹³ The DE-STRESS (Delivery of Self Training and Education for Stressful Situations) model illustrates these principles. This 8-week program combines elements of face-to-face, email, telephone, and web-based assignments and exchanges, and demonstrates a decline in posttraumatic stress disorder, depression, and anxiety scores.¹⁴ Our Michigan Depression Outreach and Collaborative Care program is another example of a self-management intervention (Box 1).

Helping patients understand when common symptoms such as impaired concentration or dizziness should be “watched”—rather than responded to aggressively—is crucial for effective long-term management.

Challenges

Health care delivery and medical educational models have been slow to embrace this change to long-term care models because doing so involves what might be uncomfortable shifts in roles and responsibilities. Effective care for long-term illness necessitates that the patient become an expert on his (her) illness, and be an active participant and partner in their treatment. Preparing health professionals for this new role as teacher, mentor, and collaborator presents a challenge to health care systems and educational programs across disciplines.

When trying to facilitate effective patient-provider partnerships, it is important to recognize the variability in patient preference for what and how information is shared, how decisions are made, and the role patients are asked to play in their care. Patients differ in their desire for an active or collaborative shared-decision model; some prefer more directive provider communication and a passive role.¹⁸ Preferences are influenced by variables such as age, sex, race, anxiety level, and education.¹⁹ Open discussion of these matters between caregivers and patients is important; studies have shown that failure to address these issues of “fit” can
impede communication, healthy behavior, and positive outcomes.

Bottom Line

Emerging care models demand that health care providers become teachers and motivators to help patients develop and implement patterns of health surveillance and intervention that will optimize their well-being and functionality. As active collaborators in their care, patients form a partnership with their care teams, allowing for regular, reciprocal exchange of information and shared decision-making. This shift to a partnership creates new, exciting roles and responsibilities for all parties.

Related Resources

• Improving Chronic Illness Care. www.improvingchroniccare.org.
• Chronic disease self-management program (Better Choices, Better Health workshop). Stanford School of Medicine. http://patienteducation.stanford.edu/programs/cdsmp.html.

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Featured Audio
Thomas E. Fluent, MD, talks about addressing patient resistance to a self-management model of care. Dr. Fluent is Clinical Assistant Professor, Department of Psychiatry, University of Michigan, Ann Arbor, Michigan.

Providing patients and their families with information and education about their psychiatric illness is a central tenet of mental health care. Discussions about diagnostic impressions, treatment options, and the risks and benefits of interventions are customary. Additionally, patients and families often receive written material or referral to other information sources, including self-help books and a growing number of online resources. Although patient education remains a useful and expected element of good care, there is evidence that, alone, it is insufficient to change health behaviors.¹

A growing body of literature and clinical experience suggests that self-management strategies complement patient education and improve treatment outcomes for patients with chronic illnesses, including psychiatric conditions.^2,3 The Cochrane Collaboration describes patient education as “teaching or training of patients concerning their own health needs,” and self-management as “the individual’s ability to manage the symptoms, treatment, physical and psychosocial consequences and lifestyle changes inherent in living with a long-term disorder.”⁴

In this article, we review:

Principles of self-management

Chronic care models must account for conditions in which the clinical course can be variable, that are not amenable to a cure, and that demand long-term treatment. Optimal health outcomes rely on patients accurately monitoring, reporting, and responding to their symptoms, while engaging in critical health-related behaviors. In addition, clinicians must teach, partner with, and motivate patients to engage in crucial disease-management activities. Although typically not considered in this light, we believe most psychiatric disorders are best approached through a long-term care model, and benefit from self-management principles.

Basics of self-management include:

Corbin and Strauss believe effective “self-managers” achieve competency in three areas:

Proficiency in these three “self-manager” domains is enhanced by mastering five key self-management skills outlined in Table 1.⁷

Successful intervention programs vary widely with regard to individual vs group formats, communication interface, and involved health professionals. However, evidence indicates that problem solving, decision making, and action planning are key components.⁷ Successful planning includes:

Table 2 illustrates elements of successful action plans in our Action Planning Worksheet. Adapted from the work of Scharzer,⁸ Prochaska,⁹ and Clark,¹⁰ we developed this self-management tool for individual and group treatment settings. It serves as a vehicle for collaborative patient-provider discussion and planning. The nuts-and-bolts nature of the discussions inevitably leads to learning new, important, and often unexpected information about our patients’ daily lives and the challenges they face as they share their dominant priorities, fears, and insecurities.

Patients provide consistently positive feedback about action planning, and clinicians often find that the process reveals fruitful areas for further psychotherapeutic intervention. Examples include identifying a range of negative automatic thoughts or catastrophic thinking impeding initiation of important activation, and exposure activities for depressed or anxious patients.

Knowing what to do is different than actually doing it. Changing behavior is difficult in the best circumstances, let alone with the strain of a chronic illness. It is critical to recognize that the presence of depressive symptoms significantly reduces the likelihood that patients will employ self-management practices.¹¹ When combined with anxiety and the impairment of motivation and executive functioning that is common in psychiatric conditions, it is not surprising that patients with a mental health condition struggle to embrace ownership of their illness and engage in critical health behaviors—which may include adhering to medication regimens; maintaining a healthy sleep cycle, nutrition, and exercise routines; vigilant symptom surveillance; and carrying out an agreed-upon action plan.

Interventions

Professionally-guided “light-touch” interventions and technology-assisted self-management interventions also can improve patient engagement in activities through individual encounters, group forums, and technology-mediated exchanges, including telephone, email, text message, tele-health, and web-based interventions.¹³ The DE-STRESS (Delivery of Self Training and Education for Stressful Situations) model illustrates these principles. This 8-week program combines elements of face-to-face, email, telephone, and web-based assignments and exchanges, and demonstrates a decline in posttraumatic stress disorder, depression, and anxiety scores.¹⁴ Our Michigan Depression Outreach and Collaborative Care program is another example of a self-management intervention (Box 1).

Helping patients understand when common symptoms such as impaired concentration or dizziness should be “watched”—rather than responded to aggressively—is crucial for effective long-term management.

Challenges

Health care delivery and medical educational models have been slow to embrace this change to long-term care models because doing so involves what might be uncomfortable shifts in roles and responsibilities. Effective care for long-term illness necessitates that the patient become an expert on his (her) illness, and be an active participant and partner in their treatment. Preparing health professionals for this new role as teacher, mentor, and collaborator presents a challenge to health care systems and educational programs across disciplines.

When trying to facilitate effective patient-provider partnerships, it is important to recognize the variability in patient preference for what and how information is shared, how decisions are made, and the role patients are asked to play in their care. Patients differ in their desire for an active or collaborative shared-decision model; some prefer more directive provider communication and a passive role.¹⁸ Preferences are influenced by variables such as age, sex, race, anxiety level, and education.¹⁹ Open discussion of these matters between caregivers and patients is important; studies have shown that failure to address these issues of “fit” can
impede communication, healthy behavior, and positive outcomes.

