Current Psychiatry

Suvorexant, FDA-approved to treat insomnia, has demonstrated efficacy in helping patients with insomnia improve their ability to fall asleep and remain asleep (Table 1).¹ This first-in-class compound represents a novel mechanism of action to promoting sleep that may avoid some prob­lems associated with other hypnotics.²

Because orexin antagonists have not been used outside of clinical trials, it is too soon to tell whether suvorexant will have the ideal real-world efficacy and safety profile to make it a first-line treatment for insomnia patients, or if it will be reserved for those who have failed a trial of several other treatments.⁴

In theory, the orexin antagonist approach to treating insomnia could represent a major advance that modulates the fundamental pathology of the disorder.⁵ The syndrome of chronic insomnia encompasses not just the nighttime sleep disturbance but also an assort­ment of daytime symptoms that can include fatigue, poor concentration, irritability, and decreased school or work performance but usually not sleepiness. This constellation of nighttime and daytime symptoms could be conceptualized as a manifestation of persis­tent CNS hyperarousal. Because the orexin system promotes and reinforces arousal, per­haps an orexin antagonist that dampens the level of orexin activity will ameliorate the full spectrum of insomnia symptoms—not sim­ply sedate patients.⁶

How suvorexant works
Suvorexant is a potent and reversible dual orexin-receptor antagonist. The orexin system, first described in 1998, has a key role in promoting and stabilizing wake­fulness.⁷ Evidence suggests that people with chronic insomnia exhibit a central hyperarousal that perpetuates their sleep difficulty. Accordingly, a targeted phar­maceutical approach that reduces orexin activity should facilitate sleep onset and sleep maintenance for these patients. It is well known that the regulation of sleep and wakefulness depends on the interaction of multiple nuclei within the hypothalamus. Orexinergic neurons in the perifornical-lateral hypothalamic region project widely in the CNS and have especially dense con­nections with wake-promoting cholinergic, serotonergic, noradrenergic, and histamin­ergic neurons.⁶

A precursor prepro-orexin peptide is split into 2 orexin neurotransmitters (orexin A and orexin B). These 2 orexins bind with 2 G-protein-coupled receptors (OX1R and OX2R) that have both overlapping and distinct distributions.⁷ Suvorexant is highly selective and has similar affinity for OX1R and OX2R, functioning as an antag­onist for both.⁸ Fundamentally, suvorexant enhances sleep by dampening the arous­ing wake drive.

Pharmacokinetics
Suvorexant is available as an immediate-release tablet with pharmacokinetic prop­erties that offer benefits for sleep onset and maintenance.⁹ Ingestion under fasting conditions results in a median time to maxi­mum concentration (T_max) of approximately 2 hours, although the T_max values vary widely from patient to patient (range 30 minutes to 6 hours). Although suvorexant can be taken with food, there is a modest absorption delay after a high-fat meal, resulting in a further T_max delay of approximately 1.5 hours.

Suvorexant is primarily metabolized through the cytochrome P450 (CYP) 3A path­way, with limited contribution by CYP2C19. There are no active metabolites. The suvorex­ant blood level and risk of side effects will be higher with concomitant use of CYP3A inhibitors. The drug should not be adminis­tered with strong CYP3A inhibitors; the ini­tial dosage should be reduced with moderate CYP3A inhibitors. Concomitant use of strong CYP3A inducers can result in a low suvorex­ant level and reduced efficacy.

Suvorexant has little effect on other med­ications, although a person taking digoxin might experience intestinal P-glycoprotein inhibition with a slight rise in the digoxin level. In a patient taking both medica­tions, monitoring of the digoxin level is recommended.

The elimination half-life of suvorexant is approximately 12 hours, with a steady state in approximately 3 days. Because the half-life of suvorexant is moderately long for a sleep-promoting medication, use of the drug might be associated with residual sleepiness the morning after bedtime dosing. The risk for next-morning sleepiness or impairment should be minimized, however, when using the recommended dosages. Elimination is approximately two-thirds through feces and one-third in the urine.

Suvorexant metabolism can be affected by sex and body mass index. Females and obese people have a modestly elevated expo­sure to suvorexant, as reflected by the area under the curve and maximum concentra­tion (C_max). These patients might not require dosage adjustments unless they are obese and female, in which case they should take a lower dosage.

Age and race have not been shown to influence suvorexant metabolism to a signifi­cant degree. Patients with renal impairment and those with mild or moderate hepatic impairment do not need dosage adjust­ment. Suvorexant has not been evaluated in patients with severe hepatic impairment.

Efficacy
Suvorexant showed significant evidence of improved sleep onset and sleep maintenance in patients with insomnia in clinical trials. The key efficacy clinical trials with insomnia patients included a phase-IIb dose-finding study,¹⁰ 2 similar 3-month phase-III studies,¹¹ and one 12-month phase-III safety study that incorporated efficacy outcomes.¹² All these trials included subjective sleep measures and all except for the long-term safety study also incorporated polysomnographic assess­ment. The specific sleep laboratory outcomes were latency to persistent sleep (LPS), wake after the onset of persistent sleep (WASO), total sleep time (TST), and sleep efficiency (SE). Subjective sleep outcomes were time to sleep onset (sTSO), wake after sleep onset (sWASO), and total sleep time (sTST). Other exploratory endpoints also were assessed. These efficacy and safety studies mostly were performed at dosages considerably higher than those approved by the FDA.

The dose-finding (phase-IIb) trial was conducted with non-geriatric (age 18 to 64) patients with insomnia in a random­ized, double-blind, crossover design of two 4-week periods with subjects given a nightly placebo or suvorexant (10 mg, 20 mg, 40 mg, or 80 mg).¹⁰ Each of the 4 groups included approximately 60 subjects. The 2 co-primary endpoints were SE at Night 1 and the end of Week 4; secondary endpoints were LPS and WASO. Suvorexant was associated with dos­age-related improvements in SE and WASO compared with placebo at both time points. Carryover effects from the period-1 active drug group complicated the analysis of LPS.

The phase-III efficacy and safety trials were performed with 40 mg high dosage (HD) and 20 mg low dosage (LD) groups for adults and with 30 mg HD and 15 mg LD groups for geriatric (age ≥65) patients.¹¹ Two similarly designed 3-month randomized, double-blind, placebo-controlled pivotal efficacy studies assessed objective and sub­jective sleep measures in 4 groups with non-geriatric (HD and LD) and geriatric (HD and LD) insomnia patients.

After baseline assessment, patients took nightly bedtime doses of placebo; suvorexant, 40 mg or 20 mg (non-geriatric individuals); or suvorexant, 30 mg or 15 mg (geriatric indi­viduals). All subjects kept a daily electronic diary and had polysomnographic recordings performed on Night 1, at the end of Month 1, and at the end of Month 3. Both the indi­vidual studies and combined analyses (2,030 subjects) showed that, in non-geriatric and geriatric patients, HD suvorexant resulted in significantly greater improvement in key subjective and objective measures through­out the study (Table 2,⁹ and Table 3,⁹), with the exception of a single LPS outcome in 1 study, compared with placebo. The LD dosages also demonstrated efficacy, but to a reduced extent.

Subjective sleep outcomes were assessed in a 1-year randomized, placebo-controlled trial with nightly placebo, suvorexant, 40 mg, for non-geriatric, or suvorexant, 30 mg, for geriatric insomnia patients.12 The 1-year phase was completed with 484 subjects. Key efficacy outcomes were sTST and sTSO changes from baseline during the first month of treatment. Compared with placebo, suvorexant dosages demonstrated significantly greater efficacy, improvements that were sustained throughout the year.

Clinical trials found suvorexant to be gen­erally safe and well tolerated.¹³ However, specific safety concerns led the FDA to approve the medication at dosages lower than those assessed in the phase-III studies.¹

Somnolence was the most common adverse event in clinical trials. In the phase- IIb dose-finding study, somnolence was reported in <1% in the placebo group, but was associated with suvorexant in 2% of the 10 mg group, 5% with 20 mg, 12% with 40 mg, and 11% with 80 mg.⁹ In the phase-III combined analysis of the 3-month studies, somnolence was reported by 3% in the placebo group and 7% of non-geriatric patients taking 20 mg or geriatric patients taking 15 mg. Somnolence was reported in 8% of women and 3% of men taking the 15 mg or 20 mg dosage in these stud­ies. The 1-year study was performed only with higher suvorexant dosages (30 mg and 40 mg), in comparison with placebo. In this long-term trial, somnolence was reported by 13% of subjects taking suvorexant and 3% taking placebo.

Additional safety issues in trials included excessive daytime sleepiness, impaired driv­ing, suicidal ideation, sleep paralysis, hyp­nagogic/hypnopompic hallucinations, and cataplexy-like symptoms.⁹ Occurrences of these events are rare but have been reported more often among patients taking suvorex­ant than among those taking placebo.

Unique clinical issues
The U.S. Drug Enforcement Agency has categorized suvorexant as a Schedule IV controlled substance. Although there is no evidence of physiological dependence or withdrawal symptoms with suvorexant, studies with recreational substance abusers have shown that the likeability rating is simi­lar to that of zolpidem.¹³

Contraindication
Suvorexant is contraindicated in patients with narcolepsy.⁹ The underlying pathol­ogy of narcolepsy involves a marked reduction in orexin functioning with corre­sponding excessive sleepiness and related symptoms, such as cataplexy, hypnago­gic hallucinations, and sleep paralysis. Although suvorexant has not been evalu­ated in patients with narcolepsy, the drug might, hypothetically, put patients at higher risk of the full spectrum of narco­lepsy symptoms.

There are no other contraindications for suvorexant.


Dosing
Suvorexant should be taken no more than once a night within 30 minutes of bedtime and with at least 7 hours before the planned wake time.⁹ The recommended starting dosage is 10 mg. If this dosage is well toler­ated but insufficiently effective, the dosage can be increased to a maximum of 20 mg. The 5-mg dosage is recommended for indi­viduals taking a moderate CYP3A inhibitor. Generally, patients should take the lowest effective dosage.

There are no specified limitations on the duration of suvorexant use. There is no evidence of withdrawal effects when discontinuing the medication. Patients tak­ing suvorexant should be educated about possible next-day effects that might impair driving or other activities that require full mental alertness, especially if they are tak­ing the 20-mg dosage.

Bottom Line
Suvorexant is FDA-approved for treating sleep onset and sleep maintenance insomnia. The drug is a dual orexin-receptor antagonist, which targets persistent CNS hyperarousal. In clinical trials, suvorexant improved the ability to fall asleep and remain asleep in patients with insomnia. It is generally safe and well tolerated. However, these studies evaluated dosages higher than those approved by the FDA.

Related Resources
• Jacobson LH, Callander GE, Hoyer D. Suvorexant for the treatment of insomnia. Expert Rev Clin Pharmacol. 2014; 7(6):711-730.
• Neubauer DN. New and emerging pharmacotherapeutic approaches for insomnia. Int Rev Psychiatry. 2014;26(2): 214-224.

Drug Brand Names
Doxepin • Silenor             Suvorexant • Belsomra
Digoxin • Lanoxin             Zaleplon • Sonata
Eszopiclone • Lunesta       Zolpidem • Ambien,
Ramelteon • Rozerem            Edluar, Intermezzo

Disclosure
Dr. Neubauer is a consultant to Ferring Pharmaceuticals and Vanda Pharmaceuticals.

Suvorexant, FDA-approved to treat insomnia, has demonstrated efficacy in helping patients with insomnia improve their ability to fall asleep and remain asleep (Table 1).¹ This first-in-class compound represents a novel mechanism of action to promoting sleep that may avoid some prob­lems associated with other hypnotics.²

Because orexin antagonists have not been used outside of clinical trials, it is too soon to tell whether suvorexant will have the ideal real-world efficacy and safety profile to make it a first-line treatment for insomnia patients, or if it will be reserved for those who have failed a trial of several other treatments.⁴

In theory, the orexin antagonist approach to treating insomnia could represent a major advance that modulates the fundamental pathology of the disorder.⁵ The syndrome of chronic insomnia encompasses not just the nighttime sleep disturbance but also an assort­ment of daytime symptoms that can include fatigue, poor concentration, irritability, and decreased school or work performance but usually not sleepiness. This constellation of nighttime and daytime symptoms could be conceptualized as a manifestation of persis­tent CNS hyperarousal. Because the orexin system promotes and reinforces arousal, per­haps an orexin antagonist that dampens the level of orexin activity will ameliorate the full spectrum of insomnia symptoms—not sim­ply sedate patients.⁶

How suvorexant works
Suvorexant is a potent and reversible dual orexin-receptor antagonist. The orexin system, first described in 1998, has a key role in promoting and stabilizing wake­fulness.⁷ Evidence suggests that people with chronic insomnia exhibit a central hyperarousal that perpetuates their sleep difficulty. Accordingly, a targeted phar­maceutical approach that reduces orexin activity should facilitate sleep onset and sleep maintenance for these patients. It is well known that the regulation of sleep and wakefulness depends on the interaction of multiple nuclei within the hypothalamus. Orexinergic neurons in the perifornical-lateral hypothalamic region project widely in the CNS and have especially dense con­nections with wake-promoting cholinergic, serotonergic, noradrenergic, and histamin­ergic neurons.⁶

A precursor prepro-orexin peptide is split into 2 orexin neurotransmitters (orexin A and orexin B). These 2 orexins bind with 2 G-protein-coupled receptors (OX1R and OX2R) that have both overlapping and distinct distributions.⁷ Suvorexant is highly selective and has similar affinity for OX1R and OX2R, functioning as an antag­onist for both.⁸ Fundamentally, suvorexant enhances sleep by dampening the arous­ing wake drive.

Pharmacokinetics
Suvorexant is available as an immediate-release tablet with pharmacokinetic prop­erties that offer benefits for sleep onset and maintenance.⁹ Ingestion under fasting conditions results in a median time to maxi­mum concentration (T_max) of approximately 2 hours, although the T_max values vary widely from patient to patient (range 30 minutes to 6 hours). Although suvorexant can be taken with food, there is a modest absorption delay after a high-fat meal, resulting in a further T_max delay of approximately 1.5 hours.

Suvorexant is primarily metabolized through the cytochrome P450 (CYP) 3A path­way, with limited contribution by CYP2C19. There are no active metabolites. The suvorex­ant blood level and risk of side effects will be higher with concomitant use of CYP3A inhibitors. The drug should not be adminis­tered with strong CYP3A inhibitors; the ini­tial dosage should be reduced with moderate CYP3A inhibitors. Concomitant use of strong CYP3A inducers can result in a low suvorex­ant level and reduced efficacy.

Suvorexant has little effect on other med­ications, although a person taking digoxin might experience intestinal P-glycoprotein inhibition with a slight rise in the digoxin level. In a patient taking both medica­tions, monitoring of the digoxin level is recommended.

The elimination half-life of suvorexant is approximately 12 hours, with a steady state in approximately 3 days. Because the half-life of suvorexant is moderately long for a sleep-promoting medication, use of the drug might be associated with residual sleepiness the morning after bedtime dosing. The risk for next-morning sleepiness or impairment should be minimized, however, when using the recommended dosages. Elimination is approximately two-thirds through feces and one-third in the urine.

Suvorexant metabolism can be affected by sex and body mass index. Females and obese people have a modestly elevated expo­sure to suvorexant, as reflected by the area under the curve and maximum concentra­tion (C_max). These patients might not require dosage adjustments unless they are obese and female, in which case they should take a lower dosage.

Age and race have not been shown to influence suvorexant metabolism to a signifi­cant degree. Patients with renal impairment and those with mild or moderate hepatic impairment do not need dosage adjust­ment. Suvorexant has not been evaluated in patients with severe hepatic impairment.

Efficacy
Suvorexant showed significant evidence of improved sleep onset and sleep maintenance in patients with insomnia in clinical trials. The key efficacy clinical trials with insomnia patients included a phase-IIb dose-finding study,¹⁰ 2 similar 3-month phase-III studies,¹¹ and one 12-month phase-III safety study that incorporated efficacy outcomes.¹² All these trials included subjective sleep measures and all except for the long-term safety study also incorporated polysomnographic assess­ment. The specific sleep laboratory outcomes were latency to persistent sleep (LPS), wake after the onset of persistent sleep (WASO), total sleep time (TST), and sleep efficiency (SE). Subjective sleep outcomes were time to sleep onset (sTSO), wake after sleep onset (sWASO), and total sleep time (sTST). Other exploratory endpoints also were assessed. These efficacy and safety studies mostly were performed at dosages considerably higher than those approved by the FDA.

The dose-finding (phase-IIb) trial was conducted with non-geriatric (age 18 to 64) patients with insomnia in a random­ized, double-blind, crossover design of two 4-week periods with subjects given a nightly placebo or suvorexant (10 mg, 20 mg, 40 mg, or 80 mg).¹⁰ Each of the 4 groups included approximately 60 subjects. The 2 co-primary endpoints were SE at Night 1 and the end of Week 4; secondary endpoints were LPS and WASO. Suvorexant was associated with dos­age-related improvements in SE and WASO compared with placebo at both time points. Carryover effects from the period-1 active drug group complicated the analysis of LPS.

