Evidence-Based Reviews

Antipsychotics are FDA-approved as a primary treatment for schizophrenia and bipolar disorder and as adjunctive therapy for major depressive disorder. In the United States, approximately 26% of antipsychotic prescriptions written for these indications are for individuals age >65.¹ Additionally, antipsychotics are widely used to treat behavioral symptoms associated with dementia.¹ The rapid expansion of the use of second-generation antipsychotics (SGAs), in particular, has been driven in part by their lower risk for extrapyramidal symptoms (EPS) compared with first-generation antipsychotics (FGAs).¹ However, a growing body of data indicates that all antipsychotics have a range of adverse effects in older patients. This focus is critical in light of demographic trends—in the next 10 to 15 years, the population age >60 will grow 3.5 times more rapidly than the general population.²

In this context, psychiatrists need information on the relative risks of antipsychotics for older patients. This 3-part series summarizes findings and recommendations on safety and tolerability when prescribing antipsychotics in older individuals with chronic psychotic disorders, such as schizophrenia, bipolar disorder, depression, and dementia. This review aims to:

• briefly summarize the major studies and analyses relevant to older patients with these diagnoses
• provide a summative opinion on safety and tolerability issues in these older adults
• highlight the gaps in the evidence base and areas that need additional research.

Part 1 focuses on older adults with schizophrenia or bipolar disorder. Subsequent articles will focus on prescribing antipsychotics to older adults with depression and those with dementia.

Schizophrenia

Summary of benefits, place in treatment armamentarium. Individuals with schizophrenia have a shorter life expectancy than that of the general population mostly as a result of suicide and comorbid physical illnesses,³ but the number of patients with schizophrenia age >55 will double over the next 2 decades.⁴ With aging, both positive and negative symptoms may be a focus of treatment (Table 1).^5,6 Antipsychotics are a first-line treatment for older patients with schizophrenia with few medication alternatives.⁷ Safety risks associated with antipsychotics in older people span a broad spectrum (Table 2).⁸

A 6-week prospective RCT evaluated paliperidone extended-release vs placebo in 114 older adults (age ≥65 years; mean age, 70 years) with schizophrenia.¹⁴ There was an optional 24-week extension of open-label treatment with paliperidone. Mean daily dose of paliperidone was 8.4 mg. Efficacy measures did not show consistent statistically significant differences between treatment groups. Discontinuation rates were similar between paliperidone (7%) vs placebo (8%). Serious adverse events occurred in 3% of paliperidone-treated vs 8% of placebo-treated patients. Elevated prolactin levels occurred in one-half of paliperidone-treated patients. There were no prolactin or glucose treatment-related adverse events or significant mean changes in body weight for either paliperidone-treated or placebo-treated patients. Safety findings in the 24-week, open-label extension group were consistent with the RCT results.

Howanitz et al¹⁵ conducted a 12-week, prospective RCT that compared clozapine (mean dose, 300 mg/d) with chlorpromazine (mean dose, 600 mg/d) in 42 older adults (mean age, 67 years) with schizophrenia. Drop-out rate prior to 5 weeks was 19% and similar between groups. Common adverse effects included sialorrhea, hematologic abnormalities, sedation, tachycardia, EPS, and weight gain. Although both drugs were effective, more patients taking clozapine had tachycardia and weight gain, while more chlorpromazine patients reported sedation.

There have been other, less rigorous studies.^7,8 Most of these studies evaluated risperidone and olanzapine, and most were conducted in “younger” geriatric patients (age <75 years). Although patients who participate in clinical trials may be healthier than “typical” patients, adverse effects such as EPS, sedation, and weight gain were still relatively common in these studies.

Other clinical data. A major consideration in treating older adults with schizophrenia is balancing the need to administer an antipsychotic dose high enough to alleviate psychotic symptoms while minimizing dose-dependent adverse effects. There is a U-shaped relationship between age and vulnerability to antipsychotic adverse effects,^16,17 wherein adverse effects are highest at younger and older ages. Evidence supports using the lowest effective antipsychotic dose for geriatric patients with schizophrenia. Positive emission tomography (PET) studies suggest that older patients develop EPS with lower doses despite lower receptor occupancy.^17,18 A recent study of 35 older patients (mean age, 60.1 years) with schizophrenia obtained PET, clinical measures, and blood pharmacokinetic measures before and after reduction of risperidone or olanzapine doses.¹⁸ A ≥40% reduction in dose was associated with reduced adverse effects, particularly EPS and elevation of prolactin levels. Moreover, the therapeutic window of striatal D2/D3 receptor occupancy appeared to be 50% to 60% in these older patients, compared with 65% to 80% in younger patients.

Long-term risks of antipsychotic treatment across the lifespan are less clear, with evidence suggesting both lower and higher mortality risk.^19,20 It is difficult to fully disentangle the long-term risks of antipsychotics from the cumulative effects of lifestyle and comorbidity among individuals who have lived with schizophrenia for decades. Large naturalistic studies that include substantial numbers of older people with schizophrenia might be a way to elicit more information on long-term safety. The Schizophrenia Outpatient Health Outcome (SOHO) study was a large naturalistic trial that recruited >10,000 individuals with schizophrenia in 10 European countries.²¹ Although the SOHO study found differences between antipsychotics and adverse effects, such as EPS, weight gain, and sexual dysfunction, because the mean age of these patients was approximately 40 years and the follow-up period was only 3 years, it is difficult to draw conclusions that could be relevant to older individuals who have had schizophrenia for decades.

Bipolar Disorder

A secondary analysis of mixed-age, RCTs examined response in older adults (age ≥55 years) with bipolar I depression who received lurasidone as monotherapy or adjunctive therapy.³² In the monotherapy study, these patients were randomized to 6 weeks of lurasidone 20 to 60 mg/d, lurasidone 80 to 120 mg/d, or placebo. In the adjunctive therapy study, they were randomized to lurasidone 20 to 120 mg/d or placebo with either lithium or valproate. There were 83 older adults (17.1% of the sample) in the monotherapy study and 53 (15.6%) in the adjunctive therapy study. Mean improvement in depression was significantly higher for both doses of lurasidone monotherapy than placebo. Adjunctive lurasidone was not associated with statistically significant improvement vs placebo. The most frequent adverse events in older patients on lurasidone monotherapy 20 to 60 mg/d or 80 to 120 mg/d were nausea (18.5% and 9.7%, respectively) and somnolence (11.1% and 0%, respectively). Akathisia (9.7%) and insomnia (9.7%) were the most common adverse events in the group receiving 80 to 120 mg/d, with the rate of akathisia exhibiting a dose-related increase. Weight change with lurasidone was similar to placebo, and there were no clinically meaningful group changes in vital signs, electrocardiography, or laboratory parameters.

A small (N = 20) open study found improvement in older adults with bipolar depression with aripiprazole (mean dose, 10.3 mg/d).³³ Adverse effects included restlessness and weight gain (n = 3, 9% each), sedation (n = 2, 10%), and drooling and diarrhea/loose stools (n = 1, 5% each). In another small study (N = 15) using asenapine (mean dose, 11.2 mg/d) in mainly older bipolar patients with depression, the most common adverse effects were gastrointestinal (GI) discomfort (n = 5, 33%) and restlessness, tremors, cognitive difficulties, and sluggishness (n = 2, 13% each).³⁴

Clinical trials: Bipolar mania. Researchers conducted a pooled analysis of two 12-week randomized trials comparing quetiapine with placebo in a mixed-age sample with bipolar mania.³⁵ In a subgroup of 59 older patients (mean age, 62.9 years), manic symptoms improved significantly more with quetiapine (modal dose, 550 mg/d) than with placebo. Adverse effects reported by >10% of older patients were dry mouth, somnolence, postural hypotension, insomnia, weight gain, and dizziness. Insomnia was reported by >10% of patients receiving placebo.

In a case series of 11 elderly patients with mania receiving asenapine, Baruch et al³⁶ reported a 63% remission rate. One patient discontinued the study because of a new rash, 1 discontinued after developing peripheral edema, and 3 patients reported mild sedation.

Beyer et al³⁷ reported on a post hoc analysis of 94 older adults (mean age, 57.1 years; range, 50.1 to 74.8 years) with acute bipolar mania receiving olanzapine (n = 47), divalproex (n = 31), or placebo (n = 16) in a pooled olanzapine clinical trials database. Patients receiving olanzapine or divalproex had improvement in mania; those receiving placebo did not improve. Safety findings were comparable with reports in younger patients with mania.

Other clinical data. Adverse effects found in mixed-age samples using secondary analyses of clinical trials need to be interpreted with caution because these types of studies usually exclude individuals with significant medical comorbidity. Medical burden, cognitive impairment, or concomitant medications generally necessitate slower drug titration and lower total daily dosing. For example, a secondary analysis of the U.S. National Institute of Health-funded Systematic Treatment Enhancement Program for Bipolar Disorder study, which had broader inclusion criteria than most clinical trials, reported that, although recovery rates in older adults with bipolar disorder were fairly good (78.5%), lower doses of risperidone were used in older vs younger patients.³⁸

Clinical considerations

Interpretation of the relative risks of antipsychotics in older people must be tempered by the caveat that there is limited high-quality data (Table 4). Antipsychotics are the first-line therapy for older patients with schizophrenia, although their use is supported by a small number of prospective RCTs. SGAs are preferred because of their lower propensity to cause EPS and other motor adverse effects. Older persons with schizophrenia have an EPS threshold lower than younger patients and determining the lowest effective dosage may minimize EPS and cognitive adverse effects. As individuals with long-standing schizophrenia get older, their antipsychotic dosages may need to be reduced, and clinicians need to monitor for adverse effects that are more common among older people, such as tardive dyskinesia and metabolic abnormalities. In healthy, “younger” geriatric patients, monitoring for adverse effects may be similar to monitoring of younger patients. Patients who are older or frail may need more frequent assessment.

Like older adults with schizophrenia, geriatric patients with bipolar disorder have reduced drug tolerability and experience more adverse effects than younger patients. There are no prospective controlled studies that evaluated using antipsychotics in older patients with bipolar disorder. In older bipolar patients, the most problematic adverse effects of antipsychotics are akathisia, parkinsonism, other EPS, sedation and dizziness (which may increase fall risk), and GI discomfort. A key tolerability and safety consideration when treating older adults with bipolar disorder is the role of antipsychotics in relation to the use of lithium and mood stabilizers. Some studies have suggested that lithium has neuroprotective effects when used long-term; however, at least 1 report suggested that long-term antipsychotic treatment may be associated with neurodegeneration.³⁹

The literature does not provide strong evidence on the many clinical variations that we see in routine practice settings, such as combinations of drug treatments or drugs prescribed to patients with specific comorbid conditions. There is a need for large cohort studies that monitor treatment course, medical comorbidity, and prognosis. Additionally, well-designed clinical trials such as the DART-AD, which investigated longer-term trajectories of people with dementia taking antipsychotics, should serve as a model for the type of research that is needed to better understand outcome variability among older people with chronic psychotic or bipolar disorders.⁴⁰

Antipsychotics are FDA-approved as a primary treatment for schizophrenia and bipolar disorder and as adjunctive therapy for major depressive disorder. In the United States, approximately 26% of antipsychotic prescriptions written for these indications are for individuals age >65.¹ Additionally, antipsychotics are widely used to treat behavioral symptoms associated with dementia.¹ The rapid expansion of the use of second-generation antipsychotics (SGAs), in particular, has been driven in part by their lower risk for extrapyramidal symptoms (EPS) compared with first-generation antipsychotics (FGAs).¹ However, a growing body of data indicates that all antipsychotics have a range of adverse effects in older patients. This focus is critical in light of demographic trends—in the next 10 to 15 years, the population age >60 will grow 3.5 times more rapidly than the general population.²

In this context, psychiatrists need information on the relative risks of antipsychotics for older patients. This 3-part series summarizes findings and recommendations on safety and tolerability when prescribing antipsychotics in older individuals with chronic psychotic disorders, such as schizophrenia, bipolar disorder, depression, and dementia. This review aims to:

• briefly summarize the major studies and analyses relevant to older patients with these diagnoses
• provide a summative opinion on safety and tolerability issues in these older adults
• highlight the gaps in the evidence base and areas that need additional research.

Part 1 focuses on older adults with schizophrenia or bipolar disorder. Subsequent articles will focus on prescribing antipsychotics to older adults with depression and those with dementia.

Schizophrenia

Summary of benefits, place in treatment armamentarium. Individuals with schizophrenia have a shorter life expectancy than that of the general population mostly as a result of suicide and comorbid physical illnesses,³ but the number of patients with schizophrenia age >55 will double over the next 2 decades.⁴ With aging, both positive and negative symptoms may be a focus of treatment (Table 1).^5,6 Antipsychotics are a first-line treatment for older patients with schizophrenia with few medication alternatives.⁷ Safety risks associated with antipsychotics in older people span a broad spectrum (Table 2).⁸

A 6-week prospective RCT evaluated paliperidone extended-release vs placebo in 114 older adults (age ≥65 years; mean age, 70 years) with schizophrenia.¹⁴ There was an optional 24-week extension of open-label treatment with paliperidone. Mean daily dose of paliperidone was 8.4 mg. Efficacy measures did not show consistent statistically significant differences between treatment groups. Discontinuation rates were similar between paliperidone (7%) vs placebo (8%). Serious adverse events occurred in 3% of paliperidone-treated vs 8% of placebo-treated patients. Elevated prolactin levels occurred in one-half of paliperidone-treated patients. There were no prolactin or glucose treatment-related adverse events or significant mean changes in body weight for either paliperidone-treated or placebo-treated patients. Safety findings in the 24-week, open-label extension group were consistent with the RCT results.

Howanitz et al¹⁵ conducted a 12-week, prospective RCT that compared clozapine (mean dose, 300 mg/d) with chlorpromazine (mean dose, 600 mg/d) in 42 older adults (mean age, 67 years) with schizophrenia. Drop-out rate prior to 5 weeks was 19% and similar between groups. Common adverse effects included sialorrhea, hematologic abnormalities, sedation, tachycardia, EPS, and weight gain. Although both drugs were effective, more patients taking clozapine had tachycardia and weight gain, while more chlorpromazine patients reported sedation.

There have been other, less rigorous studies.^7,8 Most of these studies evaluated risperidone and olanzapine, and most were conducted in “younger” geriatric patients (age <75 years). Although patients who participate in clinical trials may be healthier than “typical” patients, adverse effects such as EPS, sedation, and weight gain were still relatively common in these studies.

Other clinical data. A major consideration in treating older adults with schizophrenia is balancing the need to administer an antipsychotic dose high enough to alleviate psychotic symptoms while minimizing dose-dependent adverse effects. There is a U-shaped relationship between age and vulnerability to antipsychotic adverse effects,^16,17 wherein adverse effects are highest at younger and older ages. Evidence supports using the lowest effective antipsychotic dose for geriatric patients with schizophrenia. Positive emission tomography (PET) studies suggest that older patients develop EPS with lower doses despite lower receptor occupancy.^17,18 A recent study of 35 older patients (mean age, 60.1 years) with schizophrenia obtained PET, clinical measures, and blood pharmacokinetic measures before and after reduction of risperidone or olanzapine doses.¹⁸ A ≥40% reduction in dose was associated with reduced adverse effects, particularly EPS and elevation of prolactin levels. Moreover, the therapeutic window of striatal D2/D3 receptor occupancy appeared to be 50% to 60% in these older patients, compared with 65% to 80% in younger patients.

Long-term risks of antipsychotic treatment across the lifespan are less clear, with evidence suggesting both lower and higher mortality risk.^19,20 It is difficult to fully disentangle the long-term risks of antipsychotics from the cumulative effects of lifestyle and comorbidity among individuals who have lived with schizophrenia for decades. Large naturalistic studies that include substantial numbers of older people with schizophrenia might be a way to elicit more information on long-term safety. The Schizophrenia Outpatient Health Outcome (SOHO) study was a large naturalistic trial that recruited >10,000 individuals with schizophrenia in 10 European countries.²¹ Although the SOHO study found differences between antipsychotics and adverse effects, such as EPS, weight gain, and sexual dysfunction, because the mean age of these patients was approximately 40 years and the follow-up period was only 3 years, it is difficult to draw conclusions that could be relevant to older individuals who have had schizophrenia for decades.

Bipolar Disorder

A secondary analysis of mixed-age, RCTs examined response in older adults (age ≥55 years) with bipolar I depression who received lurasidone as monotherapy or adjunctive therapy.³² In the monotherapy study, these patients were randomized to 6 weeks of lurasidone 20 to 60 mg/d, lurasidone 80 to 120 mg/d, or placebo. In the adjunctive therapy study, they were randomized to lurasidone 20 to 120 mg/d or placebo with either lithium or valproate. There were 83 older adults (17.1% of the sample) in the monotherapy study and 53 (15.6%) in the adjunctive therapy study. Mean improvement in depression was significantly higher for both doses of lurasidone monotherapy than placebo. Adjunctive lurasidone was not associated with statistically significant improvement vs placebo. The most frequent adverse events in older patients on lurasidone monotherapy 20 to 60 mg/d or 80 to 120 mg/d were nausea (18.5% and 9.7%, respectively) and somnolence (11.1% and 0%, respectively). Akathisia (9.7%) and insomnia (9.7%) were the most common adverse events in the group receiving 80 to 120 mg/d, with the rate of akathisia exhibiting a dose-related increase. Weight change with lurasidone was similar to placebo, and there were no clinically meaningful group changes in vital signs, electrocardiography, or laboratory parameters.

A small (N = 20) open study found improvement in older adults with bipolar depression with aripiprazole (mean dose, 10.3 mg/d).³³ Adverse effects included restlessness and weight gain (n = 3, 9% each), sedation (n = 2, 10%), and drooling and diarrhea/loose stools (n = 1, 5% each). In another small study (N = 15) using asenapine (mean dose, 11.2 mg/d) in mainly older bipolar patients with depression, the most common adverse effects were gastrointestinal (GI) discomfort (n = 5, 33%) and restlessness, tremors, cognitive difficulties, and sluggishness (n = 2, 13% each).³⁴

Clinical trials: Bipolar mania. Researchers conducted a pooled analysis of two 12-week randomized trials comparing quetiapine with placebo in a mixed-age sample with bipolar mania.³⁵ In a subgroup of 59 older patients (mean age, 62.9 years), manic symptoms improved significantly more with quetiapine (modal dose, 550 mg/d) than with placebo. Adverse effects reported by >10% of older patients were dry mouth, somnolence, postural hypotension, insomnia, weight gain, and dizziness. Insomnia was reported by >10% of patients receiving placebo.

In a case series of 11 elderly patients with mania receiving asenapine, Baruch et al³⁶ reported a 63% remission rate. One patient discontinued the study because of a new rash, 1 discontinued after developing peripheral edema, and 3 patients reported mild sedation.

Beyer et al³⁷ reported on a post hoc analysis of 94 older adults (mean age, 57.1 years; range, 50.1 to 74.8 years) with acute bipolar mania receiving olanzapine (n = 47), divalproex (n = 31), or placebo (n = 16) in a pooled olanzapine clinical trials database. Patients receiving olanzapine or divalproex had improvement in mania; those receiving placebo did not improve. Safety findings were comparable with reports in younger patients with mania.

Other clinical data. Adverse effects found in mixed-age samples using secondary analyses of clinical trials need to be interpreted with caution because these types of studies usually exclude individuals with significant medical comorbidity. Medical burden, cognitive impairment, or concomitant medications generally necessitate slower drug titration and lower total daily dosing. For example, a secondary analysis of the U.S. National Institute of Health-funded Systematic Treatment Enhancement Program for Bipolar Disorder study, which had broader inclusion criteria than most clinical trials, reported that, although recovery rates in older adults with bipolar disorder were fairly good (78.5%), lower doses of risperidone were used in older vs younger patients.³⁸

Clinical considerations

Interpretation of the relative risks of antipsychotics in older people must be tempered by the caveat that there is limited high-quality data (Table 4). Antipsychotics are the first-line therapy for older patients with schizophrenia, although their use is supported by a small number of prospective RCTs. SGAs are preferred because of their lower propensity to cause EPS and other motor adverse effects. Older persons with schizophrenia have an EPS threshold lower than younger patients and determining the lowest effective dosage may minimize EPS and cognitive adverse effects. As individuals with long-standing schizophrenia get older, their antipsychotic dosages may need to be reduced, and clinicians need to monitor for adverse effects that are more common among older people, such as tardive dyskinesia and metabolic abnormalities. In healthy, “younger” geriatric patients, monitoring for adverse effects may be similar to monitoring of younger patients. Patients who are older or frail may need more frequent assessment.

Like older adults with schizophrenia, geriatric patients with bipolar disorder have reduced drug tolerability and experience more adverse effects than younger patients. There are no prospective controlled studies that evaluated using antipsychotics in older patients with bipolar disorder. In older bipolar patients, the most problematic adverse effects of antipsychotics are akathisia, parkinsonism, other EPS, sedation and dizziness (which may increase fall risk), and GI discomfort. A key tolerability and safety consideration when treating older adults with bipolar disorder is the role of antipsychotics in relation to the use of lithium and mood stabilizers. Some studies have suggested that lithium has neuroprotective effects when used long-term; however, at least 1 report suggested that long-term antipsychotic treatment may be associated with neurodegeneration.³⁹

The literature does not provide strong evidence on the many clinical variations that we see in routine practice settings, such as combinations of drug treatments or drugs prescribed to patients with specific comorbid conditions. There is a need for large cohort studies that monitor treatment course, medical comorbidity, and prognosis. Additionally, well-designed clinical trials such as the DART-AD, which investigated longer-term trajectories of people with dementia taking antipsychotics, should serve as a model for the type of research that is needed to better understand outcome variability among older people with chronic psychotic or bipolar disorders.⁴⁰

Antipsychotics are FDA-approved as a primary treatment for schizophrenia and bipolar disorder and as adjunctive therapy for major depressive disorder. In the United States, approximately 26% of antipsychotic prescriptions written for these indications are for individuals age >65.¹ Additionally, antipsychotics are widely used to treat behavioral symptoms associated with dementia.¹ The rapid expansion of the use of second-generation antipsychotics (SGAs), in particular, has been driven in part by their lower risk for extrapyramidal symptoms (EPS) compared with first-generation antipsychotics (FGAs).¹ However, a growing body of data indicates that all antipsychotics have a range of adverse effects in older patients. This focus is critical in light of demographic trends—in the next 10 to 15 years, the population age >60 will grow 3.5 times more rapidly than the general population.²

In this context, psychiatrists need information on the relative risks of antipsychotics for older patients. This 3-part series summarizes findings and recommendations on safety and tolerability when prescribing antipsychotics in older individuals with chronic psychotic disorders, such as schizophrenia, bipolar disorder, depression, and dementia. This review aims to:

• briefly summarize the major studies and analyses relevant to older patients with these diagnoses
• provide a summative opinion on safety and tolerability issues in these older adults
• highlight the gaps in the evidence base and areas that need additional research.

Part 1 focuses on older adults with schizophrenia or bipolar disorder. Subsequent articles will focus on prescribing antipsychotics to older adults with depression and those with dementia.

Schizophrenia

Summary of benefits, place in treatment armamentarium. Individuals with schizophrenia have a shorter life expectancy than that of the general population mostly as a result of suicide and comorbid physical illnesses,³ but the number of patients with schizophrenia age >55 will double over the next 2 decades.⁴ With aging, both positive and negative symptoms may be a focus of treatment (Table 1).^5,6 Antipsychotics are a first-line treatment for older patients with schizophrenia with few medication alternatives.⁷ Safety risks associated with antipsychotics in older people span a broad spectrum (Table 2).⁸

A 6-week prospective RCT evaluated paliperidone extended-release vs placebo in 114 older adults (age ≥65 years; mean age, 70 years) with schizophrenia.¹⁴ There was an optional 24-week extension of open-label treatment with paliperidone. Mean daily dose of paliperidone was 8.4 mg. Efficacy measures did not show consistent statistically significant differences between treatment groups. Discontinuation rates were similar between paliperidone (7%) vs placebo (8%). Serious adverse events occurred in 3% of paliperidone-treated vs 8% of placebo-treated patients. Elevated prolactin levels occurred in one-half of paliperidone-treated patients. There were no prolactin or glucose treatment-related adverse events or significant mean changes in body weight for either paliperidone-treated or placebo-treated patients. Safety findings in the 24-week, open-label extension group were consistent with the RCT results.

Howanitz et al¹⁵ conducted a 12-week, prospective RCT that compared clozapine (mean dose, 300 mg/d) with chlorpromazine (mean dose, 600 mg/d) in 42 older adults (mean age, 67 years) with schizophrenia. Drop-out rate prior to 5 weeks was 19% and similar between groups. Common adverse effects included sialorrhea, hematologic abnormalities, sedation, tachycardia, EPS, and weight gain. Although both drugs were effective, more patients taking clozapine had tachycardia and weight gain, while more chlorpromazine patients reported sedation.

There have been other, less rigorous studies.^7,8 Most of these studies evaluated risperidone and olanzapine, and most were conducted in “younger” geriatric patients (age <75 years). Although patients who participate in clinical trials may be healthier than “typical” patients, adverse effects such as EPS, sedation, and weight gain were still relatively common in these studies.

Other clinical data. A major consideration in treating older adults with schizophrenia is balancing the need to administer an antipsychotic dose high enough to alleviate psychotic symptoms while minimizing dose-dependent adverse effects. There is a U-shaped relationship between age and vulnerability to antipsychotic adverse effects,^16,17 wherein adverse effects are highest at younger and older ages. Evidence supports using the lowest effective antipsychotic dose for geriatric patients with schizophrenia. Positive emission tomography (PET) studies suggest that older patients develop EPS with lower doses despite lower receptor occupancy.^17,18 A recent study of 35 older patients (mean age, 60.1 years) with schizophrenia obtained PET, clinical measures, and blood pharmacokinetic measures before and after reduction of risperidone or olanzapine doses.¹⁸ A ≥40% reduction in dose was associated with reduced adverse effects, particularly EPS and elevation of prolactin levels. Moreover, the therapeutic window of striatal D2/D3 receptor occupancy appeared to be 50% to 60% in these older patients, compared with 65% to 80% in younger patients.

Long-term risks of antipsychotic treatment across the lifespan are less clear, with evidence suggesting both lower and higher mortality risk.^19,20 It is difficult to fully disentangle the long-term risks of antipsychotics from the cumulative effects of lifestyle and comorbidity among individuals who have lived with schizophrenia for decades. Large naturalistic studies that include substantial numbers of older people with schizophrenia might be a way to elicit more information on long-term safety. The Schizophrenia Outpatient Health Outcome (SOHO) study was a large naturalistic trial that recruited >10,000 individuals with schizophrenia in 10 European countries.²¹ Although the SOHO study found differences between antipsychotics and adverse effects, such as EPS, weight gain, and sexual dysfunction, because the mean age of these patients was approximately 40 years and the follow-up period was only 3 years, it is difficult to draw conclusions that could be relevant to older individuals who have had schizophrenia for decades.

Bipolar Disorder

A secondary analysis of mixed-age, RCTs examined response in older adults (age ≥55 years) with bipolar I depression who received lurasidone as monotherapy or adjunctive therapy.³² In the monotherapy study, these patients were randomized to 6 weeks of lurasidone 20 to 60 mg/d, lurasidone 80 to 120 mg/d, or placebo. In the adjunctive therapy study, they were randomized to lurasidone 20 to 120 mg/d or placebo with either lithium or valproate. There were 83 older adults (17.1% of the sample) in the monotherapy study and 53 (15.6%) in the adjunctive therapy study. Mean improvement in depression was significantly higher for both doses of lurasidone monotherapy than placebo. Adjunctive lurasidone was not associated with statistically significant improvement vs placebo. The most frequent adverse events in older patients on lurasidone monotherapy 20 to 60 mg/d or 80 to 120 mg/d were nausea (18.5% and 9.7%, respectively) and somnolence (11.1% and 0%, respectively). Akathisia (9.7%) and insomnia (9.7%) were the most common adverse events in the group receiving 80 to 120 mg/d, with the rate of akathisia exhibiting a dose-related increase. Weight change with lurasidone was similar to placebo, and there were no clinically meaningful group changes in vital signs, electrocardiography, or laboratory parameters.