Bottom Line

Emerging care models demand that health care providers become teachers and motivators to help patients develop and implement patterns of health surveillance and intervention that will optimize their well-being and functionality. As active collaborators in their care, patients form a partnership with their care teams, allowing for regular, reciprocal exchange of information and shared decision-making. This shift to a partnership creates new, exciting roles and responsibilities for all parties.

Related Resources

• Improving Chronic Illness Care. www.improvingchroniccare.org.
• Chronic disease self-management program (Better Choices, Better Health workshop). Stanford School of Medicine. http://patienteducation.stanford.edu/programs/cdsmp.html.

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Featured Audio
Thomas E. Fluent, MD, talks about addressing patient resistance to a self-management model of care. Dr. Fluent is Clinical Assistant Professor, Department of Psychiatry, University of Michigan, Ann Arbor, Michigan.

Psychiatric patients stand to benefit greatly from adhering to prescribed pharmacotherapy, but many patients typically do not follow their medication regimens.^1,2 Three months after pharmacotherapy is initiated, approximately 50% of patients with major depressive disorder (MDD) do not take their prescribed antidepressants.³ Adherence rates in patients with schizophrenia range from 50% to 60%, and patients with bipolar disorder have adherence rates as low as 35%.^4-6 One possible explanation for “treatment-resistant” depression, schizophrenia, and bipolar disorder may simply be nonadherence to prescribed pharmacotherapy.

Several strategies have been used to address this vexing problem (Table 1).^7,8 They include individual and family psychoeducation,^9,10 cognitive-behavioral therapy,¹¹ interpersonal and social rhythm therapy, and family-focused therapy. This article describes an additional strategy I call “participatory pharmacotherapy.” In this model, the patient becomes a partner in the process of treatment choices and decision-making. This encourages patients to provide their own opinions and points of view regarding medication use. The prescribing clinician makes the patient feel that he or she has been listened to and understood. This and other techniques emphasize forming a therapeutic alliance with the patient before initiating pharmacotherapy. The patient provides information on his or her family history, medical and psychiatric history, and experience with previous medications, with a specific focus on which medications worked best for the patient and family members diagnosed with a similar condition.

Getting patients to participate

One of the fundamental tasks is to encourage patients to accept a participatory role, determine their underlying diagnosis, and co-create a treatment plan that will be most compatible with their illness and their personality. There are 10 components of establishing and practicing participatory pharmacotherapy.

1. Encourage patients to share their opinion of what a desirable treatment outcome should be. Some patients have unrealistic expectations about what medications can achieve. Clarify with patients what would be a realistic expectation of pharmacotherapy, and modify the patient’s beliefs to be compatible with a more probable outcome. For example, Ms. D, a 46-year-old mother of 2, is diagnosed with MDD, recurrent type without psychotic features. She states she expects pharmacotherapy will alleviate all symptoms and allow her to achieve a new healthy, happy state in which she will be able to laugh, socialize, and have fun every day for the rest of her life. Although achieving remission is a realistic and desirable treatment goal, Ms. D’s expectations are idealistic. Helping Ms. D accept and agree to realistic and achievable outcomes will improve her adherence to prescribed medications.

2. Encourage patients to share their ideas of how a desirable outcome can be accomplished. Similar to their expectations of outcomes, some patients have an unrealistic understanding of how treatment is conducted. Some patients expect treatment to be limited to prescribed medications or a one-time injection of a curative drug. Others prefer to use herbs and supplements and want to avoid prescribed medications. Understanding the patient’s expectations of how treatment is carried out will allow clinicians to provide patients with a rational view of treatment and establish a partnership based on realistic expectations.

3. Engage patients in choosing the best medication for them. Many patients have preconceived ideas about medications and which medicine would be best for them. They get this information from various sources, including family members and friends who benefitted from a specific drug, personal experience with medications, and exposure to drug advertising.

Understanding the patient’s preference for a specific medication and why he or she made such a choice is critical because doing so can take advantage of the patient’s self-fulfilling prophecies and improve the chances of obtaining a better outcome. For example, Mr. O, a 52-year-old father of 3, has been experiencing recurrent episodes of severe panic attacks. His clinician asked him to describe medications that in his opinion were most helpful in the past. He said he preferred clonazepam because it had helped him control the panic attacks and had minimal side effects, but he discontinued it after a previous psychotherapist told him he would become addicted to it. Obtaining this information was valuable because the clinician was able to clarify guidelines for clonazepam use without the risk of dependence. Mr. O is prescribed clonazepam, which he takes consistently and responds to excellently.

4. Involve patients in setting treatment goals and targeting symptoms to be relieved. Actively listen when patients describe their symptoms, discomforts, and past experiences with treatments. I invite patients to speak uninterrupted for 5 to 10 minutes, even if they talk about issues that seem irrelevant. I then summarize the patient’s major points and ask, “And what else?” After he or she says, “That’s it,” I ask the patient to assign a priority to alleviating each symptom.

For example, Ms. J, a 38-year-old married mother of 2, was diagnosed with bipolar II disorder. She listed her highest priority as controlling her impulsive shopping rather than alleviating depression, insomnia, or overeating. She had been forced to declare bankruptcy twice, and she was determined to never do so again. She also wanted to regain her husband’s trust and her ability to manage her finances. Ensuring that Ms. J felt understood regarding this issue increased the chances of establishing a solid treatment partnership. Providing Ms. J with a menu of treatment choices and asking her to describe her previous experiences with medications helped her and the clinician choose a medication that is compatible with her desire to control her impulsive shopping.

5. Engage patients in choosing the best delivery system for the prescribed medication. For many medications, clinicians can choose from a variety of delivery systems, including pills, transdermal patches, rectal or vaginal suppositories, creams, ointments, orally disintegrating tablets, liquids, and intramuscular injections. Patients have varying beliefs about the efficacy of particular delivery systems, based on personal experiences or what they have learned from the media, their family and friends, or the Internet. For example, Ms. S, age 28, experienced recurrent, disabling anxiety attacks. When asked about the best way of providing medication to relieve her symptoms, she chose gluteal injections because, as a child, her pediatrician had treated her for an unspecified illness by injecting medication in her buttock, which rapidly relieved her symptoms. This left her with the impression that injectable medications were the best therapeutic delivery system. After discussing the practicalities and availability of fast-acting medications to control panic attacks, we agreed to use orally disintegrating clonazepam, which is absorbed swiftly and provides fast symptom relief. Ms. S reported favorable results and was pleased with the process of developing this strategy with her clinician.