The phase-III efficacy and safety trials were performed with 40 mg high dosage (HD) and 20 mg low dosage (LD) groups for adults and with 30 mg HD and 15 mg LD groups for geriatric (age ≥65) patients.¹¹ Two similarly designed 3-month randomized, double-blind, placebo-controlled pivotal efficacy studies assessed objective and sub­jective sleep measures in 4 groups with non-geriatric (HD and LD) and geriatric (HD and LD) insomnia patients.

After baseline assessment, patients took nightly bedtime doses of placebo; suvorexant, 40 mg or 20 mg (non-geriatric individuals); or suvorexant, 30 mg or 15 mg (geriatric indi­viduals). All subjects kept a daily electronic diary and had polysomnographic recordings performed on Night 1, at the end of Month 1, and at the end of Month 3. Both the indi­vidual studies and combined analyses (2,030 subjects) showed that, in non-geriatric and geriatric patients, HD suvorexant resulted in significantly greater improvement in key subjective and objective measures through­out the study (Table 2,⁹ and Table 3,⁹), with the exception of a single LPS outcome in 1 study, compared with placebo. The LD dosages also demonstrated efficacy, but to a reduced extent.

Subjective sleep outcomes were assessed in a 1-year randomized, placebo-controlled trial with nightly placebo, suvorexant, 40 mg, for non-geriatric, or suvorexant, 30 mg, for geriatric insomnia patients.12 The 1-year phase was completed with 484 subjects. Key efficacy outcomes were sTST and sTSO changes from baseline during the first month of treatment. Compared with placebo, suvorexant dosages demonstrated significantly greater efficacy, improvements that were sustained throughout the year.

Clinical trials found suvorexant to be gen­erally safe and well tolerated.¹³ However, specific safety concerns led the FDA to approve the medication at dosages lower than those assessed in the phase-III studies.¹

Somnolence was the most common adverse event in clinical trials. In the phase- IIb dose-finding study, somnolence was reported in <1% in the placebo group, but was associated with suvorexant in 2% of the 10 mg group, 5% with 20 mg, 12% with 40 mg, and 11% with 80 mg.⁹ In the phase-III combined analysis of the 3-month studies, somnolence was reported by 3% in the placebo group and 7% of non-geriatric patients taking 20 mg or geriatric patients taking 15 mg. Somnolence was reported in 8% of women and 3% of men taking the 15 mg or 20 mg dosage in these stud­ies. The 1-year study was performed only with higher suvorexant dosages (30 mg and 40 mg), in comparison with placebo. In this long-term trial, somnolence was reported by 13% of subjects taking suvorexant and 3% taking placebo.

Additional safety issues in trials included excessive daytime sleepiness, impaired driv­ing, suicidal ideation, sleep paralysis, hyp­nagogic/hypnopompic hallucinations, and cataplexy-like symptoms.⁹ Occurrences of these events are rare but have been reported more often among patients taking suvorex­ant than among those taking placebo.

Unique clinical issues
The U.S. Drug Enforcement Agency has categorized suvorexant as a Schedule IV controlled substance. Although there is no evidence of physiological dependence or withdrawal symptoms with suvorexant, studies with recreational substance abusers have shown that the likeability rating is simi­lar to that of zolpidem.¹³

Contraindication
Suvorexant is contraindicated in patients with narcolepsy.⁹ The underlying pathol­ogy of narcolepsy involves a marked reduction in orexin functioning with corre­sponding excessive sleepiness and related symptoms, such as cataplexy, hypnago­gic hallucinations, and sleep paralysis. Although suvorexant has not been evalu­ated in patients with narcolepsy, the drug might, hypothetically, put patients at higher risk of the full spectrum of narco­lepsy symptoms.

There are no other contraindications for suvorexant.


Dosing
Suvorexant should be taken no more than once a night within 30 minutes of bedtime and with at least 7 hours before the planned wake time.⁹ The recommended starting dosage is 10 mg. If this dosage is well toler­ated but insufficiently effective, the dosage can be increased to a maximum of 20 mg. The 5-mg dosage is recommended for indi­viduals taking a moderate CYP3A inhibitor. Generally, patients should take the lowest effective dosage.

There are no specified limitations on the duration of suvorexant use. There is no evidence of withdrawal effects when discontinuing the medication. Patients tak­ing suvorexant should be educated about possible next-day effects that might impair driving or other activities that require full mental alertness, especially if they are tak­ing the 20-mg dosage.

Bottom Line
Suvorexant is FDA-approved for treating sleep onset and sleep maintenance insomnia. The drug is a dual orexin-receptor antagonist, which targets persistent CNS hyperarousal. In clinical trials, suvorexant improved the ability to fall asleep and remain asleep in patients with insomnia. It is generally safe and well tolerated. However, these studies evaluated dosages higher than those approved by the FDA.

Related Resources
• Jacobson LH, Callander GE, Hoyer D. Suvorexant for the treatment of insomnia. Expert Rev Clin Pharmacol. 2014; 7(6):711-730.
• Neubauer DN. New and emerging pharmacotherapeutic approaches for insomnia. Int Rev Psychiatry. 2014;26(2): 214-224.

Drug Brand Names
Doxepin • Silenor             Suvorexant • Belsomra
Digoxin • Lanoxin             Zaleplon • Sonata
Eszopiclone • Lunesta       Zolpidem • Ambien,
Ramelteon • Rozerem            Edluar, Intermezzo

Disclosure
Dr. Neubauer is a consultant to Ferring Pharmaceuticals and Vanda Pharmaceuticals.

Suvorexant, FDA-approved to treat insomnia, has demonstrated efficacy in helping patients with insomnia improve their ability to fall asleep and remain asleep (Table 1).¹ This first-in-class compound represents a novel mechanism of action to promoting sleep that may avoid some prob­lems associated with other hypnotics.²

Because orexin antagonists have not been used outside of clinical trials, it is too soon to tell whether suvorexant will have the ideal real-world efficacy and safety profile to make it a first-line treatment for insomnia patients, or if it will be reserved for those who have failed a trial of several other treatments.⁴

In theory, the orexin antagonist approach to treating insomnia could represent a major advance that modulates the fundamental pathology of the disorder.⁵ The syndrome of chronic insomnia encompasses not just the nighttime sleep disturbance but also an assort­ment of daytime symptoms that can include fatigue, poor concentration, irritability, and decreased school or work performance but usually not sleepiness. This constellation of nighttime and daytime symptoms could be conceptualized as a manifestation of persis­tent CNS hyperarousal. Because the orexin system promotes and reinforces arousal, per­haps an orexin antagonist that dampens the level of orexin activity will ameliorate the full spectrum of insomnia symptoms—not sim­ply sedate patients.⁶

How suvorexant works
Suvorexant is a potent and reversible dual orexin-receptor antagonist. The orexin system, first described in 1998, has a key role in promoting and stabilizing wake­fulness.⁷ Evidence suggests that people with chronic insomnia exhibit a central hyperarousal that perpetuates their sleep difficulty. Accordingly, a targeted phar­maceutical approach that reduces orexin activity should facilitate sleep onset and sleep maintenance for these patients. It is well known that the regulation of sleep and wakefulness depends on the interaction of multiple nuclei within the hypothalamus. Orexinergic neurons in the perifornical-lateral hypothalamic region project widely in the CNS and have especially dense con­nections with wake-promoting cholinergic, serotonergic, noradrenergic, and histamin­ergic neurons.⁶

A precursor prepro-orexin peptide is split into 2 orexin neurotransmitters (orexin A and orexin B). These 2 orexins bind with 2 G-protein-coupled receptors (OX1R and OX2R) that have both overlapping and distinct distributions.⁷ Suvorexant is highly selective and has similar affinity for OX1R and OX2R, functioning as an antag­onist for both.⁸ Fundamentally, suvorexant enhances sleep by dampening the arous­ing wake drive.

Pharmacokinetics
Suvorexant is available as an immediate-release tablet with pharmacokinetic prop­erties that offer benefits for sleep onset and maintenance.⁹ Ingestion under fasting conditions results in a median time to maxi­mum concentration (T_max) of approximately 2 hours, although the T_max values vary widely from patient to patient (range 30 minutes to 6 hours). Although suvorexant can be taken with food, there is a modest absorption delay after a high-fat meal, resulting in a further T_max delay of approximately 1.5 hours.

Suvorexant is primarily metabolized through the cytochrome P450 (CYP) 3A path­way, with limited contribution by CYP2C19. There are no active metabolites. The suvorex­ant blood level and risk of side effects will be higher with concomitant use of CYP3A inhibitors. The drug should not be adminis­tered with strong CYP3A inhibitors; the ini­tial dosage should be reduced with moderate CYP3A inhibitors. Concomitant use of strong CYP3A inducers can result in a low suvorex­ant level and reduced efficacy.

Suvorexant has little effect on other med­ications, although a person taking digoxin might experience intestinal P-glycoprotein inhibition with a slight rise in the digoxin level. In a patient taking both medica­tions, monitoring of the digoxin level is recommended.

The elimination half-life of suvorexant is approximately 12 hours, with a steady state in approximately 3 days. Because the half-life of suvorexant is moderately long for a sleep-promoting medication, use of the drug might be associated with residual sleepiness the morning after bedtime dosing. The risk for next-morning sleepiness or impairment should be minimized, however, when using the recommended dosages. Elimination is approximately two-thirds through feces and one-third in the urine.

Suvorexant metabolism can be affected by sex and body mass index. Females and obese people have a modestly elevated expo­sure to suvorexant, as reflected by the area under the curve and maximum concentra­tion (C_max). These patients might not require dosage adjustments unless they are obese and female, in which case they should take a lower dosage.

Age and race have not been shown to influence suvorexant metabolism to a signifi­cant degree. Patients with renal impairment and those with mild or moderate hepatic impairment do not need dosage adjust­ment. Suvorexant has not been evaluated in patients with severe hepatic impairment.

Efficacy
Suvorexant showed significant evidence of improved sleep onset and sleep maintenance in patients with insomnia in clinical trials. The key efficacy clinical trials with insomnia patients included a phase-IIb dose-finding study,¹⁰ 2 similar 3-month phase-III studies,¹¹ and one 12-month phase-III safety study that incorporated efficacy outcomes.¹² All these trials included subjective sleep measures and all except for the long-term safety study also incorporated polysomnographic assess­ment. The specific sleep laboratory outcomes were latency to persistent sleep (LPS), wake after the onset of persistent sleep (WASO), total sleep time (TST), and sleep efficiency (SE). Subjective sleep outcomes were time to sleep onset (sTSO), wake after sleep onset (sWASO), and total sleep time (sTST). Other exploratory endpoints also were assessed. These efficacy and safety studies mostly were performed at dosages considerably higher than those approved by the FDA.

The dose-finding (phase-IIb) trial was conducted with non-geriatric (age 18 to 64) patients with insomnia in a random­ized, double-blind, crossover design of two 4-week periods with subjects given a nightly placebo or suvorexant (10 mg, 20 mg, 40 mg, or 80 mg).¹⁰ Each of the 4 groups included approximately 60 subjects. The 2 co-primary endpoints were SE at Night 1 and the end of Week 4; secondary endpoints were LPS and WASO. Suvorexant was associated with dos­age-related improvements in SE and WASO compared with placebo at both time points. Carryover effects from the period-1 active drug group complicated the analysis of LPS.

The phase-III efficacy and safety trials were performed with 40 mg high dosage (HD) and 20 mg low dosage (LD) groups for adults and with 30 mg HD and 15 mg LD groups for geriatric (age ≥65) patients.¹¹ Two similarly designed 3-month randomized, double-blind, placebo-controlled pivotal efficacy studies assessed objective and sub­jective sleep measures in 4 groups with non-geriatric (HD and LD) and geriatric (HD and LD) insomnia patients.

After baseline assessment, patients took nightly bedtime doses of placebo; suvorexant, 40 mg or 20 mg (non-geriatric individuals); or suvorexant, 30 mg or 15 mg (geriatric indi­viduals). All subjects kept a daily electronic diary and had polysomnographic recordings performed on Night 1, at the end of Month 1, and at the end of Month 3. Both the indi­vidual studies and combined analyses (2,030 subjects) showed that, in non-geriatric and geriatric patients, HD suvorexant resulted in significantly greater improvement in key subjective and objective measures through­out the study (Table 2,⁹ and Table 3,⁹), with the exception of a single LPS outcome in 1 study, compared with placebo. The LD dosages also demonstrated efficacy, but to a reduced extent.

Subjective sleep outcomes were assessed in a 1-year randomized, placebo-controlled trial with nightly placebo, suvorexant, 40 mg, for non-geriatric, or suvorexant, 30 mg, for geriatric insomnia patients.12 The 1-year phase was completed with 484 subjects. Key efficacy outcomes were sTST and sTSO changes from baseline during the first month of treatment. Compared with placebo, suvorexant dosages demonstrated significantly greater efficacy, improvements that were sustained throughout the year.

Clinical trials found suvorexant to be gen­erally safe and well tolerated.¹³ However, specific safety concerns led the FDA to approve the medication at dosages lower than those assessed in the phase-III studies.¹

Somnolence was the most common adverse event in clinical trials. In the phase- IIb dose-finding study, somnolence was reported in <1% in the placebo group, but was associated with suvorexant in 2% of the 10 mg group, 5% with 20 mg, 12% with 40 mg, and 11% with 80 mg.⁹ In the phase-III combined analysis of the 3-month studies, somnolence was reported by 3% in the placebo group and 7% of non-geriatric patients taking 20 mg or geriatric patients taking 15 mg. Somnolence was reported in 8% of women and 3% of men taking the 15 mg or 20 mg dosage in these stud­ies. The 1-year study was performed only with higher suvorexant dosages (30 mg and 40 mg), in comparison with placebo. In this long-term trial, somnolence was reported by 13% of subjects taking suvorexant and 3% taking placebo.

Additional safety issues in trials included excessive daytime sleepiness, impaired driv­ing, suicidal ideation, sleep paralysis, hyp­nagogic/hypnopompic hallucinations, and cataplexy-like symptoms.⁹ Occurrences of these events are rare but have been reported more often among patients taking suvorex­ant than among those taking placebo.

Unique clinical issues
The U.S. Drug Enforcement Agency has categorized suvorexant as a Schedule IV controlled substance. Although there is no evidence of physiological dependence or withdrawal symptoms with suvorexant, studies with recreational substance abusers have shown that the likeability rating is simi­lar to that of zolpidem.¹³

Contraindication
Suvorexant is contraindicated in patients with narcolepsy.⁹ The underlying pathol­ogy of narcolepsy involves a marked reduction in orexin functioning with corre­sponding excessive sleepiness and related symptoms, such as cataplexy, hypnago­gic hallucinations, and sleep paralysis. Although suvorexant has not been evalu­ated in patients with narcolepsy, the drug might, hypothetically, put patients at higher risk of the full spectrum of narco­lepsy symptoms.

There are no other contraindications for suvorexant.


Dosing
Suvorexant should be taken no more than once a night within 30 minutes of bedtime and with at least 7 hours before the planned wake time.⁹ The recommended starting dosage is 10 mg. If this dosage is well toler­ated but insufficiently effective, the dosage can be increased to a maximum of 20 mg. The 5-mg dosage is recommended for indi­viduals taking a moderate CYP3A inhibitor. Generally, patients should take the lowest effective dosage.

There are no specified limitations on the duration of suvorexant use. There is no evidence of withdrawal effects when discontinuing the medication. Patients tak­ing suvorexant should be educated about possible next-day effects that might impair driving or other activities that require full mental alertness, especially if they are tak­ing the 20-mg dosage.

Bottom Line
Suvorexant is FDA-approved for treating sleep onset and sleep maintenance insomnia. The drug is a dual orexin-receptor antagonist, which targets persistent CNS hyperarousal. In clinical trials, suvorexant improved the ability to fall asleep and remain asleep in patients with insomnia. It is generally safe and well tolerated. However, these studies evaluated dosages higher than those approved by the FDA.

Related Resources
• Jacobson LH, Callander GE, Hoyer D. Suvorexant for the treatment of insomnia. Expert Rev Clin Pharmacol. 2014; 7(6):711-730.
• Neubauer DN. New and emerging pharmacotherapeutic approaches for insomnia. Int Rev Psychiatry. 2014;26(2): 214-224.

Drug Brand Names
Doxepin • Silenor             Suvorexant • Belsomra
Digoxin • Lanoxin             Zaleplon • Sonata
Eszopiclone • Lunesta       Zolpidem • Ambien,
Ramelteon • Rozerem            Edluar, Intermezzo

Disclosure
Dr. Neubauer is a consultant to Ferring Pharmaceuticals and Vanda Pharmaceuticals.