A small (N = 20) open study found improvement in older adults with bipolar depression with aripiprazole (mean dose, 10.3 mg/d).³³ Adverse effects included restlessness and weight gain (n = 3, 9% each), sedation (n = 2, 10%), and drooling and diarrhea/loose stools (n = 1, 5% each). In another small study (N = 15) using asenapine (mean dose, 11.2 mg/d) in mainly older bipolar patients with depression, the most common adverse effects were gastrointestinal (GI) discomfort (n = 5, 33%) and restlessness, tremors, cognitive difficulties, and sluggishness (n = 2, 13% each).³⁴

Clinical trials: Bipolar mania. Researchers conducted a pooled analysis of two 12-week randomized trials comparing quetiapine with placebo in a mixed-age sample with bipolar mania.³⁵ In a subgroup of 59 older patients (mean age, 62.9 years), manic symptoms improved significantly more with quetiapine (modal dose, 550 mg/d) than with placebo. Adverse effects reported by >10% of older patients were dry mouth, somnolence, postural hypotension, insomnia, weight gain, and dizziness. Insomnia was reported by >10% of patients receiving placebo.

In a case series of 11 elderly patients with mania receiving asenapine, Baruch et al³⁶ reported a 63% remission rate. One patient discontinued the study because of a new rash, 1 discontinued after developing peripheral edema, and 3 patients reported mild sedation.

Beyer et al³⁷ reported on a post hoc analysis of 94 older adults (mean age, 57.1 years; range, 50.1 to 74.8 years) with acute bipolar mania receiving olanzapine (n = 47), divalproex (n = 31), or placebo (n = 16) in a pooled olanzapine clinical trials database. Patients receiving olanzapine or divalproex had improvement in mania; those receiving placebo did not improve. Safety findings were comparable with reports in younger patients with mania.

Other clinical data. Adverse effects found in mixed-age samples using secondary analyses of clinical trials need to be interpreted with caution because these types of studies usually exclude individuals with significant medical comorbidity. Medical burden, cognitive impairment, or concomitant medications generally necessitate slower drug titration and lower total daily dosing. For example, a secondary analysis of the U.S. National Institute of Health-funded Systematic Treatment Enhancement Program for Bipolar Disorder study, which had broader inclusion criteria than most clinical trials, reported that, although recovery rates in older adults with bipolar disorder were fairly good (78.5%), lower doses of risperidone were used in older vs younger patients.³⁸

Clinical considerations

Interpretation of the relative risks of antipsychotics in older people must be tempered by the caveat that there is limited high-quality data (Table 4). Antipsychotics are the first-line therapy for older patients with schizophrenia, although their use is supported by a small number of prospective RCTs. SGAs are preferred because of their lower propensity to cause EPS and other motor adverse effects. Older persons with schizophrenia have an EPS threshold lower than younger patients and determining the lowest effective dosage may minimize EPS and cognitive adverse effects. As individuals with long-standing schizophrenia get older, their antipsychotic dosages may need to be reduced, and clinicians need to monitor for adverse effects that are more common among older people, such as tardive dyskinesia and metabolic abnormalities. In healthy, “younger” geriatric patients, monitoring for adverse effects may be similar to monitoring of younger patients. Patients who are older or frail may need more frequent assessment.

Like older adults with schizophrenia, geriatric patients with bipolar disorder have reduced drug tolerability and experience more adverse effects than younger patients. There are no prospective controlled studies that evaluated using antipsychotics in older patients with bipolar disorder. In older bipolar patients, the most problematic adverse effects of antipsychotics are akathisia, parkinsonism, other EPS, sedation and dizziness (which may increase fall risk), and GI discomfort. A key tolerability and safety consideration when treating older adults with bipolar disorder is the role of antipsychotics in relation to the use of lithium and mood stabilizers. Some studies have suggested that lithium has neuroprotective effects when used long-term; however, at least 1 report suggested that long-term antipsychotic treatment may be associated with neurodegeneration.³⁹

The literature does not provide strong evidence on the many clinical variations that we see in routine practice settings, such as combinations of drug treatments or drugs prescribed to patients with specific comorbid conditions. There is a need for large cohort studies that monitor treatment course, medical comorbidity, and prognosis. Additionally, well-designed clinical trials such as the DART-AD, which investigated longer-term trajectories of people with dementia taking antipsychotics, should serve as a model for the type of research that is needed to better understand outcome variability among older people with chronic psychotic or bipolar disorders.⁴⁰

Borderline personality disorder (BPD) is associated with impaired psychosocial functioning,^1-4 reduced health-related quality of life,⁵ high utilization of services,^6,7 and excess mortality.^8-10 Although BPD occurs in up to 40% of psychiatric inpatients¹¹ and 10% of outpatients,¹² it is underrecognized.^13-15 Often, patients with BPD do not receive an accurate diagnosis until ≥10 years after initially seeking treatment.¹⁶ The treatment and clinical implications of failing to recognize BPD include overprescribing medication and underutilizing empirically effective psychotherapies.¹⁴

This review summarizes studies of the underdiagnosis of BPD in routine clinical practice, describes which patients should be screened, and reviews alternative approaches to screening.

Underrecognition of BPD

The Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project is an ongoing clinical research study involving the integration of research assessment methods into routine clinical practice.¹⁷ In an early report from the MIDAS project, BPD diagnoses derived from structured and unstructured clinical interviews were compared between 2 groups of psychiatric outpatients in the same practice.¹⁵ Individuals in the structured interview cohort were 35 times more often diagnosed with BPD than individuals evaluated with an unstructured clinical interview. Importantly, when the information from the structured interview was presented to the clinicians, BPD was more likely to be diagnosed clinically.

Other studies^13,16 also found that the rate of diagnosing BPD was higher when the diagnosis was based on a semi-structured diagnostic interview compared with an unstructured clinical interview, and that clinicians were reluctant to diagnose BPD during their routine intake diagnostic evaluation.

Clinicians, however, do not use semi-structured interviews in their practice, and they also do not tend to diagnose personality disorders (PDs) based on direct questioning, as they typically would when assessing a symptom-based disorder such as depression or anxiety. Rather, clinicians report that they rely on longitudinal observations to diagnose PDs.¹⁸ However, the results from the MIDAS project were inconsistent with clinicians’ reports. When clinicians were presented with the results of the semi-structured interview, they usually would diagnose BPD, even though it was the initial evaluation. If clinicians actually relied on longitudinal observations and considered information based on the direct question approach of research interviews to be irrelevant or invalid, then the results from the semi-structured interview should not have influenced the rate at which they diagnosed BPD. This suggests that the primary issue in diagnosing PDs is not the need for longi­tudinal observation but rather the need for more information, and that there is a role for screening questionnaires.

One potential criticism of studies demonstrating underrecognition of BPD in clinical practice is that patients typically were interviewed when they presented for treatment, when most were depressed or anxious. The possible pathologizing effects of psychiatric state on personality have been known for years.¹⁹ However, a large body of literature examining the treatment, prognostic, familial, and biological correlates of PDs supports the validity of diagnosing PDs in this manner. Moreover, from a clinical perspective, the sooner a clinician is aware of the presence of BPD, the more likely this information can be used for treatment planning.

Who should be screened for BPD?

BPD is underrecognized and underdiagnosed because patients with BPD often also have comorbid mood, anxiety, or substance use disorders.^20,21 The symptoms associated with these disorders are typically the chief concern of patients with undiagnosed BPD who present for treatment. Patients with BPD rarely present for an intake evaluation and state that they are struggling with abandonment fears, chronic feelings of emptiness, or an identity disturbance. If patients identified these problems as their chief concerns, BPD would be easier to recognize.

Although several studies have documented the frequency of BPD in patients with a specific psychiatric diagnosis such as major depressive disorder (MDD) or attention-deficit/hyperactivity disorder,^22-26 the MIDAS project examined the frequency of BPD in patients with various diagnoses and evaluated which disorders were associated with a significantly increased rate of BPD.²⁷ The highest rate of BPD was found in patients with bipolar disorder. Approximately 25% of patients with bipolar II disorder and one-third of those with bipolar I disorder were diagnosed with BPD; these rates were significantly higher than the rate of BPD in patients without these disorders (Table 1²⁷). The rate of BPD was second highest in patients with a principal diagnosis of posttraumatic stress disorder (PTSD) and MDD; however, the rate of BPD in these patients was not significantly elevated compared with patients who did not have these principal diagnoses. Three disorders were associated with a significantly lower rate of BPD: adjustment disorder, dysthymic disorder, and generalized anxiety disorder.

A brief review of screening statistics

Screening tests for most psychiatric disorders are based on multi-item scales in which a total score is computed from a sum of item scores, and a cutoff point is established to determine who does and does not screen positive on the test. However, sensitivity, specificity, and positive and negative predictive values are not invariant properties of a screening test with a continuous score distribution. Rather, the performance statistics of a scale can be altered by changing the threshold score to distinguish cases from non-cases. When the screening threshold is lowered, sensitivity increases and specificity decreases.

For screening, a broad net needs to be cast so that all (or almost all) cases are included. Therefore, the cutoff score should be set low to prioritize the sensitivity of the instrument. A screening scale also should have high negative predictive value so that the clinician can be confident that patients who screen negative on the test do not have the disorder.

Screening questionnaires for BPD

Several questionnaires have been developed to screen for PDs (Table 2^28-35). Some screen for each of the DSM PDs,^28,36-42 and some screen more broadly for the presence or absence of any PD.^29,43,44 The most commonly studied self-report scale for BPD is the McLean Screening Instrument for Borderline Personality Disorder (MSI-BPD),³⁰ a 10-item self-report scale derived from a subset of questions from the BPD module of a semi-structured diagnostic interview.

The initial validation study³⁰ found that the optimal cutoff score was 7, which resulted in a sensitivity of 81% and specificity of 89%. Three studies have evaluated the scale in adolescents and young adults,^45-47 and 3 studies examined the scale in adult outpatients.^48-50 Across all 6 studies, at the optimal cutoff scores determined in each study, the sensitivity of the MSI-BPD ranged from 68% to 94% (mean, 80%) and the specificity ranged from 66% to 80% (mean, 72%).

Problems with screening questionnaires. Although screening scales have been developed for many psychiatric disorders, they have not been widely used in mental health settings. In a previous commentary, I argued that the conceptual justification for using self-report screening scales for single disorders in psychiatric settings is questionable.⁵¹ Another problem with screening scales is their potential misuse as case-finding instruments. In the literature on bipolar disorder screening, several researchers misconstrued a positive screen to indicate caseness.⁵¹ Although this is not a problem with the screening measures or the selection of a cutoff score, caution must be taken to not confuse screening with diagnosis.⁵²

Screening for BPD as part of your diagnostic interview

An alternative approach to using self-administered questionnaires for screening is for clinicians to include questions in their evaluation as part of a psychiatric review of systems. When conducting a diagnostic interview, clinicians typically screen for disorders that are comorbid with the principal diagnosis by asking about the comorbid disorders’ necessary features or “gate criteria.” For example, in a patient with a principal diagnosis of MDD, the clinician would inquire about the presence of panic attacks, excessive worry, or substance use to screen for the presence of panic disorder, generalized anxiety disorder, or a substance use disorder. In contrast, for polythetically defined disorders such as BPD, there is no single gate criterion, because the disorder is diagnosed based on the presence of at least 5 of 9 criteria and no single one of these criteria is required to be present to establish the diagnosis.

As part of the MIDAS project, the psycho­metric properties of the BPD criteria were examined to determine if it was possible to identify 1 or 2 criteria that could serve as gate criteria to screen for the disorder. If the sensitivity of 1 criterion or a combination of 2 BPD criteria was sufficiently high (ie, >90%), then the assessment of this criterion (or these criteria) could be included in a psychiatric review of systems, thus potentially improving the detection of BPD. Researchers hypothesized that affective instability, considered first by Linehan⁵³ and later by other theorists⁵⁴ to be of central importance to the clinical manifestations of BPD, could function as a gate criterion. In the sample of 3,674 psychiatric outpatients who were evaluated with a semi-structured interview, the sensitivity of the affective instability criterion was 92.8%, and the negative predictive value of the criterion was 99%.

Borderline personality disorder (BPD) is associated with impaired psychosocial functioning,^1-4 reduced health-related quality of life,⁵ high utilization of services,^6,7 and excess mortality.^8-10 Although BPD occurs in up to 40% of psychiatric inpatients¹¹ and 10% of outpatients,¹² it is underrecognized.^13-15 Often, patients with BPD do not receive an accurate diagnosis until ≥10 years after initially seeking treatment.¹⁶ The treatment and clinical implications of failing to recognize BPD include overprescribing medication and underutilizing empirically effective psychotherapies.¹⁴

This review summarizes studies of the underdiagnosis of BPD in routine clinical practice, describes which patients should be screened, and reviews alternative approaches to screening.

Underrecognition of BPD

The Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project is an ongoing clinical research study involving the integration of research assessment methods into routine clinical practice.¹⁷ In an early report from the MIDAS project, BPD diagnoses derived from structured and unstructured clinical interviews were compared between 2 groups of psychiatric outpatients in the same practice.¹⁵ Individuals in the structured interview cohort were 35 times more often diagnosed with BPD than individuals evaluated with an unstructured clinical interview. Importantly, when the information from the structured interview was presented to the clinicians, BPD was more likely to be diagnosed clinically.

Other studies^13,16 also found that the rate of diagnosing BPD was higher when the diagnosis was based on a semi-structured diagnostic interview compared with an unstructured clinical interview, and that clinicians were reluctant to diagnose BPD during their routine intake diagnostic evaluation.

Clinicians, however, do not use semi-structured interviews in their practice, and they also do not tend to diagnose personality disorders (PDs) based on direct questioning, as they typically would when assessing a symptom-based disorder such as depression or anxiety. Rather, clinicians report that they rely on longitudinal observations to diagnose PDs.¹⁸ However, the results from the MIDAS project were inconsistent with clinicians’ reports. When clinicians were presented with the results of the semi-structured interview, they usually would diagnose BPD, even though it was the initial evaluation. If clinicians actually relied on longitudinal observations and considered information based on the direct question approach of research interviews to be irrelevant or invalid, then the results from the semi-structured interview should not have influenced the rate at which they diagnosed BPD. This suggests that the primary issue in diagnosing PDs is not the need for longi­tudinal observation but rather the need for more information, and that there is a role for screening questionnaires.

One potential criticism of studies demonstrating underrecognition of BPD in clinical practice is that patients typically were interviewed when they presented for treatment, when most were depressed or anxious. The possible pathologizing effects of psychiatric state on personality have been known for years.¹⁹ However, a large body of literature examining the treatment, prognostic, familial, and biological correlates of PDs supports the validity of diagnosing PDs in this manner. Moreover, from a clinical perspective, the sooner a clinician is aware of the presence of BPD, the more likely this information can be used for treatment planning.

Who should be screened for BPD?

BPD is underrecognized and underdiagnosed because patients with BPD often also have comorbid mood, anxiety, or substance use disorders.^20,21 The symptoms associated with these disorders are typically the chief concern of patients with undiagnosed BPD who present for treatment. Patients with BPD rarely present for an intake evaluation and state that they are struggling with abandonment fears, chronic feelings of emptiness, or an identity disturbance. If patients identified these problems as their chief concerns, BPD would be easier to recognize.

Although several studies have documented the frequency of BPD in patients with a specific psychiatric diagnosis such as major depressive disorder (MDD) or attention-deficit/hyperactivity disorder,^22-26 the MIDAS project examined the frequency of BPD in patients with various diagnoses and evaluated which disorders were associated with a significantly increased rate of BPD.²⁷ The highest rate of BPD was found in patients with bipolar disorder. Approximately 25% of patients with bipolar II disorder and one-third of those with bipolar I disorder were diagnosed with BPD; these rates were significantly higher than the rate of BPD in patients without these disorders (Table 1²⁷). The rate of BPD was second highest in patients with a principal diagnosis of posttraumatic stress disorder (PTSD) and MDD; however, the rate of BPD in these patients was not significantly elevated compared with patients who did not have these principal diagnoses. Three disorders were associated with a significantly lower rate of BPD: adjustment disorder, dysthymic disorder, and generalized anxiety disorder.

A brief review of screening statistics

Screening tests for most psychiatric disorders are based on multi-item scales in which a total score is computed from a sum of item scores, and a cutoff point is established to determine who does and does not screen positive on the test. However, sensitivity, specificity, and positive and negative predictive values are not invariant properties of a screening test with a continuous score distribution. Rather, the performance statistics of a scale can be altered by changing the threshold score to distinguish cases from non-cases. When the screening threshold is lowered, sensitivity increases and specificity decreases.

For screening, a broad net needs to be cast so that all (or almost all) cases are included. Therefore, the cutoff score should be set low to prioritize the sensitivity of the instrument. A screening scale also should have high negative predictive value so that the clinician can be confident that patients who screen negative on the test do not have the disorder.

Screening questionnaires for BPD

Several questionnaires have been developed to screen for PDs (Table 2^28-35). Some screen for each of the DSM PDs,^28,36-42 and some screen more broadly for the presence or absence of any PD.^29,43,44 The most commonly studied self-report scale for BPD is the McLean Screening Instrument for Borderline Personality Disorder (MSI-BPD),³⁰ a 10-item self-report scale derived from a subset of questions from the BPD module of a semi-structured diagnostic interview.

The initial validation study³⁰ found that the optimal cutoff score was 7, which resulted in a sensitivity of 81% and specificity of 89%. Three studies have evaluated the scale in adolescents and young adults,^45-47 and 3 studies examined the scale in adult outpatients.^48-50 Across all 6 studies, at the optimal cutoff scores determined in each study, the sensitivity of the MSI-BPD ranged from 68% to 94% (mean, 80%) and the specificity ranged from 66% to 80% (mean, 72%).

Problems with screening questionnaires. Although screening scales have been developed for many psychiatric disorders, they have not been widely used in mental health settings. In a previous commentary, I argued that the conceptual justification for using self-report screening scales for single disorders in psychiatric settings is questionable.⁵¹ Another problem with screening scales is their potential misuse as case-finding instruments. In the literature on bipolar disorder screening, several researchers misconstrued a positive screen to indicate caseness.⁵¹ Although this is not a problem with the screening measures or the selection of a cutoff score, caution must be taken to not confuse screening with diagnosis.⁵²

Screening for BPD as part of your diagnostic interview

An alternative approach to using self-administered questionnaires for screening is for clinicians to include questions in their evaluation as part of a psychiatric review of systems. When conducting a diagnostic interview, clinicians typically screen for disorders that are comorbid with the principal diagnosis by asking about the comorbid disorders’ necessary features or “gate criteria.” For example, in a patient with a principal diagnosis of MDD, the clinician would inquire about the presence of panic attacks, excessive worry, or substance use to screen for the presence of panic disorder, generalized anxiety disorder, or a substance use disorder. In contrast, for polythetically defined disorders such as BPD, there is no single gate criterion, because the disorder is diagnosed based on the presence of at least 5 of 9 criteria and no single one of these criteria is required to be present to establish the diagnosis.

As part of the MIDAS project, the psycho­metric properties of the BPD criteria were examined to determine if it was possible to identify 1 or 2 criteria that could serve as gate criteria to screen for the disorder. If the sensitivity of 1 criterion or a combination of 2 BPD criteria was sufficiently high (ie, >90%), then the assessment of this criterion (or these criteria) could be included in a psychiatric review of systems, thus potentially improving the detection of BPD. Researchers hypothesized that affective instability, considered first by Linehan⁵³ and later by other theorists⁵⁴ to be of central importance to the clinical manifestations of BPD, could function as a gate criterion. In the sample of 3,674 psychiatric outpatients who were evaluated with a semi-structured interview, the sensitivity of the affective instability criterion was 92.8%, and the negative predictive value of the criterion was 99%.

Borderline personality disorder (BPD) is associated with impaired psychosocial functioning,^1-4 reduced health-related quality of life,⁵ high utilization of services,^6,7 and excess mortality.^8-10 Although BPD occurs in up to 40% of psychiatric inpatients¹¹ and 10% of outpatients,¹² it is underrecognized.^13-15 Often, patients with BPD do not receive an accurate diagnosis until ≥10 years after initially seeking treatment.¹⁶ The treatment and clinical implications of failing to recognize BPD include overprescribing medication and underutilizing empirically effective psychotherapies.¹⁴

This review summarizes studies of the underdiagnosis of BPD in routine clinical practice, describes which patients should be screened, and reviews alternative approaches to screening.

Underrecognition of BPD

The Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project is an ongoing clinical research study involving the integration of research assessment methods into routine clinical practice.¹⁷ In an early report from the MIDAS project, BPD diagnoses derived from structured and unstructured clinical interviews were compared between 2 groups of psychiatric outpatients in the same practice.¹⁵ Individuals in the structured interview cohort were 35 times more often diagnosed with BPD than individuals evaluated with an unstructured clinical interview. Importantly, when the information from the structured interview was presented to the clinicians, BPD was more likely to be diagnosed clinically.

Other studies^13,16 also found that the rate of diagnosing BPD was higher when the diagnosis was based on a semi-structured diagnostic interview compared with an unstructured clinical interview, and that clinicians were reluctant to diagnose BPD during their routine intake diagnostic evaluation.

Clinicians, however, do not use semi-structured interviews in their practice, and they also do not tend to diagnose personality disorders (PDs) based on direct questioning, as they typically would when assessing a symptom-based disorder such as depression or anxiety. Rather, clinicians report that they rely on longitudinal observations to diagnose PDs.¹⁸ However, the results from the MIDAS project were inconsistent with clinicians’ reports. When clinicians were presented with the results of the semi-structured interview, they usually would diagnose BPD, even though it was the initial evaluation. If clinicians actually relied on longitudinal observations and considered information based on the direct question approach of research interviews to be irrelevant or invalid, then the results from the semi-structured interview should not have influenced the rate at which they diagnosed BPD. This suggests that the primary issue in diagnosing PDs is not the need for longi­tudinal observation but rather the need for more information, and that there is a role for screening questionnaires.

One potential criticism of studies demonstrating underrecognition of BPD in clinical practice is that patients typically were interviewed when they presented for treatment, when most were depressed or anxious. The possible pathologizing effects of psychiatric state on personality have been known for years.¹⁹ However, a large body of literature examining the treatment, prognostic, familial, and biological correlates of PDs supports the validity of diagnosing PDs in this manner. Moreover, from a clinical perspective, the sooner a clinician is aware of the presence of BPD, the more likely this information can be used for treatment planning.

Who should be screened for BPD?

BPD is underrecognized and underdiagnosed because patients with BPD often also have comorbid mood, anxiety, or substance use disorders.^20,21 The symptoms associated with these disorders are typically the chief concern of patients with undiagnosed BPD who present for treatment. Patients with BPD rarely present for an intake evaluation and state that they are struggling with abandonment fears, chronic feelings of emptiness, or an identity disturbance. If patients identified these problems as their chief concerns, BPD would be easier to recognize.

Although several studies have documented the frequency of BPD in patients with a specific psychiatric diagnosis such as major depressive disorder (MDD) or attention-deficit/hyperactivity disorder,^22-26 the MIDAS project examined the frequency of BPD in patients with various diagnoses and evaluated which disorders were associated with a significantly increased rate of BPD.²⁷ The highest rate of BPD was found in patients with bipolar disorder. Approximately 25% of patients with bipolar II disorder and one-third of those with bipolar I disorder were diagnosed with BPD; these rates were significantly higher than the rate of BPD in patients without these disorders (Table 1²⁷). The rate of BPD was second highest in patients with a principal diagnosis of posttraumatic stress disorder (PTSD) and MDD; however, the rate of BPD in these patients was not significantly elevated compared with patients who did not have these principal diagnoses. Three disorders were associated with a significantly lower rate of BPD: adjustment disorder, dysthymic disorder, and generalized anxiety disorder.

A brief review of screening statistics

Screening tests for most psychiatric disorders are based on multi-item scales in which a total score is computed from a sum of item scores, and a cutoff point is established to determine who does and does not screen positive on the test. However, sensitivity, specificity, and positive and negative predictive values are not invariant properties of a screening test with a continuous score distribution. Rather, the performance statistics of a scale can be altered by changing the threshold score to distinguish cases from non-cases. When the screening threshold is lowered, sensitivity increases and specificity decreases.

For screening, a broad net needs to be cast so that all (or almost all) cases are included. Therefore, the cutoff score should be set low to prioritize the sensitivity of the instrument. A screening scale also should have high negative predictive value so that the clinician can be confident that patients who screen negative on the test do not have the disorder.

Screening questionnaires for BPD

Several questionnaires have been developed to screen for PDs (Table 2^28-35). Some screen for each of the DSM PDs,^28,36-42 and some screen more broadly for the presence or absence of any PD.^29,43,44 The most commonly studied self-report scale for BPD is the McLean Screening Instrument for Borderline Personality Disorder (MSI-BPD),³⁰ a 10-item self-report scale derived from a subset of questions from the BPD module of a semi-structured diagnostic interview.

The initial validation study³⁰ found that the optimal cutoff score was 7, which resulted in a sensitivity of 81% and specificity of 89%. Three studies have evaluated the scale in adolescents and young adults,^45-47 and 3 studies examined the scale in adult outpatients.^48-50 Across all 6 studies, at the optimal cutoff scores determined in each study, the sensitivity of the MSI-BPD ranged from 68% to 94% (mean, 80%) and the specificity ranged from 66% to 80% (mean, 72%).

Problems with screening questionnaires. Although screening scales have been developed for many psychiatric disorders, they have not been widely used in mental health settings. In a previous commentary, I argued that the conceptual justification for using self-report screening scales for single disorders in psychiatric settings is questionable.⁵¹ Another problem with screening scales is their potential misuse as case-finding instruments. In the literature on bipolar disorder screening, several researchers misconstrued a positive screen to indicate caseness.⁵¹ Although this is not a problem with the screening measures or the selection of a cutoff score, caution must be taken to not confuse screening with diagnosis.⁵²

Screening for BPD as part of your diagnostic interview

An alternative approach to using self-administered questionnaires for screening is for clinicians to include questions in their evaluation as part of a psychiatric review of systems. When conducting a diagnostic interview, clinicians typically screen for disorders that are comorbid with the principal diagnosis by asking about the comorbid disorders’ necessary features or “gate criteria.” For example, in a patient with a principal diagnosis of MDD, the clinician would inquire about the presence of panic attacks, excessive worry, or substance use to screen for the presence of panic disorder, generalized anxiety disorder, or a substance use disorder. In contrast, for polythetically defined disorders such as BPD, there is no single gate criterion, because the disorder is diagnosed based on the presence of at least 5 of 9 criteria and no single one of these criteria is required to be present to establish the diagnosis.

As part of the MIDAS project, the psycho­metric properties of the BPD criteria were examined to determine if it was possible to identify 1 or 2 criteria that could serve as gate criteria to screen for the disorder. If the sensitivity of 1 criterion or a combination of 2 BPD criteria was sufficiently high (ie, >90%), then the assessment of this criterion (or these criteria) could be included in a psychiatric review of systems, thus potentially improving the detection of BPD. Researchers hypothesized that affective instability, considered first by Linehan⁵³ and later by other theorists⁵⁴ to be of central importance to the clinical manifestations of BPD, could function as a gate criterion. In the sample of 3,674 psychiatric outpatients who were evaluated with a semi-structured interview, the sensitivity of the affective instability criterion was 92.8%, and the negative predictive value of the criterion was 99%.

In an age when psychiatry strives to identify the biologic causes of disease, studying endocrine-related mood disorders is particularly intriguing. DSM-5 defines premenstrual dysphoric disorder (PMDD) as a depressive disorder, with a 12-month prevalence ranging from 1.8% to 5.8% among women who menstruate.^1-3 Factors that differentiate PMDD from other affective disorders include etiology, duration, and temporal relationship with the menstrual cycle.

PMDD is a disorder of consistent yet intermittent change in mental health and functionality. Therefore, it may be underdiagnosed and consequently undertreated if a psychiatric evaluation does not coincide with symptom occurrence or if patients do not understand that intermittent symptoms are treatable.