6. Involve patients in choosing the times and frequency of medication administration. The timing and frequency of medication administration can be used to enhance desirable therapeutic effects. For example, an antidepressant that causes sedation and somnolence could be taken at bedtime to help alleviate insomnia. Some studies have shown that taking a medication once a day improves adherence compared with taking the same medication in divided doses.¹³ Other patients may wish to take a medication several times a day so they can keep the medication in their purse or briefcase and feel confident that if they need a medication for immediate symptom relief, it will be readily available.

7. Teach patients to self-monitor changes and improvements in target symptoms. Engaging patients in a system of self-monitoring improves their chances of achieving successful treatment outcomes.¹⁴ Instruct patients to create a list of symptoms and monitor the intensity of each symptom using a rating scale of 1 to 5, where 1 represents the lowest intensity and 5 represents the highest. As for frequency, patients can rate each symptom from “not present” to “present most of the time.”

Self-monitoring allows patients to observe which daily behaviors and lifestyle choices make symptoms better and which make them worse. For example, Mrs. P, a 38-year-old married mother of 2, had anxiety and panic attacks associated with low self-esteem and chronic depression. Her clinician instructed her to use a 1-page form to monitor the frequency and intensity of her anxiety and panic symptoms by focusing on the physical manifestations, such as rapid heartbeat, shortness of breath, nausea, tremors, dry mouth, frequent urination, and diarrhea to see if there was any correlation between her behaviors and her symptoms.

8. Instruct patients to call you to report any changes, including minor successes. Early in my career, toward the end of each appointment after I’d prescribed medications I’d tell patients, “Please call me if you have a problem.” Frequently, patients would call with a list of problems and side effects that they believed were caused by the newly prescribed medication. Later, I realized that I may have inadvertently encouraged patients to develop problems so they would have a reason to call me. To achieve a more favorable outcome I changed the way I communicate. I now say, “Please call me next week, even if you begin to feel better with this new medication.” The phone call is now associated with the idea that they will “get better,” and internalizing such a suggestion allows patients to talk with the clinician and report favorable treatment results.

9. Tell patients to monitor their successes by relabeling and reframing their symptoms. Mr. B, age 28, has MDD and reports irritability, insomnia, short temper, and restlessness. After reviewing his desired treatment outcome, we discuss the benefits of pharmacotherapy. I tell him the new medication will improve the quality and length of his sleep, which will allow his body and mind to recharge his “internal batteries” and restore health and energy. When we discuss side effects, I tell him to expect a dry mouth, which will be his signal that the medication is working. This discussion helps patients reframe side effects and improves their ability to tolerate side effects and adhere to pharmacotherapy.     

10. Harness the placebo effect and the power of suggestion to increase chances of achieving the best treatment outcomes. In a previous article,¹² I reviewed the principles of recognizing and enhancing the placebo effect and the power of suggestion to improve the chances of achieving better pharmacotherapy outcomes. When practicing participatory pharmacotherapy, clinicians are consciously aware of the power embedded in their words and are careful to use language that enhances the placebo effect and the power of suggestion when prescribing medications. Use the patient’s own language as a way of pacing yourself to the patient’s description of his or her distress. For example, Ms. R, a 42-year-old mother of 3, describes her experiences seeking help for her anxiety and depression, stating that she has not yet found the right combination of medications that provide benefits with tolerable side effects. Her clinician responds by focusing on the word “yet” (pacing) stating, “even though you have not yet found the right combination of medications to provide the most desirable benefit of beginning healing and restoring your hope, I promise to work with you and together we will try to achieve an improvement in your overall health and well-being.” This response includes several positive words and suggestions of future success, which are referred to as leading.

Not all patients will respond to participatory pharmacotherapy. Some factors will make patients good candidates for this approach, and others should be considered exclusionary qualities (Table 2).

Bottom Line

“Participatory pharmacotherapy” involves identifying patients as partners in the process of treatment choice and decision-making, encouraging them to provide their opinions regarding medication use, and making patients feel they have been heard and understood. This technique emphasizes forming a therapeutic alliance with the patient to improve patients’ adherence to pharmacotherapy and optimize treatment outcomes.

Related Resources

• Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2008;16(2);CD000011.

• Mahone IH. Shared decision making and serious mental illness. Arch Psychiatr Nurs. 2008;22(6):334-343.

• Russel CL, Ruppar TM, Metteson M. Improving medication adherence: moving from intention and motivation to a personal systems approach. Nurs Clin North Am. 2011;46(3):271-281.

• Tibaldi G, Salvador-Carulla L, Garcia-Gutierrez JC. From treatment adherence to advanced shared decision making: New professional strategies and attitudes in mental health care. Curr Clin Pharmacol. 2011;6(2):91-99. 

Drug Brand Name

Clonazepam • Klonopin

Disclosure

Dr. Torem reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Psychiatric patients stand to benefit greatly from adhering to prescribed pharmacotherapy, but many patients typically do not follow their medication regimens.^1,2 Three months after pharmacotherapy is initiated, approximately 50% of patients with major depressive disorder (MDD) do not take their prescribed antidepressants.³ Adherence rates in patients with schizophrenia range from 50% to 60%, and patients with bipolar disorder have adherence rates as low as 35%.^4-6 One possible explanation for “treatment-resistant” depression, schizophrenia, and bipolar disorder may simply be nonadherence to prescribed pharmacotherapy.

Several strategies have been used to address this vexing problem (Table 1).^7,8 They include individual and family psychoeducation,^9,10 cognitive-behavioral therapy,¹¹ interpersonal and social rhythm therapy, and family-focused therapy. This article describes an additional strategy I call “participatory pharmacotherapy.” In this model, the patient becomes a partner in the process of treatment choices and decision-making. This encourages patients to provide their own opinions and points of view regarding medication use. The prescribing clinician makes the patient feel that he or she has been listened to and understood. This and other techniques emphasize forming a therapeutic alliance with the patient before initiating pharmacotherapy. The patient provides information on his or her family history, medical and psychiatric history, and experience with previous medications, with a specific focus on which medications worked best for the patient and family members diagnosed with a similar condition.

Getting patients to participate

One of the fundamental tasks is to encourage patients to accept a participatory role, determine their underlying diagnosis, and co-create a treatment plan that will be most compatible with their illness and their personality. There are 10 components of establishing and practicing participatory pharmacotherapy.

1. Encourage patients to share their opinion of what a desirable treatment outcome should be. Some patients have unrealistic expectations about what medications can achieve. Clarify with patients what would be a realistic expectation of pharmacotherapy, and modify the patient’s beliefs to be compatible with a more probable outcome. For example, Ms. D, a 46-year-old mother of 2, is diagnosed with MDD, recurrent type without psychotic features. She states she expects pharmacotherapy will alleviate all symptoms and allow her to achieve a new healthy, happy state in which she will be able to laugh, socialize, and have fun every day for the rest of her life. Although achieving remission is a realistic and desirable treatment goal, Ms. D’s expectations are idealistic. Helping Ms. D accept and agree to realistic and achievable outcomes will improve her adherence to prescribed medications.