Akathisia—from the Greek for “inability to sit”—is a neuropsychiatric syndrome characterized by subjective and objective psychomotor restlessness. Patients typi­cally experience feelings of unease, inner restlessness mainly involving the legs, and a compulsion to move. Most engage in repetitive movement. They might swing or cross and uncross their legs, shift from one foot to the other, continuously pace, or persistently fidget.

In clinical settings, akathisia usually is a side effect of medi­cation. Antipsychotics, serotonin reuptake inhibitors, and buspirone are common triggers, but akathisia also has been associated with some antiemetics, preoperative sedatives, calcium channel blockers, and antivertigo agents. It also can be caused by withdrawal from an antipsychotic or related to a substance use disorder, especially cocaine. Akathisia can be acute or chronic, occurring in a tardive form with symptoms that last >6 months.^1-3

Much isn’t known about drug-induced akathisia
Our understanding of the pathophysiology of akathisia is incomplete. Some have suggested that it results from an imbal­ance between the dopaminergic/cholinergic and dopaminer­gic/serotonergic systems⁴; others, that the cause is a mismatch between the core and the shell of the nucleus accumbens, due in part to overstimulation of the locus ceruleus.⁵

More recently, researchers established a positive asso­ciation between higher scores on the Liverpool University Neuroleptic Side Effects Rating Scale and D2/D3 receptor occupancy in the ventral striatum (nucleus accumbens and olfactory tubercle).⁶ The D2/D3 receptor occupancy model might explain withdrawal symptoms associated with cocaine,⁷ as well as rela­tive worsening of symptoms after tapering or discontinuing stimulants in attention-deficit/hyperactivity disorder (ADHD).

Elements of a clinical evaluation
When akathisia is suspected, evaluation by a clinician familiar with its phenom­enology is crucial. A validated tool, such as the Barnes Akathisia Rating Scale (at out cometracker.org/library/BAS.pdf) can aid in the detection and assessment of severity.⁸

In evaluating patients, keep in mind that the inner restlessness that characterizes akathisia can affect the trunk, hands, and arms, as well as the legs, and can cause dys­phoria and anxiety. Akathisia has been linked to an increased likelihood of developing sui­cidal ideation and behavior.⁹

Less common subjective symptoms include rage, fear, nausea, and worsening of psychotic symptoms. Because of its asso­ciation with aggression and agitation, drug-induced akathisia has been cited—with little success—as the basis for an insanity defense by people who have committed a violent act.¹⁰

Or is akathisia another psychiatric disorder?
Akathisia might go undetected for several reasons. One key factor: Its symptoms resem­ble and often overlap with those of other psy­chiatric disorders, such as mania, psychosis, agitated depression, and ADHD. In addition, akathisia often occurs concurrently with, and is masked by, akinesia, a common extrapy­ramidal side effect of many antipsychotics. Such patients might have the inner feeling of restlessness and urge to move but do not exhibit characteristic limb movements. In some cases, cognitive or intellectual limita­tions prevent patients from communicating the inner turmoil they feel.¹¹

Medication nonadherence further compli­cates the picture, sometimes prompting a cli­nician to increase the dosage of the drug that is causing akathisia (Box 1¹²).

Managing drug-induced akathisia
Akathisia usually resolves when the drug causing it is discontinued; decreasing the dosage might alleviate the symptoms. Whenever akathisia is detected, careful revision of the current drug regimen— substituting an antipsychotic with a lower prevalence of akathisia, for example— should be considered (Box 2^13-16). Treatment of drug-induced akathisia, which should be tailored to the patient’s psycho­pathology and comorbidities, is needed as well (Table^17-25).

Low-dose mirtazapine (15 mg/d) was found to be as effective as propranolol, 80 mg/d, in a placebo-controlled study, and to be more effective than a beta blocker in treating akathisia induced by a first-gener­ation antipsychotic. The authors concluded that both propranolol and mirtazapine should be first-line therapy.²³ Others have suggested that these results be interpreted with caution because mirtazapine (at a higher dosage) has been linked to akathi­sia.³³ Mirtazapine blocks α-adrenergic receptors, resulting in antagonism of 5-HT2 and 5-HT3 receptors and consequent enhancement of 5-HT1A serotonergic trans­mission.³⁴ In one study, it was shown to reduce binding of the D2/D3 receptor ago­nist quinpirole.³⁵

Serotonin antagonists and agonists. Blockade of 5-HT2 receptors can attenuate D2 blockade and mitigate akathisia symp­toms. Mianserin, 15 mg/d, can be helpful, and ritanserin, 5 to 20 mg/d, produced about a 50% reduction in akathisia symp­toms in 10 patients taking neuroleptics.36 Neither is available in the United States, however.

Cyproheptadine, a potent 5-HT2A and 5-HT2C antagonist with anticholinergic and antihistaminic action, improved akathisia symptoms in an open trial of 17 patients with antipsychotic-induced akathisia.³⁷ The recommended dose is 8 to 16 mg/d.

A study using the selective inverse ago­nist pimavanserin (not FDA-approved) decreased akathisia in healthy volunteers taking haloperidol.^14,24,33

Zolmitriptan, a 5-HT1D agonist, also can be used³⁸; one study found that 7.5 mg/d of zolmitriptan is as effective as propranolol.³⁹

A 2010 study showed a statistically signifi­cant improvement in 8 patients taking trazo­done, compared with 5 patients on placebo, all of whom met criteria for at least mild akathisia. Trazodone’s antiakathitic effect is attributed to its 5-HT2A antagonism.²⁵

Anticholinergics. Traditionally, benztropine, biperiden, diphenhydramine, and trihexy­phenidyl have been used for prevention and treatment of extrapyramidal side effects. A Cochrane review concluded, however, that data are insufficient to support use of anticho­linergics for akathisia.⁴⁰ Although multiple case reports have shown anticholinergics to be effective in treating drug-induced akathi­sia,^12,17,33 their association with cognitive side effects suggests a need for caution.¹⁸

Benzodiazepines. Through their sedative and anxiolytic properties, benzodiazepines are thought to partially alleviate akathisia symptoms. Two small trials found clonaz­epam helpful for akathisia symptoms^2,20; and 1 case report revealed that a patient with akathisia improved after coadministration of clonazepam and baclofen.⁴¹

Anticonvulsants. Valproic acid has not been found to be useful in antipsychotic-induced tardive akathisia.⁴² However, a case report described a patient with schizophrenia whose akathisia symptoms improved after the dosage of gabapentin was increased.⁴³ Last, carbamazepine was found to be effec­tive in reducing akathisia symptoms in 3 patients with schizophrenia who were resis­tant to beta blockers, anticholinergics, anti­histaminergics, and benzodiazepines.¹⁹

α-adrenergic agonists. In an open trial, akathisia symptoms in 6 patients improved with clonidine, 0.2 to 0.8 mg/d.¹⁷ Speculation is that strong α1 antagonism might help pre­vent akathisia, which could be why this con­dition is not associated with iloperidone.⁴⁴

D2 agonists. Akathisia and restless legs syndrome have similar pathophysiology,^1,2 and patients with akathisia could ben­efit from D2 agonists such as cabergoline, pramipexole, rotigotine, and ropinirole. One case study revealed that a patient with aripiprazole-induced akathisia improved with ropinirole.⁴⁵ D2 agonists can precipi­tate or worsen psychosis, however, and would be a relative contraindication in patients with psychotic disorders.²²


Bottom Line
Failure to detect drug-induced akathisia can increase morbidity and delay recovery in patients undergoing psychiatric care. Knowing what to look for and how to tailor treatment to the needs of a given patient is an essential component of good care.

Related Resources
• Ferrando SJ, Eisendrath SJ. Adverse neuropsychiatric effects of dopamine antagonist medications. Misdiagnosis in the medical setting. Psychosomatics. 1991;32(4):426-432.
• Vinson DR. Diphenhydramine in the treatment of akathisia induced by prochlorperazine. J Emerg Med. 2004;26(3):265-270.

Drug Brand Names
Aripiprazole • Abilify                  Haloperidol • Haldol
Baclofen • Lioresal                     Iloperidone • Fanapt
Benztropine • Cogentin              Lurasidone • Latuda
Biperiden • Akineton                  Mirtazapine • Remeron
Buspirone • BuSpar                   Pramipexole • Mirapex
Cabergoline • Dostinex              Propranolol • Inderal
Carbamazepine • Tegretol          Quetiapine • Seroquel
Clonazepam • Klonopin              Ropinirole • Requip
Clonidine • Catapres                  Rotigotine • Neupro
Clozapine • Clozaril                    Trazodone • Desyrel, Oleptro
Cyproheptadine • Periactin          Trihexyphenidyl • Artane
Diphenhydramine • Benadryl       Valproic acid • Depakene
Gabapentin • Neurontin               Zolmitriptan • Zomig


Acknowledgement
Mandy Evans, MD, assisted with editing the manuscript of this article.

Disclosure
Dr. Forcen reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Akathisia—from the Greek for “inability to sit”—is a neuropsychiatric syndrome characterized by subjective and objective psychomotor restlessness. Patients typi­cally experience feelings of unease, inner restlessness mainly involving the legs, and a compulsion to move. Most engage in repetitive movement. They might swing or cross and uncross their legs, shift from one foot to the other, continuously pace, or persistently fidget.

In clinical settings, akathisia usually is a side effect of medi­cation. Antipsychotics, serotonin reuptake inhibitors, and buspirone are common triggers, but akathisia also has been associated with some antiemetics, preoperative sedatives, calcium channel blockers, and antivertigo agents. It also can be caused by withdrawal from an antipsychotic or related to a substance use disorder, especially cocaine. Akathisia can be acute or chronic, occurring in a tardive form with symptoms that last >6 months.^1-3

Much isn’t known about drug-induced akathisia
Our understanding of the pathophysiology of akathisia is incomplete. Some have suggested that it results from an imbal­ance between the dopaminergic/cholinergic and dopaminer­gic/serotonergic systems⁴; others, that the cause is a mismatch between the core and the shell of the nucleus accumbens, due in part to overstimulation of the locus ceruleus.⁵

More recently, researchers established a positive asso­ciation between higher scores on the Liverpool University Neuroleptic Side Effects Rating Scale and D2/D3 receptor occupancy in the ventral striatum (nucleus accumbens and olfactory tubercle).⁶ The D2/D3 receptor occupancy model might explain withdrawal symptoms associated with cocaine,⁷ as well as rela­tive worsening of symptoms after tapering or discontinuing stimulants in attention-deficit/hyperactivity disorder (ADHD).

Elements of a clinical evaluation
When akathisia is suspected, evaluation by a clinician familiar with its phenom­enology is crucial. A validated tool, such as the Barnes Akathisia Rating Scale (at out cometracker.org/library/BAS.pdf) can aid in the detection and assessment of severity.⁸

In evaluating patients, keep in mind that the inner restlessness that characterizes akathisia can affect the trunk, hands, and arms, as well as the legs, and can cause dys­phoria and anxiety. Akathisia has been linked to an increased likelihood of developing sui­cidal ideation and behavior.⁹

Less common subjective symptoms include rage, fear, nausea, and worsening of psychotic symptoms. Because of its asso­ciation with aggression and agitation, drug-induced akathisia has been cited—with little success—as the basis for an insanity defense by people who have committed a violent act.¹⁰

Or is akathisia another psychiatric disorder?
Akathisia might go undetected for several reasons. One key factor: Its symptoms resem­ble and often overlap with those of other psy­chiatric disorders, such as mania, psychosis, agitated depression, and ADHD. In addition, akathisia often occurs concurrently with, and is masked by, akinesia, a common extrapy­ramidal side effect of many antipsychotics. Such patients might have the inner feeling of restlessness and urge to move but do not exhibit characteristic limb movements. In some cases, cognitive or intellectual limita­tions prevent patients from communicating the inner turmoil they feel.¹¹

Medication nonadherence further compli­cates the picture, sometimes prompting a cli­nician to increase the dosage of the drug that is causing akathisia (Box 1¹²).

Managing drug-induced akathisia
Akathisia usually resolves when the drug causing it is discontinued; decreasing the dosage might alleviate the symptoms. Whenever akathisia is detected, careful revision of the current drug regimen— substituting an antipsychotic with a lower prevalence of akathisia, for example— should be considered (Box 2^13-16). Treatment of drug-induced akathisia, which should be tailored to the patient’s psycho­pathology and comorbidities, is needed as well (Table^17-25).

Low-dose mirtazapine (15 mg/d) was found to be as effective as propranolol, 80 mg/d, in a placebo-controlled study, and to be more effective than a beta blocker in treating akathisia induced by a first-gener­ation antipsychotic. The authors concluded that both propranolol and mirtazapine should be first-line therapy.²³ Others have suggested that these results be interpreted with caution because mirtazapine (at a higher dosage) has been linked to akathi­sia.³³ Mirtazapine blocks α-adrenergic receptors, resulting in antagonism of 5-HT2 and 5-HT3 receptors and consequent enhancement of 5-HT1A serotonergic trans­mission.³⁴ In one study, it was shown to reduce binding of the D2/D3 receptor ago­nist quinpirole.³⁵

Serotonin antagonists and agonists. Blockade of 5-HT2 receptors can attenuate D2 blockade and mitigate akathisia symp­toms. Mianserin, 15 mg/d, can be helpful, and ritanserin, 5 to 20 mg/d, produced about a 50% reduction in akathisia symp­toms in 10 patients taking neuroleptics.36 Neither is available in the United States, however.

Cyproheptadine, a potent 5-HT2A and 5-HT2C antagonist with anticholinergic and antihistaminic action, improved akathisia symptoms in an open trial of 17 patients with antipsychotic-induced akathisia.³⁷ The recommended dose is 8 to 16 mg/d.

A study using the selective inverse ago­nist pimavanserin (not FDA-approved) decreased akathisia in healthy volunteers taking haloperidol.^14,24,33

Zolmitriptan, a 5-HT1D agonist, also can be used³⁸; one study found that 7.5 mg/d of zolmitriptan is as effective as propranolol.³⁹

A 2010 study showed a statistically signifi­cant improvement in 8 patients taking trazo­done, compared with 5 patients on placebo, all of whom met criteria for at least mild akathisia. Trazodone’s antiakathitic effect is attributed to its 5-HT2A antagonism.²⁵

Anticholinergics. Traditionally, benztropine, biperiden, diphenhydramine, and trihexy­phenidyl have been used for prevention and treatment of extrapyramidal side effects. A Cochrane review concluded, however, that data are insufficient to support use of anticho­linergics for akathisia.⁴⁰ Although multiple case reports have shown anticholinergics to be effective in treating drug-induced akathi­sia,^12,17,33 their association with cognitive side effects suggests a need for caution.¹⁸

Benzodiazepines. Through their sedative and anxiolytic properties, benzodiazepines are thought to partially alleviate akathisia symptoms. Two small trials found clonaz­epam helpful for akathisia symptoms^2,20; and 1 case report revealed that a patient with akathisia improved after coadministration of clonazepam and baclofen.⁴¹

Anticonvulsants. Valproic acid has not been found to be useful in antipsychotic-induced tardive akathisia.⁴² However, a case report described a patient with schizophrenia whose akathisia symptoms improved after the dosage of gabapentin was increased.⁴³ Last, carbamazepine was found to be effec­tive in reducing akathisia symptoms in 3 patients with schizophrenia who were resis­tant to beta blockers, anticholinergics, anti­histaminergics, and benzodiazepines.¹⁹

α-adrenergic agonists. In an open trial, akathisia symptoms in 6 patients improved with clonidine, 0.2 to 0.8 mg/d.¹⁷ Speculation is that strong α1 antagonism might help pre­vent akathisia, which could be why this con­dition is not associated with iloperidone.⁴⁴

D2 agonists. Akathisia and restless legs syndrome have similar pathophysiology,^1,2 and patients with akathisia could ben­efit from D2 agonists such as cabergoline, pramipexole, rotigotine, and ropinirole. One case study revealed that a patient with aripiprazole-induced akathisia improved with ropinirole.⁴⁵ D2 agonists can precipi­tate or worsen psychosis, however, and would be a relative contraindication in patients with psychotic disorders.²²


Bottom Line
Failure to detect drug-induced akathisia can increase morbidity and delay recovery in patients undergoing psychiatric care. Knowing what to look for and how to tailor treatment to the needs of a given patient is an essential component of good care.

Related Resources
• Ferrando SJ, Eisendrath SJ. Adverse neuropsychiatric effects of dopamine antagonist medications. Misdiagnosis in the medical setting. Psychosomatics. 1991;32(4):426-432.
• Vinson DR. Diphenhydramine in the treatment of akathisia induced by prochlorperazine. J Emerg Med. 2004;26(3):265-270.