This article summarizes what is known about the etiology of PMDD. Although there are several treatments for PMDD, many women experience adverse effects or incomplete effectiveness. Further understanding of this disorder may lead to more efficacious treatments. Additionally, understanding the pathophysiology of PMDD might shed a light on the etiology of other disorders that are temporally related to reproductive life changes, such as pregnancy-, postpartum-, or menopause-related affective dysregulation.

Making the diagnosis

The diagnosis of PMDD is made when a patient has at least 5 of 11 specific symptoms that occur during the week before onset of menses, improve within a few days after the onset of menses (shown as the “PMDD Hazard Zone” in Figure 1), and are minimal or absent post-menses.³ Symptoms should be tracked prospectively for at least 2 menstrual cycles in order to confirm the diagnosis (one must be an affective symptom and another must be a behavioral/cognitive symptom).³

The affective symptoms are:

• lability of affect (eg, sudden sadness, tearfulness, or sensitivity to rejection)
• irritability, anger, or increased inter­personal conflicts
• depressed mood, hopelessness, or self- deprecating thoughts
• anxiety or tension, feeling “keyed up” or “on edge.”

The behavioral/cognitive symptoms are:

• decreased interest in usual activities (eg, work, hobbies, friends, school)
• difficulty concentrating
• lethargy, low energy, easy fatigability
• change in appetite, overeating, food cravings
• hypersomnia or insomnia
• feeling overwhelmed or out of control
• physical symptoms (breast tenderness or swelling, headache, joint or muscle pain, bloating, weight gain).

Ruling out premenstrual exacerbation (PME). Perhaps the most common cause for misdiagnosis of PMDD is failing to rule out PME of another underlying or comorbid condition (Figure 2). In many women who have a primary mood or anxiety disorder, the late luteal phase is a vulnerable time. A patient might be coping with untreated anxiety, for example, but the symptoms become unbearable the week before menstruation begins, which is likely when she seeks help. At this stage, a diagnosis of PMDD should be provisional at best. Often, PME is treated by treating the underlying condition. Therefore, a full diagnostic psychiatric interview is important to first rule out other underlying psychiatric disorders. PMDD is diagnosed if the premenstrual symptoms persist for 2 consecutive months after treating the suspected mood or anxiety disorder. Patients can use one of many PMDD daily symptom charts available online. Alternatively, they can use a cycle-tracking mobile phone application to correlate their symptoms with their cycle and share this information with their providers.

Consider these 5 interwoven pieces

The many variables that contribute to the pathophysiology of PMDD overlap and should be considered connecting pieces in the puzzle that is the etiology of this disorder (Figure 3). In reviewing the literature, we have identified 5 topics likely to be major contributors to this disorder:

1. genetic susceptibility
2. progesterone and allopregnanolone (ALLO)
3. estrogen, serotonin, and brain-derived neurotrophic factor (BDNF)
4. putative brain structural and functional differences
5. further involvement of the hypothalamic–pituitary–adrenal (HPA) axis and hypothalamic–pituitary–gonadal (HPG) axis: trauma, resiliency, and inflammation.

Genetic susceptibility. PMDD is thought to have a heritability range between 30% to 80%.³ This is demonstrated by family and twin studies^4-7 and specific genetic studies.⁸ The involvement of genetics means an underlying neurobiologic pathophysiology is in place.

Estrogen receptor alpha (ESR1) gene. Huo et al⁸ found an associated variation in ESR1 in women with PMDD compared with controls. They speculated that because ESR1 is important for arousal, if dysfunctional, this gene could be implicated in somatic as well as affective and cognitive deficits in PMDD patients. In another study, investigators reported a relationship between PMDD and heritable personality traits, as well as a link between these traits and ESR1 polymorphic variants.¹ They suggested that personality traits (independent of affective state) might be used to distinguish patients with PMDD from controls.¹

Studies on serotonin gene polymorphism and serotonin transporter genotype. Although a study of serotonin gene polymorphism did not find an association between serotonin1A gene polymorphism and PMDD, it did show that the presence of at least 1 C allele was associated with a 2.5-fold increased risk of PMDD.⁹ Another study did not find an association between the serotonin transporter genotype 5-HTTLPR and PMDD.¹⁰ However, it showed lower frontocingulate cortex activation during the luteal phase of PMDD patients compared with controls, suggesting that PMDD is linked to impaired frontocingulate cortex activation induced by emotions during the luteal phase.¹⁰ 

Seasonal affective disorder (SAD) and PMDD have shared clinical features. A polymorphism in the serotonin transporter promoter gene 5-HTTLPR has been associated with SAD. One study found that patients with comorbid SAD and PMDD are genetically more vulnerable to comorbid affective disorders compared with patients who have SAD only.¹¹

Progesterone and ALLO. Chronic exposure to progesterone and ALLO (a main progesterone metabolite) and rapid withdrawal from ovarian hormones may play a role in the etiology of PMDD. Much like alcohol or benzodiazepines, ALLO is a potent positive allosteric modulator of GABA_A receptors and has sedative, anesthetic, and anxiolytic properties. In times of acute stress, increased ALLO is known to provide relief.^12,13 However, in women with PMDD, this typical ALLO increase might not occur.¹⁴

Patients with PMDD have been reported to have decreased levels of ALLO in the luteal phase.^15-17 In one study, women with highly symptomatic PMDD had lower levels of ALLO compared with women with less symptomatic PMDD.¹⁴ A gonadotropin-releasing hormone challenge study showed the increase in ALLO response was less in PMDD patients compared with controls.¹⁷ Luteal-phase ALLO concentrations are reported to be lower in women with premenstrual syndrome (PMS), a milder form of PMDD.^14,17

The efficacy of selective serotonin reuptake inhibitors (SSRIs) for treating PMDD could be the result of the interaction of these medications with neuroactive steroids,¹⁸ possibly because SSRIs enhance the sensitivity of GABA_A receptors or promote the formation of more ALLO (Figure 4).^19-21

Estrogen, serotonin, and BDNF. Estrogen affects multiple neurotransmitter systems that regulate mood, cognition, sleep, and eating.²² Studying estrogen in context of PMDD is important because women with PMDD can have low mood, specific food cravings, and impaired cognitive function.

Estrogen–serotonin interactions are thought to be involved in hormone-related mood disorders such as perimenopausal depression and PMDD.^23,24 However, the nature of their relationship is not yet fully understood. Ovariectomized animals have shown estrogen-induced changes related to serotonin metabolism, binding, and transmission in the regions of the brain involved in regulation of affect and cognition. Research in menopausal women also has provided some support for this interaction.²⁴

Positron emission tomography studies in humans have found increased cortical serotonin binding modulated by levels of estrogen, similar to those previously seen in rat studies.^24-27 One study showed an increased binding potential of serotonin in the cerebral cortex with estrogen treatment. This study further showed an even greater binding potential with estrogen plus progesterone, signaling a synergistic effect of the 2 hormones.²⁸

SSRIs are an effective treatment for the irritability, anxiety, and mood swings of PMDD.^29-30 Although the exact mechanism of action is unknown, the serotonergic properties are certainly of primary attention. For some PMDD patients, SSRIs work within hours to days, as opposed to days or weeks for patients with depression or anxiety, which suggests a separate or co-occurring mechanism of action is in place. In a double-blind, placebo-controlled crossover study, researchers administered the serotonin receptor antagonist metergoline to women with PMDD whose symptoms had remitted during treatment with fluoxetine and a group of healthy controls who were not receiving any medication.³¹ The women with PMDD experienced a return of symptoms 24 hours after treatment with metergoline but not with placebo; the controls experienced no mood changes.³¹

BDNF is a neurotransmitter linked to estrogen and likely related to PMDD. BDNF is critical for neurogenesis and is expressed in brain regions involved in learning and memory and also affects regulation.³² BDNF levels are increased by serotonergic antidepressants, affected by estradiol, and have cyclicity throughout the menstrual cycle.^33-35

Putative brain structural and functional differences. Imaging studies have suggested differences in brain structure in women with PMDD, with a focus on the amygdala and the prefrontal cortex. Women with PMDD have greater gray matter volume in the posterior cerebellum,³⁶ greater gray matter density of hippocampal cortex, and lower gray matter density in the parahippocampal cortex.³⁷

Some studies have shown a functional variability of the amygdala’s response to stress in women with PMDD vs healthy controls.^38,39 A proton magnetic resonance spectroscopy (1H-MRS) study of the displays the possibility of an altered GABAergic function in patients with PMDD.⁴⁰

Patients will PMDD have enhanced dorsolateral prefrontal cortex reactivity when anticipating negative stimuli (but not to the actual exposure) during the luteal phase. A positive correlation between this reactivity and progesterone levels also was observed.⁴¹ Some researchers have suggested that prefrontal cortex dysfunction may be a risk factor for PMDD.⁴²

HPA axis and HPG axis: Trauma, resiliency, inflammation. Altered cortisol levels (higher during the luteal phase⁴³ and lower during times of stress^14,44) suggest a possibly altered HPA axis in some women with PMDD. However, studies on this topic have been few and inconsistent.

Dysregulation of the HPG axis could cause vasomotor symptoms, sleep dysregulation, and mood symptoms during menopause; women with PMDD can also experience these symptoms. The influence of estrogen and progesterone on mood is also highly dependent on this axis.

Ultimately, the interplay between the HPA axis and the HPG axis is important. One study found that women with PMDD who had high serum ALLO levels (HPG-related) had blunted nocturnal cortisol levels (HPA-related) compared with healthy controls who had low ALLO levels.⁴⁵

Significant stress and trauma exposure have been associated with PMDD. A study of 3,968 women found a history of trauma and PTSD were independently associated with PMDD.⁴⁶ Another study of approximately 3,000 women found a strong correlation between abuse and PMS.⁴⁷ However, a third study found no correlations between PMDD and trauma.⁴⁸

Patients with a predisposition to PMDD may be more vulnerable to develop a posttraumatic stress-related disorder, perhaps due to decreased biologic resiliency. For example, the startle response (hyper­vigilance) has been shown to be different in women with PMDD. One study suggested that suboptimal production of premenstrual ALLO may lead to increased arousal and increased stress reactivity to psycho­social or environmental triggers.⁴⁹

The possible role of inflammation in PMDD deserves further investigation. The luteal phase entails an increase in the production of proinflammatory markers.^50,51 A 10-fold increase in progesterone is correlated with a 20% to 23% increase in C-reactive protein levels.^52,53 Women with inflammatory diseases (eg, gingivitis or irritable bowel syndrome) show worsening of symptoms prior to menstruation.^54-56 One study found increased levels of proinflammatory markers in women with PMDD compared with controls.⁵⁷

Because PMDD is heritable, it must have an underlying neurobiologic pathophysiology. Brain imaging studies show differences in structure and function in women with PMDD across the menstrual cycle. Conversion of progesterone to ALLO and the GABAergic influence of this metabolite is a topic of interest in current research. Similarly, the role of estrogen and its connection to serotonin and other neurotransmitters such as BDNF have been implicated.

The link between a history of stress, trauma, and PMDD raises the question of biologic resiliency and illness in these patients, as it connects to the HPA and HPG axis and production of inflammatory stress hormones and steroid hormones and their metabolites. PMDD can be conceptualized as variable sensitivity to hormonal response to stress,⁵⁸ thus contextualizing biochemical and psychological resiliency.

Further research is needed to clarify the possibility of a shared pathophysiology between endocrine-related mood disorders such as postpartum depression (PPD) and PMDD because current research is controversial.^59,60 In PPD, women who are exposed to high levels of progesterone and estrogen during pregnancy (just like in the mid-luteal phase) have a sudden drop in these hormones postpartum.

The ‘withdrawal theory.’ The affective symptoms of PMDD resolve almost instantaneously after the start of menstruation. Perhaps this type of immediate relief is akin to substance use disorders and symptoms of withdrawal. It could be that reinstatement of a certain amount of gonadal steroids in the follicular phase of the cycle diminishes a withdrawal-like response to these steroids.

Currently, the main leading theory is that PMDD is a result of “an abnormal response to normal hormonal changes.”⁶¹ A new study also has shown that the change in estradiol/progesterone levels (vs the steady state) was associated with PMDD symptoms.⁶² Thinking of PMDD as a disorder of withdrawal offers an alternative (yet complementary) perspective to the current theory: PMDD may be caused by the absence or diminishing of the above-named hormones and their metabolites in the late luteal phase (in the context of developed “tolerance” during the early- to mid-luteal phase).

Considering the interplay between neuro­transmitters and neurosteroids, both a “serotonin withdrawal theory” (caused by a drop in steroid hormones) and a “GABAergic withdrawal theory” (due to the decline in progesterone) could be proposed. This theory would be supported by the fact that SSRIs seem to mitigate symptoms of PMDD as well as the genetic association between PMDD and ESR1. It is more than likely that the “withdrawal” is caused by the interactions between estrogen-serotonin, progesterone-ALLO, and GABA receptors, and the complementary fashion in which progesterone and estrogen influence each other.

In an age when psychiatry strives to identify the biologic causes of disease, studying endocrine-related mood disorders is particularly intriguing. DSM-5 defines premenstrual dysphoric disorder (PMDD) as a depressive disorder, with a 12-month prevalence ranging from 1.8% to 5.8% among women who menstruate.^1-3 Factors that differentiate PMDD from other affective disorders include etiology, duration, and temporal relationship with the menstrual cycle.

PMDD is a disorder of consistent yet intermittent change in mental health and functionality. Therefore, it may be underdiagnosed and consequently undertreated if a psychiatric evaluation does not coincide with symptom occurrence or if patients do not understand that intermittent symptoms are treatable.

This article summarizes what is known about the etiology of PMDD. Although there are several treatments for PMDD, many women experience adverse effects or incomplete effectiveness. Further understanding of this disorder may lead to more efficacious treatments. Additionally, understanding the pathophysiology of PMDD might shed a light on the etiology of other disorders that are temporally related to reproductive life changes, such as pregnancy-, postpartum-, or menopause-related affective dysregulation.

Making the diagnosis

The diagnosis of PMDD is made when a patient has at least 5 of 11 specific symptoms that occur during the week before onset of menses, improve within a few days after the onset of menses (shown as the “PMDD Hazard Zone” in Figure 1), and are minimal or absent post-menses.³ Symptoms should be tracked prospectively for at least 2 menstrual cycles in order to confirm the diagnosis (one must be an affective symptom and another must be a behavioral/cognitive symptom).³

The affective symptoms are:

• lability of affect (eg, sudden sadness, tearfulness, or sensitivity to rejection)
• irritability, anger, or increased inter­personal conflicts
• depressed mood, hopelessness, or self- deprecating thoughts
• anxiety or tension, feeling “keyed up” or “on edge.”

The behavioral/cognitive symptoms are:

• decreased interest in usual activities (eg, work, hobbies, friends, school)
• difficulty concentrating
• lethargy, low energy, easy fatigability
• change in appetite, overeating, food cravings
• hypersomnia or insomnia
• feeling overwhelmed or out of control
• physical symptoms (breast tenderness or swelling, headache, joint or muscle pain, bloating, weight gain).

Ruling out premenstrual exacerbation (PME). Perhaps the most common cause for misdiagnosis of PMDD is failing to rule out PME of another underlying or comorbid condition (Figure 2). In many women who have a primary mood or anxiety disorder, the late luteal phase is a vulnerable time. A patient might be coping with untreated anxiety, for example, but the symptoms become unbearable the week before menstruation begins, which is likely when she seeks help. At this stage, a diagnosis of PMDD should be provisional at best. Often, PME is treated by treating the underlying condition. Therefore, a full diagnostic psychiatric interview is important to first rule out other underlying psychiatric disorders. PMDD is diagnosed if the premenstrual symptoms persist for 2 consecutive months after treating the suspected mood or anxiety disorder. Patients can use one of many PMDD daily symptom charts available online. Alternatively, they can use a cycle-tracking mobile phone application to correlate their symptoms with their cycle and share this information with their providers.

Consider these 5 interwoven pieces

The many variables that contribute to the pathophysiology of PMDD overlap and should be considered connecting pieces in the puzzle that is the etiology of this disorder (Figure 3). In reviewing the literature, we have identified 5 topics likely to be major contributors to this disorder:

1. genetic susceptibility
2. progesterone and allopregnanolone (ALLO)
3. estrogen, serotonin, and brain-derived neurotrophic factor (BDNF)
4. putative brain structural and functional differences
5. further involvement of the hypothalamic–pituitary–adrenal (HPA) axis and hypothalamic–pituitary–gonadal (HPG) axis: trauma, resiliency, and inflammation.

Genetic susceptibility. PMDD is thought to have a heritability range between 30% to 80%.³ This is demonstrated by family and twin studies^4-7 and specific genetic studies.⁸ The involvement of genetics means an underlying neurobiologic pathophysiology is in place.

Estrogen receptor alpha (ESR1) gene. Huo et al⁸ found an associated variation in ESR1 in women with PMDD compared with controls. They speculated that because ESR1 is important for arousal, if dysfunctional, this gene could be implicated in somatic as well as affective and cognitive deficits in PMDD patients. In another study, investigators reported a relationship between PMDD and heritable personality traits, as well as a link between these traits and ESR1 polymorphic variants.¹ They suggested that personality traits (independent of affective state) might be used to distinguish patients with PMDD from controls.¹

Studies on serotonin gene polymorphism and serotonin transporter genotype. Although a study of serotonin gene polymorphism did not find an association between serotonin1A gene polymorphism and PMDD, it did show that the presence of at least 1 C allele was associated with a 2.5-fold increased risk of PMDD.⁹ Another study did not find an association between the serotonin transporter genotype 5-HTTLPR and PMDD.¹⁰ However, it showed lower frontocingulate cortex activation during the luteal phase of PMDD patients compared with controls, suggesting that PMDD is linked to impaired frontocingulate cortex activation induced by emotions during the luteal phase.¹⁰ 

Seasonal affective disorder (SAD) and PMDD have shared clinical features. A polymorphism in the serotonin transporter promoter gene 5-HTTLPR has been associated with SAD. One study found that patients with comorbid SAD and PMDD are genetically more vulnerable to comorbid affective disorders compared with patients who have SAD only.¹¹

Progesterone and ALLO. Chronic exposure to progesterone and ALLO (a main progesterone metabolite) and rapid withdrawal from ovarian hormones may play a role in the etiology of PMDD. Much like alcohol or benzodiazepines, ALLO is a potent positive allosteric modulator of GABA_A receptors and has sedative, anesthetic, and anxiolytic properties. In times of acute stress, increased ALLO is known to provide relief.^12,13 However, in women with PMDD, this typical ALLO increase might not occur.¹⁴

Patients with PMDD have been reported to have decreased levels of ALLO in the luteal phase.^15-17 In one study, women with highly symptomatic PMDD had lower levels of ALLO compared with women with less symptomatic PMDD.¹⁴ A gonadotropin-releasing hormone challenge study showed the increase in ALLO response was less in PMDD patients compared with controls.¹⁷ Luteal-phase ALLO concentrations are reported to be lower in women with premenstrual syndrome (PMS), a milder form of PMDD.^14,17

The efficacy of selective serotonin reuptake inhibitors (SSRIs) for treating PMDD could be the result of the interaction of these medications with neuroactive steroids,¹⁸ possibly because SSRIs enhance the sensitivity of GABA_A receptors or promote the formation of more ALLO (Figure 4).^19-21

Estrogen, serotonin, and BDNF. Estrogen affects multiple neurotransmitter systems that regulate mood, cognition, sleep, and eating.²² Studying estrogen in context of PMDD is important because women with PMDD can have low mood, specific food cravings, and impaired cognitive function.

Estrogen–serotonin interactions are thought to be involved in hormone-related mood disorders such as perimenopausal depression and PMDD.^23,24 However, the nature of their relationship is not yet fully understood. Ovariectomized animals have shown estrogen-induced changes related to serotonin metabolism, binding, and transmission in the regions of the brain involved in regulation of affect and cognition. Research in menopausal women also has provided some support for this interaction.²⁴

Positron emission tomography studies in humans have found increased cortical serotonin binding modulated by levels of estrogen, similar to those previously seen in rat studies.^24-27 One study showed an increased binding potential of serotonin in the cerebral cortex with estrogen treatment. This study further showed an even greater binding potential with estrogen plus progesterone, signaling a synergistic effect of the 2 hormones.²⁸

SSRIs are an effective treatment for the irritability, anxiety, and mood swings of PMDD.^29-30 Although the exact mechanism of action is unknown, the serotonergic properties are certainly of primary attention. For some PMDD patients, SSRIs work within hours to days, as opposed to days or weeks for patients with depression or anxiety, which suggests a separate or co-occurring mechanism of action is in place. In a double-blind, placebo-controlled crossover study, researchers administered the serotonin receptor antagonist metergoline to women with PMDD whose symptoms had remitted during treatment with fluoxetine and a group of healthy controls who were not receiving any medication.³¹ The women with PMDD experienced a return of symptoms 24 hours after treatment with metergoline but not with placebo; the controls experienced no mood changes.³¹

BDNF is a neurotransmitter linked to estrogen and likely related to PMDD. BDNF is critical for neurogenesis and is expressed in brain regions involved in learning and memory and also affects regulation.³² BDNF levels are increased by serotonergic antidepressants, affected by estradiol, and have cyclicity throughout the menstrual cycle.^33-35

Putative brain structural and functional differences. Imaging studies have suggested differences in brain structure in women with PMDD, with a focus on the amygdala and the prefrontal cortex. Women with PMDD have greater gray matter volume in the posterior cerebellum,³⁶ greater gray matter density of hippocampal cortex, and lower gray matter density in the parahippocampal cortex.³⁷

Some studies have shown a functional variability of the amygdala’s response to stress in women with PMDD vs healthy controls.^38,39 A proton magnetic resonance spectroscopy (1H-MRS) study of the displays the possibility of an altered GABAergic function in patients with PMDD.⁴⁰

Patients will PMDD have enhanced dorsolateral prefrontal cortex reactivity when anticipating negative stimuli (but not to the actual exposure) during the luteal phase. A positive correlation between this reactivity and progesterone levels also was observed.⁴¹ Some researchers have suggested that prefrontal cortex dysfunction may be a risk factor for PMDD.⁴²

HPA axis and HPG axis: Trauma, resiliency, inflammation. Altered cortisol levels (higher during the luteal phase⁴³ and lower during times of stress^14,44) suggest a possibly altered HPA axis in some women with PMDD. However, studies on this topic have been few and inconsistent.

Dysregulation of the HPG axis could cause vasomotor symptoms, sleep dysregulation, and mood symptoms during menopause; women with PMDD can also experience these symptoms. The influence of estrogen and progesterone on mood is also highly dependent on this axis.

Ultimately, the interplay between the HPA axis and the HPG axis is important. One study found that women with PMDD who had high serum ALLO levels (HPG-related) had blunted nocturnal cortisol levels (HPA-related) compared with healthy controls who had low ALLO levels.⁴⁵

Significant stress and trauma exposure have been associated with PMDD. A study of 3,968 women found a history of trauma and PTSD were independently associated with PMDD.⁴⁶ Another study of approximately 3,000 women found a strong correlation between abuse and PMS.⁴⁷ However, a third study found no correlations between PMDD and trauma.⁴⁸

Patients with a predisposition to PMDD may be more vulnerable to develop a posttraumatic stress-related disorder, perhaps due to decreased biologic resiliency. For example, the startle response (hyper­vigilance) has been shown to be different in women with PMDD. One study suggested that suboptimal production of premenstrual ALLO may lead to increased arousal and increased stress reactivity to psycho­social or environmental triggers.⁴⁹

The possible role of inflammation in PMDD deserves further investigation. The luteal phase entails an increase in the production of proinflammatory markers.^50,51 A 10-fold increase in progesterone is correlated with a 20% to 23% increase in C-reactive protein levels.^52,53 Women with inflammatory diseases (eg, gingivitis or irritable bowel syndrome) show worsening of symptoms prior to menstruation.^54-56 One study found increased levels of proinflammatory markers in women with PMDD compared with controls.⁵⁷

Because PMDD is heritable, it must have an underlying neurobiologic pathophysiology. Brain imaging studies show differences in structure and function in women with PMDD across the menstrual cycle. Conversion of progesterone to ALLO and the GABAergic influence of this metabolite is a topic of interest in current research. Similarly, the role of estrogen and its connection to serotonin and other neurotransmitters such as BDNF have been implicated.

The link between a history of stress, trauma, and PMDD raises the question of biologic resiliency and illness in these patients, as it connects to the HPA and HPG axis and production of inflammatory stress hormones and steroid hormones and their metabolites. PMDD can be conceptualized as variable sensitivity to hormonal response to stress,⁵⁸ thus contextualizing biochemical and psychological resiliency.

Further research is needed to clarify the possibility of a shared pathophysiology between endocrine-related mood disorders such as postpartum depression (PPD) and PMDD because current research is controversial.^59,60 In PPD, women who are exposed to high levels of progesterone and estrogen during pregnancy (just like in the mid-luteal phase) have a sudden drop in these hormones postpartum.

The ‘withdrawal theory.’ The affective symptoms of PMDD resolve almost instantaneously after the start of menstruation. Perhaps this type of immediate relief is akin to substance use disorders and symptoms of withdrawal. It could be that reinstatement of a certain amount of gonadal steroids in the follicular phase of the cycle diminishes a withdrawal-like response to these steroids.

Currently, the main leading theory is that PMDD is a result of “an abnormal response to normal hormonal changes.”⁶¹ A new study also has shown that the change in estradiol/progesterone levels (vs the steady state) was associated with PMDD symptoms.⁶² Thinking of PMDD as a disorder of withdrawal offers an alternative (yet complementary) perspective to the current theory: PMDD may be caused by the absence or diminishing of the above-named hormones and their metabolites in the late luteal phase (in the context of developed “tolerance” during the early- to mid-luteal phase).

Considering the interplay between neuro­transmitters and neurosteroids, both a “serotonin withdrawal theory” (caused by a drop in steroid hormones) and a “GABAergic withdrawal theory” (due to the decline in progesterone) could be proposed. This theory would be supported by the fact that SSRIs seem to mitigate symptoms of PMDD as well as the genetic association between PMDD and ESR1. It is more than likely that the “withdrawal” is caused by the interactions between estrogen-serotonin, progesterone-ALLO, and GABA receptors, and the complementary fashion in which progesterone and estrogen influence each other.

In an age when psychiatry strives to identify the biologic causes of disease, studying endocrine-related mood disorders is particularly intriguing. DSM-5 defines premenstrual dysphoric disorder (PMDD) as a depressive disorder, with a 12-month prevalence ranging from 1.8% to 5.8% among women who menstruate.^1-3 Factors that differentiate PMDD from other affective disorders include etiology, duration, and temporal relationship with the menstrual cycle.

PMDD is a disorder of consistent yet intermittent change in mental health and functionality. Therefore, it may be underdiagnosed and consequently undertreated if a psychiatric evaluation does not coincide with symptom occurrence or if patients do not understand that intermittent symptoms are treatable.

This article summarizes what is known about the etiology of PMDD. Although there are several treatments for PMDD, many women experience adverse effects or incomplete effectiveness. Further understanding of this disorder may lead to more efficacious treatments. Additionally, understanding the pathophysiology of PMDD might shed a light on the etiology of other disorders that are temporally related to reproductive life changes, such as pregnancy-, postpartum-, or menopause-related affective dysregulation.

Making the diagnosis

The diagnosis of PMDD is made when a patient has at least 5 of 11 specific symptoms that occur during the week before onset of menses, improve within a few days after the onset of menses (shown as the “PMDD Hazard Zone” in Figure 1), and are minimal or absent post-menses.³ Symptoms should be tracked prospectively for at least 2 menstrual cycles in order to confirm the diagnosis (one must be an affective symptom and another must be a behavioral/cognitive symptom).³

The affective symptoms are:

• lability of affect (eg, sudden sadness, tearfulness, or sensitivity to rejection)
• irritability, anger, or increased inter­personal conflicts
• depressed mood, hopelessness, or self- deprecating thoughts
• anxiety or tension, feeling “keyed up” or “on edge.”