2. Encourage patients to share their ideas of how a desirable outcome can be accomplished. Similar to their expectations of outcomes, some patients have an unrealistic understanding of how treatment is conducted. Some patients expect treatment to be limited to prescribed medications or a one-time injection of a curative drug. Others prefer to use herbs and supplements and want to avoid prescribed medications. Understanding the patient’s expectations of how treatment is carried out will allow clinicians to provide patients with a rational view of treatment and establish a partnership based on realistic expectations.

3. Engage patients in choosing the best medication for them. Many patients have preconceived ideas about medications and which medicine would be best for them. They get this information from various sources, including family members and friends who benefitted from a specific drug, personal experience with medications, and exposure to drug advertising.

Understanding the patient’s preference for a specific medication and why he or she made such a choice is critical because doing so can take advantage of the patient’s self-fulfilling prophecies and improve the chances of obtaining a better outcome. For example, Mr. O, a 52-year-old father of 3, has been experiencing recurrent episodes of severe panic attacks. His clinician asked him to describe medications that in his opinion were most helpful in the past. He said he preferred clonazepam because it had helped him control the panic attacks and had minimal side effects, but he discontinued it after a previous psychotherapist told him he would become addicted to it. Obtaining this information was valuable because the clinician was able to clarify guidelines for clonazepam use without the risk of dependence. Mr. O is prescribed clonazepam, which he takes consistently and responds to excellently.

4. Involve patients in setting treatment goals and targeting symptoms to be relieved. Actively listen when patients describe their symptoms, discomforts, and past experiences with treatments. I invite patients to speak uninterrupted for 5 to 10 minutes, even if they talk about issues that seem irrelevant. I then summarize the patient’s major points and ask, “And what else?” After he or she says, “That’s it,” I ask the patient to assign a priority to alleviating each symptom.

For example, Ms. J, a 38-year-old married mother of 2, was diagnosed with bipolar II disorder. She listed her highest priority as controlling her impulsive shopping rather than alleviating depression, insomnia, or overeating. She had been forced to declare bankruptcy twice, and she was determined to never do so again. She also wanted to regain her husband’s trust and her ability to manage her finances. Ensuring that Ms. J felt understood regarding this issue increased the chances of establishing a solid treatment partnership. Providing Ms. J with a menu of treatment choices and asking her to describe her previous experiences with medications helped her and the clinician choose a medication that is compatible with her desire to control her impulsive shopping.

5. Engage patients in choosing the best delivery system for the prescribed medication. For many medications, clinicians can choose from a variety of delivery systems, including pills, transdermal patches, rectal or vaginal suppositories, creams, ointments, orally disintegrating tablets, liquids, and intramuscular injections. Patients have varying beliefs about the efficacy of particular delivery systems, based on personal experiences or what they have learned from the media, their family and friends, or the Internet. For example, Ms. S, age 28, experienced recurrent, disabling anxiety attacks. When asked about the best way of providing medication to relieve her symptoms, she chose gluteal injections because, as a child, her pediatrician had treated her for an unspecified illness by injecting medication in her buttock, which rapidly relieved her symptoms. This left her with the impression that injectable medications were the best therapeutic delivery system. After discussing the practicalities and availability of fast-acting medications to control panic attacks, we agreed to use orally disintegrating clonazepam, which is absorbed swiftly and provides fast symptom relief. Ms. S reported favorable results and was pleased with the process of developing this strategy with her clinician.

6. Involve patients in choosing the times and frequency of medication administration. The timing and frequency of medication administration can be used to enhance desirable therapeutic effects. For example, an antidepressant that causes sedation and somnolence could be taken at bedtime to help alleviate insomnia. Some studies have shown that taking a medication once a day improves adherence compared with taking the same medication in divided doses.¹³ Other patients may wish to take a medication several times a day so they can keep the medication in their purse or briefcase and feel confident that if they need a medication for immediate symptom relief, it will be readily available.

7. Teach patients to self-monitor changes and improvements in target symptoms. Engaging patients in a system of self-monitoring improves their chances of achieving successful treatment outcomes.¹⁴ Instruct patients to create a list of symptoms and monitor the intensity of each symptom using a rating scale of 1 to 5, where 1 represents the lowest intensity and 5 represents the highest. As for frequency, patients can rate each symptom from “not present” to “present most of the time.”

Self-monitoring allows patients to observe which daily behaviors and lifestyle choices make symptoms better and which make them worse. For example, Mrs. P, a 38-year-old married mother of 2, had anxiety and panic attacks associated with low self-esteem and chronic depression. Her clinician instructed her to use a 1-page form to monitor the frequency and intensity of her anxiety and panic symptoms by focusing on the physical manifestations, such as rapid heartbeat, shortness of breath, nausea, tremors, dry mouth, frequent urination, and diarrhea to see if there was any correlation between her behaviors and her symptoms.

8. Instruct patients to call you to report any changes, including minor successes. Early in my career, toward the end of each appointment after I’d prescribed medications I’d tell patients, “Please call me if you have a problem.” Frequently, patients would call with a list of problems and side effects that they believed were caused by the newly prescribed medication. Later, I realized that I may have inadvertently encouraged patients to develop problems so they would have a reason to call me. To achieve a more favorable outcome I changed the way I communicate. I now say, “Please call me next week, even if you begin to feel better with this new medication.” The phone call is now associated with the idea that they will “get better,” and internalizing such a suggestion allows patients to talk with the clinician and report favorable treatment results.

9. Tell patients to monitor their successes by relabeling and reframing their symptoms. Mr. B, age 28, has MDD and reports irritability, insomnia, short temper, and restlessness. After reviewing his desired treatment outcome, we discuss the benefits of pharmacotherapy. I tell him the new medication will improve the quality and length of his sleep, which will allow his body and mind to recharge his “internal batteries” and restore health and energy. When we discuss side effects, I tell him to expect a dry mouth, which will be his signal that the medication is working. This discussion helps patients reframe side effects and improves their ability to tolerate side effects and adhere to pharmacotherapy.     

10. Harness the placebo effect and the power of suggestion to increase chances of achieving the best treatment outcomes. In a previous article,¹² I reviewed the principles of recognizing and enhancing the placebo effect and the power of suggestion to improve the chances of achieving better pharmacotherapy outcomes. When practicing participatory pharmacotherapy, clinicians are consciously aware of the power embedded in their words and are careful to use language that enhances the placebo effect and the power of suggestion when prescribing medications. Use the patient’s own language as a way of pacing yourself to the patient’s description of his or her distress. For example, Ms. R, a 42-year-old mother of 3, describes her experiences seeking help for her anxiety and depression, stating that she has not yet found the right combination of medications that provide benefits with tolerable side effects. Her clinician responds by focusing on the word “yet” (pacing) stating, “even though you have not yet found the right combination of medications to provide the most desirable benefit of beginning healing and restoring your hope, I promise to work with you and together we will try to achieve an improvement in your overall health and well-being.” This response includes several positive words and suggestions of future success, which are referred to as leading.