Drug Brand Names
Aripiprazole • Abilify                  Haloperidol • Haldol
Baclofen • Lioresal                     Iloperidone • Fanapt
Benztropine • Cogentin              Lurasidone • Latuda
Biperiden • Akineton                  Mirtazapine • Remeron
Buspirone • BuSpar                   Pramipexole • Mirapex
Cabergoline • Dostinex              Propranolol • Inderal
Carbamazepine • Tegretol          Quetiapine • Seroquel
Clonazepam • Klonopin              Ropinirole • Requip
Clonidine • Catapres                  Rotigotine • Neupro
Clozapine • Clozaril                    Trazodone • Desyrel, Oleptro
Cyproheptadine • Periactin          Trihexyphenidyl • Artane
Diphenhydramine • Benadryl       Valproic acid • Depakene
Gabapentin • Neurontin               Zolmitriptan • Zomig


Acknowledgement
Mandy Evans, MD, assisted with editing the manuscript of this article.

Disclosure
Dr. Forcen reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

Akathisia—from the Greek for “inability to sit”—is a neuropsychiatric syndrome characterized by subjective and objective psychomotor restlessness. Patients typi­cally experience feelings of unease, inner restlessness mainly involving the legs, and a compulsion to move. Most engage in repetitive movement. They might swing or cross and uncross their legs, shift from one foot to the other, continuously pace, or persistently fidget.

In clinical settings, akathisia usually is a side effect of medi­cation. Antipsychotics, serotonin reuptake inhibitors, and buspirone are common triggers, but akathisia also has been associated with some antiemetics, preoperative sedatives, calcium channel blockers, and antivertigo agents. It also can be caused by withdrawal from an antipsychotic or related to a substance use disorder, especially cocaine. Akathisia can be acute or chronic, occurring in a tardive form with symptoms that last >6 months.^1-3

Much isn’t known about drug-induced akathisia
Our understanding of the pathophysiology of akathisia is incomplete. Some have suggested that it results from an imbal­ance between the dopaminergic/cholinergic and dopaminer­gic/serotonergic systems⁴; others, that the cause is a mismatch between the core and the shell of the nucleus accumbens, due in part to overstimulation of the locus ceruleus.⁵

More recently, researchers established a positive asso­ciation between higher scores on the Liverpool University Neuroleptic Side Effects Rating Scale and D2/D3 receptor occupancy in the ventral striatum (nucleus accumbens and olfactory tubercle).⁶ The D2/D3 receptor occupancy model might explain withdrawal symptoms associated with cocaine,⁷ as well as rela­tive worsening of symptoms after tapering or discontinuing stimulants in attention-deficit/hyperactivity disorder (ADHD).

Elements of a clinical evaluation
When akathisia is suspected, evaluation by a clinician familiar with its phenom­enology is crucial. A validated tool, such as the Barnes Akathisia Rating Scale (at out cometracker.org/library/BAS.pdf) can aid in the detection and assessment of severity.⁸

In evaluating patients, keep in mind that the inner restlessness that characterizes akathisia can affect the trunk, hands, and arms, as well as the legs, and can cause dys­phoria and anxiety. Akathisia has been linked to an increased likelihood of developing sui­cidal ideation and behavior.⁹

Less common subjective symptoms include rage, fear, nausea, and worsening of psychotic symptoms. Because of its asso­ciation with aggression and agitation, drug-induced akathisia has been cited—with little success—as the basis for an insanity defense by people who have committed a violent act.¹⁰

Or is akathisia another psychiatric disorder?
Akathisia might go undetected for several reasons. One key factor: Its symptoms resem­ble and often overlap with those of other psy­chiatric disorders, such as mania, psychosis, agitated depression, and ADHD. In addition, akathisia often occurs concurrently with, and is masked by, akinesia, a common extrapy­ramidal side effect of many antipsychotics. Such patients might have the inner feeling of restlessness and urge to move but do not exhibit characteristic limb movements. In some cases, cognitive or intellectual limita­tions prevent patients from communicating the inner turmoil they feel.¹¹

Medication nonadherence further compli­cates the picture, sometimes prompting a cli­nician to increase the dosage of the drug that is causing akathisia (Box 1¹²).

Managing drug-induced akathisia
Akathisia usually resolves when the drug causing it is discontinued; decreasing the dosage might alleviate the symptoms. Whenever akathisia is detected, careful revision of the current drug regimen— substituting an antipsychotic with a lower prevalence of akathisia, for example— should be considered (Box 2^13-16). Treatment of drug-induced akathisia, which should be tailored to the patient’s psycho­pathology and comorbidities, is needed as well (Table^17-25).

Low-dose mirtazapine (15 mg/d) was found to be as effective as propranolol, 80 mg/d, in a placebo-controlled study, and to be more effective than a beta blocker in treating akathisia induced by a first-gener­ation antipsychotic. The authors concluded that both propranolol and mirtazapine should be first-line therapy.²³ Others have suggested that these results be interpreted with caution because mirtazapine (at a higher dosage) has been linked to akathi­sia.³³ Mirtazapine blocks α-adrenergic receptors, resulting in antagonism of 5-HT2 and 5-HT3 receptors and consequent enhancement of 5-HT1A serotonergic trans­mission.³⁴ In one study, it was shown to reduce binding of the D2/D3 receptor ago­nist quinpirole.³⁵

Serotonin antagonists and agonists. Blockade of 5-HT2 receptors can attenuate D2 blockade and mitigate akathisia symp­toms. Mianserin, 15 mg/d, can be helpful, and ritanserin, 5 to 20 mg/d, produced about a 50% reduction in akathisia symp­toms in 10 patients taking neuroleptics.36 Neither is available in the United States, however.

Cyproheptadine, a potent 5-HT2A and 5-HT2C antagonist with anticholinergic and antihistaminic action, improved akathisia symptoms in an open trial of 17 patients with antipsychotic-induced akathisia.³⁷ The recommended dose is 8 to 16 mg/d.

A study using the selective inverse ago­nist pimavanserin (not FDA-approved) decreased akathisia in healthy volunteers taking haloperidol.^14,24,33

Zolmitriptan, a 5-HT1D agonist, also can be used³⁸; one study found that 7.5 mg/d of zolmitriptan is as effective as propranolol.³⁹

A 2010 study showed a statistically signifi­cant improvement in 8 patients taking trazo­done, compared with 5 patients on placebo, all of whom met criteria for at least mild akathisia. Trazodone’s antiakathitic effect is attributed to its 5-HT2A antagonism.²⁵

Anticholinergics. Traditionally, benztropine, biperiden, diphenhydramine, and trihexy­phenidyl have been used for prevention and treatment of extrapyramidal side effects. A Cochrane review concluded, however, that data are insufficient to support use of anticho­linergics for akathisia.⁴⁰ Although multiple case reports have shown anticholinergics to be effective in treating drug-induced akathi­sia,^12,17,33 their association with cognitive side effects suggests a need for caution.¹⁸

Benzodiazepines. Through their sedative and anxiolytic properties, benzodiazepines are thought to partially alleviate akathisia symptoms. Two small trials found clonaz­epam helpful for akathisia symptoms^2,20; and 1 case report revealed that a patient with akathisia improved after coadministration of clonazepam and baclofen.⁴¹

Anticonvulsants. Valproic acid has not been found to be useful in antipsychotic-induced tardive akathisia.⁴² However, a case report described a patient with schizophrenia whose akathisia symptoms improved after the dosage of gabapentin was increased.⁴³ Last, carbamazepine was found to be effec­tive in reducing akathisia symptoms in 3 patients with schizophrenia who were resis­tant to beta blockers, anticholinergics, anti­histaminergics, and benzodiazepines.¹⁹

α-adrenergic agonists. In an open trial, akathisia symptoms in 6 patients improved with clonidine, 0.2 to 0.8 mg/d.¹⁷ Speculation is that strong α1 antagonism might help pre­vent akathisia, which could be why this con­dition is not associated with iloperidone.⁴⁴

D2 agonists. Akathisia and restless legs syndrome have similar pathophysiology,^1,2 and patients with akathisia could ben­efit from D2 agonists such as cabergoline, pramipexole, rotigotine, and ropinirole. One case study revealed that a patient with aripiprazole-induced akathisia improved with ropinirole.⁴⁵ D2 agonists can precipi­tate or worsen psychosis, however, and would be a relative contraindication in patients with psychotic disorders.²²


Bottom Line
Failure to detect drug-induced akathisia can increase morbidity and delay recovery in patients undergoing psychiatric care. Knowing what to look for and how to tailor treatment to the needs of a given patient is an essential component of good care.

Related Resources
• Ferrando SJ, Eisendrath SJ. Adverse neuropsychiatric effects of dopamine antagonist medications. Misdiagnosis in the medical setting. Psychosomatics. 1991;32(4):426-432.
• Vinson DR. Diphenhydramine in the treatment of akathisia induced by prochlorperazine. J Emerg Med. 2004;26(3):265-270.

Drug Brand Names
Aripiprazole • Abilify                  Haloperidol • Haldol
Baclofen • Lioresal                     Iloperidone • Fanapt
Benztropine • Cogentin              Lurasidone • Latuda
Biperiden • Akineton                  Mirtazapine • Remeron
Buspirone • BuSpar                   Pramipexole • Mirapex
Cabergoline • Dostinex              Propranolol • Inderal
Carbamazepine • Tegretol          Quetiapine • Seroquel
Clonazepam • Klonopin              Ropinirole • Requip
Clonidine • Catapres                  Rotigotine • Neupro
Clozapine • Clozaril                    Trazodone • Desyrel, Oleptro
Cyproheptadine • Periactin          Trihexyphenidyl • Artane
Diphenhydramine • Benadryl       Valproic acid • Depakene
Gabapentin • Neurontin               Zolmitriptan • Zomig


Acknowledgement
Mandy Evans, MD, assisted with editing the manuscript of this article.

Disclosure
Dr. Forcen reports no financial relationships with any company whose products are mentioned in this article or with manufacturers of competing products.

CASE Seeing friends
Mr. B, age 91, presents to the emergency room (ER) for hip pain. As he is being evalu­ated, he asks a nurse to tell the “other people” around her to leave so that he can have pri­vacy. As clarification, Mr. B reports visual hal­lucinations, which prompts the ER physician to request a psychiatry consult.

Mr. B is alert and oriented to time, place, and person when he is evaluated by the on-call psychiatry resident. He reports that he has been seeing several unusual things for the last 4 to 5 months. Asked to elaborate, Mr. B admits seeing colorful and vivid images of people around him. These people come and go as they like; rarely, they talk to him. He describes the conversations as “a constant chatter” in the background and adds that it is difficult to understand what they are talking about.

Mr. B states that he has been “seeing” a cou­ple of people on a regular basis, and they are “sort of like my friends.” He endorses that these people often sing songs or dance for him. He states that, sometimes, these “friends” bring 3 or 4 friends and, although he could not make out their faces clearly, “they all are around me.” He describes the people he sees as “nice people” and does not report being scared or frightened by them.

Mr. B does not report paranoia, and denies command-type hallucinations. He and his family report no unusual changes in behavior in recent months. The medical history is remarkable for atrial fibrillation, coronary artery disease, chronic obstructive pulmonary disease, age-related macular degeneration, and glaucoma.

Mr. B denies having any ongoing mood or anxiety symptoms. He states that he knows these people are “probably not real,” and they do not bother him and just keep him company.

What could be causing Mr. B’s hallucinations?
   a) a stroke
   b) late-onset schizophrenia
   c) dementia
   d) Charles Bonnet syndrome

The authors’ observations
Visual hallucinations among geriatric pa-tients are a common and confusing pre­sentation. In addition to several medical causes for this presentation (Table 1), con­sider Charles Bonnet syndrome in patients with visual loss, presenting as visual hal­lucinations with intact insight and absence of a mental illness. Other conditions to con­sider in the differential diagnosis include Parkinson’s disease, dementia with Lewy bodies, schizophrenia, seizures, migraine, and stroke, including lesions of the thala­mus or brain stem.

Charles Bonnet syndrome was first described by Swiss philosopher Charles Bonnet in the 18th century. He reported vivid visual hallucinations in his visually impaired grandfather (bilateral cataracts).¹

It is important to recognize this syn­drome because patients can present across different specialties, including psychia­try, ophthalmology, neurology, geriat­ric medicine, and family medicine.² As life expectancy increases, this condition might be seen more often. It is prudent to identify, intervene, and refer as appropri­ate, in addition to educating patients and caregivers about the nature and course of the condition. 

EVALUATION Not psychotic
Mr. B reports good sleep and appetite. He denies using alcohol or illicit drugs. He states he slipped in the bathroom the day before coming to the ER, but denies other recent falls or injuries. Other than hip pain, he has no other physical complaints. His medication regimen includes aspirin, lisinopril, lovastatin, and metoprolol.

The ER team diagnoses a hip fracture. Mr. B is transferred to the orthopedic service; the psychiatry consult team continues to fol­low him. Mental status examination is unre­markable other than the visual hallucinations. His speech is clear, non-pressured, with goal-directed thought processing. Mini-Mental State Examination score is 23/30 with Mr. B having difficulty with object drawing and 3-object recall. Brief cognitive examination in the ER is unremarkable.

The orthopedic team decides on conserva­tive management of the hip fracture. There is no evidence of infection. Mr. B is afebrile with clear sensorium; complete blood cell count and normal liver function tests are normal; urinalysis and urine drug screen are negative; and chest radiography is unremarkable. CT and MRI of the head are unremarkable.

After 1 week in the hospital, Mr. B contin­ues to experience vivid visual imagery. No signs of active infection are found. An oph­thalmologist is consulted, who confirms Mr. B’s earlier diagnosis of glaucoma and age-related macular degeneration but does not recommend further treatment. Visual field test by confrontation is normal, with normal visual reflexes.

The authors’ observations
The reported prevalence of Charles Bonnet syndrome among visually impaired peo­ple varies from study to study—from as low as 0.4% to as high as 63%.^3-6 The rea­son for such variation can be attributed to several variables:
   • underdiagnosis
   • misdiagnosis
   • underreporting by patients because of the benign nature of the hallucinations
   • patients’ reluctance to report visual hallucinations because of fear of being labeled “mentally ill.”^7,8

Symptoms
There are no specific diagnostic criteria for Charles Bonnet syndrome (Table 2). However, the following are generally accepted for diagnosis⁹:
   • grossly intact cognition, although mild cognitive impairment may be present in some cases¹⁰
    • underlying visual disorder, usually acquired, such as glaucoma, age-related macular degeneration, diabetic retinopathy, central retinal artery occlusion, and optic neuritis^3,4,11
   • no hallucinations or perceptive difficul­ties in other sensory modalities
   • generally intact insight
   • absence of delusions
   • absence of other neurologic, psychiat­ric, toxic, or metabolic conditions; medical causes of delirium must be ruled out.

Hallucinations might not be disturb­ing to the patient. Hallucinations could be simple (light flashes, lines, or geomet­ric shapes) or complex (faces, figures, or scenes),¹² and perceived as in color or in black and white. Hallucinations mostly are pleasant and rarely have any emotional impact or meaning. Although hallucina­tions are almost exclusively visual, they can be accompanied by noise or auditory hallucinations.^13,14

Other characteristics of Charles Bonnet syndrome include:  
   • typical age of onset is approximately 72 years (range, 70 to 92 years)  
   • no sex distinction has been identified  
   • episodes can last from a few seconds to few hours; the syndrome may last a few days or a few years⁵     
   • it is not uncommon for episodes to occur in clusters, followed by symptom-free intervals and recurrences  
   • symptoms tend to fade away as patients progress to complete loss of sight.¹⁵

The course of Charles Bonnet syndrome is uncertain and unpredictable and the epi­sodic nature can be frustrating for both patient and clinician. The syndrome could be misdiagnosed as a psychiatric condition.

Pathophysiology
The precise mechanism behind simple or complex vivid hallucinations in persons with Charles Bonnet syndrome is unclear. Several theories have been proposed.

Release theory proposes a loss of input to the primary visual areas, which decreases cortical inhibition and further causes disin­hibition of visual association areas, thereby “releasing” visual hallucinations.¹⁶ Research suggests that this might be an attempt by surviving neurons to recover vision. Loss of input somehow causes surviving neurons to adapt by increased sensitivity to residual visual stimuli.

Deafferentation theory. This relatively new theory proposes deafferentation of the visual sensory pathway, which, in turn, causes disinhibition of neurons in the visual cortical regions, thereby caus­ing them to fire spontaneously. This could cause a sensation analogous to phantom limb pain, which would be called “phan­tom vision presence of brain activity in the absence of an actual visual input.” Further, biochemical and molecular changes have been proposed to explain the deafferenta­tion theory.¹⁷

Neurobiological evidence. Limited data are available for a neurobiological basis to visual hallucinations in Charles Bonnet syndrome. A few studies have used func­tional MRI and single-photon emission CT and reported possible association of visual hallucinations to specific visual areas.^18,19

Risk factors
Social or physical isolation, loneliness, low extraversion, and shyness are risk factors for Charles Bonnet syndrome in visually impaired people.²⁰ Sensory deprivation and low level of arousal favor the occur­rence of hallucinations.⁵ Rate of vision loss—not the nature of pathology or sever­ity of visual impairment—has been sug­gested to increase the risk of developing Charles Bonnet syndrome.²¹

What are the treatment options for Charles Bonnet syndrome?
   a) begin an antipsychotic
   b) do nothing; there is no cure
   c) educate the patient about the nature of the hallucinations
   d) refer the patient to an ophthalmologist for evaluation of vision loss

Treatment
There are several modalities to manage visual hallucinations in a patient with Charles Bonnet syndrome (Table 3). After ruling out medical and other psychiat­ric causes of visual hallucinations, treat­ment might not be indicated if the patient is not disturbed by the hallucinations. In most cases, reassurance and educating the patient and family about the benign nature of the visual hallucinations is all that is needed.