The behavioral/cognitive symptoms are:

• decreased interest in usual activities (eg, work, hobbies, friends, school)
• difficulty concentrating
• lethargy, low energy, easy fatigability
• change in appetite, overeating, food cravings
• hypersomnia or insomnia
• feeling overwhelmed or out of control
• physical symptoms (breast tenderness or swelling, headache, joint or muscle pain, bloating, weight gain).

Ruling out premenstrual exacerbation (PME). Perhaps the most common cause for misdiagnosis of PMDD is failing to rule out PME of another underlying or comorbid condition (Figure 2). In many women who have a primary mood or anxiety disorder, the late luteal phase is a vulnerable time. A patient might be coping with untreated anxiety, for example, but the symptoms become unbearable the week before menstruation begins, which is likely when she seeks help. At this stage, a diagnosis of PMDD should be provisional at best. Often, PME is treated by treating the underlying condition. Therefore, a full diagnostic psychiatric interview is important to first rule out other underlying psychiatric disorders. PMDD is diagnosed if the premenstrual symptoms persist for 2 consecutive months after treating the suspected mood or anxiety disorder. Patients can use one of many PMDD daily symptom charts available online. Alternatively, they can use a cycle-tracking mobile phone application to correlate their symptoms with their cycle and share this information with their providers.

Consider these 5 interwoven pieces

The many variables that contribute to the pathophysiology of PMDD overlap and should be considered connecting pieces in the puzzle that is the etiology of this disorder (Figure 3). In reviewing the literature, we have identified 5 topics likely to be major contributors to this disorder:

1. genetic susceptibility
2. progesterone and allopregnanolone (ALLO)
3. estrogen, serotonin, and brain-derived neurotrophic factor (BDNF)
4. putative brain structural and functional differences
5. further involvement of the hypothalamic–pituitary–adrenal (HPA) axis and hypothalamic–pituitary–gonadal (HPG) axis: trauma, resiliency, and inflammation.

Genetic susceptibility. PMDD is thought to have a heritability range between 30% to 80%.³ This is demonstrated by family and twin studies^4-7 and specific genetic studies.⁸ The involvement of genetics means an underlying neurobiologic pathophysiology is in place.

Estrogen receptor alpha (ESR1) gene. Huo et al⁸ found an associated variation in ESR1 in women with PMDD compared with controls. They speculated that because ESR1 is important for arousal, if dysfunctional, this gene could be implicated in somatic as well as affective and cognitive deficits in PMDD patients. In another study, investigators reported a relationship between PMDD and heritable personality traits, as well as a link between these traits and ESR1 polymorphic variants.¹ They suggested that personality traits (independent of affective state) might be used to distinguish patients with PMDD from controls.¹

Studies on serotonin gene polymorphism and serotonin transporter genotype. Although a study of serotonin gene polymorphism did not find an association between serotonin1A gene polymorphism and PMDD, it did show that the presence of at least 1 C allele was associated with a 2.5-fold increased risk of PMDD.⁹ Another study did not find an association between the serotonin transporter genotype 5-HTTLPR and PMDD.¹⁰ However, it showed lower frontocingulate cortex activation during the luteal phase of PMDD patients compared with controls, suggesting that PMDD is linked to impaired frontocingulate cortex activation induced by emotions during the luteal phase.¹⁰ 

Seasonal affective disorder (SAD) and PMDD have shared clinical features. A polymorphism in the serotonin transporter promoter gene 5-HTTLPR has been associated with SAD. One study found that patients with comorbid SAD and PMDD are genetically more vulnerable to comorbid affective disorders compared with patients who have SAD only.¹¹

Progesterone and ALLO. Chronic exposure to progesterone and ALLO (a main progesterone metabolite) and rapid withdrawal from ovarian hormones may play a role in the etiology of PMDD. Much like alcohol or benzodiazepines, ALLO is a potent positive allosteric modulator of GABA_A receptors and has sedative, anesthetic, and anxiolytic properties. In times of acute stress, increased ALLO is known to provide relief.^12,13 However, in women with PMDD, this typical ALLO increase might not occur.¹⁴

Patients with PMDD have been reported to have decreased levels of ALLO in the luteal phase.^15-17 In one study, women with highly symptomatic PMDD had lower levels of ALLO compared with women with less symptomatic PMDD.¹⁴ A gonadotropin-releasing hormone challenge study showed the increase in ALLO response was less in PMDD patients compared with controls.¹⁷ Luteal-phase ALLO concentrations are reported to be lower in women with premenstrual syndrome (PMS), a milder form of PMDD.^14,17

The efficacy of selective serotonin reuptake inhibitors (SSRIs) for treating PMDD could be the result of the interaction of these medications with neuroactive steroids,¹⁸ possibly because SSRIs enhance the sensitivity of GABA_A receptors or promote the formation of more ALLO (Figure 4).^19-21

Estrogen, serotonin, and BDNF. Estrogen affects multiple neurotransmitter systems that regulate mood, cognition, sleep, and eating.²² Studying estrogen in context of PMDD is important because women with PMDD can have low mood, specific food cravings, and impaired cognitive function.

Estrogen–serotonin interactions are thought to be involved in hormone-related mood disorders such as perimenopausal depression and PMDD.^23,24 However, the nature of their relationship is not yet fully understood. Ovariectomized animals have shown estrogen-induced changes related to serotonin metabolism, binding, and transmission in the regions of the brain involved in regulation of affect and cognition. Research in menopausal women also has provided some support for this interaction.²⁴

Positron emission tomography studies in humans have found increased cortical serotonin binding modulated by levels of estrogen, similar to those previously seen in rat studies.^24-27 One study showed an increased binding potential of serotonin in the cerebral cortex with estrogen treatment. This study further showed an even greater binding potential with estrogen plus progesterone, signaling a synergistic effect of the 2 hormones.²⁸

SSRIs are an effective treatment for the irritability, anxiety, and mood swings of PMDD.^29-30 Although the exact mechanism of action is unknown, the serotonergic properties are certainly of primary attention. For some PMDD patients, SSRIs work within hours to days, as opposed to days or weeks for patients with depression or anxiety, which suggests a separate or co-occurring mechanism of action is in place. In a double-blind, placebo-controlled crossover study, researchers administered the serotonin receptor antagonist metergoline to women with PMDD whose symptoms had remitted during treatment with fluoxetine and a group of healthy controls who were not receiving any medication.³¹ The women with PMDD experienced a return of symptoms 24 hours after treatment with metergoline but not with placebo; the controls experienced no mood changes.³¹

BDNF is a neurotransmitter linked to estrogen and likely related to PMDD. BDNF is critical for neurogenesis and is expressed in brain regions involved in learning and memory and also affects regulation.³² BDNF levels are increased by serotonergic antidepressants, affected by estradiol, and have cyclicity throughout the menstrual cycle.^33-35

Putative brain structural and functional differences. Imaging studies have suggested differences in brain structure in women with PMDD, with a focus on the amygdala and the prefrontal cortex. Women with PMDD have greater gray matter volume in the posterior cerebellum,³⁶ greater gray matter density of hippocampal cortex, and lower gray matter density in the parahippocampal cortex.³⁷

Some studies have shown a functional variability of the amygdala’s response to stress in women with PMDD vs healthy controls.^38,39 A proton magnetic resonance spectroscopy (1H-MRS) study of the displays the possibility of an altered GABAergic function in patients with PMDD.⁴⁰

Patients will PMDD have enhanced dorsolateral prefrontal cortex reactivity when anticipating negative stimuli (but not to the actual exposure) during the luteal phase. A positive correlation between this reactivity and progesterone levels also was observed.⁴¹ Some researchers have suggested that prefrontal cortex dysfunction may be a risk factor for PMDD.⁴²

HPA axis and HPG axis: Trauma, resiliency, inflammation. Altered cortisol levels (higher during the luteal phase⁴³ and lower during times of stress^14,44) suggest a possibly altered HPA axis in some women with PMDD. However, studies on this topic have been few and inconsistent.

Dysregulation of the HPG axis could cause vasomotor symptoms, sleep dysregulation, and mood symptoms during menopause; women with PMDD can also experience these symptoms. The influence of estrogen and progesterone on mood is also highly dependent on this axis.

Ultimately, the interplay between the HPA axis and the HPG axis is important. One study found that women with PMDD who had high serum ALLO levels (HPG-related) had blunted nocturnal cortisol levels (HPA-related) compared with healthy controls who had low ALLO levels.⁴⁵

Significant stress and trauma exposure have been associated with PMDD. A study of 3,968 women found a history of trauma and PTSD were independently associated with PMDD.⁴⁶ Another study of approximately 3,000 women found a strong correlation between abuse and PMS.⁴⁷ However, a third study found no correlations between PMDD and trauma.⁴⁸

Patients with a predisposition to PMDD may be more vulnerable to develop a posttraumatic stress-related disorder, perhaps due to decreased biologic resiliency. For example, the startle response (hyper­vigilance) has been shown to be different in women with PMDD. One study suggested that suboptimal production of premenstrual ALLO may lead to increased arousal and increased stress reactivity to psycho­social or environmental triggers.⁴⁹

The possible role of inflammation in PMDD deserves further investigation. The luteal phase entails an increase in the production of proinflammatory markers.^50,51 A 10-fold increase in progesterone is correlated with a 20% to 23% increase in C-reactive protein levels.^52,53 Women with inflammatory diseases (eg, gingivitis or irritable bowel syndrome) show worsening of symptoms prior to menstruation.^54-56 One study found increased levels of proinflammatory markers in women with PMDD compared with controls.⁵⁷

Because PMDD is heritable, it must have an underlying neurobiologic pathophysiology. Brain imaging studies show differences in structure and function in women with PMDD across the menstrual cycle. Conversion of progesterone to ALLO and the GABAergic influence of this metabolite is a topic of interest in current research. Similarly, the role of estrogen and its connection to serotonin and other neurotransmitters such as BDNF have been implicated.

The link between a history of stress, trauma, and PMDD raises the question of biologic resiliency and illness in these patients, as it connects to the HPA and HPG axis and production of inflammatory stress hormones and steroid hormones and their metabolites. PMDD can be conceptualized as variable sensitivity to hormonal response to stress,⁵⁸ thus contextualizing biochemical and psychological resiliency.

Further research is needed to clarify the possibility of a shared pathophysiology between endocrine-related mood disorders such as postpartum depression (PPD) and PMDD because current research is controversial.^59,60 In PPD, women who are exposed to high levels of progesterone and estrogen during pregnancy (just like in the mid-luteal phase) have a sudden drop in these hormones postpartum.

The ‘withdrawal theory.’ The affective symptoms of PMDD resolve almost instantaneously after the start of menstruation. Perhaps this type of immediate relief is akin to substance use disorders and symptoms of withdrawal. It could be that reinstatement of a certain amount of gonadal steroids in the follicular phase of the cycle diminishes a withdrawal-like response to these steroids.

Currently, the main leading theory is that PMDD is a result of “an abnormal response to normal hormonal changes.”⁶¹ A new study also has shown that the change in estradiol/progesterone levels (vs the steady state) was associated with PMDD symptoms.⁶² Thinking of PMDD as a disorder of withdrawal offers an alternative (yet complementary) perspective to the current theory: PMDD may be caused by the absence or diminishing of the above-named hormones and their metabolites in the late luteal phase (in the context of developed “tolerance” during the early- to mid-luteal phase).

Considering the interplay between neuro­transmitters and neurosteroids, both a “serotonin withdrawal theory” (caused by a drop in steroid hormones) and a “GABAergic withdrawal theory” (due to the decline in progesterone) could be proposed. This theory would be supported by the fact that SSRIs seem to mitigate symptoms of PMDD as well as the genetic association between PMDD and ESR1. It is more than likely that the “withdrawal” is caused by the interactions between estrogen-serotonin, progesterone-ALLO, and GABA receptors, and the complementary fashion in which progesterone and estrogen influence each other.

Like all physicians, psychiatrists practice in an increasingly complex health care environment, with escalating demands for productivity, rising threats of malpractice, expanding clinical oversight, and growing concerns about income. Additionally, psychiatric practice presents its own challenges, including limited resources and concerns about patient violence and suicide. These concerns can make it difficult to establish a healthy work–life balance.

Physicians, including psychiatrists, are at risk for alcohol or substance abuse/dependency, burnout, and suicide. As psychiatrists, we need to attend to our own personal and professional health so that we can best help our patients. This review focuses on the challenges psychiatrists face that can adversely affect their well-being and offers strategies to reduce the risk of burnout and enhance wellness.

The challenges of medicine and their impact on psychiatrists

The practice of medicine is inherently challenging. It requires hard work, discipline, dedication, and faithfulness to high ethical standards. Additional challenges include declining autonomy and opportunities for social support, increasing accountability, and a growing interest in reducing the cost of care by employing more non-physician health professionals—which in psychiatry typically include psychologists, nurse practitioners, and social workers. The uncertainty of the Affordable Care Act, declining income, and concerns about the nature of future medical practice are also stressors.^1,2

Factors that contribute to psychiatrists’ stress include:

• limited resources
• concerns about patient violence and suicide
• crowded inpatient units
• changing culture in mental health services
• high work demands
• poorly defined roles of consultants
• declining authority
• frustration with the inability to impact systemic change
• conflict between responsibility toward employers vs the patient
• isolation.³

Concern about patient suicide is a significant stressor.^4,5 Some evidence suggests that the impact of a patient’s suicide on a physician is more severe when it occurs during training than after graduation and is inversely correlated with the clinician’s perceived social integration into their professional network.⁵

Impediments to a physician’s well-being

Alcohol abuse/dependence. Approximately 13% of male physicians and 21% of female physicians meet Alcohol Use Disorders Identification Test Version C criteria for alcohol abuse or dependence, according to a study of approximately 7,300 U.S. physicians from all specialties.⁶ (In this study, prescription drug abuse and use of illicit drugs were rare.) Age, hours worked, male sex, being married or partnered, having children, and being in a specialty other than internal medicine were independently associated with alcohol abuse or dependence.

Fortunately, psychiatrists were among the specialties with below average likelihood to meet diagnostic criteria for alcohol abuse/dependency.⁶ However, alcohol abuse or dependency was associated with burnout, depression, suicidal ideation, lower quality of life, lower career satisfaction, and medical errors.

Burnout is a long-term stress reaction consisting of:

• physical and emotional exhaustion (feeling depleted)
• depersonalization (cynicism, lack of engagement with or negative attitudes toward patients)
• reduced sense of personal accomplishment (lack of a sense of purpose).⁷

In a 2017 survey of >14,000 U.S. physicians from 27 specialties, 42% of psychiatrists reported burnout.⁸ In another survey of approximately 300 resident physicians across all specialties in a tertiary academic hospital, 69% met criteria for burnout.⁹ This condition affects resident physicians as well as those in practice. Residents and program directors cited a lack of work–life balance and feeling unappreciated as factors contributing to burnout.

Among physicians, factors that contribute to burnout include loss of autonomy, diminished status as physicians, and increased work pressures. Burnout has a negative impact on both patients and health care systems. It is associated with an increased risk of depression and can contribute to:

• broken relationships
• alcohol abuse
• physician suicide
• decreased quality of care, including patient safety and satisfaction
• increased risk of malpractice suits
• reduced patient adherence to medical recommendations.^5,10-12

Physicians who embrace medicine as a calling (ie, committing one’s life to personally meaningful work that serves a prosocial purpose) experience less burnout. According to a survey of approximately 900 primary care physicians and 300 psychiatrists, 42% of psychiatrists strongly agreed that medicine is a calling.¹³ Overall, physicians with a high sense of calling reported less burnout than those with a lower sense of calling (17% vs 31%, respectively).¹³

Depression and suicide. Gold et al¹² analyzed a database that included information on approximately 31,600 adult suicide victims, and 203 of these victims were physicians. Compared with others, physicians were more likely to have a diagnosed mental illness or an occupation-related problem that contributed to suicide. Toxicology results also showed that physician suicide victims were significantly more likely than non-physician victims to test positive for benzodiazepines and barbiturates, but not antidepressants, which suggests that physicians with depression may not have been receiving adequate treatment.¹²

Although occupation-related stress and inadequate mental health treatment may be modifiable risk factors to reduce suicide deaths among physicians, stigma and fear of medical staff and licensure issues may deter physicians from seeking treatment.¹⁴

Steps to avoid burnout

Evidence-based interventions. There is limited evidence-based data regarding specific interventions for preventing burnout and reducing stress among physicians, particularly among psychiatrists.⁴

A randomized controlled trial of 74 practicing physicians at the Mayo Clinic in Rochester, Minnesota, evaluated the effectiveness of 19 biweekly physician-facilitated discussion groups.¹⁵ The groups covered topics such as elements of mindfulness, reflection, shared experience, and small-group learning. The institution provided 1 hour of paid time every other week for physicians to participate in this program. Physicians in the control group could schedule and use this time as they chose. Researchers also collected data on 350 non-trial participants.

The proportion of participants who strongly agreed that their work was meaningful increased 6.3% in the intervention group but decreased 6.3% in the control group and 13.4% among non-trial participants (P = .04).¹⁵ Rates of depersonalization, emotional exhaustion, and overall burnout decreased substantially in the intervention group, decreased slightly in the control group, and increased in the non-trial cohort. Results were sustained at 12 months after the study. There were no statistically significant differences in stress, symptoms of depression, overall quality of life, or job satisfaction.¹⁵

Preliminary evidence suggests that residents and fellows would find a wellness or suicide prevention program helpful. One study found that the use of one such program, which provided individual counseling, psychiatric evaluation, and wellness workshops for residents, fellows, and faculty in an academic health center, increased from 5% to 25% of eligible participants, and participants reported high levels of satisfaction with the program.¹⁶ Such programs would require institutional support for space and clinical staff.¹⁵

Empathy. As psychiatrists, we are taught to be empathetic. Yet, with the numerous challenges we face, it is not always easy. Stressors such as an increased workload or burnout can adversely affect a psychiatrist’s ability to provide empathetic care.¹⁷ However, empathetic treatment has clear benefits for both physicians and patients. Empathic skills can lead to more professional satisfaction and outcomes, which are important components of accountability, and can:

• promote patient satisfaction
• establish trust
• reduce anxiety
• increase adherence to treatment regimens
• improve health outcomes
• decrease the likelihood of malpractice suits.¹⁷

Mindfulness is a “flexible state of mind in which we are actively engaged in the present, noticing new things and sensitive to context.”^18,19 It may sound mundane to cling to phrases such as “living in the present,” but mindfulness can be a valuable tool for psychiatrists who struggle to maintain well-being in medicine’s challenging milieu. The process of mindfulness—actively drawing distinctions and noticing new things, “seeing the familiar in the novel and the novel in the familiar”—can ensure that we have active minds, that we are involved, and that we are capturing the joy of living in the stimulating present.¹⁸

Focus on issues you can control

Many of the factors that negatively influence professional satisfaction and well-being, such as loss of autonomy, demand for increased patient care volume, and increasing scrutiny on the quality of care, are beyond a psychiatrist’s control. Medical administrators can help reduce some of these issues by increasing physician autonomy, offering physicians the opportunity to work part-time, offering medical staff workshops to enhance positive communication, or addressing leader­ship problems. However, psychiatrists may benefit most by identifying modifiable issues under their own control, such as prioritizing a work–life balance, applying the fundamentals of a health prevention strategy to their own lives (Box^20,21), approaching medicine as a calling, embracing an empathetic approach to patient care, and bringing mindfulness to medical practice.

Like all physicians, psychiatrists practice in an increasingly complex health care environment, with escalating demands for productivity, rising threats of malpractice, expanding clinical oversight, and growing concerns about income. Additionally, psychiatric practice presents its own challenges, including limited resources and concerns about patient violence and suicide. These concerns can make it difficult to establish a healthy work–life balance.

Physicians, including psychiatrists, are at risk for alcohol or substance abuse/dependency, burnout, and suicide. As psychiatrists, we need to attend to our own personal and professional health so that we can best help our patients. This review focuses on the challenges psychiatrists face that can adversely affect their well-being and offers strategies to reduce the risk of burnout and enhance wellness.

The challenges of medicine and their impact on psychiatrists

The practice of medicine is inherently challenging. It requires hard work, discipline, dedication, and faithfulness to high ethical standards. Additional challenges include declining autonomy and opportunities for social support, increasing accountability, and a growing interest in reducing the cost of care by employing more non-physician health professionals—which in psychiatry typically include psychologists, nurse practitioners, and social workers. The uncertainty of the Affordable Care Act, declining income, and concerns about the nature of future medical practice are also stressors.^1,2

Factors that contribute to psychiatrists’ stress include:

• limited resources
• concerns about patient violence and suicide
• crowded inpatient units
• changing culture in mental health services
• high work demands
• poorly defined roles of consultants
• declining authority
• frustration with the inability to impact systemic change
• conflict between responsibility toward employers vs the patient
• isolation.³

Concern about patient suicide is a significant stressor.^4,5 Some evidence suggests that the impact of a patient’s suicide on a physician is more severe when it occurs during training than after graduation and is inversely correlated with the clinician’s perceived social integration into their professional network.⁵

Impediments to a physician’s well-being

Alcohol abuse/dependence. Approximately 13% of male physicians and 21% of female physicians meet Alcohol Use Disorders Identification Test Version C criteria for alcohol abuse or dependence, according to a study of approximately 7,300 U.S. physicians from all specialties.⁶ (In this study, prescription drug abuse and use of illicit drugs were rare.) Age, hours worked, male sex, being married or partnered, having children, and being in a specialty other than internal medicine were independently associated with alcohol abuse or dependence.

Fortunately, psychiatrists were among the specialties with below average likelihood to meet diagnostic criteria for alcohol abuse/dependency.⁶ However, alcohol abuse or dependency was associated with burnout, depression, suicidal ideation, lower quality of life, lower career satisfaction, and medical errors.

Burnout is a long-term stress reaction consisting of:

• physical and emotional exhaustion (feeling depleted)
• depersonalization (cynicism, lack of engagement with or negative attitudes toward patients)
• reduced sense of personal accomplishment (lack of a sense of purpose).⁷

In a 2017 survey of >14,000 U.S. physicians from 27 specialties, 42% of psychiatrists reported burnout.⁸ In another survey of approximately 300 resident physicians across all specialties in a tertiary academic hospital, 69% met criteria for burnout.⁹ This condition affects resident physicians as well as those in practice. Residents and program directors cited a lack of work–life balance and feeling unappreciated as factors contributing to burnout.

Among physicians, factors that contribute to burnout include loss of autonomy, diminished status as physicians, and increased work pressures. Burnout has a negative impact on both patients and health care systems. It is associated with an increased risk of depression and can contribute to:

• broken relationships
• alcohol abuse
• physician suicide
• decreased quality of care, including patient safety and satisfaction
• increased risk of malpractice suits
• reduced patient adherence to medical recommendations.^5,10-12

Physicians who embrace medicine as a calling (ie, committing one’s life to personally meaningful work that serves a prosocial purpose) experience less burnout. According to a survey of approximately 900 primary care physicians and 300 psychiatrists, 42% of psychiatrists strongly agreed that medicine is a calling.¹³ Overall, physicians with a high sense of calling reported less burnout than those with a lower sense of calling (17% vs 31%, respectively).¹³

Depression and suicide. Gold et al¹² analyzed a database that included information on approximately 31,600 adult suicide victims, and 203 of these victims were physicians. Compared with others, physicians were more likely to have a diagnosed mental illness or an occupation-related problem that contributed to suicide. Toxicology results also showed that physician suicide victims were significantly more likely than non-physician victims to test positive for benzodiazepines and barbiturates, but not antidepressants, which suggests that physicians with depression may not have been receiving adequate treatment.¹²

Although occupation-related stress and inadequate mental health treatment may be modifiable risk factors to reduce suicide deaths among physicians, stigma and fear of medical staff and licensure issues may deter physicians from seeking treatment.¹⁴

Steps to avoid burnout

Evidence-based interventions. There is limited evidence-based data regarding specific interventions for preventing burnout and reducing stress among physicians, particularly among psychiatrists.⁴

A randomized controlled trial of 74 practicing physicians at the Mayo Clinic in Rochester, Minnesota, evaluated the effectiveness of 19 biweekly physician-facilitated discussion groups.¹⁵ The groups covered topics such as elements of mindfulness, reflection, shared experience, and small-group learning. The institution provided 1 hour of paid time every other week for physicians to participate in this program. Physicians in the control group could schedule and use this time as they chose. Researchers also collected data on 350 non-trial participants.

The proportion of participants who strongly agreed that their work was meaningful increased 6.3% in the intervention group but decreased 6.3% in the control group and 13.4% among non-trial participants (P = .04).¹⁵ Rates of depersonalization, emotional exhaustion, and overall burnout decreased substantially in the intervention group, decreased slightly in the control group, and increased in the non-trial cohort. Results were sustained at 12 months after the study. There were no statistically significant differences in stress, symptoms of depression, overall quality of life, or job satisfaction.¹⁵

Preliminary evidence suggests that residents and fellows would find a wellness or suicide prevention program helpful. One study found that the use of one such program, which provided individual counseling, psychiatric evaluation, and wellness workshops for residents, fellows, and faculty in an academic health center, increased from 5% to 25% of eligible participants, and participants reported high levels of satisfaction with the program.¹⁶ Such programs would require institutional support for space and clinical staff.¹⁵

Empathy. As psychiatrists, we are taught to be empathetic. Yet, with the numerous challenges we face, it is not always easy. Stressors such as an increased workload or burnout can adversely affect a psychiatrist’s ability to provide empathetic care.¹⁷ However, empathetic treatment has clear benefits for both physicians and patients. Empathic skills can lead to more professional satisfaction and outcomes, which are important components of accountability, and can:

• promote patient satisfaction
• establish trust
• reduce anxiety
• increase adherence to treatment regimens
• improve health outcomes
• decrease the likelihood of malpractice suits.¹⁷

Mindfulness is a “flexible state of mind in which we are actively engaged in the present, noticing new things and sensitive to context.”^18,19 It may sound mundane to cling to phrases such as “living in the present,” but mindfulness can be a valuable tool for psychiatrists who struggle to maintain well-being in medicine’s challenging milieu. The process of mindfulness—actively drawing distinctions and noticing new things, “seeing the familiar in the novel and the novel in the familiar”—can ensure that we have active minds, that we are involved, and that we are capturing the joy of living in the stimulating present.¹⁸

Focus on issues you can control

Many of the factors that negatively influence professional satisfaction and well-being, such as loss of autonomy, demand for increased patient care volume, and increasing scrutiny on the quality of care, are beyond a psychiatrist’s control. Medical administrators can help reduce some of these issues by increasing physician autonomy, offering physicians the opportunity to work part-time, offering medical staff workshops to enhance positive communication, or addressing leader­ship problems. However, psychiatrists may benefit most by identifying modifiable issues under their own control, such as prioritizing a work–life balance, applying the fundamentals of a health prevention strategy to their own lives (Box^20,21), approaching medicine as a calling, embracing an empathetic approach to patient care, and bringing mindfulness to medical practice.

Like all physicians, psychiatrists practice in an increasingly complex health care environment, with escalating demands for productivity, rising threats of malpractice, expanding clinical oversight, and growing concerns about income. Additionally, psychiatric practice presents its own challenges, including limited resources and concerns about patient violence and suicide. These concerns can make it difficult to establish a healthy work–life balance.

Physicians, including psychiatrists, are at risk for alcohol or substance abuse/dependency, burnout, and suicide. As psychiatrists, we need to attend to our own personal and professional health so that we can best help our patients. This review focuses on the challenges psychiatrists face that can adversely affect their well-being and offers strategies to reduce the risk of burnout and enhance wellness.