Not all patients will respond to participatory pharmacotherapy. Some factors will make patients good candidates for this approach, and others should be considered exclusionary qualities (Table 2).

Bottom Line

“Participatory pharmacotherapy” involves identifying patients as partners in the process of treatment choice and decision-making, encouraging them to provide their opinions regarding medication use, and making patients feel they have been heard and understood. This technique emphasizes forming a therapeutic alliance with the patient to improve patients’ adherence to pharmacotherapy and optimize treatment outcomes.

Related Resources

• Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2008;16(2);CD000011.

• Mahone IH. Shared decision making and serious mental illness. Arch Psychiatr Nurs. 2008;22(6):334-343.

• Russel CL, Ruppar TM, Metteson M. Improving medication adherence: moving from intention and motivation to a personal systems approach. Nurs Clin North Am. 2011;46(3):271-281.

• Tibaldi G, Salvador-Carulla L, Garcia-Gutierrez JC. From treatment adherence to advanced shared decision making: New professional strategies and attitudes in mental health care. Curr Clin Pharmacol. 2011;6(2):91-99. 

Drug Brand Name

Clonazepam • Klonopin

Disclosure

Dr. Torem reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Psychiatric patients stand to benefit greatly from adhering to prescribed pharmacotherapy, but many patients typically do not follow their medication regimens.^1,2 Three months after pharmacotherapy is initiated, approximately 50% of patients with major depressive disorder (MDD) do not take their prescribed antidepressants.³ Adherence rates in patients with schizophrenia range from 50% to 60%, and patients with bipolar disorder have adherence rates as low as 35%.^4-6 One possible explanation for “treatment-resistant” depression, schizophrenia, and bipolar disorder may simply be nonadherence to prescribed pharmacotherapy.

Several strategies have been used to address this vexing problem (Table 1).^7,8 They include individual and family psychoeducation,^9,10 cognitive-behavioral therapy,¹¹ interpersonal and social rhythm therapy, and family-focused therapy. This article describes an additional strategy I call “participatory pharmacotherapy.” In this model, the patient becomes a partner in the process of treatment choices and decision-making. This encourages patients to provide their own opinions and points of view regarding medication use. The prescribing clinician makes the patient feel that he or she has been listened to and understood. This and other techniques emphasize forming a therapeutic alliance with the patient before initiating pharmacotherapy. The patient provides information on his or her family history, medical and psychiatric history, and experience with previous medications, with a specific focus on which medications worked best for the patient and family members diagnosed with a similar condition.

Getting patients to participate

One of the fundamental tasks is to encourage patients to accept a participatory role, determine their underlying diagnosis, and co-create a treatment plan that will be most compatible with their illness and their personality. There are 10 components of establishing and practicing participatory pharmacotherapy.

1. Encourage patients to share their opinion of what a desirable treatment outcome should be. Some patients have unrealistic expectations about what medications can achieve. Clarify with patients what would be a realistic expectation of pharmacotherapy, and modify the patient’s beliefs to be compatible with a more probable outcome. For example, Ms. D, a 46-year-old mother of 2, is diagnosed with MDD, recurrent type without psychotic features. She states she expects pharmacotherapy will alleviate all symptoms and allow her to achieve a new healthy, happy state in which she will be able to laugh, socialize, and have fun every day for the rest of her life. Although achieving remission is a realistic and desirable treatment goal, Ms. D’s expectations are idealistic. Helping Ms. D accept and agree to realistic and achievable outcomes will improve her adherence to prescribed medications.

2. Encourage patients to share their ideas of how a desirable outcome can be accomplished. Similar to their expectations of outcomes, some patients have an unrealistic understanding of how treatment is conducted. Some patients expect treatment to be limited to prescribed medications or a one-time injection of a curative drug. Others prefer to use herbs and supplements and want to avoid prescribed medications. Understanding the patient’s expectations of how treatment is carried out will allow clinicians to provide patients with a rational view of treatment and establish a partnership based on realistic expectations.

3. Engage patients in choosing the best medication for them. Many patients have preconceived ideas about medications and which medicine would be best for them. They get this information from various sources, including family members and friends who benefitted from a specific drug, personal experience with medications, and exposure to drug advertising.

Understanding the patient’s preference for a specific medication and why he or she made such a choice is critical because doing so can take advantage of the patient’s self-fulfilling prophecies and improve the chances of obtaining a better outcome. For example, Mr. O, a 52-year-old father of 3, has been experiencing recurrent episodes of severe panic attacks. His clinician asked him to describe medications that in his opinion were most helpful in the past. He said he preferred clonazepam because it had helped him control the panic attacks and had minimal side effects, but he discontinued it after a previous psychotherapist told him he would become addicted to it. Obtaining this information was valuable because the clinician was able to clarify guidelines for clonazepam use without the risk of dependence. Mr. O is prescribed clonazepam, which he takes consistently and responds to excellently.

4. Involve patients in setting treatment goals and targeting symptoms to be relieved. Actively listen when patients describe their symptoms, discomforts, and past experiences with treatments. I invite patients to speak uninterrupted for 5 to 10 minutes, even if they talk about issues that seem irrelevant. I then summarize the patient’s major points and ask, “And what else?” After he or she says, “That’s it,” I ask the patient to assign a priority to alleviating each symptom.

For example, Ms. J, a 38-year-old married mother of 2, was diagnosed with bipolar II disorder. She listed her highest priority as controlling her impulsive shopping rather than alleviating depression, insomnia, or overeating. She had been forced to declare bankruptcy twice, and she was determined to never do so again. She also wanted to regain her husband’s trust and her ability to manage her finances. Ensuring that Ms. J felt understood regarding this issue increased the chances of establishing a solid treatment partnership. Providing Ms. J with a menu of treatment choices and asking her to describe her previous experiences with medications helped her and the clinician choose a medication that is compatible with her desire to control her impulsive shopping.

5. Engage patients in choosing the best delivery system for the prescribed medication. For many medications, clinicians can choose from a variety of delivery systems, including pills, transdermal patches, rectal or vaginal suppositories, creams, ointments, orally disintegrating tablets, liquids, and intramuscular injections. Patients have varying beliefs about the efficacy of particular delivery systems, based on personal experiences or what they have learned from the media, their family and friends, or the Internet. For example, Ms. S, age 28, experienced recurrent, disabling anxiety attacks. When asked about the best way of providing medication to relieve her symptoms, she chose gluteal injections because, as a child, her pediatrician had treated her for an unspecified illness by injecting medication in her buttock, which rapidly relieved her symptoms. This left her with the impression that injectable medications were the best therapeutic delivery system. After discussing the practicalities and availability of fast-acting medications to control panic attacks, we agreed to use orally disintegrating clonazepam, which is absorbed swiftly and provides fast symptom relief. Ms. S reported favorable results and was pleased with the process of developing this strategy with her clinician.