For patients who are disturbed by these visions or for whom there is a treatable cause, treatment could include cataract removal, medical therapy to reduce intra­ocular pressure in glaucoma, treatment of diabetic retinopathy, or laser photoco­agulation. These treatments are associ­ated with a reduction in hallucinations.²²

In some cases, hallucinations disappear as visual acuity deteriorates. Psychotropics have been used to treat Charles Bonnet syndrome, including:
   • antipsychotics, including haloperi­dol, risperidone, and olanzapine
   • anticonvulsants, including valproic acid, gabapentin, and carbamazepine
   • antidepressants, including mirtazap­ine and venlafaxine.^23-30

Some experts recommend a conserva­tive approach, which might be justified because some cases of Charles Bonnet syndrome are episodic and remit sponta­neously.³¹ Again, however, consider phar­macotherapy if a patient is disturbed by hallucinations or if hallucinations impair overall functioning.

TREATMENT Education
After discussion with Mr. B and his family, he is started on risperidone, 1 mg at bedtime, and the psychiatric team provides informa­tion about the nature of Charles Bonnet syndrome. Mr. B reportedly takes this medi­cation for a few days and then stops because he does not want the visual hallucinations to go away.

The psychiatry team sees Mr. B before dis­charge. He and his family are educated about the benign nature of the syndrome, the need for continued family support, and the fact that hallucinations will have minimal or no impli­cations for his life.

The authors’ observations
It is important to remember that a visual description of hallucinations in Charles Bonnet syndrome can be quite vivid, and that the patient might not identify his hal­lucinations as such or consider them as a problem. Be careful not to dismiss the patient’s complaints as a primary psychi­atric condition. It also is important to be mindful of the patient’s concerns with a psychiatric diagnosis; detailed discussion with the patient is helpful in most cases. A more comprehensive and empathetic approach to care could go a long way to sustain quality of life for these patients.

Bottom Line
Charles Bonnet syndrome is characterized by visual hallucinations in patients with visual impairment who have intact insight and an absence of mental illness. Taking a thorough history can help rule out medical and psychiatric causes of visual hallucinations. Educate patients and family about the nature of the hallucinations. In some cases, a psychotropic may be indicated.

Related Resources
• Nguyen ND, Osterweil D, Hoffman J. Charles Bonnet syn­drome: treating nonpsychiatric hallucinations. Consult Pharm. 2013;28(3):184-188.
• Lapid MI, Burton MC, Chang MT, et al. Clinical phenomenology and mortality in Charles Bonnet syndrome. J Geriatr Psychiatry Neurol. 2013;26(1):3-9.

Drug Brand Names
Carbamazepine • Tegretol                          Mirtazapine • Remeron
Gabapentin • Neurontin                              Olanzapine • Zyprexa
Haloperidol • Haldol                                   Risperidone • Risperdal
Lisinopril • Prinivil, Zestril                           Valproic acid • Depakene
Lovastatin • Mevacor                                  Venlafaxine • Effexor
Metoprolol • Lopressor

Acknowledgement
The authors acknowledge Barry Liskow, MD, Vice Chair of Psychiatry, Kansas University Medical Center, Kansas City, Kansas, for providing both insight into the topic and useful feedback on the manuscript.

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

CASE Seeing friends
Mr. B, age 91, presents to the emergency room (ER) for hip pain. As he is being evalu­ated, he asks a nurse to tell the “other people” around her to leave so that he can have pri­vacy. As clarification, Mr. B reports visual hal­lucinations, which prompts the ER physician to request a psychiatry consult.

Mr. B is alert and oriented to time, place, and person when he is evaluated by the on-call psychiatry resident. He reports that he has been seeing several unusual things for the last 4 to 5 months. Asked to elaborate, Mr. B admits seeing colorful and vivid images of people around him. These people come and go as they like; rarely, they talk to him. He describes the conversations as “a constant chatter” in the background and adds that it is difficult to understand what they are talking about.

Mr. B states that he has been “seeing” a cou­ple of people on a regular basis, and they are “sort of like my friends.” He endorses that these people often sing songs or dance for him. He states that, sometimes, these “friends” bring 3 or 4 friends and, although he could not make out their faces clearly, “they all are around me.” He describes the people he sees as “nice people” and does not report being scared or frightened by them.

Mr. B does not report paranoia, and denies command-type hallucinations. He and his family report no unusual changes in behavior in recent months. The medical history is remarkable for atrial fibrillation, coronary artery disease, chronic obstructive pulmonary disease, age-related macular degeneration, and glaucoma.

Mr. B denies having any ongoing mood or anxiety symptoms. He states that he knows these people are “probably not real,” and they do not bother him and just keep him company.

What could be causing Mr. B’s hallucinations?
   a) a stroke
   b) late-onset schizophrenia
   c) dementia
   d) Charles Bonnet syndrome

The authors’ observations
Visual hallucinations among geriatric pa-tients are a common and confusing pre­sentation. In addition to several medical causes for this presentation (Table 1), con­sider Charles Bonnet syndrome in patients with visual loss, presenting as visual hal­lucinations with intact insight and absence of a mental illness. Other conditions to con­sider in the differential diagnosis include Parkinson’s disease, dementia with Lewy bodies, schizophrenia, seizures, migraine, and stroke, including lesions of the thala­mus or brain stem.

Charles Bonnet syndrome was first described by Swiss philosopher Charles Bonnet in the 18th century. He reported vivid visual hallucinations in his visually impaired grandfather (bilateral cataracts).¹

It is important to recognize this syn­drome because patients can present across different specialties, including psychia­try, ophthalmology, neurology, geriat­ric medicine, and family medicine.² As life expectancy increases, this condition might be seen more often. It is prudent to identify, intervene, and refer as appropri­ate, in addition to educating patients and caregivers about the nature and course of the condition. 

EVALUATION Not psychotic
Mr. B reports good sleep and appetite. He denies using alcohol or illicit drugs. He states he slipped in the bathroom the day before coming to the ER, but denies other recent falls or injuries. Other than hip pain, he has no other physical complaints. His medication regimen includes aspirin, lisinopril, lovastatin, and metoprolol.

The ER team diagnoses a hip fracture. Mr. B is transferred to the orthopedic service; the psychiatry consult team continues to fol­low him. Mental status examination is unre­markable other than the visual hallucinations. His speech is clear, non-pressured, with goal-directed thought processing. Mini-Mental State Examination score is 23/30 with Mr. B having difficulty with object drawing and 3-object recall. Brief cognitive examination in the ER is unremarkable.

The orthopedic team decides on conserva­tive management of the hip fracture. There is no evidence of infection. Mr. B is afebrile with clear sensorium; complete blood cell count and normal liver function tests are normal; urinalysis and urine drug screen are negative; and chest radiography is unremarkable. CT and MRI of the head are unremarkable.

After 1 week in the hospital, Mr. B contin­ues to experience vivid visual imagery. No signs of active infection are found. An oph­thalmologist is consulted, who confirms Mr. B’s earlier diagnosis of glaucoma and age-related macular degeneration but does not recommend further treatment. Visual field test by confrontation is normal, with normal visual reflexes.

The authors’ observations
The reported prevalence of Charles Bonnet syndrome among visually impaired peo­ple varies from study to study—from as low as 0.4% to as high as 63%.^3-6 The rea­son for such variation can be attributed to several variables:
   • underdiagnosis
   • misdiagnosis
   • underreporting by patients because of the benign nature of the hallucinations
   • patients’ reluctance to report visual hallucinations because of fear of being labeled “mentally ill.”^7,8

Symptoms
There are no specific diagnostic criteria for Charles Bonnet syndrome (Table 2). However, the following are generally accepted for diagnosis⁹:
   • grossly intact cognition, although mild cognitive impairment may be present in some cases¹⁰
    • underlying visual disorder, usually acquired, such as glaucoma, age-related macular degeneration, diabetic retinopathy, central retinal artery occlusion, and optic neuritis^3,4,11
   • no hallucinations or perceptive difficul­ties in other sensory modalities
   • generally intact insight
   • absence of delusions
   • absence of other neurologic, psychiat­ric, toxic, or metabolic conditions; medical causes of delirium must be ruled out.

Hallucinations might not be disturb­ing to the patient. Hallucinations could be simple (light flashes, lines, or geomet­ric shapes) or complex (faces, figures, or scenes),¹² and perceived as in color or in black and white. Hallucinations mostly are pleasant and rarely have any emotional impact or meaning. Although hallucina­tions are almost exclusively visual, they can be accompanied by noise or auditory hallucinations.^13,14

Other characteristics of Charles Bonnet syndrome include:  
   • typical age of onset is approximately 72 years (range, 70 to 92 years)  
   • no sex distinction has been identified  
   • episodes can last from a few seconds to few hours; the syndrome may last a few days or a few years⁵     
   • it is not uncommon for episodes to occur in clusters, followed by symptom-free intervals and recurrences  
   • symptoms tend to fade away as patients progress to complete loss of sight.¹⁵

The course of Charles Bonnet syndrome is uncertain and unpredictable and the epi­sodic nature can be frustrating for both patient and clinician. The syndrome could be misdiagnosed as a psychiatric condition.

Pathophysiology
The precise mechanism behind simple or complex vivid hallucinations in persons with Charles Bonnet syndrome is unclear. Several theories have been proposed.

Release theory proposes a loss of input to the primary visual areas, which decreases cortical inhibition and further causes disin­hibition of visual association areas, thereby “releasing” visual hallucinations.¹⁶ Research suggests that this might be an attempt by surviving neurons to recover vision. Loss of input somehow causes surviving neurons to adapt by increased sensitivity to residual visual stimuli.

Deafferentation theory. This relatively new theory proposes deafferentation of the visual sensory pathway, which, in turn, causes disinhibition of neurons in the visual cortical regions, thereby caus­ing them to fire spontaneously. This could cause a sensation analogous to phantom limb pain, which would be called “phan­tom vision presence of brain activity in the absence of an actual visual input.” Further, biochemical and molecular changes have been proposed to explain the deafferenta­tion theory.¹⁷

Neurobiological evidence. Limited data are available for a neurobiological basis to visual hallucinations in Charles Bonnet syndrome. A few studies have used func­tional MRI and single-photon emission CT and reported possible association of visual hallucinations to specific visual areas.^18,19

Risk factors
Social or physical isolation, loneliness, low extraversion, and shyness are risk factors for Charles Bonnet syndrome in visually impaired people.²⁰ Sensory deprivation and low level of arousal favor the occur­rence of hallucinations.⁵ Rate of vision loss—not the nature of pathology or sever­ity of visual impairment—has been sug­gested to increase the risk of developing Charles Bonnet syndrome.²¹

What are the treatment options for Charles Bonnet syndrome?
   a) begin an antipsychotic
   b) do nothing; there is no cure
   c) educate the patient about the nature of the hallucinations
   d) refer the patient to an ophthalmologist for evaluation of vision loss

Treatment
There are several modalities to manage visual hallucinations in a patient with Charles Bonnet syndrome (Table 3). After ruling out medical and other psychiat­ric causes of visual hallucinations, treat­ment might not be indicated if the patient is not disturbed by the hallucinations. In most cases, reassurance and educating the patient and family about the benign nature of the visual hallucinations is all that is needed.

For patients who are disturbed by these visions or for whom there is a treatable cause, treatment could include cataract removal, medical therapy to reduce intra­ocular pressure in glaucoma, treatment of diabetic retinopathy, or laser photoco­agulation. These treatments are associ­ated with a reduction in hallucinations.²²

In some cases, hallucinations disappear as visual acuity deteriorates. Psychotropics have been used to treat Charles Bonnet syndrome, including:
   • antipsychotics, including haloperi­dol, risperidone, and olanzapine
   • anticonvulsants, including valproic acid, gabapentin, and carbamazepine
   • antidepressants, including mirtazap­ine and venlafaxine.^23-30

Some experts recommend a conserva­tive approach, which might be justified because some cases of Charles Bonnet syndrome are episodic and remit sponta­neously.³¹ Again, however, consider phar­macotherapy if a patient is disturbed by hallucinations or if hallucinations impair overall functioning.

TREATMENT Education
After discussion with Mr. B and his family, he is started on risperidone, 1 mg at bedtime, and the psychiatric team provides informa­tion about the nature of Charles Bonnet syndrome. Mr. B reportedly takes this medi­cation for a few days and then stops because he does not want the visual hallucinations to go away.

The psychiatry team sees Mr. B before dis­charge. He and his family are educated about the benign nature of the syndrome, the need for continued family support, and the fact that hallucinations will have minimal or no impli­cations for his life.

The authors’ observations
It is important to remember that a visual description of hallucinations in Charles Bonnet syndrome can be quite vivid, and that the patient might not identify his hal­lucinations as such or consider them as a problem. Be careful not to dismiss the patient’s complaints as a primary psychi­atric condition. It also is important to be mindful of the patient’s concerns with a psychiatric diagnosis; detailed discussion with the patient is helpful in most cases. A more comprehensive and empathetic approach to care could go a long way to sustain quality of life for these patients.

Bottom Line
Charles Bonnet syndrome is characterized by visual hallucinations in patients with visual impairment who have intact insight and an absence of mental illness. Taking a thorough history can help rule out medical and psychiatric causes of visual hallucinations. Educate patients and family about the nature of the hallucinations. In some cases, a psychotropic may be indicated.

Related Resources
• Nguyen ND, Osterweil D, Hoffman J. Charles Bonnet syn­drome: treating nonpsychiatric hallucinations. Consult Pharm. 2013;28(3):184-188.
• Lapid MI, Burton MC, Chang MT, et al. Clinical phenomenology and mortality in Charles Bonnet syndrome. J Geriatr Psychiatry Neurol. 2013;26(1):3-9.

Drug Brand Names
Carbamazepine • Tegretol                          Mirtazapine • Remeron
Gabapentin • Neurontin                              Olanzapine • Zyprexa
Haloperidol • Haldol                                   Risperidone • Risperdal
Lisinopril • Prinivil, Zestril                           Valproic acid • Depakene
Lovastatin • Mevacor                                  Venlafaxine • Effexor
Metoprolol • Lopressor

Acknowledgement
The authors acknowledge Barry Liskow, MD, Vice Chair of Psychiatry, Kansas University Medical Center, Kansas City, Kansas, for providing both insight into the topic and useful feedback on the manuscript.

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

CASE Seeing friends
Mr. B, age 91, presents to the emergency room (ER) for hip pain. As he is being evalu­ated, he asks a nurse to tell the “other people” around her to leave so that he can have pri­vacy. As clarification, Mr. B reports visual hal­lucinations, which prompts the ER physician to request a psychiatry consult.

Mr. B is alert and oriented to time, place, and person when he is evaluated by the on-call psychiatry resident. He reports that he has been seeing several unusual things for the last 4 to 5 months. Asked to elaborate, Mr. B admits seeing colorful and vivid images of people around him. These people come and go as they like; rarely, they talk to him. He describes the conversations as “a constant chatter” in the background and adds that it is difficult to understand what they are talking about.

Mr. B states that he has been “seeing” a cou­ple of people on a regular basis, and they are “sort of like my friends.” He endorses that these people often sing songs or dance for him. He states that, sometimes, these “friends” bring 3 or 4 friends and, although he could not make out their faces clearly, “they all are around me.” He describes the people he sees as “nice people” and does not report being scared or frightened by them.

Mr. B does not report paranoia, and denies command-type hallucinations. He and his family report no unusual changes in behavior in recent months. The medical history is remarkable for atrial fibrillation, coronary artery disease, chronic obstructive pulmonary disease, age-related macular degeneration, and glaucoma.

Mr. B denies having any ongoing mood or anxiety symptoms. He states that he knows these people are “probably not real,” and they do not bother him and just keep him company.

What could be causing Mr. B’s hallucinations?
   a) a stroke
   b) late-onset schizophrenia
   c) dementia
   d) Charles Bonnet syndrome

The authors’ observations
Visual hallucinations among geriatric pa-tients are a common and confusing pre­sentation. In addition to several medical causes for this presentation (Table 1), con­sider Charles Bonnet syndrome in patients with visual loss, presenting as visual hal­lucinations with intact insight and absence of a mental illness. Other conditions to con­sider in the differential diagnosis include Parkinson’s disease, dementia with Lewy bodies, schizophrenia, seizures, migraine, and stroke, including lesions of the thala­mus or brain stem.