The challenges of medicine and their impact on psychiatrists

The practice of medicine is inherently challenging. It requires hard work, discipline, dedication, and faithfulness to high ethical standards. Additional challenges include declining autonomy and opportunities for social support, increasing accountability, and a growing interest in reducing the cost of care by employing more non-physician health professionals—which in psychiatry typically include psychologists, nurse practitioners, and social workers. The uncertainty of the Affordable Care Act, declining income, and concerns about the nature of future medical practice are also stressors.^1,2

Factors that contribute to psychiatrists’ stress include:

• limited resources
• concerns about patient violence and suicide
• crowded inpatient units
• changing culture in mental health services
• high work demands
• poorly defined roles of consultants
• declining authority
• frustration with the inability to impact systemic change
• conflict between responsibility toward employers vs the patient
• isolation.³

Concern about patient suicide is a significant stressor.^4,5 Some evidence suggests that the impact of a patient’s suicide on a physician is more severe when it occurs during training than after graduation and is inversely correlated with the clinician’s perceived social integration into their professional network.⁵

Impediments to a physician’s well-being

Alcohol abuse/dependence. Approximately 13% of male physicians and 21% of female physicians meet Alcohol Use Disorders Identification Test Version C criteria for alcohol abuse or dependence, according to a study of approximately 7,300 U.S. physicians from all specialties.⁶ (In this study, prescription drug abuse and use of illicit drugs were rare.) Age, hours worked, male sex, being married or partnered, having children, and being in a specialty other than internal medicine were independently associated with alcohol abuse or dependence.

Fortunately, psychiatrists were among the specialties with below average likelihood to meet diagnostic criteria for alcohol abuse/dependency.⁶ However, alcohol abuse or dependency was associated with burnout, depression, suicidal ideation, lower quality of life, lower career satisfaction, and medical errors.

Burnout is a long-term stress reaction consisting of:

• physical and emotional exhaustion (feeling depleted)
• depersonalization (cynicism, lack of engagement with or negative attitudes toward patients)
• reduced sense of personal accomplishment (lack of a sense of purpose).⁷

In a 2017 survey of >14,000 U.S. physicians from 27 specialties, 42% of psychiatrists reported burnout.⁸ In another survey of approximately 300 resident physicians across all specialties in a tertiary academic hospital, 69% met criteria for burnout.⁹ This condition affects resident physicians as well as those in practice. Residents and program directors cited a lack of work–life balance and feeling unappreciated as factors contributing to burnout.

Among physicians, factors that contribute to burnout include loss of autonomy, diminished status as physicians, and increased work pressures. Burnout has a negative impact on both patients and health care systems. It is associated with an increased risk of depression and can contribute to:

• broken relationships
• alcohol abuse
• physician suicide
• decreased quality of care, including patient safety and satisfaction
• increased risk of malpractice suits
• reduced patient adherence to medical recommendations.^5,10-12

Physicians who embrace medicine as a calling (ie, committing one’s life to personally meaningful work that serves a prosocial purpose) experience less burnout. According to a survey of approximately 900 primary care physicians and 300 psychiatrists, 42% of psychiatrists strongly agreed that medicine is a calling.¹³ Overall, physicians with a high sense of calling reported less burnout than those with a lower sense of calling (17% vs 31%, respectively).¹³

Depression and suicide. Gold et al¹² analyzed a database that included information on approximately 31,600 adult suicide victims, and 203 of these victims were physicians. Compared with others, physicians were more likely to have a diagnosed mental illness or an occupation-related problem that contributed to suicide. Toxicology results also showed that physician suicide victims were significantly more likely than non-physician victims to test positive for benzodiazepines and barbiturates, but not antidepressants, which suggests that physicians with depression may not have been receiving adequate treatment.¹²

Although occupation-related stress and inadequate mental health treatment may be modifiable risk factors to reduce suicide deaths among physicians, stigma and fear of medical staff and licensure issues may deter physicians from seeking treatment.¹⁴

Steps to avoid burnout

Evidence-based interventions. There is limited evidence-based data regarding specific interventions for preventing burnout and reducing stress among physicians, particularly among psychiatrists.⁴

A randomized controlled trial of 74 practicing physicians at the Mayo Clinic in Rochester, Minnesota, evaluated the effectiveness of 19 biweekly physician-facilitated discussion groups.¹⁵ The groups covered topics such as elements of mindfulness, reflection, shared experience, and small-group learning. The institution provided 1 hour of paid time every other week for physicians to participate in this program. Physicians in the control group could schedule and use this time as they chose. Researchers also collected data on 350 non-trial participants.

The proportion of participants who strongly agreed that their work was meaningful increased 6.3% in the intervention group but decreased 6.3% in the control group and 13.4% among non-trial participants (P = .04).¹⁵ Rates of depersonalization, emotional exhaustion, and overall burnout decreased substantially in the intervention group, decreased slightly in the control group, and increased in the non-trial cohort. Results were sustained at 12 months after the study. There were no statistically significant differences in stress, symptoms of depression, overall quality of life, or job satisfaction.¹⁵

Preliminary evidence suggests that residents and fellows would find a wellness or suicide prevention program helpful. One study found that the use of one such program, which provided individual counseling, psychiatric evaluation, and wellness workshops for residents, fellows, and faculty in an academic health center, increased from 5% to 25% of eligible participants, and participants reported high levels of satisfaction with the program.¹⁶ Such programs would require institutional support for space and clinical staff.¹⁵

Empathy. As psychiatrists, we are taught to be empathetic. Yet, with the numerous challenges we face, it is not always easy. Stressors such as an increased workload or burnout can adversely affect a psychiatrist’s ability to provide empathetic care.¹⁷ However, empathetic treatment has clear benefits for both physicians and patients. Empathic skills can lead to more professional satisfaction and outcomes, which are important components of accountability, and can:

• promote patient satisfaction
• establish trust
• reduce anxiety
• increase adherence to treatment regimens
• improve health outcomes
• decrease the likelihood of malpractice suits.¹⁷

Mindfulness is a “flexible state of mind in which we are actively engaged in the present, noticing new things and sensitive to context.”^18,19 It may sound mundane to cling to phrases such as “living in the present,” but mindfulness can be a valuable tool for psychiatrists who struggle to maintain well-being in medicine’s challenging milieu. The process of mindfulness—actively drawing distinctions and noticing new things, “seeing the familiar in the novel and the novel in the familiar”—can ensure that we have active minds, that we are involved, and that we are capturing the joy of living in the stimulating present.¹⁸

Focus on issues you can control

Many of the factors that negatively influence professional satisfaction and well-being, such as loss of autonomy, demand for increased patient care volume, and increasing scrutiny on the quality of care, are beyond a psychiatrist’s control. Medical administrators can help reduce some of these issues by increasing physician autonomy, offering physicians the opportunity to work part-time, offering medical staff workshops to enhance positive communication, or addressing leader­ship problems. However, psychiatrists may benefit most by identifying modifiable issues under their own control, such as prioritizing a work–life balance, applying the fundamentals of a health prevention strategy to their own lives (Box^20,21), approaching medicine as a calling, embracing an empathetic approach to patient care, and bringing mindfulness to medical practice.

Opioid abuse and overdose are large and growing problems, and in recent years the numbers have been staggering. Overdose deaths related to opioids increased from 28,647 in 2014 to 33,091 in 2015 (Figure).¹ More than 2 million individuals in the United States had opioid use disorder in 2015,² and approximately 80% of them received no treatment,³ even though effective treatment could reduce the scope of abuse.^4,5

Although psychiatrists typically are not the primary prescribers of opioid medications, they often treat psychiatric disorders in patients with chronic pain who take prescription opioids. A recent study found that, despite representing only 16% of the adult population, adults with mental health disorders receive more than one-half of all opioid prescriptions distributed each year in the United States.⁶ Therefore, psychiatrists must be aware of risk assessment strategies for patients receiving opioids.

In this article, we provide recommendations for managing individuals with opioid use disorder, including:

• how to identify risk factors for opioid use disorder and use screening tools
• how to evaluate a patient with suspected opioid use disorder and make the diagnosis
• how to treat a patient with opioid use disorder, including a review of approved pharmaceutical agents.

Risk factors for opioid abuse and overdose

Patients with a history of mental health and/or substance use disorders or at least 3 months of prescribed opioid treatment are at risk for opioid abuse. Those taking a high daily dose of opioids or who have a history of overdose are at risk for overdose from opioid abuse (Table 1).^7-12 Standardized tools, such as the Opioid Risk Tool, can be used to screen to assess risk for opioid abuse among individuals prescribed opioids for treatment of chronic pain.¹² However, clinicians must be aware that even patients without characteristic risk factors can become dependent on opioids and/or be at risk for an accidental or intentional overdose. For example, opioid therapy following surgical procedures, even in patients who do not have a history of opioid use, increases risk of developing opioid use disorder.¹³

Evaluation and diagnosis

DSM-5 criteria define 3 degrees of opioid use disorder, depending on how many of the following traits a patient exhibits (mild, 2 to 3; moderate, 4 to 5; and severe, ≥6 )¹⁴:

• taking more than the initially intended quantities of opioids or for a longer period of time than intended
• continuous attempts to reduce or otherwise manage opioid use or desires to do so
• a great deal of time using, recovering from, or acquiring opioids
• reports of strong cravings to use opioids
• failing to meet personal objectives at home, work, or school
• continued opioid use even though it causes recurrent social problems
• reduction or elimination of activities the patient once considered important due to opioid use
• opioid use in situations where it is physically dangerous
• continued opioid use despite persistent psychological or physiologic problems despite knowing that continued use is causing or worsening those problems
• tolerance to opioids (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision)
• withdrawal or use of opioids (or related substances) to prevent withdrawal (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision).

Clinicians should be vigilant for symptoms of opioid use or withdrawal, such as needle marks and weight loss, during the interview (Table 2). High-risk populations that require regular screening include individuals with a history of opioid use disorder, patients taking chronic pain medication, and psychiatric patients.¹⁵ During the interview, clinicians should take an nonjudgmental approach and avoid “shame and blame.”

Standardized scales. Various standardized scales can be used to evaluate patients for opioid withdrawal and risk for substance use disorder. Scales for assessing opioid withdrawal include:

• Clinical Opiate Withdrawal Scale
• Subjective Opiate Withdrawal Scale.

Substance use disorder screening tools include:

• Drug Abuse Screen Test-10
• Alcohol Use Disorders Identification Test
• National Institute on Drug Abuse (NIDA) Drug Screening Tool.¹⁷

Examination findings. A brief physical examination is necessary to document key findings (Table 2). Patients should undergo a urine drug screen; gas chromatography/mass spectroscopy can confirm positive results. During the examination, clinicians should look for signs and symptoms of co-occurring substance use (eg, benzo­diazepines, marijuana, alcohol, cocaine) or mental disorders (mood, anxiety, attention-deficit).^18-21 Because nonprescription opioid use is associated with increased risk of suicide attempts and ideation,²² a suicide risk assessment is necessary.

Managing opioid use disorder

Detoxification is a 3-tiered approach that requires judicious prescription of medication, psychosocial support, and supervision to relieve opioid withdrawal symptoms. In both inpatient and outpatient settings, medications used for opioid detoxification include buprenorphine, clonidine, and methadone administered in doses tapered over 5 to 7 days. Appropriate detoxification increases treatment retention for continuing care.^23,24

Buprenorphine or buprenorphine/naloxone is the first-line option for outpatient and inpatient detoxification. Short-term detoxification schedules include starting doses between 4 and 16 mg/d, tapered over 5 to 7 days. Compared with methadone, buprenorphine has a lower risk of overdose²⁵ and abuse potential and can be given in an office-based setting. Clonidine, 0.3 to 1.2 mg/d in divided doses, is an alternative to buprenorphine and can be used in in­patient settings.²⁶

Clonidine is not as effective as buprenorphine for detoxification, but it may be used when buprenorphine is contraindicated. Clonidine may require adjuvant symptomatic treatment for insomnia (eg, trazodone, 100 mg at bedtime), anxiety (eg, hydroxyzine, 25 mg, twice a day), or diarrhea (loperamide, 2 mg/d). If a patient needs more structure and monitoring, he (she) should be referred for inpatient detoxification or to a methadone program.²⁷

Medication-assisted therapies

Detoxification alone often is not sufficient treatment. Medication-assisted therapy (MAT) is typically recommended by federal guidelines provided by the Substance Abuse and Mental Health Administration (SAMHSA) for patients with opioid use disorders.³ Patients can be directly transitioned from currently abused opioids to MAT on an outpatient basis. FDA-approved medications for MAT for opioid use disorder include buprenorphine, naltrexone (oral and long-acting injectable), and methadone (Table 3). Choice of MAT depends on several factors, including cost, patient preference, and availability of methadone programs and buprenorphine providers.²⁸

MAT should include psychosocial support^29-33 and active monitoring with urine drug screens. Maintenance therapy with medications is usually long-term and has been shown to have better outcomes than detoxification alone or short-term treatment.³⁴ Relapse during MAT should not be cause to discontinue treatment; instead, the patient should be referred to a higher level of care.

Some patients require individualized treatment approaches. For example, the SAMHSA has developed specific treatment improvement protocols to tailor treatments to address specific needs of adolescents.³² The American Academy of Pediatrics recommends MAT with buprenorphine in adolescents with opioid use disorder.³³ Although methadone has been approved for pregnant, opioid-dependent patients, recent data indicate buprenorphine is as effective with lower intensity of neonatal abstinence syndrome.³⁴

Buprenorphine. This long-acting (half-life of 24 to 42 hours) opioid partial agonist is approved for treating opioid use disorder in office-based settings according to the Drug Abuse Treatment Act of 2000. Buprenorphine is administered in doses of 8 to 16 mg/d in film or tablet form (sublingual or buccal) and is available in various formulations (Table 4). It is well tolerated; constipation and unpleasant taste are the most common adverse effects. Physicians are required to have a federal waiver to obtain the Drug Enforcement Administration license to prescribe buprenorphine for opioid use disorder in an office setting.

Buprenorphine reduces or eliminates cravings and withdrawal symptoms and helps improve outcomes of abstinence from opioids and retention in treatment.³¹ Formulations of naloxone combined with buprenorphine reduce the risk of abuse via injection.³⁵ Buprenorphine is safe; however, overdoses can occur when it is combined with benzodiazepines and/or other opiates.

Methadone. This long-acting (half-life 8 to 59 hours), full opioid agonist is approved to treat opioid addiction in federal- and state-regulated opioid treatment programs, also known as methadone maintenance programs. These programs are highly structured and include intensive counseling, monitoring, and dispensing to reduce relapse. Methadone is administered orally either via powder, liquid concentrate, tablet, or solution of diskette. Typically, methadone is dispensed daily in doses of 60 to 100 mg, although higher doses are sometimes necessary. Patients who meet certain criteria for stability may be allowed to take home supplies of methadone.

Methadone has a “black-box” warning for overdose, QT prolongation, and risk for respiratory depression when used in combination with benzodiazepines. Because of its long and unpredictable half-life and tissue accumulation, methadone carries a high overdose risk, particularly with rapidly titrated doses during therapy initiation.³⁵ However, most overdose deaths have occurred with methadone prescribed for pain management. When prescribed and monitored in an opioid treatment program, methadone has shown a high safety profile with respect to overdoses.³⁶

Injectable and oral naltrexone. Used for prevention of relapse to opioid dependence, naltrexone is a pure opioid antagonist that is available as an oral or IM form. Naltrexone has high affinity for the opioid receptors and in therapeutic doses provides an effective blockade for heroin or opioids. Compliance with oral naltrexone has been poor, leading to development of an IM form of naltrexone that can be administered as a single 380-mg dose once every 4 weeks for 6 months or sometimes longer. Naltrexone is also approved for alcohol dependence.

To avoid precipitated withdrawal, patients should be detoxified from opioids for 7 to 10 days before they begin naltrexone, which has no potential for abuse. Common adverse effects include fatigue, nausea, headache, and, for the IM formulation, injection site reactions. There is a “black-box” warning for liver toxicity; therefore, baseline and periodic liver function tests are necessary.

A NIDA review reported poor compliance with oral naltrexone compared with methadone.³⁵ However, naltrexone has been shown to be effective in highly motivated patients (eg, impaired physicians) and the criminal justice population and for preventing relapse following taper from buprenorphine or methadone.^37,38

Treatment for opioid overdose

Naloxone is a highly effective treatment to reverse opioid overdose that is delivered via IM or IV injection or by nasal application. Naloxone has no abuse potential. In doses of 0.4 to 2 mg, naloxone reverses overdose within 2 minutes and is effective for 30 to 90 minutes.³⁹ One should call 911 as soon as possible after naloxone is administered. In several states, naloxone is available without a prescription for patients and family members to combat opioid overdoses. The CDC recommends offering naloxone to patients who have risk factors for opioid overdose.⁴⁰

Opioid abuse and overdose are large and growing problems, and in recent years the numbers have been staggering. Overdose deaths related to opioids increased from 28,647 in 2014 to 33,091 in 2015 (Figure).¹ More than 2 million individuals in the United States had opioid use disorder in 2015,² and approximately 80% of them received no treatment,³ even though effective treatment could reduce the scope of abuse.^4,5

Although psychiatrists typically are not the primary prescribers of opioid medications, they often treat psychiatric disorders in patients with chronic pain who take prescription opioids. A recent study found that, despite representing only 16% of the adult population, adults with mental health disorders receive more than one-half of all opioid prescriptions distributed each year in the United States.⁶ Therefore, psychiatrists must be aware of risk assessment strategies for patients receiving opioids.

In this article, we provide recommendations for managing individuals with opioid use disorder, including:

• how to identify risk factors for opioid use disorder and use screening tools
• how to evaluate a patient with suspected opioid use disorder and make the diagnosis
• how to treat a patient with opioid use disorder, including a review of approved pharmaceutical agents.

Risk factors for opioid abuse and overdose

Patients with a history of mental health and/or substance use disorders or at least 3 months of prescribed opioid treatment are at risk for opioid abuse. Those taking a high daily dose of opioids or who have a history of overdose are at risk for overdose from opioid abuse (Table 1).^7-12 Standardized tools, such as the Opioid Risk Tool, can be used to screen to assess risk for opioid abuse among individuals prescribed opioids for treatment of chronic pain.¹² However, clinicians must be aware that even patients without characteristic risk factors can become dependent on opioids and/or be at risk for an accidental or intentional overdose. For example, opioid therapy following surgical procedures, even in patients who do not have a history of opioid use, increases risk of developing opioid use disorder.¹³

Evaluation and diagnosis

DSM-5 criteria define 3 degrees of opioid use disorder, depending on how many of the following traits a patient exhibits (mild, 2 to 3; moderate, 4 to 5; and severe, ≥6 )¹⁴:

• taking more than the initially intended quantities of opioids or for a longer period of time than intended
• continuous attempts to reduce or otherwise manage opioid use or desires to do so
• a great deal of time using, recovering from, or acquiring opioids
• reports of strong cravings to use opioids
• failing to meet personal objectives at home, work, or school
• continued opioid use even though it causes recurrent social problems
• reduction or elimination of activities the patient once considered important due to opioid use
• opioid use in situations where it is physically dangerous
• continued opioid use despite persistent psychological or physiologic problems despite knowing that continued use is causing or worsening those problems
• tolerance to opioids (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision)
• withdrawal or use of opioids (or related substances) to prevent withdrawal (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision).

Clinicians should be vigilant for symptoms of opioid use or withdrawal, such as needle marks and weight loss, during the interview (Table 2). High-risk populations that require regular screening include individuals with a history of opioid use disorder, patients taking chronic pain medication, and psychiatric patients.¹⁵ During the interview, clinicians should take an nonjudgmental approach and avoid “shame and blame.”

Standardized scales. Various standardized scales can be used to evaluate patients for opioid withdrawal and risk for substance use disorder. Scales for assessing opioid withdrawal include:

• Clinical Opiate Withdrawal Scale
• Subjective Opiate Withdrawal Scale.

Substance use disorder screening tools include:

• Drug Abuse Screen Test-10
• Alcohol Use Disorders Identification Test
• National Institute on Drug Abuse (NIDA) Drug Screening Tool.¹⁷

Examination findings. A brief physical examination is necessary to document key findings (Table 2). Patients should undergo a urine drug screen; gas chromatography/mass spectroscopy can confirm positive results. During the examination, clinicians should look for signs and symptoms of co-occurring substance use (eg, benzo­diazepines, marijuana, alcohol, cocaine) or mental disorders (mood, anxiety, attention-deficit).^18-21 Because nonprescription opioid use is associated with increased risk of suicide attempts and ideation,²² a suicide risk assessment is necessary.

Managing opioid use disorder

Detoxification is a 3-tiered approach that requires judicious prescription of medication, psychosocial support, and supervision to relieve opioid withdrawal symptoms. In both inpatient and outpatient settings, medications used for opioid detoxification include buprenorphine, clonidine, and methadone administered in doses tapered over 5 to 7 days. Appropriate detoxification increases treatment retention for continuing care.^23,24

Buprenorphine or buprenorphine/naloxone is the first-line option for outpatient and inpatient detoxification. Short-term detoxification schedules include starting doses between 4 and 16 mg/d, tapered over 5 to 7 days. Compared with methadone, buprenorphine has a lower risk of overdose²⁵ and abuse potential and can be given in an office-based setting. Clonidine, 0.3 to 1.2 mg/d in divided doses, is an alternative to buprenorphine and can be used in in­patient settings.²⁶

Clonidine is not as effective as buprenorphine for detoxification, but it may be used when buprenorphine is contraindicated. Clonidine may require adjuvant symptomatic treatment for insomnia (eg, trazodone, 100 mg at bedtime), anxiety (eg, hydroxyzine, 25 mg, twice a day), or diarrhea (loperamide, 2 mg/d). If a patient needs more structure and monitoring, he (she) should be referred for inpatient detoxification or to a methadone program.²⁷

Medication-assisted therapies

Detoxification alone often is not sufficient treatment. Medication-assisted therapy (MAT) is typically recommended by federal guidelines provided by the Substance Abuse and Mental Health Administration (SAMHSA) for patients with opioid use disorders.³ Patients can be directly transitioned from currently abused opioids to MAT on an outpatient basis. FDA-approved medications for MAT for opioid use disorder include buprenorphine, naltrexone (oral and long-acting injectable), and methadone (Table 3). Choice of MAT depends on several factors, including cost, patient preference, and availability of methadone programs and buprenorphine providers.²⁸

MAT should include psychosocial support^29-33 and active monitoring with urine drug screens. Maintenance therapy with medications is usually long-term and has been shown to have better outcomes than detoxification alone or short-term treatment.³⁴ Relapse during MAT should not be cause to discontinue treatment; instead, the patient should be referred to a higher level of care.

Some patients require individualized treatment approaches. For example, the SAMHSA has developed specific treatment improvement protocols to tailor treatments to address specific needs of adolescents.³² The American Academy of Pediatrics recommends MAT with buprenorphine in adolescents with opioid use disorder.³³ Although methadone has been approved for pregnant, opioid-dependent patients, recent data indicate buprenorphine is as effective with lower intensity of neonatal abstinence syndrome.³⁴

Buprenorphine. This long-acting (half-life of 24 to 42 hours) opioid partial agonist is approved for treating opioid use disorder in office-based settings according to the Drug Abuse Treatment Act of 2000. Buprenorphine is administered in doses of 8 to 16 mg/d in film or tablet form (sublingual or buccal) and is available in various formulations (Table 4). It is well tolerated; constipation and unpleasant taste are the most common adverse effects. Physicians are required to have a federal waiver to obtain the Drug Enforcement Administration license to prescribe buprenorphine for opioid use disorder in an office setting.

Buprenorphine reduces or eliminates cravings and withdrawal symptoms and helps improve outcomes of abstinence from opioids and retention in treatment.³¹ Formulations of naloxone combined with buprenorphine reduce the risk of abuse via injection.³⁵ Buprenorphine is safe; however, overdoses can occur when it is combined with benzodiazepines and/or other opiates.

Methadone. This long-acting (half-life 8 to 59 hours), full opioid agonist is approved to treat opioid addiction in federal- and state-regulated opioid treatment programs, also known as methadone maintenance programs. These programs are highly structured and include intensive counseling, monitoring, and dispensing to reduce relapse. Methadone is administered orally either via powder, liquid concentrate, tablet, or solution of diskette. Typically, methadone is dispensed daily in doses of 60 to 100 mg, although higher doses are sometimes necessary. Patients who meet certain criteria for stability may be allowed to take home supplies of methadone.

Methadone has a “black-box” warning for overdose, QT prolongation, and risk for respiratory depression when used in combination with benzodiazepines. Because of its long and unpredictable half-life and tissue accumulation, methadone carries a high overdose risk, particularly with rapidly titrated doses during therapy initiation.³⁵ However, most overdose deaths have occurred with methadone prescribed for pain management. When prescribed and monitored in an opioid treatment program, methadone has shown a high safety profile with respect to overdoses.³⁶

Injectable and oral naltrexone. Used for prevention of relapse to opioid dependence, naltrexone is a pure opioid antagonist that is available as an oral or IM form. Naltrexone has high affinity for the opioid receptors and in therapeutic doses provides an effective blockade for heroin or opioids. Compliance with oral naltrexone has been poor, leading to development of an IM form of naltrexone that can be administered as a single 380-mg dose once every 4 weeks for 6 months or sometimes longer. Naltrexone is also approved for alcohol dependence.

To avoid precipitated withdrawal, patients should be detoxified from opioids for 7 to 10 days before they begin naltrexone, which has no potential for abuse. Common adverse effects include fatigue, nausea, headache, and, for the IM formulation, injection site reactions. There is a “black-box” warning for liver toxicity; therefore, baseline and periodic liver function tests are necessary.

A NIDA review reported poor compliance with oral naltrexone compared with methadone.³⁵ However, naltrexone has been shown to be effective in highly motivated patients (eg, impaired physicians) and the criminal justice population and for preventing relapse following taper from buprenorphine or methadone.^37,38

Treatment for opioid overdose

Naloxone is a highly effective treatment to reverse opioid overdose that is delivered via IM or IV injection or by nasal application. Naloxone has no abuse potential. In doses of 0.4 to 2 mg, naloxone reverses overdose within 2 minutes and is effective for 30 to 90 minutes.³⁹ One should call 911 as soon as possible after naloxone is administered. In several states, naloxone is available without a prescription for patients and family members to combat opioid overdoses. The CDC recommends offering naloxone to patients who have risk factors for opioid overdose.⁴⁰

Opioid abuse and overdose are large and growing problems, and in recent years the numbers have been staggering. Overdose deaths related to opioids increased from 28,647 in 2014 to 33,091 in 2015 (Figure).¹ More than 2 million individuals in the United States had opioid use disorder in 2015,² and approximately 80% of them received no treatment,³ even though effective treatment could reduce the scope of abuse.^4,5

Although psychiatrists typically are not the primary prescribers of opioid medications, they often treat psychiatric disorders in patients with chronic pain who take prescription opioids. A recent study found that, despite representing only 16% of the adult population, adults with mental health disorders receive more than one-half of all opioid prescriptions distributed each year in the United States.⁶ Therefore, psychiatrists must be aware of risk assessment strategies for patients receiving opioids.

In this article, we provide recommendations for managing individuals with opioid use disorder, including:

• how to identify risk factors for opioid use disorder and use screening tools
• how to evaluate a patient with suspected opioid use disorder and make the diagnosis
• how to treat a patient with opioid use disorder, including a review of approved pharmaceutical agents.

Risk factors for opioid abuse and overdose

Patients with a history of mental health and/or substance use disorders or at least 3 months of prescribed opioid treatment are at risk for opioid abuse. Those taking a high daily dose of opioids or who have a history of overdose are at risk for overdose from opioid abuse (Table 1).^7-12 Standardized tools, such as the Opioid Risk Tool, can be used to screen to assess risk for opioid abuse among individuals prescribed opioids for treatment of chronic pain.¹² However, clinicians must be aware that even patients without characteristic risk factors can become dependent on opioids and/or be at risk for an accidental or intentional overdose. For example, opioid therapy following surgical procedures, even in patients who do not have a history of opioid use, increases risk of developing opioid use disorder.¹³

Evaluation and diagnosis

DSM-5 criteria define 3 degrees of opioid use disorder, depending on how many of the following traits a patient exhibits (mild, 2 to 3; moderate, 4 to 5; and severe, ≥6 )¹⁴:

• taking more than the initially intended quantities of opioids or for a longer period of time than intended
• continuous attempts to reduce or otherwise manage opioid use or desires to do so
• a great deal of time using, recovering from, or acquiring opioids
• reports of strong cravings to use opioids
• failing to meet personal objectives at home, work, or school
• continued opioid use even though it causes recurrent social problems
• reduction or elimination of activities the patient once considered important due to opioid use
• opioid use in situations where it is physically dangerous
• continued opioid use despite persistent psychological or physiologic problems despite knowing that continued use is causing or worsening those problems
• tolerance to opioids (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision)
• withdrawal or use of opioids (or related substances) to prevent withdrawal (not consequential for the diagnosis if the patient is taking opioids under appropriate medical supervision).