6. Involve patients in choosing the times and frequency of medication administration. The timing and frequency of medication administration can be used to enhance desirable therapeutic effects. For example, an antidepressant that causes sedation and somnolence could be taken at bedtime to help alleviate insomnia. Some studies have shown that taking a medication once a day improves adherence compared with taking the same medication in divided doses.¹³ Other patients may wish to take a medication several times a day so they can keep the medication in their purse or briefcase and feel confident that if they need a medication for immediate symptom relief, it will be readily available.

7. Teach patients to self-monitor changes and improvements in target symptoms. Engaging patients in a system of self-monitoring improves their chances of achieving successful treatment outcomes.¹⁴ Instruct patients to create a list of symptoms and monitor the intensity of each symptom using a rating scale of 1 to 5, where 1 represents the lowest intensity and 5 represents the highest. As for frequency, patients can rate each symptom from “not present” to “present most of the time.”

Self-monitoring allows patients to observe which daily behaviors and lifestyle choices make symptoms better and which make them worse. For example, Mrs. P, a 38-year-old married mother of 2, had anxiety and panic attacks associated with low self-esteem and chronic depression. Her clinician instructed her to use a 1-page form to monitor the frequency and intensity of her anxiety and panic symptoms by focusing on the physical manifestations, such as rapid heartbeat, shortness of breath, nausea, tremors, dry mouth, frequent urination, and diarrhea to see if there was any correlation between her behaviors and her symptoms.

8. Instruct patients to call you to report any changes, including minor successes. Early in my career, toward the end of each appointment after I’d prescribed medications I’d tell patients, “Please call me if you have a problem.” Frequently, patients would call with a list of problems and side effects that they believed were caused by the newly prescribed medication. Later, I realized that I may have inadvertently encouraged patients to develop problems so they would have a reason to call me. To achieve a more favorable outcome I changed the way I communicate. I now say, “Please call me next week, even if you begin to feel better with this new medication.” The phone call is now associated with the idea that they will “get better,” and internalizing such a suggestion allows patients to talk with the clinician and report favorable treatment results.

9. Tell patients to monitor their successes by relabeling and reframing their symptoms. Mr. B, age 28, has MDD and reports irritability, insomnia, short temper, and restlessness. After reviewing his desired treatment outcome, we discuss the benefits of pharmacotherapy. I tell him the new medication will improve the quality and length of his sleep, which will allow his body and mind to recharge his “internal batteries” and restore health and energy. When we discuss side effects, I tell him to expect a dry mouth, which will be his signal that the medication is working. This discussion helps patients reframe side effects and improves their ability to tolerate side effects and adhere to pharmacotherapy.     

10. Harness the placebo effect and the power of suggestion to increase chances of achieving the best treatment outcomes. In a previous article,¹² I reviewed the principles of recognizing and enhancing the placebo effect and the power of suggestion to improve the chances of achieving better pharmacotherapy outcomes. When practicing participatory pharmacotherapy, clinicians are consciously aware of the power embedded in their words and are careful to use language that enhances the placebo effect and the power of suggestion when prescribing medications. Use the patient’s own language as a way of pacing yourself to the patient’s description of his or her distress. For example, Ms. R, a 42-year-old mother of 3, describes her experiences seeking help for her anxiety and depression, stating that she has not yet found the right combination of medications that provide benefits with tolerable side effects. Her clinician responds by focusing on the word “yet” (pacing) stating, “even though you have not yet found the right combination of medications to provide the most desirable benefit of beginning healing and restoring your hope, I promise to work with you and together we will try to achieve an improvement in your overall health and well-being.” This response includes several positive words and suggestions of future success, which are referred to as leading.

Not all patients will respond to participatory pharmacotherapy. Some factors will make patients good candidates for this approach, and others should be considered exclusionary qualities (Table 2).

Bottom Line

“Participatory pharmacotherapy” involves identifying patients as partners in the process of treatment choice and decision-making, encouraging them to provide their opinions regarding medication use, and making patients feel they have been heard and understood. This technique emphasizes forming a therapeutic alliance with the patient to improve patients’ adherence to pharmacotherapy and optimize treatment outcomes.

Related Resources

• Haynes RB, Ackloo E, Sahota N, et al. Interventions for enhancing medication adherence. Cochrane Database Syst Rev. 2008;16(2);CD000011.

• Mahone IH. Shared decision making and serious mental illness. Arch Psychiatr Nurs. 2008;22(6):334-343.

• Russel CL, Ruppar TM, Metteson M. Improving medication adherence: moving from intention and motivation to a personal systems approach. Nurs Clin North Am. 2011;46(3):271-281.

• Tibaldi G, Salvador-Carulla L, Garcia-Gutierrez JC. From treatment adherence to advanced shared decision making: New professional strategies and attitudes in mental health care. Curr Clin Pharmacol. 2011;6(2):91-99. 

Drug Brand Name

Clonazepam • Klonopin

Disclosure

Dr. Torem reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Although clinicians and patients generally are cautious when prescribing or using antipsychotics during pregnancy, inadequately controlled psychiatric illness poses risks to both mother and child. Calculating the risks and benefits of antipsychotic use during pregnancy is limited by an incomplete understanding of the true effectiveness and full spectrum of risks of these medications. Ethical principles prohibit the type of rigorous research that would be needed to achieve clarity on this issue. This article reviews studies that might help guide clinicians who are considering prescribing an atypical antipsychotic to manage psychiatric illness in a pregnant woman.

Antipsychotic efficacy in pregnancy

All atypical antipsychotics available in the United States are FDA-approved for treating schizophrenia; some also have been approved for treating bipolar disorder, unipolar depression, or symptoms associated with autism (Table 1). Atypical antipsychotics frequently are used off-label for these and other categories of psychiatric illness, including unipolar depression, generalized anxiety disorder, and obsessive-compulsive disorder.

Studies of pharmacotherapy in pregnant women tend to focus more on safety rather than efficacy. Clinical decisions for an individual patient are best made based on knowledge about which medications have been effective for that patient in the past (Algorithm).

However, safety concerns in pregnancy may require modifying an existing regimen. In other cases, new symptoms arise during pregnancy and necessitate new medications. Additionally, a drug’s effectiveness may be affected by physiologic changes of pregnancy that can alter drug metabolism,¹ potentially necessitating dose changes (Box 1).

Risks of treatment vs illness

Complete safety data on the use of any psychotropic medication during pregnancy are not available. To date, studies of atypical anti­psychotics do not support any increased risk for congenital malformations large enough to be detected in medium-sized samples,^2-4 although it is possible that there are increases in risk that are below the detection limit of these studies. Data regarding delivery outcomes are conflicting and difficult to interpret.