Charles Bonnet syndrome was first described by Swiss philosopher Charles Bonnet in the 18th century. He reported vivid visual hallucinations in his visually impaired grandfather (bilateral cataracts).¹

It is important to recognize this syn­drome because patients can present across different specialties, including psychia­try, ophthalmology, neurology, geriat­ric medicine, and family medicine.² As life expectancy increases, this condition might be seen more often. It is prudent to identify, intervene, and refer as appropri­ate, in addition to educating patients and caregivers about the nature and course of the condition. 

EVALUATION Not psychotic
Mr. B reports good sleep and appetite. He denies using alcohol or illicit drugs. He states he slipped in the bathroom the day before coming to the ER, but denies other recent falls or injuries. Other than hip pain, he has no other physical complaints. His medication regimen includes aspirin, lisinopril, lovastatin, and metoprolol.

The ER team diagnoses a hip fracture. Mr. B is transferred to the orthopedic service; the psychiatry consult team continues to fol­low him. Mental status examination is unre­markable other than the visual hallucinations. His speech is clear, non-pressured, with goal-directed thought processing. Mini-Mental State Examination score is 23/30 with Mr. B having difficulty with object drawing and 3-object recall. Brief cognitive examination in the ER is unremarkable.

The orthopedic team decides on conserva­tive management of the hip fracture. There is no evidence of infection. Mr. B is afebrile with clear sensorium; complete blood cell count and normal liver function tests are normal; urinalysis and urine drug screen are negative; and chest radiography is unremarkable. CT and MRI of the head are unremarkable.

After 1 week in the hospital, Mr. B contin­ues to experience vivid visual imagery. No signs of active infection are found. An oph­thalmologist is consulted, who confirms Mr. B’s earlier diagnosis of glaucoma and age-related macular degeneration but does not recommend further treatment. Visual field test by confrontation is normal, with normal visual reflexes.

The authors’ observations
The reported prevalence of Charles Bonnet syndrome among visually impaired peo­ple varies from study to study—from as low as 0.4% to as high as 63%.^3-6 The rea­son for such variation can be attributed to several variables:
   • underdiagnosis
   • misdiagnosis
   • underreporting by patients because of the benign nature of the hallucinations
   • patients’ reluctance to report visual hallucinations because of fear of being labeled “mentally ill.”^7,8

Symptoms
There are no specific diagnostic criteria for Charles Bonnet syndrome (Table 2). However, the following are generally accepted for diagnosis⁹:
   • grossly intact cognition, although mild cognitive impairment may be present in some cases¹⁰
    • underlying visual disorder, usually acquired, such as glaucoma, age-related macular degeneration, diabetic retinopathy, central retinal artery occlusion, and optic neuritis^3,4,11
   • no hallucinations or perceptive difficul­ties in other sensory modalities
   • generally intact insight
   • absence of delusions
   • absence of other neurologic, psychiat­ric, toxic, or metabolic conditions; medical causes of delirium must be ruled out.

Hallucinations might not be disturb­ing to the patient. Hallucinations could be simple (light flashes, lines, or geomet­ric shapes) or complex (faces, figures, or scenes),¹² and perceived as in color or in black and white. Hallucinations mostly are pleasant and rarely have any emotional impact or meaning. Although hallucina­tions are almost exclusively visual, they can be accompanied by noise or auditory hallucinations.^13,14

Other characteristics of Charles Bonnet syndrome include:  
   • typical age of onset is approximately 72 years (range, 70 to 92 years)  
   • no sex distinction has been identified  
   • episodes can last from a few seconds to few hours; the syndrome may last a few days or a few years⁵     
   • it is not uncommon for episodes to occur in clusters, followed by symptom-free intervals and recurrences  
   • symptoms tend to fade away as patients progress to complete loss of sight.¹⁵

The course of Charles Bonnet syndrome is uncertain and unpredictable and the epi­sodic nature can be frustrating for both patient and clinician. The syndrome could be misdiagnosed as a psychiatric condition.

Pathophysiology
The precise mechanism behind simple or complex vivid hallucinations in persons with Charles Bonnet syndrome is unclear. Several theories have been proposed.

Release theory proposes a loss of input to the primary visual areas, which decreases cortical inhibition and further causes disin­hibition of visual association areas, thereby “releasing” visual hallucinations.¹⁶ Research suggests that this might be an attempt by surviving neurons to recover vision. Loss of input somehow causes surviving neurons to adapt by increased sensitivity to residual visual stimuli.

Deafferentation theory. This relatively new theory proposes deafferentation of the visual sensory pathway, which, in turn, causes disinhibition of neurons in the visual cortical regions, thereby caus­ing them to fire spontaneously. This could cause a sensation analogous to phantom limb pain, which would be called “phan­tom vision presence of brain activity in the absence of an actual visual input.” Further, biochemical and molecular changes have been proposed to explain the deafferenta­tion theory.¹⁷

Neurobiological evidence. Limited data are available for a neurobiological basis to visual hallucinations in Charles Bonnet syndrome. A few studies have used func­tional MRI and single-photon emission CT and reported possible association of visual hallucinations to specific visual areas.^18,19

Risk factors
Social or physical isolation, loneliness, low extraversion, and shyness are risk factors for Charles Bonnet syndrome in visually impaired people.²⁰ Sensory deprivation and low level of arousal favor the occur­rence of hallucinations.⁵ Rate of vision loss—not the nature of pathology or sever­ity of visual impairment—has been sug­gested to increase the risk of developing Charles Bonnet syndrome.²¹

What are the treatment options for Charles Bonnet syndrome?
   a) begin an antipsychotic
   b) do nothing; there is no cure
   c) educate the patient about the nature of the hallucinations
   d) refer the patient to an ophthalmologist for evaluation of vision loss

Treatment
There are several modalities to manage visual hallucinations in a patient with Charles Bonnet syndrome (Table 3). After ruling out medical and other psychiat­ric causes of visual hallucinations, treat­ment might not be indicated if the patient is not disturbed by the hallucinations. In most cases, reassurance and educating the patient and family about the benign nature of the visual hallucinations is all that is needed.

For patients who are disturbed by these visions or for whom there is a treatable cause, treatment could include cataract removal, medical therapy to reduce intra­ocular pressure in glaucoma, treatment of diabetic retinopathy, or laser photoco­agulation. These treatments are associ­ated with a reduction in hallucinations.²²

In some cases, hallucinations disappear as visual acuity deteriorates. Psychotropics have been used to treat Charles Bonnet syndrome, including:
   • antipsychotics, including haloperi­dol, risperidone, and olanzapine
   • anticonvulsants, including valproic acid, gabapentin, and carbamazepine
   • antidepressants, including mirtazap­ine and venlafaxine.^23-30

Some experts recommend a conserva­tive approach, which might be justified because some cases of Charles Bonnet syndrome are episodic and remit sponta­neously.³¹ Again, however, consider phar­macotherapy if a patient is disturbed by hallucinations or if hallucinations impair overall functioning.

TREATMENT Education
After discussion with Mr. B and his family, he is started on risperidone, 1 mg at bedtime, and the psychiatric team provides informa­tion about the nature of Charles Bonnet syndrome. Mr. B reportedly takes this medi­cation for a few days and then stops because he does not want the visual hallucinations to go away.

The psychiatry team sees Mr. B before dis­charge. He and his family are educated about the benign nature of the syndrome, the need for continued family support, and the fact that hallucinations will have minimal or no impli­cations for his life.

The authors’ observations
It is important to remember that a visual description of hallucinations in Charles Bonnet syndrome can be quite vivid, and that the patient might not identify his hal­lucinations as such or consider them as a problem. Be careful not to dismiss the patient’s complaints as a primary psychi­atric condition. It also is important to be mindful of the patient’s concerns with a psychiatric diagnosis; detailed discussion with the patient is helpful in most cases. A more comprehensive and empathetic approach to care could go a long way to sustain quality of life for these patients.

Bottom Line
Charles Bonnet syndrome is characterized by visual hallucinations in patients with visual impairment who have intact insight and an absence of mental illness. Taking a thorough history can help rule out medical and psychiatric causes of visual hallucinations. Educate patients and family about the nature of the hallucinations. In some cases, a psychotropic may be indicated.

Related Resources
• Nguyen ND, Osterweil D, Hoffman J. Charles Bonnet syn­drome: treating nonpsychiatric hallucinations. Consult Pharm. 2013;28(3):184-188.
• Lapid MI, Burton MC, Chang MT, et al. Clinical phenomenology and mortality in Charles Bonnet syndrome. J Geriatr Psychiatry Neurol. 2013;26(1):3-9.

Drug Brand Names
Carbamazepine • Tegretol                          Mirtazapine • Remeron
Gabapentin • Neurontin                              Olanzapine • Zyprexa
Haloperidol • Haldol                                   Risperidone • Risperdal
Lisinopril • Prinivil, Zestril                           Valproic acid • Depakene
Lovastatin • Mevacor                                  Venlafaxine • Effexor
Metoprolol • Lopressor

Acknowledgement
The authors acknowledge Barry Liskow, MD, Vice Chair of Psychiatry, Kansas University Medical Center, Kansas City, Kansas, for providing both insight into the topic and useful feedback on the manuscript.

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

Police bring Ms. R, age 35, to the psychiat­ric ER after they find her asleep in a park. She is awake but drowsy, and states that she has a history of bipolar disorder. She claims that she had been stable on valproic acid (VPA), 1,500 mg/d, bupropion XL, 300 mg/d, quetiap­ine, 400 mg/d, and trazodone, 100 mg/d, until 2 weeks ago, when her best friend died and she stopped taking her medications all together. The previous evening, feeling “alone, hopeless, and sad,” she attempted suicide by ingesting a handful of VPA and clonazepam, obtained from a friend, and 2 liters of vodka. She complains of nausea, vomiting, and abdominal pain. Ele­vated laboratory chemistries included aspartate aminotransferase (AST), 220 U/L; alanine ami­notransferase (ALT), 182 U/L; alkaline phospha­tase (AP), 75 U/L; γ-glutamyltransferase (GGT), 104 U/L; total bilirubin, 1.4 mg/dL; and an ele­vated VPA serum concentration of 152 μg/mL.

Drug-induced hepatotoxicity accounts for approximately 50% of acute liver failure cases, and almost 10% of liver transplants in some facilities.¹ The incidence of drug-induced hepatotoxicity is between 0.001% and 0.1% in patients on standard medication doses.² Drug-induced hepatotoxicity is char­acterized by:
   • abnormalities in laboratory parameters (hepatocellular, cholestatic, or mixed)
   • mechanisms of toxicity (direct, immune-mediated, idiosyncratic, mito­chondrial toxicity)
   • liver biopsy histology (steatosis, sinu­soidal obstruction syndrome).3

Liver function test results of hepatocel­lular injury are characterized by ALT ele­vation and minimal AP elevation, whereas cholestatic injury manifests as high AP. Table 1³ categorizes psychotropics based on type of liver injury and how each injury manifest in liver function tests. Delayed idiosyncratic reactions occur after tak­ing the drug, whereas direct toxicities are dose-dependent and more predictable. By definition, a clinically significant hepato­toxicity is associated with an ALT >3 times the upper limit of normal.³

VPA-induced liver injury occurs in approximately 1 in 37,000 persons taking the drug.⁴ Patients at an increased risk of VPA-induced liver injury include:
   • children
   • patients with mitochondrial enzyme deficiencies
   • Reye’s syndrome
   • Friedreich’s ataxia
   • polypharmacy patients
   • patients with a sibling who has experi­enced VPA toxicity.⁴

Benign enzyme elevations occur in approximately 20% of patients taking VPA.⁵ In Ms. R’s case, concomitant VPA, clonazepam, and alcohol may have led to elevations in ALT, AST, and GGT. Her nausea, vomiting, and abdominal pain are consistent with hepatic dysfunction.

Carnitine is effective in increasing sur­vival of patients with VPA-induced hepa­totoxicity.⁴ Because Ms. R is symptomatic, discontinuing VPA and administering IV L-carnitine is warranted.⁵ L-carnitine can be initiated at 100 mg/kg as an IV bolus, followed by 50 mg/kg as an IV infusion every 8 hours, with a maximum dosage of 3,000 mg.⁶ Patients may require sev­eral days of therapy based on symptoms. L-carnitine should be continued until a patient shows clinical improvement, such as decreases in ALT and AST.

Ms. R experienced a VPA-induced hepa­totoxic reaction. However, continuous mon­itoring is appropriate for all patients who are prescribed any potentially hepatotoxic psychotropic, especially after hepatic inju­ries resolve. This includes mood stabilizers, antipsychotics, benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), and serotonin-norepinephrine reuptake inhibi­tors, especially when given concomitantly with other hepatotoxic agents.

Table 2 lists dosing recommen­dations for commonly used psychotro­pics in patients with hepatic impairment. Among mood stabilizers, carbamazepine and VPA are associated with the highest incidence of hepatotoxicity.² A follow-up study of more than 1,000,000 VPA prescrip­tions found 29 cases of fatal hepatotoxicity in a 7-year period.⁷ Although there are case reports of hepatotoxicity with oxcarbaze­pine, it may have a better liver safety profile than carbamazepine.² Hepatotoxicity with lamotrigine is rare, although fatal cases have been reported.⁵

When initiating an antipsychotic, a tem­porary, benign increase in liver enzymes can be expected, but typically discontinuation is unnecessary.² Phenothiazines in particular can cause increases in liver enzymes in 20% of patients.² Hepatotoxicity with benzodi­azepines is infrequent, with a few cases of cholestatic injury reported with diazepam, chlordiazepoxide, and flurazepam.²

SSRIs are relatively safe; incidents of hepatic injury are rare. Among SSRIs, parox­etine is most frequently associated with hep­atotoxicity. Abnormal liver function tests have been observed with fluoxetine (0.5% of long-term recipients) and other SSRIs.^1,2,4

Among antidepressants with dual serotonergic action, nefazodone carries a black-box warning for hepatotoxicity and is used rarely, whereas trazodone is not regarded as hepatotoxic.² Antidepressants with dual norepinephrine and serotonin reuptake inhibitor properties carry a higher risk of liver injury, especially duloxetine. Hepatocellular, cholestatic, and mixed types of hepatotoxicity are associated with duloxetine-induced hepatotoxicity.²

Monitoring recommendations
Before prescribing potentially hepatotoxic medications, order baseline liver function tests. During therapy, periodic liver func­tion monitoring is recommended. Elevated transaminase concentrations (>3 × the upper limit of normal), bilirubin (>2 × the upper limit of normal), and prolonged pro­thrombin times are indicators of hepatic injury.² Caution should be taken to prevent polypharmacy with multiple hepatotoxic medications and over-the-counter use of hepatotoxic drugs and supplements.

When choosing a psychotropic, take into account patient-specific factors, such as underlying liver disease and alcohol con­sumption. Patients on potentially hepato­toxic medications should be counseled to recognize and report symptoms of liver dysfunction, including nausea, vomiting, jaundice, and lower-extremity edema.² If liver injury occurs, modify therapy with the potential offending agent and check liver function periodically.

Related Resourcesa
• Bleibel W, Kim S, D’Silva K, et al. Drug-induced liver injury: review article. Dig Dis Sci. 2007;52(10):2463-2471.
• U.S. National Library of Medicine. LiverTox. National Institute of Health. www.livertox.nih.gov.

Drug Brand Names
Amitriptyline • Elavil                                       Lurasidone • Latuda
Molindone • Moban                                         Molindone • Moban
Aripiprazole • Abilify                                       Nefazodone • Serzone
Asenapine • Saphris                                       Nortriptyline • Pamelor
Bupropion XL • Wellbutrin XL                          Olanzapine • Zyprexa
Citalopram • Celexa                                       Oxcarbazepine • Trileptal
Carbamazepine • Tegretol                               Paroxetine • Paxil
Chlordiazepoxide • Librium                              Perphenazine • Trilafon
Chlorpromazine • Thorazine                             Phenobarbital • Luminal
Clonazepam • Klonopin                                   Phenytoin • Dilantin
Clozapine • Clozaril                                         Quetiapine • Seroquel
Desvenlafaxine • Pristiq                                   Risperidone • Risperdal
Diazepam • Valium                                         Sertraline • Zoloft
Duloxetine • Cymbalta                                    Thiothixene • Navane
Escitalopram • Lexapro                                   Trazodone • Desyrel
Fluoxetine • Prozac                                         Trifluoperazine • Stelazine
Fluphenazine • Prolixin                                    Topiramate • Topamax
Flurazepam • Dalmane                                    Valproic acid • Depakote
Haloperidol • Haldol                                        Venlafaxine • Effexor
Iloperidone • Fanapt                                       Ziprasidone • Geodon
Lamotrigine • Lamictal
Levocarnitine • L-carnitine

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

Police bring Ms. R, age 35, to the psychiat­ric ER after they find her asleep in a park. She is awake but drowsy, and states that she has a history of bipolar disorder. She claims that she had been stable on valproic acid (VPA), 1,500 mg/d, bupropion XL, 300 mg/d, quetiap­ine, 400 mg/d, and trazodone, 100 mg/d, until 2 weeks ago, when her best friend died and she stopped taking her medications all together. The previous evening, feeling “alone, hopeless, and sad,” she attempted suicide by ingesting a handful of VPA and clonazepam, obtained from a friend, and 2 liters of vodka. She complains of nausea, vomiting, and abdominal pain. Ele­vated laboratory chemistries included aspartate aminotransferase (AST), 220 U/L; alanine ami­notransferase (ALT), 182 U/L; alkaline phospha­tase (AP), 75 U/L; γ-glutamyltransferase (GGT), 104 U/L; total bilirubin, 1.4 mg/dL; and an ele­vated VPA serum concentration of 152 μg/mL.