Clinicians should be vigilant for symptoms of opioid use or withdrawal, such as needle marks and weight loss, during the interview (Table 2). High-risk populations that require regular screening include individuals with a history of opioid use disorder, patients taking chronic pain medication, and psychiatric patients.¹⁵ During the interview, clinicians should take an nonjudgmental approach and avoid “shame and blame.”

Standardized scales. Various standardized scales can be used to evaluate patients for opioid withdrawal and risk for substance use disorder. Scales for assessing opioid withdrawal include:

• Clinical Opiate Withdrawal Scale
• Subjective Opiate Withdrawal Scale.

Substance use disorder screening tools include:

• Drug Abuse Screen Test-10
• Alcohol Use Disorders Identification Test
• National Institute on Drug Abuse (NIDA) Drug Screening Tool.¹⁷

Examination findings. A brief physical examination is necessary to document key findings (Table 2). Patients should undergo a urine drug screen; gas chromatography/mass spectroscopy can confirm positive results. During the examination, clinicians should look for signs and symptoms of co-occurring substance use (eg, benzo­diazepines, marijuana, alcohol, cocaine) or mental disorders (mood, anxiety, attention-deficit).^18-21 Because nonprescription opioid use is associated with increased risk of suicide attempts and ideation,²² a suicide risk assessment is necessary.

Managing opioid use disorder

Detoxification is a 3-tiered approach that requires judicious prescription of medication, psychosocial support, and supervision to relieve opioid withdrawal symptoms. In both inpatient and outpatient settings, medications used for opioid detoxification include buprenorphine, clonidine, and methadone administered in doses tapered over 5 to 7 days. Appropriate detoxification increases treatment retention for continuing care.^23,24

Buprenorphine or buprenorphine/naloxone is the first-line option for outpatient and inpatient detoxification. Short-term detoxification schedules include starting doses between 4 and 16 mg/d, tapered over 5 to 7 days. Compared with methadone, buprenorphine has a lower risk of overdose²⁵ and abuse potential and can be given in an office-based setting. Clonidine, 0.3 to 1.2 mg/d in divided doses, is an alternative to buprenorphine and can be used in in­patient settings.²⁶

Clonidine is not as effective as buprenorphine for detoxification, but it may be used when buprenorphine is contraindicated. Clonidine may require adjuvant symptomatic treatment for insomnia (eg, trazodone, 100 mg at bedtime), anxiety (eg, hydroxyzine, 25 mg, twice a day), or diarrhea (loperamide, 2 mg/d). If a patient needs more structure and monitoring, he (she) should be referred for inpatient detoxification or to a methadone program.²⁷

Medication-assisted therapies

Detoxification alone often is not sufficient treatment. Medication-assisted therapy (MAT) is typically recommended by federal guidelines provided by the Substance Abuse and Mental Health Administration (SAMHSA) for patients with opioid use disorders.³ Patients can be directly transitioned from currently abused opioids to MAT on an outpatient basis. FDA-approved medications for MAT for opioid use disorder include buprenorphine, naltrexone (oral and long-acting injectable), and methadone (Table 3). Choice of MAT depends on several factors, including cost, patient preference, and availability of methadone programs and buprenorphine providers.²⁸

MAT should include psychosocial support^29-33 and active monitoring with urine drug screens. Maintenance therapy with medications is usually long-term and has been shown to have better outcomes than detoxification alone or short-term treatment.³⁴ Relapse during MAT should not be cause to discontinue treatment; instead, the patient should be referred to a higher level of care.

Some patients require individualized treatment approaches. For example, the SAMHSA has developed specific treatment improvement protocols to tailor treatments to address specific needs of adolescents.³² The American Academy of Pediatrics recommends MAT with buprenorphine in adolescents with opioid use disorder.³³ Although methadone has been approved for pregnant, opioid-dependent patients, recent data indicate buprenorphine is as effective with lower intensity of neonatal abstinence syndrome.³⁴

Buprenorphine. This long-acting (half-life of 24 to 42 hours) opioid partial agonist is approved for treating opioid use disorder in office-based settings according to the Drug Abuse Treatment Act of 2000. Buprenorphine is administered in doses of 8 to 16 mg/d in film or tablet form (sublingual or buccal) and is available in various formulations (Table 4). It is well tolerated; constipation and unpleasant taste are the most common adverse effects. Physicians are required to have a federal waiver to obtain the Drug Enforcement Administration license to prescribe buprenorphine for opioid use disorder in an office setting.

Buprenorphine reduces or eliminates cravings and withdrawal symptoms and helps improve outcomes of abstinence from opioids and retention in treatment.³¹ Formulations of naloxone combined with buprenorphine reduce the risk of abuse via injection.³⁵ Buprenorphine is safe; however, overdoses can occur when it is combined with benzodiazepines and/or other opiates.

Methadone. This long-acting (half-life 8 to 59 hours), full opioid agonist is approved to treat opioid addiction in federal- and state-regulated opioid treatment programs, also known as methadone maintenance programs. These programs are highly structured and include intensive counseling, monitoring, and dispensing to reduce relapse. Methadone is administered orally either via powder, liquid concentrate, tablet, or solution of diskette. Typically, methadone is dispensed daily in doses of 60 to 100 mg, although higher doses are sometimes necessary. Patients who meet certain criteria for stability may be allowed to take home supplies of methadone.

Methadone has a “black-box” warning for overdose, QT prolongation, and risk for respiratory depression when used in combination with benzodiazepines. Because of its long and unpredictable half-life and tissue accumulation, methadone carries a high overdose risk, particularly with rapidly titrated doses during therapy initiation.³⁵ However, most overdose deaths have occurred with methadone prescribed for pain management. When prescribed and monitored in an opioid treatment program, methadone has shown a high safety profile with respect to overdoses.³⁶

Injectable and oral naltrexone. Used for prevention of relapse to opioid dependence, naltrexone is a pure opioid antagonist that is available as an oral or IM form. Naltrexone has high affinity for the opioid receptors and in therapeutic doses provides an effective blockade for heroin or opioids. Compliance with oral naltrexone has been poor, leading to development of an IM form of naltrexone that can be administered as a single 380-mg dose once every 4 weeks for 6 months or sometimes longer. Naltrexone is also approved for alcohol dependence.

To avoid precipitated withdrawal, patients should be detoxified from opioids for 7 to 10 days before they begin naltrexone, which has no potential for abuse. Common adverse effects include fatigue, nausea, headache, and, for the IM formulation, injection site reactions. There is a “black-box” warning for liver toxicity; therefore, baseline and periodic liver function tests are necessary.

A NIDA review reported poor compliance with oral naltrexone compared with methadone.³⁵ However, naltrexone has been shown to be effective in highly motivated patients (eg, impaired physicians) and the criminal justice population and for preventing relapse following taper from buprenorphine or methadone.^37,38

Treatment for opioid overdose

Naloxone is a highly effective treatment to reverse opioid overdose that is delivered via IM or IV injection or by nasal application. Naloxone has no abuse potential. In doses of 0.4 to 2 mg, naloxone reverses overdose within 2 minutes and is effective for 30 to 90 minutes.³⁹ One should call 911 as soon as possible after naloxone is administered. In several states, naloxone is available without a prescription for patients and family members to combat opioid overdoses. The CDC recommends offering naloxone to patients who have risk factors for opioid overdose.⁴⁰

Patients with chronic, severe mental illness live much shorter lives than the general population. The 25-year loss in life expectancy for people with chronic mental illness has been attributed to higher rates of cardiovascular disease driven by increased smoking, obesity, poverty, and poor nutrition.¹ These individuals also face the added burden of struggling with a psychiatric condition that often interferes with their ability to make optimal preventative health decisions, including staying up to date on vaccinations.² A recent review from Toronto, Canada, found that the influenza vaccination rates among homeless adults with mental illness—a population at high risk of respiratory illness—was only 6.7% compared with 31.1% for the general population of Ontario.³

Mental health professionals may serve as the only contacts to offer medical care to this vulnerable population, leading some psychiatric leaders to advocate that psychiatrists be considered primary care providers within accountable care organizations. Because most vaccines are easily available, mental health professionals should know about key immunizations to guide their patients accordingly.

In the United States, approximately 45,000 adults die annually from vaccine-preventable diseases, the majority from influenza.⁴ When combined with the most recent Adult Immunization Schedule and general recommendations adapted from the CDC,^5,6 the mnemonic ARM SHOT allows for a quick assessment of risk factors to guide administration and education about most vaccinations (Table 1). ARM SHOT involves assessing the following components of an individual’s health status and living arrangements to determine one’s risk of contracting communicable diseases:

• Age
• Risk of exposure
• Medical conditions (comorbidities)
• Substance use history
• HIV status or other immunocompromised states
• Occupancy, or living arrangements
• Tobacco use.

We recommend keeping a copy of the Adult Immunization Schedule (age ≥19) and/or the immunization schedule for children and adolescents (age ≤18) close for quick reference. Here, we provide a case and then explore how each component of the ARM SHOT mnemonic applies in decision-making.

Case Evaluating risk, assess needs

Ms. W, age 24, has bipolar I disorder, most recently manic with psychotic features. She presents for follow-up in clinic after a 5-day hospitalization for mania and comorbid alcohol use disorder. Her medical comorbidities include asthma and active tobacco use. She is taking lurasidone, 20 mg/d, and lithium, 900 mg/d. Her case manager is working to place Ms. W in a residential substance use disorder treatment program. Ms. W is on a waiting list to establish care with a primary care physician and has a history of poor engagement with medical services in general; prior attempts to place her with a primary care physician failed.

In advance of Ms. W’s transfer to a residential treatment facility, you have been asked to place a Mantoux screening test for tuberculosis (purified protein derivative), which raises the important question about her susceptibility to infectious diseases in general. To protect Ms. W from preventable diseases for which vaccines are available, you review the ARM SHOT mnemonic to broadly assess her candidacy for vaccinations.

Age

Age may be the most important determinant of a patient’s need for vaccination (Table 2). The CDC immunization schedules account for age-specific risks for diseases, complications, and responses to vaccination (Figure 1).⁶

Influenza vaccination. Adults can have an intramuscular or intradermal inactivated influenza vaccination yearly in the fall or winter, unless they have an allergy to a vaccine component such as egg protein. Those with such an allergy can receive a recombinant influenza vaccine. Until the 2016 to 2017 flu season, an intranasal mist of live, attenuated influenza vaccine was available to healthy, non-pregnant women, ages 2 to 49, without high-risk medical conditions. However, the CDC dropped its recommendation for this vaccine because data showed it did not effectively prevent the flu.⁷ Individuals age ≥65 can receive either the standard- or high-dose inactivated influenza vaccination. The latter contains 4 times the amount of antigen with the intention of triggering a stronger immune response in older adults.

Human papillomavirus (HPV) immunization. The ACIP recommendation⁹ has been for children to receive routine vaccination for the 4 major strains of HPV—strains 6, 11, 16, and 18—starting at ages 11 to 12 to confer protection from HPV-associated diseases, such as genital warts, oropharyngeal cancer, and anal cancer; cancers of the cervix, vulva, and vagina in women; and penile cancer in men. Ideally, the vaccines are administered prior to HPV exposure from sexual contact. The quadrivalent HPV vaccine is safe and is administered as a 3-dose series, with the second and third doses given 2 and 6 months, respectively, after the initial dose. Adolescent girls also have the option of a bivalent HPV vaccine.

In 2016, the FDA approved a 9-valent HPV vaccine, a simpler 2-dose schedule for children ages 9 to 14 (2 doses at least 6 months apart). Leading cancer centers have endorsed this vaccine based on strong comparative data with the 3-dose regimen.¹⁰ For those not previously vaccinated, the HPV vaccine is available for women ages 13 to 26 and for men ages 13 to 21 (although men ages 22 to 26 can receive the vaccine, and it is recommended for men who have sex with men [MSM]). Women do not require Papanicolaou, serum pregnancy, HPV DNA, or HPV antibody tests prior to vaccination. If a woman becomes pregnant, remaining doses of the vaccine should be postponed until after delivery. Women still need to follow recommendations for cervical cancer screening because the HPV vaccine does not cover all genital strains of the virus. For sexually active individuals who might have HPV or genital warts, immunization has no clinical effect except to prevent other HPV strains.

Measles, mumps, and rubella (MMR) vaccine. All adults should receive, at minimum, 1 dose of MMR vaccination unless serological immunity can be verified or if contraindicated. Two doses of the vaccine are recommended for students attending post-high school institutions, health care personnel, and international travelers because they are at higher risk for exposure and transmission of measles and mumps. Individuals born before 1957 are considered immune to measles and mumps. A measles outbreak from December 2014 to February 2015¹¹ highlighted the importance of maintaining one’s immunity status for MMR.

Case continued

Based on Ms. W’s age, she should be offered vaccinations for influenza and opportunities to receive vaccinations for HPV, Tdap (the primary series, a Tdap or Td booster), and MMR, if appropriate and not completed previously.

Risk of exposure

Certain behaviors will increase the risk of exposure to and transmission of diseases communicable by blood and other bodily fluids (Table 3). These behaviors include needle injections (eg, during use of illicit drugs) and sexual activity with multiple partners, including MSM or promiscuity/impulsivity during a manic episode. A common consequence of risky behaviors is comorbid infection of HIV and viral hepatitis for those with substance use disorder or those who engage in high-risk sexual practices.^12,13

Hepatitis B virus (HBV) immunization. Vaccination is one of the most effective ways to prevent HBV infection, which is why it is offered to all health care workers. HBV immunization is a 3-dose series in which the second and third doses are given 1 and 6 months after the initial doses, respectively. In addition to certain medical risk factors or conditions that indicate HBV vaccination, people should be offered the vaccine if they are in a higher risk occupation, travel, are of Asian or Pacific Islander ethnicity from an endemic area, or have any present or suspected sexually transmitted diseases.

Hepatitis A virus (HAV) vaccination. HAV is transmitted via fecal–oral routes, often from contaminated water or food, or through household or sexual contact with an infected person. Individuals should receive the HAV vaccine if they use illicit drugs by any route of administration, work with primates infected with HAV, travel to countries with unknown or high rates of HAV, or have chronic liver disease (ie, hepatitis, alcohol use disorder, or non-alcoholic fatty liver disease) or clotting deficiencies. The CDC Health Information for International Travel, commonly called the “Yellow Book,” publishes vaccination recommendations for those who plan travel to specific countries.¹⁴

Case continued

Ms. W’s history of mania (if such episodes included increased sexual activity) and substance use would make her a candidate for the HBV and HAV vaccinations and could also strengthen our previous recommendation that she receive the HPV vaccination.

Medical conditions

Patients with certain medical conditions may have difficulty fighting infections or become more susceptible to morbidity and mortality from coinfection with vaccine-preventable illnesses. Secondary effects of psychotropic medications that may carry implications for vaccine recommendations (eg, risk of agranulocytosis and impaired cell-medicated immunity with mirtazapine and clozapine or renal impairment from lithium use) are of particular concern in psychiatric patients.²

To help care for these patients, the CDC has developed a “medical conditions” schedule (Figure 2). This schedule makes vaccination recommendations for those with a weakened immune system, including patients with HIV, chronic obstructive pulmonary disease (COPD), diabetes, hepatitis, asplenia, end-stage renal disease, cardiac disease, and pregnancy.

Case continued

Because of Ms. W’s asthma, the CDC schedule recommends ensuring she is up to date on her influenza, pneumococcal, and Tdap vaccinations.

Substance use

Patients with combined psychiatric and substance use disorders (“dual diagnosis”) have lower rates of receiving preventive care than patients with either condition alone.¹⁵ Substance use can be behaviorally disinhibiting, leading to increased risk of exposures from sexual contact or other risky activities. The use of illicit substances can provide a nidus for infection depending on the route of administration and can result in negative effects on organ systems, compromising one’s ability to ward off infection.

Patients who use any illicit drugs, regardless of the method of delivery, should be recommended for HAV vaccination. For those with alcohol use disorder and/or chronic liver disease, and/or seeking treatment for substance use, hepatitis B screening and vaccination is recommended.

Case continued

From a substance use perspective, discussion of vaccination status for both hepatitis A and B would be important for Ms. W.

HIV or immunocompromised

Persons with severe mental illness have high rates of HIV, with almost 8 times the risk of exposure, compared with the general population due to myriad reasons, including greater rates of substance abuse, higher risk sexual behavior, and lack of awareness of HIV transmission.^12,13 Patients with mental illness are also at risk of leukopenia and agranulocytosis from certain drugs used to treat their conditions, such as clozapine.

Pregnancy is a challenge for women with mental illness because of the pharmacologic risk and immune-system compromise to the mother and baby. A pregnant woman who has HIV with a CD4 count <200, or has a weakened immune system from an organ transplant or a similar condition, is a candidate for certain vaccines based on the Adult Immunization Schedule (Figure 2). However, these patients should avoid live vaccines, such as the intranasal mist of live influenza, MMR, VZV, and varicella, to avoid illness from these inoculations.
 

Case continued

Ms. W should undergo testing for pregnancy and HIV (and preferably other sexually transmitted infections per general preventive health guidelines) before receiving any live vaccinations.

Occupancy

Aside from direct transmission of bodily fluids, infectious diseases also can spread through droplets/secretions from the throat and respiratory tract. Close quarters or lengthy contact enhances communicability by droplets, and therefore people who reside in a communal living space (eg, individuals in substance use treatment facilities or those who reside in a nursing home) are most susceptible.

The bacterial disease Neisseria meningitidis (meningococcus) can spread through droplets and can cause pneumonia, bacteremia, and meningitis. Vaccination is indicated, and in some states is mandated, for college students who live in residence halls and missed routine vaccination by age 16. Meningococcus conjugate vaccine is administered in 2 doses; each dose may be given at least 2 months apart for those with HIV, asplenia, or persistent complement-related disorders. A single dose may be recommended for travelers to areas where meningococcal disease is hyperendemic or epidemic, military recruits, or microbiologists. For those age ≥55 and older, meningococcal polysaccharide vaccine is recommended over meningococcal conjugate vaccine.

Influenza, MMR, diphtheria, pertussis, and pneumococcus also spread through droplet contact.

Case continued

If Ms. W had not previously received the meningococcus vaccine as part of adolescent immunizations, she could benefit from this vaccine because she plans to enter a residential substance use disorder treatment program.

Tobacco use

Patients with psychiatric illness are twice as likely to smoke compared with the general population.¹⁶ Adult smokers, especially those with chronic lung disease, are at higher risk for influenza and pneumococcal-related illness; they should be vaccinated against these illnesses regardless of age (as discussed in the “Age” section).

Case continued

Because she smokes, Ms. W should receive counseling on vaccinations, such as influenza and pneumonia, to lessen her risk of respiratory illnesses and downstream sepsis.

Conclusion

Ms. W’s case represents an unfortunately all-too-common scenario where her multifaceted biopsychosocial circumstances place her at high risk for vaccine-preventable conditions. Her weight is recorded and laboratory work ordered to evaluate her pregnancy status, blood counts, lipids, complete metabolic panel, lithium level, and HIV status. Fortunately, she had received her series of MMR, meningococcal, and Tdap vaccinations when she was younger. Influenza, HPV, HAV, HBV, and pneumococcal vaccinations were all recommended to her, all of which can be given on the same day (HAV and HBV often are available as a combined vaccine). Ms. W receives a renewal of her psychiatric medications and counseling on healthy living habits (eg, diet and exercise, quitting tobacco and alcohol use, and safe sex practices) and the importance of immunizations.

Vaccination is 1 of the 10 great public health achievements of the 20th century when one considers how immunization of vaccine-preventable diseases has reduced morbidity, mortality, and health-associated costs.¹⁷ As mental health professionals, we can help pass on the direct and indirect benefits of immunizations to an often underserved and undertreated population to help improve their health outcomes and quality of life.

Patients with chronic, severe mental illness live much shorter lives than the general population. The 25-year loss in life expectancy for people with chronic mental illness has been attributed to higher rates of cardiovascular disease driven by increased smoking, obesity, poverty, and poor nutrition.¹ These individuals also face the added burden of struggling with a psychiatric condition that often interferes with their ability to make optimal preventative health decisions, including staying up to date on vaccinations.² A recent review from Toronto, Canada, found that the influenza vaccination rates among homeless adults with mental illness—a population at high risk of respiratory illness—was only 6.7% compared with 31.1% for the general population of Ontario.³

Mental health professionals may serve as the only contacts to offer medical care to this vulnerable population, leading some psychiatric leaders to advocate that psychiatrists be considered primary care providers within accountable care organizations. Because most vaccines are easily available, mental health professionals should know about key immunizations to guide their patients accordingly.

In the United States, approximately 45,000 adults die annually from vaccine-preventable diseases, the majority from influenza.⁴ When combined with the most recent Adult Immunization Schedule and general recommendations adapted from the CDC,^5,6 the mnemonic ARM SHOT allows for a quick assessment of risk factors to guide administration and education about most vaccinations (Table 1). ARM SHOT involves assessing the following components of an individual’s health status and living arrangements to determine one’s risk of contracting communicable diseases:

• Age
• Risk of exposure
• Medical conditions (comorbidities)
• Substance use history
• HIV status or other immunocompromised states
• Occupancy, or living arrangements
• Tobacco use.

We recommend keeping a copy of the Adult Immunization Schedule (age ≥19) and/or the immunization schedule for children and adolescents (age ≤18) close for quick reference. Here, we provide a case and then explore how each component of the ARM SHOT mnemonic applies in decision-making.

Case Evaluating risk, assess needs

Ms. W, age 24, has bipolar I disorder, most recently manic with psychotic features. She presents for follow-up in clinic after a 5-day hospitalization for mania and comorbid alcohol use disorder. Her medical comorbidities include asthma and active tobacco use. She is taking lurasidone, 20 mg/d, and lithium, 900 mg/d. Her case manager is working to place Ms. W in a residential substance use disorder treatment program. Ms. W is on a waiting list to establish care with a primary care physician and has a history of poor engagement with medical services in general; prior attempts to place her with a primary care physician failed.

In advance of Ms. W’s transfer to a residential treatment facility, you have been asked to place a Mantoux screening test for tuberculosis (purified protein derivative), which raises the important question about her susceptibility to infectious diseases in general. To protect Ms. W from preventable diseases for which vaccines are available, you review the ARM SHOT mnemonic to broadly assess her candidacy for vaccinations.

Age

Age may be the most important determinant of a patient’s need for vaccination (Table 2). The CDC immunization schedules account for age-specific risks for diseases, complications, and responses to vaccination (Figure 1).⁶

Influenza vaccination. Adults can have an intramuscular or intradermal inactivated influenza vaccination yearly in the fall or winter, unless they have an allergy to a vaccine component such as egg protein. Those with such an allergy can receive a recombinant influenza vaccine. Until the 2016 to 2017 flu season, an intranasal mist of live, attenuated influenza vaccine was available to healthy, non-pregnant women, ages 2 to 49, without high-risk medical conditions. However, the CDC dropped its recommendation for this vaccine because data showed it did not effectively prevent the flu.⁷ Individuals age ≥65 can receive either the standard- or high-dose inactivated influenza vaccination. The latter contains 4 times the amount of antigen with the intention of triggering a stronger immune response in older adults.

Human papillomavirus (HPV) immunization. The ACIP recommendation⁹ has been for children to receive routine vaccination for the 4 major strains of HPV—strains 6, 11, 16, and 18—starting at ages 11 to 12 to confer protection from HPV-associated diseases, such as genital warts, oropharyngeal cancer, and anal cancer; cancers of the cervix, vulva, and vagina in women; and penile cancer in men. Ideally, the vaccines are administered prior to HPV exposure from sexual contact. The quadrivalent HPV vaccine is safe and is administered as a 3-dose series, with the second and third doses given 2 and 6 months, respectively, after the initial dose. Adolescent girls also have the option of a bivalent HPV vaccine.

In 2016, the FDA approved a 9-valent HPV vaccine, a simpler 2-dose schedule for children ages 9 to 14 (2 doses at least 6 months apart). Leading cancer centers have endorsed this vaccine based on strong comparative data with the 3-dose regimen.¹⁰ For those not previously vaccinated, the HPV vaccine is available for women ages 13 to 26 and for men ages 13 to 21 (although men ages 22 to 26 can receive the vaccine, and it is recommended for men who have sex with men [MSM]). Women do not require Papanicolaou, serum pregnancy, HPV DNA, or HPV antibody tests prior to vaccination. If a woman becomes pregnant, remaining doses of the vaccine should be postponed until after delivery. Women still need to follow recommendations for cervical cancer screening because the HPV vaccine does not cover all genital strains of the virus. For sexually active individuals who might have HPV or genital warts, immunization has no clinical effect except to prevent other HPV strains.

Measles, mumps, and rubella (MMR) vaccine. All adults should receive, at minimum, 1 dose of MMR vaccination unless serological immunity can be verified or if contraindicated. Two doses of the vaccine are recommended for students attending post-high school institutions, health care personnel, and international travelers because they are at higher risk for exposure and transmission of measles and mumps. Individuals born before 1957 are considered immune to measles and mumps. A measles outbreak from December 2014 to February 2015¹¹ highlighted the importance of maintaining one’s immunity status for MMR.

Case continued

Based on Ms. W’s age, she should be offered vaccinations for influenza and opportunities to receive vaccinations for HPV, Tdap (the primary series, a Tdap or Td booster), and MMR, if appropriate and not completed previously.

Risk of exposure

Certain behaviors will increase the risk of exposure to and transmission of diseases communicable by blood and other bodily fluids (Table 3). These behaviors include needle injections (eg, during use of illicit drugs) and sexual activity with multiple partners, including MSM or promiscuity/impulsivity during a manic episode. A common consequence of risky behaviors is comorbid infection of HIV and viral hepatitis for those with substance use disorder or those who engage in high-risk sexual practices.^12,13

Hepatitis B virus (HBV) immunization. Vaccination is one of the most effective ways to prevent HBV infection, which is why it is offered to all health care workers. HBV immunization is a 3-dose series in which the second and third doses are given 1 and 6 months after the initial doses, respectively. In addition to certain medical risk factors or conditions that indicate HBV vaccination, people should be offered the vaccine if they are in a higher risk occupation, travel, are of Asian or Pacific Islander ethnicity from an endemic area, or have any present or suspected sexually transmitted diseases.

Hepatitis A virus (HAV) vaccination. HAV is transmitted via fecal–oral routes, often from contaminated water or food, or through household or sexual contact with an infected person. Individuals should receive the HAV vaccine if they use illicit drugs by any route of administration, work with primates infected with HAV, travel to countries with unknown or high rates of HAV, or have chronic liver disease (ie, hepatitis, alcohol use disorder, or non-alcoholic fatty liver disease) or clotting deficiencies. The CDC Health Information for International Travel, commonly called the “Yellow Book,” publishes vaccination recommendations for those who plan travel to specific countries.¹⁴

Case continued

Ms. W’s history of mania (if such episodes included increased sexual activity) and substance use would make her a candidate for the HBV and HAV vaccinations and could also strengthen our previous recommendation that she receive the HPV vaccination.

Medical conditions

Patients with certain medical conditions may have difficulty fighting infections or become more susceptible to morbidity and mortality from coinfection with vaccine-preventable illnesses. Secondary effects of psychotropic medications that may carry implications for vaccine recommendations (eg, risk of agranulocytosis and impaired cell-medicated immunity with mirtazapine and clozapine or renal impairment from lithium use) are of particular concern in psychiatric patients.²

To help care for these patients, the CDC has developed a “medical conditions” schedule (Figure 2). This schedule makes vaccination recommendations for those with a weakened immune system, including patients with HIV, chronic obstructive pulmonary disease (COPD), diabetes, hepatitis, asplenia, end-stage renal disease, cardiac disease, and pregnancy.