Several studies^2-4 have yielded inconsistent results, including:

•  risks for increased birth weight and large for gestational age³

•  risks for low birth weight and small for gestational age²

•  no significant differences from controls.⁴

Atypical antipsychotics increase the risk of gestational diabetes, whereas typical antipsychotics do not appear to increase this risk.⁴

Until recently, research has been limited by difficulties in separating the effects of treatment from the effects of psychiatric illness, which include intrauterine growth retardation, prematurity, preterm birth, low Apgar scores, and congenital defects.⁵ In addition, most studies address early and easily measurable outcomes such as preterm labor, birth weight, and congenital malformations. Researchers are just beginning to investigate more subtle and long-term potential behavioral effects.

Several recent studies have explored outcomes associated with antipsychotic use during pregnancy while attempting to separate the effects of treatment from those of disease (Box 2).

Data on atypicals

Aripiprazole. Case reports of aripiprazole use during pregnancy have reported difficulties including transient unexplained fetal tachycardia that required emergent caesarean section⁶ and transient respiratory distress.⁷ Several small case series were not powered to detect risks related to aripiprazole.^8,9

Animal data suggest teratogenic potential at dosages 3 and 10 times the maximum recommended human dose.^10,11 Two studies^7,12 that measured placental transfer of aripiprazole found cord-to-maternal serum concentration ratios ranging from 0.47 to 0.63, which is similar to the ratios for quetiapine and risperidone and lower than those for olanzapine and haloperidol.¹³

There are insufficient data to identify risks related to aripiprazole compared with other drugs in its class, and fewer reports are available than for other atypical antipsychotics such as quetiapine and olanzapine. Placental transfer appears to be on the lower end of the spectrum for drugs in this class. Aripiprazole would be an acceptable choice for a woman who had a history of response to aripiprazole but likely would not be a first choice for a woman requiring a new medication during pregnancy.

Clozapine. In case reports, adverse effects associated with clozapine exposure during pregnancy include major malformations, gestational metabolic complications, poor pregnancy outcome, and perinatal adverse reactions. In one case, neonatal plasma clozapine concentrations were found to be twice that found in maternal plasma.¹⁴ Animal data have shown no evidence of increased teratogenicity at 2 to 4 times the maximum recommended human dosages.¹⁵ Boden et al⁸ found an increased risk for gestational diabetes and macrocephaly with clozapine (11 exposures). Four other series^2-4,16 were underpowered to detect concerns related specifically to clozapine.

There are insufficient data to identify risks related specifically to clozapine use during pregnancy. However, the rare but severe adverse effects associated with clozapine in other patient populations—including agranulocytosis and severe constipation¹⁷—could be devastating in a pregnant patient, which suggests this medication would not be a first-line treatment.

Olanzapine. In postmarketing surveillance studies and case reports, there have been have anecdotal cases of fetal malformations related to olanzapine use during pregnancy. Several larger studies^2-4,8,16 did not find higher rates of congenital malformations or any pattern of malformation types, although none were designed or powered to examine rare events. Animal data show no evidence of teratogenicity.¹⁸ A study comparing rates of placental passage of antipsychotics¹³ found higher rates for olanzapine than for quetiapine and risperidone, as well as higher prevalence of low birth weight and perinatal complications. A neonatal withdrawal syndrome has also been reported.¹⁹ Boden et al⁸ found an increased risk for gestational diabetes and macrocephaly with olanzapine.

Data suggest that olanzapine may be associated with somewhat higher rates of the adverse effects attributable to atypical antipsychotics (gestational diabetes and possibly macrocephaly), which could be related to olanzapine’s relatively higher rate of placental passage. Olanzapine could be a reasonable choice in a woman who had a history of good response to this medication, but would be lower priority than quetiapine when a new drug is indicated during pregnancy.

Quetiapine. In clinical trials, quetiapine had lower rates of placental passage compared with risperidone and olanzapine.¹³ One case report found only small changes in quetiapine serum levels during pregnancy.²⁰ Prospective studies (90 exposures,⁸ 36 exposures,² 7 exposures,¹⁶ 4 exposures,³ and 4 exposures⁴) show no increase in fetal malformations or adverse neonatal health outcomes related to quetiapine, and manufacturer safety data reveal no teratogenic effect, although delays in fetal skeletal ossification were seen in rats and rabbits at doses comparable to the recommended human range.²¹

Quetiapine is a reasonable first choice when a new atypical antipsychotic is indicated for a pregnant patient.

Risperidone. Rates of placental passage of risperidone are higher compared with quetiapine.¹³ Postmarketing surveillance data (265 exposures²² and 10 exposures²³) and prospective studies (including 72 exposures,⁸ 49 exposures,² 51 exposures,⁴ 16 exposures,¹⁶ and 5 exposures³) suggest risperidone has no major teratogenic effect. When malformations were present, they were similar to expected rates and types of malformations, and no specific malformation type was overrepresented. However, in some cases, researchers noted a withdrawal-emergent syndrome that included various combinations of tremors, irritability, poor feeding, and somnolence.²² Animal data are similarly reassuring, although increases in early fetal mortality and (potentially related) changes in maternal behavior have been observed in rats.^24,25 A major caveat with risperidone is its propensity to cause hyperprolactinemia, which is detrimental to efforts to conceive and maintain a pregnancy.^26,27

Risperidone is not associated with higher rates of adverse events in pregnancy than other atypical antipsychotics. It would not be a first choice for a woman trying to conceive or in the early stages of pregnancy, but would be a reasonable choice for a woman already well into pregnancy.

Ziprasidone. Available reports are few and generally do not report findings on ziprasidone separately.^8,28 Manufacturer data includes 5 spontaneous abortions, one malformation, and one stillbirth among 57 exposures,⁴ and available animal data suggest significant developmental toxicity and impaired fertility.²⁹ In pregnant rats, ziprasidone dosed as low as 0.5 times the maximum human recommended dose resulted in delayed fetal skeletal ossification, increased stillbirths, and decreased fetal weight and postnatal survival, and ziprasidone dosed as low as 0.2 times the maximum recommended human dose resulted in developmental delays and neurobehavioral impairments in offspring. In pregnant rabbits, ziprasidone dosed at 3 times the maximum recommended human dose resulted in cardiac and renal malformations.²⁹

Although available data are too sparse to draw reliable conclusions, the small amount of human data plus animal data suggest that ziprasidone should be less preferred than other atypical antipsychotics during pregnancy.