Drug-induced hepatotoxicity accounts for approximately 50% of acute liver failure cases, and almost 10% of liver transplants in some facilities.¹ The incidence of drug-induced hepatotoxicity is between 0.001% and 0.1% in patients on standard medication doses.² Drug-induced hepatotoxicity is char­acterized by:
   • abnormalities in laboratory parameters (hepatocellular, cholestatic, or mixed)
   • mechanisms of toxicity (direct, immune-mediated, idiosyncratic, mito­chondrial toxicity)
   • liver biopsy histology (steatosis, sinu­soidal obstruction syndrome).3

Liver function test results of hepatocel­lular injury are characterized by ALT ele­vation and minimal AP elevation, whereas cholestatic injury manifests as high AP. Table 1³ categorizes psychotropics based on type of liver injury and how each injury manifest in liver function tests. Delayed idiosyncratic reactions occur after tak­ing the drug, whereas direct toxicities are dose-dependent and more predictable. By definition, a clinically significant hepato­toxicity is associated with an ALT >3 times the upper limit of normal.³

VPA-induced liver injury occurs in approximately 1 in 37,000 persons taking the drug.⁴ Patients at an increased risk of VPA-induced liver injury include:
   • children
   • patients with mitochondrial enzyme deficiencies
   • Reye’s syndrome
   • Friedreich’s ataxia
   • polypharmacy patients
   • patients with a sibling who has experi­enced VPA toxicity.⁴

Benign enzyme elevations occur in approximately 20% of patients taking VPA.⁵ In Ms. R’s case, concomitant VPA, clonazepam, and alcohol may have led to elevations in ALT, AST, and GGT. Her nausea, vomiting, and abdominal pain are consistent with hepatic dysfunction.

Carnitine is effective in increasing sur­vival of patients with VPA-induced hepa­totoxicity.⁴ Because Ms. R is symptomatic, discontinuing VPA and administering IV L-carnitine is warranted.⁵ L-carnitine can be initiated at 100 mg/kg as an IV bolus, followed by 50 mg/kg as an IV infusion every 8 hours, with a maximum dosage of 3,000 mg.⁶ Patients may require sev­eral days of therapy based on symptoms. L-carnitine should be continued until a patient shows clinical improvement, such as decreases in ALT and AST.

Ms. R experienced a VPA-induced hepa­totoxic reaction. However, continuous mon­itoring is appropriate for all patients who are prescribed any potentially hepatotoxic psychotropic, especially after hepatic inju­ries resolve. This includes mood stabilizers, antipsychotics, benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), and serotonin-norepinephrine reuptake inhibi­tors, especially when given concomitantly with other hepatotoxic agents.

Table 2 lists dosing recommen­dations for commonly used psychotro­pics in patients with hepatic impairment. Among mood stabilizers, carbamazepine and VPA are associated with the highest incidence of hepatotoxicity.² A follow-up study of more than 1,000,000 VPA prescrip­tions found 29 cases of fatal hepatotoxicity in a 7-year period.⁷ Although there are case reports of hepatotoxicity with oxcarbaze­pine, it may have a better liver safety profile than carbamazepine.² Hepatotoxicity with lamotrigine is rare, although fatal cases have been reported.⁵

When initiating an antipsychotic, a tem­porary, benign increase in liver enzymes can be expected, but typically discontinuation is unnecessary.² Phenothiazines in particular can cause increases in liver enzymes in 20% of patients.² Hepatotoxicity with benzodi­azepines is infrequent, with a few cases of cholestatic injury reported with diazepam, chlordiazepoxide, and flurazepam.²

SSRIs are relatively safe; incidents of hepatic injury are rare. Among SSRIs, parox­etine is most frequently associated with hep­atotoxicity. Abnormal liver function tests have been observed with fluoxetine (0.5% of long-term recipients) and other SSRIs.^1,2,4

Among antidepressants with dual serotonergic action, nefazodone carries a black-box warning for hepatotoxicity and is used rarely, whereas trazodone is not regarded as hepatotoxic.² Antidepressants with dual norepinephrine and serotonin reuptake inhibitor properties carry a higher risk of liver injury, especially duloxetine. Hepatocellular, cholestatic, and mixed types of hepatotoxicity are associated with duloxetine-induced hepatotoxicity.²

Monitoring recommendations
Before prescribing potentially hepatotoxic medications, order baseline liver function tests. During therapy, periodic liver func­tion monitoring is recommended. Elevated transaminase concentrations (>3 × the upper limit of normal), bilirubin (>2 × the upper limit of normal), and prolonged pro­thrombin times are indicators of hepatic injury.² Caution should be taken to prevent polypharmacy with multiple hepatotoxic medications and over-the-counter use of hepatotoxic drugs and supplements.

When choosing a psychotropic, take into account patient-specific factors, such as underlying liver disease and alcohol con­sumption. Patients on potentially hepato­toxic medications should be counseled to recognize and report symptoms of liver dysfunction, including nausea, vomiting, jaundice, and lower-extremity edema.² If liver injury occurs, modify therapy with the potential offending agent and check liver function periodically.

Related Resourcesa
• Bleibel W, Kim S, D’Silva K, et al. Drug-induced liver injury: review article. Dig Dis Sci. 2007;52(10):2463-2471.
• U.S. National Library of Medicine. LiverTox. National Institute of Health. www.livertox.nih.gov.

Drug Brand Names
Amitriptyline • Elavil                                       Lurasidone • Latuda
Molindone • Moban                                         Molindone • Moban
Aripiprazole • Abilify                                       Nefazodone • Serzone
Asenapine • Saphris                                       Nortriptyline • Pamelor
Bupropion XL • Wellbutrin XL                          Olanzapine • Zyprexa
Citalopram • Celexa                                       Oxcarbazepine • Trileptal
Carbamazepine • Tegretol                               Paroxetine • Paxil
Chlordiazepoxide • Librium                              Perphenazine • Trilafon
Chlorpromazine • Thorazine                             Phenobarbital • Luminal
Clonazepam • Klonopin                                   Phenytoin • Dilantin
Clozapine • Clozaril                                         Quetiapine • Seroquel
Desvenlafaxine • Pristiq                                   Risperidone • Risperdal
Diazepam • Valium                                         Sertraline • Zoloft
Duloxetine • Cymbalta                                    Thiothixene • Navane
Escitalopram • Lexapro                                   Trazodone • Desyrel
Fluoxetine • Prozac                                         Trifluoperazine • Stelazine
Fluphenazine • Prolixin                                    Topiramate • Topamax
Flurazepam • Dalmane                                    Valproic acid • Depakote
Haloperidol • Haldol                                        Venlafaxine • Effexor
Iloperidone • Fanapt                                       Ziprasidone • Geodon
Lamotrigine • Lamictal
Levocarnitine • L-carnitine

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

Police bring Ms. R, age 35, to the psychiat­ric ER after they find her asleep in a park. She is awake but drowsy, and states that she has a history of bipolar disorder. She claims that she had been stable on valproic acid (VPA), 1,500 mg/d, bupropion XL, 300 mg/d, quetiap­ine, 400 mg/d, and trazodone, 100 mg/d, until 2 weeks ago, when her best friend died and she stopped taking her medications all together. The previous evening, feeling “alone, hopeless, and sad,” she attempted suicide by ingesting a handful of VPA and clonazepam, obtained from a friend, and 2 liters of vodka. She complains of nausea, vomiting, and abdominal pain. Ele­vated laboratory chemistries included aspartate aminotransferase (AST), 220 U/L; alanine ami­notransferase (ALT), 182 U/L; alkaline phospha­tase (AP), 75 U/L; γ-glutamyltransferase (GGT), 104 U/L; total bilirubin, 1.4 mg/dL; and an ele­vated VPA serum concentration of 152 μg/mL.

Drug-induced hepatotoxicity accounts for approximately 50% of acute liver failure cases, and almost 10% of liver transplants in some facilities.¹ The incidence of drug-induced hepatotoxicity is between 0.001% and 0.1% in patients on standard medication doses.² Drug-induced hepatotoxicity is char­acterized by:
   • abnormalities in laboratory parameters (hepatocellular, cholestatic, or mixed)
   • mechanisms of toxicity (direct, immune-mediated, idiosyncratic, mito­chondrial toxicity)
   • liver biopsy histology (steatosis, sinu­soidal obstruction syndrome).3

Liver function test results of hepatocel­lular injury are characterized by ALT ele­vation and minimal AP elevation, whereas cholestatic injury manifests as high AP. Table 1³ categorizes psychotropics based on type of liver injury and how each injury manifest in liver function tests. Delayed idiosyncratic reactions occur after tak­ing the drug, whereas direct toxicities are dose-dependent and more predictable. By definition, a clinically significant hepato­toxicity is associated with an ALT >3 times the upper limit of normal.³

VPA-induced liver injury occurs in approximately 1 in 37,000 persons taking the drug.⁴ Patients at an increased risk of VPA-induced liver injury include:
   • children
   • patients with mitochondrial enzyme deficiencies
   • Reye’s syndrome
   • Friedreich’s ataxia
   • polypharmacy patients
   • patients with a sibling who has experi­enced VPA toxicity.⁴

Benign enzyme elevations occur in approximately 20% of patients taking VPA.⁵ In Ms. R’s case, concomitant VPA, clonazepam, and alcohol may have led to elevations in ALT, AST, and GGT. Her nausea, vomiting, and abdominal pain are consistent with hepatic dysfunction.

Carnitine is effective in increasing sur­vival of patients with VPA-induced hepa­totoxicity.⁴ Because Ms. R is symptomatic, discontinuing VPA and administering IV L-carnitine is warranted.⁵ L-carnitine can be initiated at 100 mg/kg as an IV bolus, followed by 50 mg/kg as an IV infusion every 8 hours, with a maximum dosage of 3,000 mg.⁶ Patients may require sev­eral days of therapy based on symptoms. L-carnitine should be continued until a patient shows clinical improvement, such as decreases in ALT and AST.

Ms. R experienced a VPA-induced hepa­totoxic reaction. However, continuous mon­itoring is appropriate for all patients who are prescribed any potentially hepatotoxic psychotropic, especially after hepatic inju­ries resolve. This includes mood stabilizers, antipsychotics, benzodiazepines, selective serotonin reuptake inhibitors (SSRIs), and serotonin-norepinephrine reuptake inhibi­tors, especially when given concomitantly with other hepatotoxic agents.

Table 2 lists dosing recommen­dations for commonly used psychotro­pics in patients with hepatic impairment. Among mood stabilizers, carbamazepine and VPA are associated with the highest incidence of hepatotoxicity.² A follow-up study of more than 1,000,000 VPA prescrip­tions found 29 cases of fatal hepatotoxicity in a 7-year period.⁷ Although there are case reports of hepatotoxicity with oxcarbaze­pine, it may have a better liver safety profile than carbamazepine.² Hepatotoxicity with lamotrigine is rare, although fatal cases have been reported.⁵

When initiating an antipsychotic, a tem­porary, benign increase in liver enzymes can be expected, but typically discontinuation is unnecessary.² Phenothiazines in particular can cause increases in liver enzymes in 20% of patients.² Hepatotoxicity with benzodi­azepines is infrequent, with a few cases of cholestatic injury reported with diazepam, chlordiazepoxide, and flurazepam.²

SSRIs are relatively safe; incidents of hepatic injury are rare. Among SSRIs, parox­etine is most frequently associated with hep­atotoxicity. Abnormal liver function tests have been observed with fluoxetine (0.5% of long-term recipients) and other SSRIs.^1,2,4

Among antidepressants with dual serotonergic action, nefazodone carries a black-box warning for hepatotoxicity and is used rarely, whereas trazodone is not regarded as hepatotoxic.² Antidepressants with dual norepinephrine and serotonin reuptake inhibitor properties carry a higher risk of liver injury, especially duloxetine. Hepatocellular, cholestatic, and mixed types of hepatotoxicity are associated with duloxetine-induced hepatotoxicity.²

Monitoring recommendations
Before prescribing potentially hepatotoxic medications, order baseline liver function tests. During therapy, periodic liver func­tion monitoring is recommended. Elevated transaminase concentrations (>3 × the upper limit of normal), bilirubin (>2 × the upper limit of normal), and prolonged pro­thrombin times are indicators of hepatic injury.² Caution should be taken to prevent polypharmacy with multiple hepatotoxic medications and over-the-counter use of hepatotoxic drugs and supplements.

When choosing a psychotropic, take into account patient-specific factors, such as underlying liver disease and alcohol con­sumption. Patients on potentially hepato­toxic medications should be counseled to recognize and report symptoms of liver dysfunction, including nausea, vomiting, jaundice, and lower-extremity edema.² If liver injury occurs, modify therapy with the potential offending agent and check liver function periodically.

Related Resourcesa
• Bleibel W, Kim S, D’Silva K, et al. Drug-induced liver injury: review article. Dig Dis Sci. 2007;52(10):2463-2471.
• U.S. National Library of Medicine. LiverTox. National Institute of Health. www.livertox.nih.gov.

Drug Brand Names
Amitriptyline • Elavil                                       Lurasidone • Latuda
Molindone • Moban                                         Molindone • Moban
Aripiprazole • Abilify                                       Nefazodone • Serzone
Asenapine • Saphris                                       Nortriptyline • Pamelor
Bupropion XL • Wellbutrin XL                          Olanzapine • Zyprexa
Citalopram • Celexa                                       Oxcarbazepine • Trileptal
Carbamazepine • Tegretol                               Paroxetine • Paxil
Chlordiazepoxide • Librium                              Perphenazine • Trilafon
Chlorpromazine • Thorazine                             Phenobarbital • Luminal
Clonazepam • Klonopin                                   Phenytoin • Dilantin
Clozapine • Clozaril                                         Quetiapine • Seroquel
Desvenlafaxine • Pristiq                                   Risperidone • Risperdal
Diazepam • Valium                                         Sertraline • Zoloft
Duloxetine • Cymbalta                                    Thiothixene • Navane
Escitalopram • Lexapro                                   Trazodone • Desyrel
Fluoxetine • Prozac                                         Trifluoperazine • Stelazine
Fluphenazine • Prolixin                                    Topiramate • Topamax
Flurazepam • Dalmane                                    Valproic acid • Depakote
Haloperidol • Haldol                                        Venlafaxine • Effexor
Iloperidone • Fanapt                                       Ziprasidone • Geodon
Lamotrigine • Lamictal
Levocarnitine • L-carnitine

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

Dear Dr. Mossman,

Where I practice, most health care plans won’t pay for certain medications without giving prior authorization (PA). Completing PA forms and making telephone calls take up time that could be better spent treat­ing patients. I’m tempted to set a new policy of not doing PAs. If I do, might I face legal trouble?

Submitted by “Dr. A”

If you provide clinical care, you’ve prob­ably dealt with third-party payers who require prior authorization (PA) before they will pay for certain treatments. Dr. A is not alone in feeling exasperated about the time it takes to complete a PA.¹ After spend­ing several hours each month waiting on hold and wading through stacks of paper­work, you may feel like Dr. A and consider refusing to do any more PAs.

But is Dr. A’s proposed solution a good idea? To address this question and the frus­tration that lies behind it, we’ll take a cue from Italian film director Sergio Leone and discuss:
   • how PAs affect psychiatric care: the good, the bad, and the ugly
   • potential exposure to professional liabil­ity and ethics complaints that might result from refusing or failing to seek PA
   • strategies to reduce the burden of PAs while providing efficient, effective care.