Case continued

Because of Ms. W’s asthma, the CDC schedule recommends ensuring she is up to date on her influenza, pneumococcal, and Tdap vaccinations.

Substance use

Patients with combined psychiatric and substance use disorders (“dual diagnosis”) have lower rates of receiving preventive care than patients with either condition alone.¹⁵ Substance use can be behaviorally disinhibiting, leading to increased risk of exposures from sexual contact or other risky activities. The use of illicit substances can provide a nidus for infection depending on the route of administration and can result in negative effects on organ systems, compromising one’s ability to ward off infection.

Patients who use any illicit drugs, regardless of the method of delivery, should be recommended for HAV vaccination. For those with alcohol use disorder and/or chronic liver disease, and/or seeking treatment for substance use, hepatitis B screening and vaccination is recommended.

Case continued

From a substance use perspective, discussion of vaccination status for both hepatitis A and B would be important for Ms. W.

HIV or immunocompromised

Persons with severe mental illness have high rates of HIV, with almost 8 times the risk of exposure, compared with the general population due to myriad reasons, including greater rates of substance abuse, higher risk sexual behavior, and lack of awareness of HIV transmission.^12,13 Patients with mental illness are also at risk of leukopenia and agranulocytosis from certain drugs used to treat their conditions, such as clozapine.

Pregnancy is a challenge for women with mental illness because of the pharmacologic risk and immune-system compromise to the mother and baby. A pregnant woman who has HIV with a CD4 count <200, or has a weakened immune system from an organ transplant or a similar condition, is a candidate for certain vaccines based on the Adult Immunization Schedule (Figure 2). However, these patients should avoid live vaccines, such as the intranasal mist of live influenza, MMR, VZV, and varicella, to avoid illness from these inoculations.
 

Case continued

Ms. W should undergo testing for pregnancy and HIV (and preferably other sexually transmitted infections per general preventive health guidelines) before receiving any live vaccinations.

Occupancy

Aside from direct transmission of bodily fluids, infectious diseases also can spread through droplets/secretions from the throat and respiratory tract. Close quarters or lengthy contact enhances communicability by droplets, and therefore people who reside in a communal living space (eg, individuals in substance use treatment facilities or those who reside in a nursing home) are most susceptible.

The bacterial disease Neisseria meningitidis (meningococcus) can spread through droplets and can cause pneumonia, bacteremia, and meningitis. Vaccination is indicated, and in some states is mandated, for college students who live in residence halls and missed routine vaccination by age 16. Meningococcus conjugate vaccine is administered in 2 doses; each dose may be given at least 2 months apart for those with HIV, asplenia, or persistent complement-related disorders. A single dose may be recommended for travelers to areas where meningococcal disease is hyperendemic or epidemic, military recruits, or microbiologists. For those age ≥55 and older, meningococcal polysaccharide vaccine is recommended over meningococcal conjugate vaccine.

Influenza, MMR, diphtheria, pertussis, and pneumococcus also spread through droplet contact.

Case continued

If Ms. W had not previously received the meningococcus vaccine as part of adolescent immunizations, she could benefit from this vaccine because she plans to enter a residential substance use disorder treatment program.

Tobacco use

Patients with psychiatric illness are twice as likely to smoke compared with the general population.¹⁶ Adult smokers, especially those with chronic lung disease, are at higher risk for influenza and pneumococcal-related illness; they should be vaccinated against these illnesses regardless of age (as discussed in the “Age” section).

Case continued

Because she smokes, Ms. W should receive counseling on vaccinations, such as influenza and pneumonia, to lessen her risk of respiratory illnesses and downstream sepsis.

Conclusion

Ms. W’s case represents an unfortunately all-too-common scenario where her multifaceted biopsychosocial circumstances place her at high risk for vaccine-preventable conditions. Her weight is recorded and laboratory work ordered to evaluate her pregnancy status, blood counts, lipids, complete metabolic panel, lithium level, and HIV status. Fortunately, she had received her series of MMR, meningococcal, and Tdap vaccinations when she was younger. Influenza, HPV, HAV, HBV, and pneumococcal vaccinations were all recommended to her, all of which can be given on the same day (HAV and HBV often are available as a combined vaccine). Ms. W receives a renewal of her psychiatric medications and counseling on healthy living habits (eg, diet and exercise, quitting tobacco and alcohol use, and safe sex practices) and the importance of immunizations.

Vaccination is 1 of the 10 great public health achievements of the 20th century when one considers how immunization of vaccine-preventable diseases has reduced morbidity, mortality, and health-associated costs.¹⁷ As mental health professionals, we can help pass on the direct and indirect benefits of immunizations to an often underserved and undertreated population to help improve their health outcomes and quality of life.

Patients with chronic, severe mental illness live much shorter lives than the general population. The 25-year loss in life expectancy for people with chronic mental illness has been attributed to higher rates of cardiovascular disease driven by increased smoking, obesity, poverty, and poor nutrition.¹ These individuals also face the added burden of struggling with a psychiatric condition that often interferes with their ability to make optimal preventative health decisions, including staying up to date on vaccinations.² A recent review from Toronto, Canada, found that the influenza vaccination rates among homeless adults with mental illness—a population at high risk of respiratory illness—was only 6.7% compared with 31.1% for the general population of Ontario.³

Mental health professionals may serve as the only contacts to offer medical care to this vulnerable population, leading some psychiatric leaders to advocate that psychiatrists be considered primary care providers within accountable care organizations. Because most vaccines are easily available, mental health professionals should know about key immunizations to guide their patients accordingly.

In the United States, approximately 45,000 adults die annually from vaccine-preventable diseases, the majority from influenza.⁴ When combined with the most recent Adult Immunization Schedule and general recommendations adapted from the CDC,^5,6 the mnemonic ARM SHOT allows for a quick assessment of risk factors to guide administration and education about most vaccinations (Table 1). ARM SHOT involves assessing the following components of an individual’s health status and living arrangements to determine one’s risk of contracting communicable diseases:

• Age
• Risk of exposure
• Medical conditions (comorbidities)
• Substance use history
• HIV status or other immunocompromised states
• Occupancy, or living arrangements
• Tobacco use.

We recommend keeping a copy of the Adult Immunization Schedule (age ≥19) and/or the immunization schedule for children and adolescents (age ≤18) close for quick reference. Here, we provide a case and then explore how each component of the ARM SHOT mnemonic applies in decision-making.

Case Evaluating risk, assess needs

Ms. W, age 24, has bipolar I disorder, most recently manic with psychotic features. She presents for follow-up in clinic after a 5-day hospitalization for mania and comorbid alcohol use disorder. Her medical comorbidities include asthma and active tobacco use. She is taking lurasidone, 20 mg/d, and lithium, 900 mg/d. Her case manager is working to place Ms. W in a residential substance use disorder treatment program. Ms. W is on a waiting list to establish care with a primary care physician and has a history of poor engagement with medical services in general; prior attempts to place her with a primary care physician failed.

In advance of Ms. W’s transfer to a residential treatment facility, you have been asked to place a Mantoux screening test for tuberculosis (purified protein derivative), which raises the important question about her susceptibility to infectious diseases in general. To protect Ms. W from preventable diseases for which vaccines are available, you review the ARM SHOT mnemonic to broadly assess her candidacy for vaccinations.

Age

Age may be the most important determinant of a patient’s need for vaccination (Table 2). The CDC immunization schedules account for age-specific risks for diseases, complications, and responses to vaccination (Figure 1).⁶

Influenza vaccination. Adults can have an intramuscular or intradermal inactivated influenza vaccination yearly in the fall or winter, unless they have an allergy to a vaccine component such as egg protein. Those with such an allergy can receive a recombinant influenza vaccine. Until the 2016 to 2017 flu season, an intranasal mist of live, attenuated influenza vaccine was available to healthy, non-pregnant women, ages 2 to 49, without high-risk medical conditions. However, the CDC dropped its recommendation for this vaccine because data showed it did not effectively prevent the flu.⁷ Individuals age ≥65 can receive either the standard- or high-dose inactivated influenza vaccination. The latter contains 4 times the amount of antigen with the intention of triggering a stronger immune response in older adults.

Human papillomavirus (HPV) immunization. The ACIP recommendation⁹ has been for children to receive routine vaccination for the 4 major strains of HPV—strains 6, 11, 16, and 18—starting at ages 11 to 12 to confer protection from HPV-associated diseases, such as genital warts, oropharyngeal cancer, and anal cancer; cancers of the cervix, vulva, and vagina in women; and penile cancer in men. Ideally, the vaccines are administered prior to HPV exposure from sexual contact. The quadrivalent HPV vaccine is safe and is administered as a 3-dose series, with the second and third doses given 2 and 6 months, respectively, after the initial dose. Adolescent girls also have the option of a bivalent HPV vaccine.

In 2016, the FDA approved a 9-valent HPV vaccine, a simpler 2-dose schedule for children ages 9 to 14 (2 doses at least 6 months apart). Leading cancer centers have endorsed this vaccine based on strong comparative data with the 3-dose regimen.¹⁰ For those not previously vaccinated, the HPV vaccine is available for women ages 13 to 26 and for men ages 13 to 21 (although men ages 22 to 26 can receive the vaccine, and it is recommended for men who have sex with men [MSM]). Women do not require Papanicolaou, serum pregnancy, HPV DNA, or HPV antibody tests prior to vaccination. If a woman becomes pregnant, remaining doses of the vaccine should be postponed until after delivery. Women still need to follow recommendations for cervical cancer screening because the HPV vaccine does not cover all genital strains of the virus. For sexually active individuals who might have HPV or genital warts, immunization has no clinical effect except to prevent other HPV strains.

Measles, mumps, and rubella (MMR) vaccine. All adults should receive, at minimum, 1 dose of MMR vaccination unless serological immunity can be verified or if contraindicated. Two doses of the vaccine are recommended for students attending post-high school institutions, health care personnel, and international travelers because they are at higher risk for exposure and transmission of measles and mumps. Individuals born before 1957 are considered immune to measles and mumps. A measles outbreak from December 2014 to February 2015¹¹ highlighted the importance of maintaining one’s immunity status for MMR.

Case continued

Based on Ms. W’s age, she should be offered vaccinations for influenza and opportunities to receive vaccinations for HPV, Tdap (the primary series, a Tdap or Td booster), and MMR, if appropriate and not completed previously.

Risk of exposure

Certain behaviors will increase the risk of exposure to and transmission of diseases communicable by blood and other bodily fluids (Table 3). These behaviors include needle injections (eg, during use of illicit drugs) and sexual activity with multiple partners, including MSM or promiscuity/impulsivity during a manic episode. A common consequence of risky behaviors is comorbid infection of HIV and viral hepatitis for those with substance use disorder or those who engage in high-risk sexual practices.^12,13

Hepatitis B virus (HBV) immunization. Vaccination is one of the most effective ways to prevent HBV infection, which is why it is offered to all health care workers. HBV immunization is a 3-dose series in which the second and third doses are given 1 and 6 months after the initial doses, respectively. In addition to certain medical risk factors or conditions that indicate HBV vaccination, people should be offered the vaccine if they are in a higher risk occupation, travel, are of Asian or Pacific Islander ethnicity from an endemic area, or have any present or suspected sexually transmitted diseases.

Hepatitis A virus (HAV) vaccination. HAV is transmitted via fecal–oral routes, often from contaminated water or food, or through household or sexual contact with an infected person. Individuals should receive the HAV vaccine if they use illicit drugs by any route of administration, work with primates infected with HAV, travel to countries with unknown or high rates of HAV, or have chronic liver disease (ie, hepatitis, alcohol use disorder, or non-alcoholic fatty liver disease) or clotting deficiencies. The CDC Health Information for International Travel, commonly called the “Yellow Book,” publishes vaccination recommendations for those who plan travel to specific countries.¹⁴

Case continued

Ms. W’s history of mania (if such episodes included increased sexual activity) and substance use would make her a candidate for the HBV and HAV vaccinations and could also strengthen our previous recommendation that she receive the HPV vaccination.

Medical conditions

Patients with certain medical conditions may have difficulty fighting infections or become more susceptible to morbidity and mortality from coinfection with vaccine-preventable illnesses. Secondary effects of psychotropic medications that may carry implications for vaccine recommendations (eg, risk of agranulocytosis and impaired cell-medicated immunity with mirtazapine and clozapine or renal impairment from lithium use) are of particular concern in psychiatric patients.²

To help care for these patients, the CDC has developed a “medical conditions” schedule (Figure 2). This schedule makes vaccination recommendations for those with a weakened immune system, including patients with HIV, chronic obstructive pulmonary disease (COPD), diabetes, hepatitis, asplenia, end-stage renal disease, cardiac disease, and pregnancy.

Case continued

Because of Ms. W’s asthma, the CDC schedule recommends ensuring she is up to date on her influenza, pneumococcal, and Tdap vaccinations.

Substance use

Patients with combined psychiatric and substance use disorders (“dual diagnosis”) have lower rates of receiving preventive care than patients with either condition alone.¹⁵ Substance use can be behaviorally disinhibiting, leading to increased risk of exposures from sexual contact or other risky activities. The use of illicit substances can provide a nidus for infection depending on the route of administration and can result in negative effects on organ systems, compromising one’s ability to ward off infection.

Patients who use any illicit drugs, regardless of the method of delivery, should be recommended for HAV vaccination. For those with alcohol use disorder and/or chronic liver disease, and/or seeking treatment for substance use, hepatitis B screening and vaccination is recommended.

Case continued

From a substance use perspective, discussion of vaccination status for both hepatitis A and B would be important for Ms. W.

HIV or immunocompromised

Persons with severe mental illness have high rates of HIV, with almost 8 times the risk of exposure, compared with the general population due to myriad reasons, including greater rates of substance abuse, higher risk sexual behavior, and lack of awareness of HIV transmission.^12,13 Patients with mental illness are also at risk of leukopenia and agranulocytosis from certain drugs used to treat their conditions, such as clozapine.

Pregnancy is a challenge for women with mental illness because of the pharmacologic risk and immune-system compromise to the mother and baby. A pregnant woman who has HIV with a CD4 count <200, or has a weakened immune system from an organ transplant or a similar condition, is a candidate for certain vaccines based on the Adult Immunization Schedule (Figure 2). However, these patients should avoid live vaccines, such as the intranasal mist of live influenza, MMR, VZV, and varicella, to avoid illness from these inoculations.
 

Case continued

Ms. W should undergo testing for pregnancy and HIV (and preferably other sexually transmitted infections per general preventive health guidelines) before receiving any live vaccinations.

Occupancy

Aside from direct transmission of bodily fluids, infectious diseases also can spread through droplets/secretions from the throat and respiratory tract. Close quarters or lengthy contact enhances communicability by droplets, and therefore people who reside in a communal living space (eg, individuals in substance use treatment facilities or those who reside in a nursing home) are most susceptible.

The bacterial disease Neisseria meningitidis (meningococcus) can spread through droplets and can cause pneumonia, bacteremia, and meningitis. Vaccination is indicated, and in some states is mandated, for college students who live in residence halls and missed routine vaccination by age 16. Meningococcus conjugate vaccine is administered in 2 doses; each dose may be given at least 2 months apart for those with HIV, asplenia, or persistent complement-related disorders. A single dose may be recommended for travelers to areas where meningococcal disease is hyperendemic or epidemic, military recruits, or microbiologists. For those age ≥55 and older, meningococcal polysaccharide vaccine is recommended over meningococcal conjugate vaccine.

Influenza, MMR, diphtheria, pertussis, and pneumococcus also spread through droplet contact.

Case continued

If Ms. W had not previously received the meningococcus vaccine as part of adolescent immunizations, she could benefit from this vaccine because she plans to enter a residential substance use disorder treatment program.

Tobacco use

Patients with psychiatric illness are twice as likely to smoke compared with the general population.¹⁶ Adult smokers, especially those with chronic lung disease, are at higher risk for influenza and pneumococcal-related illness; they should be vaccinated against these illnesses regardless of age (as discussed in the “Age” section).

Case continued

Because she smokes, Ms. W should receive counseling on vaccinations, such as influenza and pneumonia, to lessen her risk of respiratory illnesses and downstream sepsis.

Conclusion

Ms. W’s case represents an unfortunately all-too-common scenario where her multifaceted biopsychosocial circumstances place her at high risk for vaccine-preventable conditions. Her weight is recorded and laboratory work ordered to evaluate her pregnancy status, blood counts, lipids, complete metabolic panel, lithium level, and HIV status. Fortunately, she had received her series of MMR, meningococcal, and Tdap vaccinations when she was younger. Influenza, HPV, HAV, HBV, and pneumococcal vaccinations were all recommended to her, all of which can be given on the same day (HAV and HBV often are available as a combined vaccine). Ms. W receives a renewal of her psychiatric medications and counseling on healthy living habits (eg, diet and exercise, quitting tobacco and alcohol use, and safe sex practices) and the importance of immunizations.

Vaccination is 1 of the 10 great public health achievements of the 20th century when one considers how immunization of vaccine-preventable diseases has reduced morbidity, mortality, and health-associated costs.¹⁷ As mental health professionals, we can help pass on the direct and indirect benefits of immunizations to an often underserved and undertreated population to help improve their health outcomes and quality of life.

Excessive daytime sleepiness (EDS) is “the inability to maintain wakefulness and alertness during the major waking periods of the day, with sleep occurring unintentionally or at inappropriate times, almost daily for at least 3 months,” according to the American Academy of Sleep Medicine.¹ EDS is common, with a prevalence up to 25% to 30% in the general population.^1-4 The prevalence rate varies in different studies, primarily because of inconsistent definitions of EDS, and therefore differences in diagnosis and assessment.^1,2,4 In a study of 300 psychiatric outpatients, 34% had EDS.³ However, studies and evidence reviewing EDS in psychiatric patients are limited.

EDS vs fatigue

Many patients describe EDS as “fatigue”¹; however, a patient’s report of fatigue could be mistaken for EDS.⁴ Although there is overlap, it is important for physicians to distinguish between these 2 entities for accurate identification and treatment.^1,4

Risk of inadequate screening

A study of 117 patients with symptomatic coronary artery disease showed that EDS is associated with significantly greater incidence of cardiovascular adverse events at 16-month follow up.² This study had limitations such as small sample size; therefore, more studies are needed. Because of these risks, timely and accurate diagnosis not only improves the patient’s quality of life and reduces polypharmacy but also can be life-saving.

Common causes of EDS in psychiatric patients

Because of the high prevalence and severity of impairments caused by EDS, it is essential for psychiatrists to be informed about causes of EDS and thoroughly assess for the potential underlying etiology before concluding that the sleep problem is a manifestation of the psychiatric disorder and prescribing psychotropic medication for it.

Some common causes of EDS in psychiatric patients include:

Sleep-disordered breathing.⁸ Obstructive sleep apnea (OSA) is often underdiagnosed,^6,7 and considering how common it is,⁶ psychiatrists likely will see many patients with OSA in their practice.⁵ OSA has a higher prevalence among patients with psychiatric disorders such as depression^6,9 and schizophrenia. Additionally, there is evidence suggesting that patients with OSA are more likely to suffer from depression and EDS than healthy controls^6,9,10; some of the proposed mechanisms are sleep fragmentation and hypoxemia.^6,9-11 OSA is the most common form of sleep-disordered breathing and is a common cause of EDS.^1,2,12 Also, undiagnosed and untreated OSA in patients with depression could cause refractoriness to pharmacological treatment of depression.^6,9,10

When unrecognized and untreated, OSA can be life-threatening. Despite this, OSA is not regularly screened for in clinical psychiatric practice.^6,10 Therefore, it is imperative that psychiatrists be well-acquainted with measures to identify at-risk patients and refer to a sleep specialist when appropriate.

OSA is accompanied by irritability, cognitive difficulties, and poor sleep, creating an overlap with symptoms of depressive disorders.^6,10 Use of sedative hypnotic medications, such as benzodiazepines, which further reduces muscle tone in the airway and suppresses respiratory effort, can worsen OSA symptoms^5,6,10 and pose cerebrovascular, cardiovascular, and potentially life-threatening risks, and therefore is not indicated in this population.^9,13

Obesity is a risk factor for OSA.⁶ Patients with mood disorders or schizophrenia or other psychotic disorders are at higher risk of obesity because of psychotropic-induced weight gain, stress-induced mechanisms, and/or lower levels of self-care. When these patients have unrecognized or untreated OSA and are prescribed sedative medications at night or stimulant medications during the day, they could be at increased cardiac or respiratory risks without resolving their underlying condition. A diligent psychiatrist can dramatically reduce the risks by referring a patient for nocturnal polysomnography,¹ helping the patient implement lifestyle modifications (eg, exercise, weight loss, and healthy nutrition), prescribing judiciously, and monitoring closely for such risks. An accurate diagnosis of and treatment for OSA can improve sleep⁶ dramatically and help depressive symptoms through better sleep, more daytime energy and concentration, and adequate oxygenation of the brain while sleeping.

Psychiatrists can screen for OSA using the STOP-Bang (Snoring, Tired, Observed apnea, Pressure, Body mass index, Age, Neck circumference, Gender) Questionnaire, which is a quick, 8-item screening scale that helps to categorize OSA risk as mild, moderate, or severe.¹² Hypertension, snoring, and/or gasping for breath (“observed apnea”)—a history which often is provided by spouses or significant others—daytime dozing and/or tiredness, having a large neck circumference or volume, body mass index, male sex, and age are items on the STOP-Bang Questionnaire and also are features that should raise high clinical suspicion of OSA.¹² Referral for nocturnal polysomnography in at-risk patients should be the next step^1,5 in any sleep-related breathing disorder.

Treatment for OSA involves continuous positive airway pressure (CPAP) therapy, which has been shown to relieve OSA and decrease related EDS.^5,6 Other treatment modalities, such as oral appliances and surgery, may be used⁵ in some cases, but more studies are needed for conclusive results.

Several studies have shown improved depression, mood, and cognition after administering treatment such as CPAP^6,9,14 in patients with OSA and depression. Considering the significant risks of cardiovascular,⁸ cerebrovascular,⁸ and overall morbidity and mortality associated with untreated OSA,¹² it is important to routinely screen for sleep-disordered breathing in patients with depression⁹ or other psychiatric disorders and refer for specialized sleep evaluation and treatment, when indicated.

Medications. EDS can result from some prescription and over-the-counter medications.^1,2,5,7 Sedating antidepressants, antihistamines, antipsychotics, anticonvulsants,^1,8 and beta blockers² could cause sedation, which can persist during daytime, although a few studies did not find an association between antipsychotic use and EDS.³ Benzodiazepines and other sedative-hypnotics,^1,7 especially long-acting agents or higher dosages,⁵ can lead to EDS and decreased alertness. Non-psychotropics, such as opioid pain medications,^1,7 anti­tussives, and skeletal muscle relaxants, also can contribute to or cause daytime sedation.⁷ When using these agents, psychiatrists should monitor and routinely assess patients while aiming for the lowest effective dosage when feasible.

This strategy creates a framework for psychiatrists to routinely educate patients about these commonly encountered side effects, reduce polypharmacy when possible, and help patients effectively manage or prevent these adverse effects.

Depression.¹ Some studies found >45% patients with depression had EDS.^3,13,15 Besides an association between depression and EDS,^13,16 Chellappa and Araújo¹³ also found a significant association between EDS and suicidal ideation. The causes of EDS in patients with depression may be varied, ranging from restless legs syndrome, residual depressive symptoms,¹⁵ to OSA. Depression is often comorbid with OSA,⁶ with up to 20% of patients with depression suffering from OSA,¹⁰ creating higher risk for EDS. Depressive disorders are routinely assessed during an evaluation of OSA at sleep centers, but OSA often is not screened in psychiatric practice.¹⁰

There is a strong need for regular screening for OSA in patients with depression, particularly because most studies show a link between the 2 conditions.¹⁰ Both depression and OSA have some common risk factors, such as obesity, hypertension, and metabolic syndrome.¹⁰ Patients with these conditions are at greater risk for OSA, and therefore a psychiatrist should proactively screen and refer such patients for nocturnal polysomnography when they suspect OSA. Patients with OSA and depression often present to the psychiatrist with depressive symptoms that appear to be resistant to pharmacological treatment,¹⁰ therefore underscoring the importance of screening and ruling out OSA in patients with depression.

Circadian rhythm disorders, restless legs syndrome, alcohol and other substance use, and use of prescription sedative-hypnotics are more common in patients with depression; therefore, this population is at high risk for EDS.

Circadian rhythm disorders and insufficient sleep syndrome. Insufficient sleep syndrome^1,2,8 frequently causes EDS and occurs more commonly in busy people who try to get by with less sleep.⁸ Over time, the effect of sleep loss is cumulative and can be accompanied by mood symptoms, such as irritability, fatigue, and problems with concentration.⁸ Shift workers^1,8 commonly experience insufficient sleep as well as circadian rhythm disorders and EDS. Modafinil is FDA-approved for EDS in shift work sleep disorder.

Geriatric patients may experience advanced sleep phase syndrome involving early awakenings.⁸ Adolescents, on the other hand, often suffer from delayed sleep phase syndrome, which is a type of circadian rhythm disorder, related to increasing academic and social pressures, natural pubertal shift to later sleep onset, pervading technology use, and often nebulous bedtime routines. This can be a cause of sleep persisting into daytime.⁸ Taking a careful history and a sleep diary may be useful because this disorder might be confused for insomnia. Treatment involves gradual shifting of the time of sleep onset through bright light exposure and other modalities.⁸

Adolescents might not be forthcoming about the severity of their sleep problems; therefore, psychiatrists should screen proactively through clinical interviews of patients and parents and consider this possibility when encountering an adolescent with recent-onset attention or cognitive difficulties.

Treatment for circadian rhythm disorders usually includes planned or prescribed sleep scheduling, timed light exposure,⁸ and occasional use of melatonin or other sedative agents.¹⁷

Hypersomnia of central origin, which includes narcolepsy, idiopathic hypersomnia, and recurrent hypersomnia, can present with EDS.^1,18,19 Narcolepsy is a rare, debilitating sleep disorder that manifests as EDS or sleep attacks, with or without cataplexy, and sleep paralysis.^5,8,18,19 The Multiple Sleep Latency Test and polysomnography are used for diagnosis.^1,5 Shortened REM latency is a classic finding often noted on polysomnography. Treatment involves pharmacologic and behavioral strategies and education.^5,8 Modafinil is FDA-approved for EDS associated with narcolepsy. Stimulant medications have been used for narcolepsy in the past; further studies are needed to establish benefit–risk ratio of use in this population.¹⁸

Kleine-Levin syndrome is a form of recurrent hypersomnia, a less common sleep disorder, characterized by episodes of excessive sleepiness accompanied by hyperphagia and hypersexuality.^5,18,19

Other medical conditions,¹ such as the rare familial fatal insomnia, neurological conditions¹ such as encephalitis,⁸ epilepsy,⁸ Alzheimer’s disease or other types of dementia,⁸ Parkinson’s disease,¹ or multiple sclerosis,^1,18 can cause excessive daytime fatigue by causing secondary insomnia or hypersomnia.

Treating the underlying disorder is an important first step in these cases. In addition, coordinating with neurologists or other specialists involved in caring for patients with these conditions is important. Regularly reviewing and simplifying the often complex medication regimen, when possible, can go a long way in mitigating EDS in this population.

Other disorders affecting sleep. Restless legs syndrome and periodic limb movement disorder are other causes of EDS.³ Treatment involves lifestyle modifications, iron supplementation in certain patients, and use of dopaminergic agents such as ropinirole, pramipexole, and other medications, depending on severity of the condition, comorbidities, and other factors.²⁰

Alcohol or substance use. Substance use or withdrawal can be associated with sleep disorders, such as hypersomnia,¹⁹ insomnia,¹⁹ and related EDS.⁵ For example, alcohol use disorder affects REM sleep, and can cause EDS. Secondary central apnea can be the result of long-standing opioid use¹⁹ and can present like EDS.

Insomnia. Primary insomnia rarely causes EDS.⁵ Insomnia due to a medical or psychiatric condition may be an indirect cause of EDS by causing sleep deprivation.