Lurasidone. No data addressing lurasidone use in humans during pregnancy are available. Material submitted to the FDA includes no evidence of teratogenicity or embryo-fetal toxicity in rat and rabbit studies using 3 and 12 times the maximum recommended human dose (80 mg) based on a body surface area comparison.³⁰

Asenapine. No data specifically addressing asenapine use in humans during pregnancy are available. Studies in rats and rabbits found no increase in teratogenicity, but did find increases in postimplantation loss and decreases in pup survival and weight gain with maternal doses equivalent to less than the maximum recommended human dose.³¹

Iloperidone. No data specifically addressing iloperidone use in humans during pregnancy are available. Animal studies of iloperidone found multiple developmental toxicities when iloperidone was administered during gestation.³² In one study, pregnant rats were given up to 26 times the maximum recommended human dose of 24 mg/d during the period of organogenesis. The highest dose caused increased early intrauterine deaths, decreased fetal weight and length, decreased fetal skeletal ossification, and increased minor fetal skeletal anomalies and variations. In a similar study using pregnant rabbits, the highest dose caused increased early intrauterine deaths and decreased fetal viability at term.

Paliperidone. In animal studies, there were no increases in fetal abnormalities when pregnant rats and rabbits were treated with up to 8 times the maximum recommended human dose of paliperidone during the period of organogenesis.³³

A single case report³⁴ measured levels of risperidone and its 9-hydroxy metabolite, paliperidone, in the breast milk of a mother who had taken risperidone during pregnancy and in the serum of her child. 9-OH-risperidone dose in breast milk was calculated as 4.7% of the weight-adjusted maternal dose, and serum levels in the infant were undetectable. No ill effects on the child were observed.

It is not possible to draw solid conclusions about atypical antipsychotics’ potential effects on human development from animal studies. Because of the lack of human data for the newer atypical antipsychotics—asenapine, iloperidone, lurasidone, paliperidone—in general these agents would not be advisable as first-line medications for treating pregnant women.

A few caveats

All atypical antipsychotics share the propensity to trigger or worsen glucose intolerance, which can have significant negative consequences in a pregnant patient. When deciding to use an atypical antipsychotic during pregnancy, blood glucose should be monitored carefully and regularly.

Because all atypical antipsychotics (except clozapine) are FDA pregnancy class C—indicating that animal reproduction studies have shown an adverse effect on the fetus and there are no adequate and well-controlled studies in humans, but potential benefits may warrant use of the drug in pregnant women despite potential risks—the decision to use these medications must be based on an individualized assessment of risks and benefits. Patients and their providers together should make a fully informed decision.

There is an urgent need for larger and better-designed investigations that will be sufficiently powered to detect differences in outcomes—particularly major malformations, preterm delivery, adverse events in labor and delivery, metabolic and anthropometric effects on the newborn, and neurodevelopmental and psychiatric outcomes for individuals exposed in utero—between women without mental illness, untreated women with mental illness, and women receiving atypical antipsychotics during pregnancy. Further research into the pharmacokinetics and clinical efficacy of antipsychotics in pregnant women also would be useful. Clinicians can assist with these efforts by submitting their patient data to a pregnancy registry maintained by the Massachusetts General Hospital (see Related Resources).

Bottom Line
Treatment with atypical antipsychotics during pregnancy may increase the risk of adverse birth outcomes, but inadequately controlled mental illness also carries some degree of risk. The decision to use any atypical antipsychotic during pregnancy must be based on an individualized assessment of risks and benefits and made by the pregnant patient and her provider.

Related Resources

• Gentile S. Antipsychotic therapy during early and late pregnancy. A systematic review. Schizophr Bull. 2010;36(3):518-544. www.ncbi.nlm.nih.gov/pmc/articles/PMC2879689.

• Massachusetts General Hospital National Pregnancy Registry for Atypical Antipsychotics. www.womensmentalhealth.org/clinical-and-research-programs/pregnancyregistry.

Drug Brand Names

Aripiprazole • Abilify                         Olanzapine • Zyprexa

Asenapine • Saphris                         Paliperidone • Invega

Clozapine • Clozaril                           Quetiapine • Seroquel

Haloperidol • Haldol                          Risperidone • Risperdal

Iloperidone • Fanapt                         Ziprasidone • Geodon

Lurasidone • Latuda

Disclosures

The authors report no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Although clinicians and patients generally are cautious when prescribing or using antipsychotics during pregnancy, inadequately controlled psychiatric illness poses risks to both mother and child. Calculating the risks and benefits of antipsychotic use during pregnancy is limited by an incomplete understanding of the true effectiveness and full spectrum of risks of these medications. Ethical principles prohibit the type of rigorous research that would be needed to achieve clarity on this issue. This article reviews studies that might help guide clinicians who are considering prescribing an atypical antipsychotic to manage psychiatric illness in a pregnant woman.

Antipsychotic efficacy in pregnancy

All atypical antipsychotics available in the United States are FDA-approved for treating schizophrenia; some also have been approved for treating bipolar disorder, unipolar depression, or symptoms associated with autism (Table 1). Atypical antipsychotics frequently are used off-label for these and other categories of psychiatric illness, including unipolar depression, generalized anxiety disorder, and obsessive-compulsive disorder.

Studies of pharmacotherapy in pregnant women tend to focus more on safety rather than efficacy. Clinical decisions for an individual patient are best made based on knowledge about which medications have been effective for that patient in the past (Algorithm).

However, safety concerns in pregnancy may require modifying an existing regimen. In other cases, new symptoms arise during pregnancy and necessitate new medications. Additionally, a drug’s effectiveness may be affected by physiologic changes of pregnancy that can alter drug metabolism,¹ potentially necessitating dose changes (Box 1).

Risks of treatment vs illness

Complete safety data on the use of any psychotropic medication during pregnancy are not available. To date, studies of atypical anti­psychotics do not support any increased risk for congenital malformations large enough to be detected in medium-sized samples,^2-4 although it is possible that there are increases in risk that are below the detection limit of these studies. Data regarding delivery outcomes are conflicting and difficult to interpret.

Several studies^2-4 have yielded inconsistent results, including:

•  risks for increased birth weight and large for gestational age³

•  risks for low birth weight and small for gestational age²

•  no significant differences from controls.⁴

Atypical antipsychotics increase the risk of gestational diabetes, whereas typical antipsychotics do not appear to increase this risk.⁴

Until recently, research has been limited by difficulties in separating the effects of treatment from the effects of psychiatric illness, which include intrauterine growth retardation, prematurity, preterm birth, low Apgar scores, and congenital defects.⁵ In addition, most studies address early and easily measurable outcomes such as preterm labor, birth weight, and congenital malformations. Researchers are just beginning to investigate more subtle and long-term potential behavioral effects.

Several recent studies have explored outcomes associated with antipsychotic use during pregnancy while attempting to separate the effects of treatment from those of disease (Box 2).

Data on atypicals

Aripiprazole. Case reports of aripiprazole use during pregnancy have reported difficulties including transient unexplained fetal tachycardia that required emergent caesarean section⁶ and transient respiratory distress.⁷ Several small case series were not powered to detect risks related to aripiprazole.^8,9