Steps for timely and accurate diagnosis

Utilize the following steps for facilitating timely diagnosis and treatment of EDS:

Thorough history. Patients often describe “tiredness” instead of sleepiness.⁸ Therefore, the astute psychiatrist should explore further when patients are presenting with this concern, especially by asking more specific questions such as the tendency to doze off during daytime.⁸

Family members can be vital sources for obtaining a complete history,⁵ especially because patients might deny,⁸ minimize, or not be fully aware¹ of the extent of their symptoms. Asking family members about patient’s snoring, irregular breathing, or gasping at night can be particularly valuable.⁵ Obtaining a family history of sleep disorders can be particularly important, especially in conditions such as OSA and narcolepsy.

Asking about any history of safety issues,⁸ including sleepiness during driving, cooking, or other activities, is also important.

Use of scales and other screening measures. Psychiatrists can use initial screening measures in the office setting. Epworth Sleepiness Scale^15,21 is a validated,² short, self-administered measure to assess the level of daytime sleepiness; however, it has some limitations such as not being able to measure changes in sleepiness from hour to hour or day to day. Because of its limitations, the Epworth Sleepiness Scale should not be used by itself as a diagnostic tool.³ It has been commonly used for detecting OSA² and narcolepsy. The Stanford Sleepiness Scale is a self-rating scale that measures the subjective degree of sleepiness and alertness; it has limitations as well, such as having little correlation with chronic sleep loss.⁸ Other tools such as visual analogue scales also could be helpful.⁸ For more specialized testing, such as Multiple Sleep Latency Test or polysomnography, referral to a sleep specialist is ideal.⁸

Education. The assessment is an opportunity for the psychiatrist to educate patients about sleep hygiene, the importance of regular bedtimes, and getting adequate sleep to avoid accumulating a sleep deficit.

Urgent referral of at-risk populations. Prompt or urgent referral of at-risk populations, such as geriatric patients or those with a history of dozing off during driving, is invaluable in preventing morbidity and mortality from untreated sleep disorders.

Patients with severe daytime sleepiness should be advised to not drive or operate heavy machinery until this condition is adequately controlled.¹⁸

Bottom Line

Excessive daytime sleepiness (EDS) is “the inability to maintain wakefulness and alertness during the major waking periods of the day, with sleep occurring unintentionally or at inappropriate times, almost daily for at least 3 months,” according to the American Academy of Sleep Medicine.¹ EDS is common, with a prevalence up to 25% to 30% in the general population.^1-4 The prevalence rate varies in different studies, primarily because of inconsistent definitions of EDS, and therefore differences in diagnosis and assessment.^1,2,4 In a study of 300 psychiatric outpatients, 34% had EDS.³ However, studies and evidence reviewing EDS in psychiatric patients are limited.

EDS vs fatigue

Many patients describe EDS as “fatigue”¹; however, a patient’s report of fatigue could be mistaken for EDS.⁴ Although there is overlap, it is important for physicians to distinguish between these 2 entities for accurate identification and treatment.^1,4

Risk of inadequate screening

A study of 117 patients with symptomatic coronary artery disease showed that EDS is associated with significantly greater incidence of cardiovascular adverse events at 16-month follow up.² This study had limitations such as small sample size; therefore, more studies are needed. Because of these risks, timely and accurate diagnosis not only improves the patient’s quality of life and reduces polypharmacy but also can be life-saving.

Common causes of EDS in psychiatric patients

Because of the high prevalence and severity of impairments caused by EDS, it is essential for psychiatrists to be informed about causes of EDS and thoroughly assess for the potential underlying etiology before concluding that the sleep problem is a manifestation of the psychiatric disorder and prescribing psychotropic medication for it.

Some common causes of EDS in psychiatric patients include:

Sleep-disordered breathing.⁸ Obstructive sleep apnea (OSA) is often underdiagnosed,^6,7 and considering how common it is,⁶ psychiatrists likely will see many patients with OSA in their practice.⁵ OSA has a higher prevalence among patients with psychiatric disorders such as depression^6,9 and schizophrenia. Additionally, there is evidence suggesting that patients with OSA are more likely to suffer from depression and EDS than healthy controls^6,9,10; some of the proposed mechanisms are sleep fragmentation and hypoxemia.^6,9-11 OSA is the most common form of sleep-disordered breathing and is a common cause of EDS.^1,2,12 Also, undiagnosed and untreated OSA in patients with depression could cause refractoriness to pharmacological treatment of depression.^6,9,10

When unrecognized and untreated, OSA can be life-threatening. Despite this, OSA is not regularly screened for in clinical psychiatric practice.^6,10 Therefore, it is imperative that psychiatrists be well-acquainted with measures to identify at-risk patients and refer to a sleep specialist when appropriate.

OSA is accompanied by irritability, cognitive difficulties, and poor sleep, creating an overlap with symptoms of depressive disorders.^6,10 Use of sedative hypnotic medications, such as benzodiazepines, which further reduces muscle tone in the airway and suppresses respiratory effort, can worsen OSA symptoms^5,6,10 and pose cerebrovascular, cardiovascular, and potentially life-threatening risks, and therefore is not indicated in this population.^9,13

Obesity is a risk factor for OSA.⁶ Patients with mood disorders or schizophrenia or other psychotic disorders are at higher risk of obesity because of psychotropic-induced weight gain, stress-induced mechanisms, and/or lower levels of self-care. When these patients have unrecognized or untreated OSA and are prescribed sedative medications at night or stimulant medications during the day, they could be at increased cardiac or respiratory risks without resolving their underlying condition. A diligent psychiatrist can dramatically reduce the risks by referring a patient for nocturnal polysomnography,¹ helping the patient implement lifestyle modifications (eg, exercise, weight loss, and healthy nutrition), prescribing judiciously, and monitoring closely for such risks. An accurate diagnosis of and treatment for OSA can improve sleep⁶ dramatically and help depressive symptoms through better sleep, more daytime energy and concentration, and adequate oxygenation of the brain while sleeping.

Psychiatrists can screen for OSA using the STOP-Bang (Snoring, Tired, Observed apnea, Pressure, Body mass index, Age, Neck circumference, Gender) Questionnaire, which is a quick, 8-item screening scale that helps to categorize OSA risk as mild, moderate, or severe.¹² Hypertension, snoring, and/or gasping for breath (“observed apnea”)—a history which often is provided by spouses or significant others—daytime dozing and/or tiredness, having a large neck circumference or volume, body mass index, male sex, and age are items on the STOP-Bang Questionnaire and also are features that should raise high clinical suspicion of OSA.¹² Referral for nocturnal polysomnography in at-risk patients should be the next step^1,5 in any sleep-related breathing disorder.

Treatment for OSA involves continuous positive airway pressure (CPAP) therapy, which has been shown to relieve OSA and decrease related EDS.^5,6 Other treatment modalities, such as oral appliances and surgery, may be used⁵ in some cases, but more studies are needed for conclusive results.

Several studies have shown improved depression, mood, and cognition after administering treatment such as CPAP^6,9,14 in patients with OSA and depression. Considering the significant risks of cardiovascular,⁸ cerebrovascular,⁸ and overall morbidity and mortality associated with untreated OSA,¹² it is important to routinely screen for sleep-disordered breathing in patients with depression⁹ or other psychiatric disorders and refer for specialized sleep evaluation and treatment, when indicated.

Medications. EDS can result from some prescription and over-the-counter medications.^1,2,5,7 Sedating antidepressants, antihistamines, antipsychotics, anticonvulsants,^1,8 and beta blockers² could cause sedation, which can persist during daytime, although a few studies did not find an association between antipsychotic use and EDS.³ Benzodiazepines and other sedative-hypnotics,^1,7 especially long-acting agents or higher dosages,⁵ can lead to EDS and decreased alertness. Non-psychotropics, such as opioid pain medications,^1,7 anti­tussives, and skeletal muscle relaxants, also can contribute to or cause daytime sedation.⁷ When using these agents, psychiatrists should monitor and routinely assess patients while aiming for the lowest effective dosage when feasible.

This strategy creates a framework for psychiatrists to routinely educate patients about these commonly encountered side effects, reduce polypharmacy when possible, and help patients effectively manage or prevent these adverse effects.

Depression.¹ Some studies found >45% patients with depression had EDS.^3,13,15 Besides an association between depression and EDS,^13,16 Chellappa and Araújo¹³ also found a significant association between EDS and suicidal ideation. The causes of EDS in patients with depression may be varied, ranging from restless legs syndrome, residual depressive symptoms,¹⁵ to OSA. Depression is often comorbid with OSA,⁶ with up to 20% of patients with depression suffering from OSA,¹⁰ creating higher risk for EDS. Depressive disorders are routinely assessed during an evaluation of OSA at sleep centers, but OSA often is not screened in psychiatric practice.¹⁰

There is a strong need for regular screening for OSA in patients with depression, particularly because most studies show a link between the 2 conditions.¹⁰ Both depression and OSA have some common risk factors, such as obesity, hypertension, and metabolic syndrome.¹⁰ Patients with these conditions are at greater risk for OSA, and therefore a psychiatrist should proactively screen and refer such patients for nocturnal polysomnography when they suspect OSA. Patients with OSA and depression often present to the psychiatrist with depressive symptoms that appear to be resistant to pharmacological treatment,¹⁰ therefore underscoring the importance of screening and ruling out OSA in patients with depression.

Circadian rhythm disorders, restless legs syndrome, alcohol and other substance use, and use of prescription sedative-hypnotics are more common in patients with depression; therefore, this population is at high risk for EDS.

Circadian rhythm disorders and insufficient sleep syndrome. Insufficient sleep syndrome^1,2,8 frequently causes EDS and occurs more commonly in busy people who try to get by with less sleep.⁸ Over time, the effect of sleep loss is cumulative and can be accompanied by mood symptoms, such as irritability, fatigue, and problems with concentration.⁸ Shift workers^1,8 commonly experience insufficient sleep as well as circadian rhythm disorders and EDS. Modafinil is FDA-approved for EDS in shift work sleep disorder.

Geriatric patients may experience advanced sleep phase syndrome involving early awakenings.⁸ Adolescents, on the other hand, often suffer from delayed sleep phase syndrome, which is a type of circadian rhythm disorder, related to increasing academic and social pressures, natural pubertal shift to later sleep onset, pervading technology use, and often nebulous bedtime routines. This can be a cause of sleep persisting into daytime.⁸ Taking a careful history and a sleep diary may be useful because this disorder might be confused for insomnia. Treatment involves gradual shifting of the time of sleep onset through bright light exposure and other modalities.⁸

Adolescents might not be forthcoming about the severity of their sleep problems; therefore, psychiatrists should screen proactively through clinical interviews of patients and parents and consider this possibility when encountering an adolescent with recent-onset attention or cognitive difficulties.

Treatment for circadian rhythm disorders usually includes planned or prescribed sleep scheduling, timed light exposure,⁸ and occasional use of melatonin or other sedative agents.¹⁷

Hypersomnia of central origin, which includes narcolepsy, idiopathic hypersomnia, and recurrent hypersomnia, can present with EDS.^1,18,19 Narcolepsy is a rare, debilitating sleep disorder that manifests as EDS or sleep attacks, with or without cataplexy, and sleep paralysis.^5,8,18,19 The Multiple Sleep Latency Test and polysomnography are used for diagnosis.^1,5 Shortened REM latency is a classic finding often noted on polysomnography. Treatment involves pharmacologic and behavioral strategies and education.^5,8 Modafinil is FDA-approved for EDS associated with narcolepsy. Stimulant medications have been used for narcolepsy in the past; further studies are needed to establish benefit–risk ratio of use in this population.¹⁸

Kleine-Levin syndrome is a form of recurrent hypersomnia, a less common sleep disorder, characterized by episodes of excessive sleepiness accompanied by hyperphagia and hypersexuality.^5,18,19

Other medical conditions,¹ such as the rare familial fatal insomnia, neurological conditions¹ such as encephalitis,⁸ epilepsy,⁸ Alzheimer’s disease or other types of dementia,⁸ Parkinson’s disease,¹ or multiple sclerosis,^1,18 can cause excessive daytime fatigue by causing secondary insomnia or hypersomnia.

Treating the underlying disorder is an important first step in these cases. In addition, coordinating with neurologists or other specialists involved in caring for patients with these conditions is important. Regularly reviewing and simplifying the often complex medication regimen, when possible, can go a long way in mitigating EDS in this population.

Other disorders affecting sleep. Restless legs syndrome and periodic limb movement disorder are other causes of EDS.³ Treatment involves lifestyle modifications, iron supplementation in certain patients, and use of dopaminergic agents such as ropinirole, pramipexole, and other medications, depending on severity of the condition, comorbidities, and other factors.²⁰

Alcohol or substance use. Substance use or withdrawal can be associated with sleep disorders, such as hypersomnia,¹⁹ insomnia,¹⁹ and related EDS.⁵ For example, alcohol use disorder affects REM sleep, and can cause EDS. Secondary central apnea can be the result of long-standing opioid use¹⁹ and can present like EDS.

Insomnia. Primary insomnia rarely causes EDS.⁵ Insomnia due to a medical or psychiatric condition may be an indirect cause of EDS by causing sleep deprivation.

Steps for timely and accurate diagnosis

Utilize the following steps for facilitating timely diagnosis and treatment of EDS:

Thorough history. Patients often describe “tiredness” instead of sleepiness.⁸ Therefore, the astute psychiatrist should explore further when patients are presenting with this concern, especially by asking more specific questions such as the tendency to doze off during daytime.⁸

Family members can be vital sources for obtaining a complete history,⁵ especially because patients might deny,⁸ minimize, or not be fully aware¹ of the extent of their symptoms. Asking family members about patient’s snoring, irregular breathing, or gasping at night can be particularly valuable.⁵ Obtaining a family history of sleep disorders can be particularly important, especially in conditions such as OSA and narcolepsy.

Asking about any history of safety issues,⁸ including sleepiness during driving, cooking, or other activities, is also important.

Use of scales and other screening measures. Psychiatrists can use initial screening measures in the office setting. Epworth Sleepiness Scale^15,21 is a validated,² short, self-administered measure to assess the level of daytime sleepiness; however, it has some limitations such as not being able to measure changes in sleepiness from hour to hour or day to day. Because of its limitations, the Epworth Sleepiness Scale should not be used by itself as a diagnostic tool.³ It has been commonly used for detecting OSA² and narcolepsy. The Stanford Sleepiness Scale is a self-rating scale that measures the subjective degree of sleepiness and alertness; it has limitations as well, such as having little correlation with chronic sleep loss.⁸ Other tools such as visual analogue scales also could be helpful.⁸ For more specialized testing, such as Multiple Sleep Latency Test or polysomnography, referral to a sleep specialist is ideal.⁸

Education. The assessment is an opportunity for the psychiatrist to educate patients about sleep hygiene, the importance of regular bedtimes, and getting adequate sleep to avoid accumulating a sleep deficit.

Urgent referral of at-risk populations. Prompt or urgent referral of at-risk populations, such as geriatric patients or those with a history of dozing off during driving, is invaluable in preventing morbidity and mortality from untreated sleep disorders.

Patients with severe daytime sleepiness should be advised to not drive or operate heavy machinery until this condition is adequately controlled.¹⁸

Bottom Line

Excessive daytime sleepiness (EDS) is “the inability to maintain wakefulness and alertness during the major waking periods of the day, with sleep occurring unintentionally or at inappropriate times, almost daily for at least 3 months,” according to the American Academy of Sleep Medicine.¹ EDS is common, with a prevalence up to 25% to 30% in the general population.^1-4 The prevalence rate varies in different studies, primarily because of inconsistent definitions of EDS, and therefore differences in diagnosis and assessment.^1,2,4 In a study of 300 psychiatric outpatients, 34% had EDS.³ However, studies and evidence reviewing EDS in psychiatric patients are limited.

EDS vs fatigue

Many patients describe EDS as “fatigue”¹; however, a patient’s report of fatigue could be mistaken for EDS.⁴ Although there is overlap, it is important for physicians to distinguish between these 2 entities for accurate identification and treatment.^1,4

Risk of inadequate screening

A study of 117 patients with symptomatic coronary artery disease showed that EDS is associated with significantly greater incidence of cardiovascular adverse events at 16-month follow up.² This study had limitations such as small sample size; therefore, more studies are needed. Because of these risks, timely and accurate diagnosis not only improves the patient’s quality of life and reduces polypharmacy but also can be life-saving.

Common causes of EDS in psychiatric patients

Because of the high prevalence and severity of impairments caused by EDS, it is essential for psychiatrists to be informed about causes of EDS and thoroughly assess for the potential underlying etiology before concluding that the sleep problem is a manifestation of the psychiatric disorder and prescribing psychotropic medication for it.

Some common causes of EDS in psychiatric patients include:

Sleep-disordered breathing.⁸ Obstructive sleep apnea (OSA) is often underdiagnosed,^6,7 and considering how common it is,⁶ psychiatrists likely will see many patients with OSA in their practice.⁵ OSA has a higher prevalence among patients with psychiatric disorders such as depression^6,9 and schizophrenia. Additionally, there is evidence suggesting that patients with OSA are more likely to suffer from depression and EDS than healthy controls^6,9,10; some of the proposed mechanisms are sleep fragmentation and hypoxemia.^6,9-11 OSA is the most common form of sleep-disordered breathing and is a common cause of EDS.^1,2,12 Also, undiagnosed and untreated OSA in patients with depression could cause refractoriness to pharmacological treatment of depression.^6,9,10

When unrecognized and untreated, OSA can be life-threatening. Despite this, OSA is not regularly screened for in clinical psychiatric practice.^6,10 Therefore, it is imperative that psychiatrists be well-acquainted with measures to identify at-risk patients and refer to a sleep specialist when appropriate.

OSA is accompanied by irritability, cognitive difficulties, and poor sleep, creating an overlap with symptoms of depressive disorders.^6,10 Use of sedative hypnotic medications, such as benzodiazepines, which further reduces muscle tone in the airway and suppresses respiratory effort, can worsen OSA symptoms^5,6,10 and pose cerebrovascular, cardiovascular, and potentially life-threatening risks, and therefore is not indicated in this population.^9,13

Obesity is a risk factor for OSA.⁶ Patients with mood disorders or schizophrenia or other psychotic disorders are at higher risk of obesity because of psychotropic-induced weight gain, stress-induced mechanisms, and/or lower levels of self-care. When these patients have unrecognized or untreated OSA and are prescribed sedative medications at night or stimulant medications during the day, they could be at increased cardiac or respiratory risks without resolving their underlying condition. A diligent psychiatrist can dramatically reduce the risks by referring a patient for nocturnal polysomnography,¹ helping the patient implement lifestyle modifications (eg, exercise, weight loss, and healthy nutrition), prescribing judiciously, and monitoring closely for such risks. An accurate diagnosis of and treatment for OSA can improve sleep⁶ dramatically and help depressive symptoms through better sleep, more daytime energy and concentration, and adequate oxygenation of the brain while sleeping.

Psychiatrists can screen for OSA using the STOP-Bang (Snoring, Tired, Observed apnea, Pressure, Body mass index, Age, Neck circumference, Gender) Questionnaire, which is a quick, 8-item screening scale that helps to categorize OSA risk as mild, moderate, or severe.¹² Hypertension, snoring, and/or gasping for breath (“observed apnea”)—a history which often is provided by spouses or significant others—daytime dozing and/or tiredness, having a large neck circumference or volume, body mass index, male sex, and age are items on the STOP-Bang Questionnaire and also are features that should raise high clinical suspicion of OSA.¹² Referral for nocturnal polysomnography in at-risk patients should be the next step^1,5 in any sleep-related breathing disorder.

Treatment for OSA involves continuous positive airway pressure (CPAP) therapy, which has been shown to relieve OSA and decrease related EDS.^5,6 Other treatment modalities, such as oral appliances and surgery, may be used⁵ in some cases, but more studies are needed for conclusive results.

Several studies have shown improved depression, mood, and cognition after administering treatment such as CPAP^6,9,14 in patients with OSA and depression. Considering the significant risks of cardiovascular,⁸ cerebrovascular,⁸ and overall morbidity and mortality associated with untreated OSA,¹² it is important to routinely screen for sleep-disordered breathing in patients with depression⁹ or other psychiatric disorders and refer for specialized sleep evaluation and treatment, when indicated.

Medications. EDS can result from some prescription and over-the-counter medications.^1,2,5,7 Sedating antidepressants, antihistamines, antipsychotics, anticonvulsants,^1,8 and beta blockers² could cause sedation, which can persist during daytime, although a few studies did not find an association between antipsychotic use and EDS.³ Benzodiazepines and other sedative-hypnotics,^1,7 especially long-acting agents or higher dosages,⁵ can lead to EDS and decreased alertness. Non-psychotropics, such as opioid pain medications,^1,7 anti­tussives, and skeletal muscle relaxants, also can contribute to or cause daytime sedation.⁷ When using these agents, psychiatrists should monitor and routinely assess patients while aiming for the lowest effective dosage when feasible.

This strategy creates a framework for psychiatrists to routinely educate patients about these commonly encountered side effects, reduce polypharmacy when possible, and help patients effectively manage or prevent these adverse effects.

Depression.¹ Some studies found >45% patients with depression had EDS.^3,13,15 Besides an association between depression and EDS,^13,16 Chellappa and Araújo¹³ also found a significant association between EDS and suicidal ideation. The causes of EDS in patients with depression may be varied, ranging from restless legs syndrome, residual depressive symptoms,¹⁵ to OSA. Depression is often comorbid with OSA,⁶ with up to 20% of patients with depression suffering from OSA,¹⁰ creating higher risk for EDS. Depressive disorders are routinely assessed during an evaluation of OSA at sleep centers, but OSA often is not screened in psychiatric practice.¹⁰

There is a strong need for regular screening for OSA in patients with depression, particularly because most studies show a link between the 2 conditions.¹⁰ Both depression and OSA have some common risk factors, such as obesity, hypertension, and metabolic syndrome.¹⁰ Patients with these conditions are at greater risk for OSA, and therefore a psychiatrist should proactively screen and refer such patients for nocturnal polysomnography when they suspect OSA. Patients with OSA and depression often present to the psychiatrist with depressive symptoms that appear to be resistant to pharmacological treatment,¹⁰ therefore underscoring the importance of screening and ruling out OSA in patients with depression.

Circadian rhythm disorders, restless legs syndrome, alcohol and other substance use, and use of prescription sedative-hypnotics are more common in patients with depression; therefore, this population is at high risk for EDS.

Circadian rhythm disorders and insufficient sleep syndrome. Insufficient sleep syndrome^1,2,8 frequently causes EDS and occurs more commonly in busy people who try to get by with less sleep.⁸ Over time, the effect of sleep loss is cumulative and can be accompanied by mood symptoms, such as irritability, fatigue, and problems with concentration.⁸ Shift workers^1,8 commonly experience insufficient sleep as well as circadian rhythm disorders and EDS. Modafinil is FDA-approved for EDS in shift work sleep disorder.

Geriatric patients may experience advanced sleep phase syndrome involving early awakenings.⁸ Adolescents, on the other hand, often suffer from delayed sleep phase syndrome, which is a type of circadian rhythm disorder, related to increasing academic and social pressures, natural pubertal shift to later sleep onset, pervading technology use, and often nebulous bedtime routines. This can be a cause of sleep persisting into daytime.⁸ Taking a careful history and a sleep diary may be useful because this disorder might be confused for insomnia. Treatment involves gradual shifting of the time of sleep onset through bright light exposure and other modalities.⁸

Adolescents might not be forthcoming about the severity of their sleep problems; therefore, psychiatrists should screen proactively through clinical interviews of patients and parents and consider this possibility when encountering an adolescent with recent-onset attention or cognitive difficulties.

Treatment for circadian rhythm disorders usually includes planned or prescribed sleep scheduling, timed light exposure,⁸ and occasional use of melatonin or other sedative agents.¹⁷

Hypersomnia of central origin, which includes narcolepsy, idiopathic hypersomnia, and recurrent hypersomnia, can present with EDS.^1,18,19 Narcolepsy is a rare, debilitating sleep disorder that manifests as EDS or sleep attacks, with or without cataplexy, and sleep paralysis.^5,8,18,19 The Multiple Sleep Latency Test and polysomnography are used for diagnosis.^1,5 Shortened REM latency is a classic finding often noted on polysomnography. Treatment involves pharmacologic and behavioral strategies and education.^5,8 Modafinil is FDA-approved for EDS associated with narcolepsy. Stimulant medications have been used for narcolepsy in the past; further studies are needed to establish benefit–risk ratio of use in this population.¹⁸

Kleine-Levin syndrome is a form of recurrent hypersomnia, a less common sleep disorder, characterized by episodes of excessive sleepiness accompanied by hyperphagia and hypersexuality.^5,18,19

Other medical conditions,¹ such as the rare familial fatal insomnia, neurological conditions¹ such as encephalitis,⁸ epilepsy,⁸ Alzheimer’s disease or other types of dementia,⁸ Parkinson’s disease,¹ or multiple sclerosis,^1,18 can cause excessive daytime fatigue by causing secondary insomnia or hypersomnia.

Treating the underlying disorder is an important first step in these cases. In addition, coordinating with neurologists or other specialists involved in caring for patients with these conditions is important. Regularly reviewing and simplifying the often complex medication regimen, when possible, can go a long way in mitigating EDS in this population.

Other disorders affecting sleep. Restless legs syndrome and periodic limb movement disorder are other causes of EDS.³ Treatment involves lifestyle modifications, iron supplementation in certain patients, and use of dopaminergic agents such as ropinirole, pramipexole, and other medications, depending on severity of the condition, comorbidities, and other factors.²⁰

Alcohol or substance use. Substance use or withdrawal can be associated with sleep disorders, such as hypersomnia,¹⁹ insomnia,¹⁹ and related EDS.⁵ For example, alcohol use disorder affects REM sleep, and can cause EDS. Secondary central apnea can be the result of long-standing opioid use¹⁹ and can present like EDS.

Insomnia. Primary insomnia rarely causes EDS.⁵ Insomnia due to a medical or psychiatric condition may be an indirect cause of EDS by causing sleep deprivation.

Steps for timely and accurate diagnosis

Utilize the following steps for facilitating timely diagnosis and treatment of EDS:

Thorough history. Patients often describe “tiredness” instead of sleepiness.⁸ Therefore, the astute psychiatrist should explore further when patients are presenting with this concern, especially by asking more specific questions such as the tendency to doze off during daytime.⁸

Family members can be vital sources for obtaining a complete history,⁵ especially because patients might deny,⁸ minimize, or not be fully aware¹ of the extent of their symptoms. Asking family members about patient’s snoring, irregular breathing, or gasping at night can be particularly valuable.⁵ Obtaining a family history of sleep disorders can be particularly important, especially in conditions such as OSA and narcolepsy.

Asking about any history of safety issues,⁸ including sleepiness during driving, cooking, or other activities, is also important.

Use of scales and other screening measures. Psychiatrists can use initial screening measures in the office setting. Epworth Sleepiness Scale^15,21 is a validated,² short, self-administered measure to assess the level of daytime sleepiness; however, it has some limitations such as not being able to measure changes in sleepiness from hour to hour or day to day. Because of its limitations, the Epworth Sleepiness Scale should not be used by itself as a diagnostic tool.³ It has been commonly used for detecting OSA² and narcolepsy. The Stanford Sleepiness Scale is a self-rating scale that measures the subjective degree of sleepiness and alertness; it has limitations as well, such as having little correlation with chronic sleep loss.⁸ Other tools such as visual analogue scales also could be helpful.⁸ For more specialized testing, such as Multiple Sleep Latency Test or polysomnography, referral to a sleep specialist is ideal.⁸

Education. The assessment is an opportunity for the psychiatrist to educate patients about sleep hygiene, the importance of regular bedtimes, and getting adequate sleep to avoid accumulating a sleep deficit.

Urgent referral of at-risk populations. Prompt or urgent referral of at-risk populations, such as geriatric patients or those with a history of dozing off during driving, is invaluable in preventing morbidity and mortality from untreated sleep disorders.

Patients with severe daytime sleepiness should be advised to not drive or operate heavy machinery until this condition is adequately controlled.¹⁸

Bottom Line