The good
Recent decades have witnessed huge increases in spending on prescription medication. Psychotropics are no excep­tion; state Medicaid spending for anti-psychotic medication grew from <$1 bil­lion in 1995 to >$5.5 billion in 2005.²

Requiring a PA for expensive drugs is one way that third-party payers try to rein in costs and hold down insurance premi­ums. Imposing financial constraints often is just one aim of a pharmacy benefit man­agement (PBM) program. Insurers also jus­tify PBMs by pointing out that feedback to practitioners whose prescribing falls well outside the norm—in the form of mailed warnings, physician second opinions, or pharmacist consultation—can improve patient safety and encourage appropriate treatment options for enrolled patients.^3,4 Examples of such benefits include reduc­ing overuse of prescription opioids⁵ and antipsychotics among children,³ misuse of buprenorphine,⁶ and adverse effects from potentially inappropriate prescriptions.⁷

The bad
The bad news for doctors: Cost savings for payers come at the expense of pro­viders and their practices, in the form of time spent doing paperwork and talking on the phone to complete PAs or contest PA decisions.⁸ Addressing PA requests costs an estimated $83,000 per physician per year. The total administrative burden for all 835,000 physicians who practice in the United States therefore is 868,000,000 hours, or $69 billion annually.⁹

To make matters worse, PA requirements may increase the overall cost of care. After Georgia Medicaid instituted PA require­ments for second-generation antipsychotics (SGAs), average monthly per member drug costs fell $19.62, but average monthly outpa­tient treatment costs rose $31.59 per mem­ber.10 Pharmacy savings that result from requiring PAs for SGAs can be offset quickly by small increases in the hospitalization rate or emergency department visits.^9,11

The ugly
Many physicians believe that the PA pro­cess undermines patient care by decreasing time devoted to direct patient contact, incen­tivizing suboptimal treatment, and limit­ing medication access.^1,12,13 But do any data support this belief? Do PAs impede treat­ment for vulnerable persons with severe mental illnesses?

The answer, some studies suggest, is “Yes.” A Maine Medicaid PA policy slowed initiation of treatment for bipolar disor­der by reducing the rate of starting non-preferred medications, although the same policy had no impact on patients already receiving treatment.¹⁴ Another study exam­ined the effect of PA processes for inpatient psychiatry treatment and found that patients were less likely to be admitted on weekends, probably because PA review was not avail­able on those days.¹⁵ A third study showed that PA requirements and resulting impedi­ments to getting refills were correlated with medication discontinuation by patients with schizophrenia or bipolar disorder, which can increase the risk of decompensation, work-related problems, and hospitalization.¹⁶

Problems with PAs
Whether they are helpful or counterpro­ductive, PAs are a practice reality. Dr. A’s proposed solution sounds appealing, but it might create ethical and legal problems.

Among the fundamental elements of ethi­cal medical practice is physicians’ obliga­tion to give patients “guidance … as to the optimal course of action” and to “advocate for patients in dealing with third parties when appropriate.”¹⁷ It’s fine for psychia­trists to consider prescribing treatments that patients’ health care coverage favors, but we also have to help patients weigh and evaluate costs, particularly when patients’ circumstances and medical interests militate strongly for options that third-party payers balk at paying for. Patients’ interests—not what’s expedient—are always physicians’ foremost concern.¹⁸

Beyond purely ethical considerations, you might face legal consequences if you refuse or fail to seek PAs for what you think is the proper medication. As Table 1 shows, one key factor is whether you are under contract with the patient’s insurance carrier; if you are, failure to seek a PA when appropriate may constitute a breach of the contract (Donna Vanderpool, written communication, October 5, 2014).

If the prescribed medication does not meet the standard of care and your patient suffers some harm, a licensing board complaint and investigation are possible. You also face exposure to a medical malpractice action. Although we do not know of any instances in which such an action has succeeded, 2 recent court decisions suggest that harm to a patient stemmed from failing to seek PA for a medication could constitute grounds for a lawsuit.^19,20 Efforts to contain medical costs have been around for decades, and courts have held that physicians, third-party pay­ers, and utilization review intermediaries are bound by “the standard of reasonable com­munity practice”²¹ and should not let cost limitations “corrupt medical judgment.”²² Physicians who do not appeal limitations at odds with their medical judgment might bear responsibility for any injuries that occur.^18,22


Managing PA requests
Given the inevitability of encountering PA requests and your ethical and professional obligations to help patients, what can you do (Table 2^9,23,27)?

Some practitioners charge patients for time spent completing PAs.²³ Although phy­sicians should “complete without charge the appropriate ‘simplified’ insurance claim form as a part of service to the patient;” they also may consider “a charge for more complex or multiple forms … in conformity with local custom.”²⁴ Legally, physicians’ contracts with insurance panels may pre­clude charging such fees, but if a patient is being seen out of network, the physician does not have a contractual obligation and may charge.⁹ If your practice setting lets you choose which insurances you accept, the impact and burden of seeking PAs is a factor to consider when deciding whether to par­ticipate in a particular panel.²³

In an interesting twist, an Ohio physi­cian successfully sued a medical insur­ance administrator for the cost of his time responding to PA inquiries.²⁵ Reasoning that the insurance administrator “should expect to pay for the reasonable value of” the doctor’s time because the PAs “were solely intended for the benefit of the insur­ance administrator” or parties whom the administrator served, the judge awarded the doctor $187.50 plus 8% interest.

Considerations that are more practi­cal relate to how to triage and address the volume of PA requests. Some large medi­cal practices centralize PAs and try to set up pre-approved plans of care or blanket approvals for frequently encountered con­ditions. Centralization also allows one key administrative assistant to develop skills in processing PA requests and to build rela­tionships with payers.²⁶

The administrative assistant also can compile lists of preferred alternative medica­tions, PA forms, and payer Web sites. Using and submitting requests through payer Web sites can speed up PA processing, which saves time and money.²⁷ As electronic health records improve, they may incorporate patients’ formularies and provide automatic alerts for required PAs.²³

Patients should be involved, too. They can help to obtain relevant formulary infor­mation and to weigh alternative therapies. You can help them understand your role in the PA process, the reasoning behind your treatment recommendations, and the delays in picking up prescribed medications that waiting for PA approval can create.

It’s easy to get angry about PAs
Your best response, however, is to practice prudent and—within reason— cost-effective medicine. When generic or insurer-preferred medications are clini­cally appropriate and meet treatment guidelines, trying them first is sensible and defensible. If your patient fails the initial low-cost treatment, or if a low-cost choice isn’t appropriate, document this clearly and seek approval for a costlier treatment.⁹
 

BOTTOM LINE
Physicians have ethical and legal obligations to advocate for their patients’ needs and best interests. This sometimes includes completing prior authorization requests. Find strategies that minimize hassle and make sense in your practice, and seek efficient ways to document the medical necessity of requested tests, procedures, or therapies.
 

Acknowledgment
Drs. Marett and Mossman thanks Donna Vanderpool, MBA, JD, and Annette Reynolds, MD, for their helpful input in preparing this article.

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

Dear Dr. Mossman,

Where I practice, most health care plans won’t pay for certain medications without giving prior authorization (PA). Completing PA forms and making telephone calls take up time that could be better spent treat­ing patients. I’m tempted to set a new policy of not doing PAs. If I do, might I face legal trouble?

Submitted by “Dr. A”

If you provide clinical care, you’ve prob­ably dealt with third-party payers who require prior authorization (PA) before they will pay for certain treatments. Dr. A is not alone in feeling exasperated about the time it takes to complete a PA.¹ After spend­ing several hours each month waiting on hold and wading through stacks of paper­work, you may feel like Dr. A and consider refusing to do any more PAs.

But is Dr. A’s proposed solution a good idea? To address this question and the frus­tration that lies behind it, we’ll take a cue from Italian film director Sergio Leone and discuss:
   • how PAs affect psychiatric care: the good, the bad, and the ugly
   • potential exposure to professional liabil­ity and ethics complaints that might result from refusing or failing to seek PA
   • strategies to reduce the burden of PAs while providing efficient, effective care.

The good
Recent decades have witnessed huge increases in spending on prescription medication. Psychotropics are no excep­tion; state Medicaid spending for anti-psychotic medication grew from <$1 bil­lion in 1995 to >$5.5 billion in 2005.²

Requiring a PA for expensive drugs is one way that third-party payers try to rein in costs and hold down insurance premi­ums. Imposing financial constraints often is just one aim of a pharmacy benefit man­agement (PBM) program. Insurers also jus­tify PBMs by pointing out that feedback to practitioners whose prescribing falls well outside the norm—in the form of mailed warnings, physician second opinions, or pharmacist consultation—can improve patient safety and encourage appropriate treatment options for enrolled patients.^3,4 Examples of such benefits include reduc­ing overuse of prescription opioids⁵ and antipsychotics among children,³ misuse of buprenorphine,⁶ and adverse effects from potentially inappropriate prescriptions.⁷

The bad
The bad news for doctors: Cost savings for payers come at the expense of pro­viders and their practices, in the form of time spent doing paperwork and talking on the phone to complete PAs or contest PA decisions.⁸ Addressing PA requests costs an estimated $83,000 per physician per year. The total administrative burden for all 835,000 physicians who practice in the United States therefore is 868,000,000 hours, or $69 billion annually.⁹

To make matters worse, PA requirements may increase the overall cost of care. After Georgia Medicaid instituted PA require­ments for second-generation antipsychotics (SGAs), average monthly per member drug costs fell $19.62, but average monthly outpa­tient treatment costs rose $31.59 per mem­ber.10 Pharmacy savings that result from requiring PAs for SGAs can be offset quickly by small increases in the hospitalization rate or emergency department visits.^9,11

The ugly
Many physicians believe that the PA pro­cess undermines patient care by decreasing time devoted to direct patient contact, incen­tivizing suboptimal treatment, and limit­ing medication access.^1,12,13 But do any data support this belief? Do PAs impede treat­ment for vulnerable persons with severe mental illnesses?

The answer, some studies suggest, is “Yes.” A Maine Medicaid PA policy slowed initiation of treatment for bipolar disor­der by reducing the rate of starting non-preferred medications, although the same policy had no impact on patients already receiving treatment.¹⁴ Another study exam­ined the effect of PA processes for inpatient psychiatry treatment and found that patients were less likely to be admitted on weekends, probably because PA review was not avail­able on those days.¹⁵ A third study showed that PA requirements and resulting impedi­ments to getting refills were correlated with medication discontinuation by patients with schizophrenia or bipolar disorder, which can increase the risk of decompensation, work-related problems, and hospitalization.¹⁶

Problems with PAs
Whether they are helpful or counterpro­ductive, PAs are a practice reality. Dr. A’s proposed solution sounds appealing, but it might create ethical and legal problems.

Among the fundamental elements of ethi­cal medical practice is physicians’ obliga­tion to give patients “guidance … as to the optimal course of action” and to “advocate for patients in dealing with third parties when appropriate.”¹⁷ It’s fine for psychia­trists to consider prescribing treatments that patients’ health care coverage favors, but we also have to help patients weigh and evaluate costs, particularly when patients’ circumstances and medical interests militate strongly for options that third-party payers balk at paying for. Patients’ interests—not what’s expedient—are always physicians’ foremost concern.¹⁸

Beyond purely ethical considerations, you might face legal consequences if you refuse or fail to seek PAs for what you think is the proper medication. As Table 1 shows, one key factor is whether you are under contract with the patient’s insurance carrier; if you are, failure to seek a PA when appropriate may constitute a breach of the contract (Donna Vanderpool, written communication, October 5, 2014).

If the prescribed medication does not meet the standard of care and your patient suffers some harm, a licensing board complaint and investigation are possible. You also face exposure to a medical malpractice action. Although we do not know of any instances in which such an action has succeeded, 2 recent court decisions suggest that harm to a patient stemmed from failing to seek PA for a medication could constitute grounds for a lawsuit.^19,20 Efforts to contain medical costs have been around for decades, and courts have held that physicians, third-party pay­ers, and utilization review intermediaries are bound by “the standard of reasonable com­munity practice”²¹ and should not let cost limitations “corrupt medical judgment.”²² Physicians who do not appeal limitations at odds with their medical judgment might bear responsibility for any injuries that occur.^18,22


Managing PA requests
Given the inevitability of encountering PA requests and your ethical and professional obligations to help patients, what can you do (Table 2^9,23,27)?

Some practitioners charge patients for time spent completing PAs.²³ Although phy­sicians should “complete without charge the appropriate ‘simplified’ insurance claim form as a part of service to the patient;” they also may consider “a charge for more complex or multiple forms … in conformity with local custom.”²⁴ Legally, physicians’ contracts with insurance panels may pre­clude charging such fees, but if a patient is being seen out of network, the physician does not have a contractual obligation and may charge.⁹ If your practice setting lets you choose which insurances you accept, the impact and burden of seeking PAs is a factor to consider when deciding whether to par­ticipate in a particular panel.²³

In an interesting twist, an Ohio physi­cian successfully sued a medical insur­ance administrator for the cost of his time responding to PA inquiries.²⁵ Reasoning that the insurance administrator “should expect to pay for the reasonable value of” the doctor’s time because the PAs “were solely intended for the benefit of the insur­ance administrator” or parties whom the administrator served, the judge awarded the doctor $187.50 plus 8% interest.

Considerations that are more practi­cal relate to how to triage and address the volume of PA requests. Some large medi­cal practices centralize PAs and try to set up pre-approved plans of care or blanket approvals for frequently encountered con­ditions. Centralization also allows one key administrative assistant to develop skills in processing PA requests and to build rela­tionships with payers.²⁶

The administrative assistant also can compile lists of preferred alternative medica­tions, PA forms, and payer Web sites. Using and submitting requests through payer Web sites can speed up PA processing, which saves time and money.²⁷ As electronic health records improve, they may incorporate patients’ formularies and provide automatic alerts for required PAs.²³

Patients should be involved, too. They can help to obtain relevant formulary infor­mation and to weigh alternative therapies. You can help them understand your role in the PA process, the reasoning behind your treatment recommendations, and the delays in picking up prescribed medications that waiting for PA approval can create.

It’s easy to get angry about PAs
Your best response, however, is to practice prudent and—within reason— cost-effective medicine. When generic or insurer-preferred medications are clini­cally appropriate and meet treatment guidelines, trying them first is sensible and defensible. If your patient fails the initial low-cost treatment, or if a low-cost choice isn’t appropriate, document this clearly and seek approval for a costlier treatment.⁹
 

BOTTOM LINE
Physicians have ethical and legal obligations to advocate for their patients’ needs and best interests. This sometimes includes completing prior authorization requests. Find strategies that minimize hassle and make sense in your practice, and seek efficient ways to document the medical necessity of requested tests, procedures, or therapies.
 

Acknowledgment
Drs. Marett and Mossman thanks Donna Vanderpool, MBA, JD, and Annette Reynolds, MD, for their helpful input in preparing this article.

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

Dear Dr. Mossman,

Where I practice, most health care plans won’t pay for certain medications without giving prior authorization (PA). Completing PA forms and making telephone calls take up time that could be better spent treat­ing patients. I’m tempted to set a new policy of not doing PAs. If I do, might I face legal trouble?

Submitted by “Dr. A”

If you provide clinical care, you’ve prob­ably dealt with third-party payers who require prior authorization (PA) before they will pay for certain treatments. Dr. A is not alone in feeling exasperated about the time it takes to complete a PA.¹ After spend­ing several hours each month waiting on hold and wading through stacks of paper­work, you may feel like Dr. A and consider refusing to do any more PAs.

But is Dr. A’s proposed solution a good idea? To address this question and the frus­tration that lies behind it, we’ll take a cue from Italian film director Sergio Leone and discuss:
   • how PAs affect psychiatric care: the good, the bad, and the ugly
   • potential exposure to professional liabil­ity and ethics complaints that might result from refusing or failing to seek PA
   • strategies to reduce the burden of PAs while providing efficient, effective care.

The good
Recent decades have witnessed huge increases in spending on prescription medication. Psychotropics are no excep­tion; state Medicaid spending for anti-psychotic medication grew from <$1 bil­lion in 1995 to >$5.5 billion in 2005.²

Requiring a PA for expensive drugs is one way that third-party payers try to rein in costs and hold down insurance premi­ums. Imposing financial constraints often is just one aim of a pharmacy benefit man­agement (PBM) program. Insurers also jus­tify PBMs by pointing out that feedback to practitioners whose prescribing falls well outside the norm—in the form of mailed warnings, physician second opinions, or pharmacist consultation—can improve patient safety and encourage appropriate treatment options for enrolled patients.^3,4 Examples of such benefits include reduc­ing overuse of prescription opioids⁵ and antipsychotics among children,³ misuse of buprenorphine,⁶ and adverse effects from potentially inappropriate prescriptions.⁷

The bad
The bad news for doctors: Cost savings for payers come at the expense of pro­viders and their practices, in the form of time spent doing paperwork and talking on the phone to complete PAs or contest PA decisions.⁸ Addressing PA requests costs an estimated $83,000 per physician per year. The total administrative burden for all 835,000 physicians who practice in the United States therefore is 868,000,000 hours, or $69 billion annually.⁹

To make matters worse, PA requirements may increase the overall cost of care. After Georgia Medicaid instituted PA require­ments for second-generation antipsychotics (SGAs), average monthly per member drug costs fell $19.62, but average monthly outpa­tient treatment costs rose $31.59 per mem­ber.10 Pharmacy savings that result from requiring PAs for SGAs can be offset quickly by small increases in the hospitalization rate or emergency department visits.^9,11

The ugly
Many physicians believe that the PA pro­cess undermines patient care by decreasing time devoted to direct patient contact, incen­tivizing suboptimal treatment, and limit­ing medication access.^1,12,13 But do any data support this belief? Do PAs impede treat­ment for vulnerable persons with severe mental illnesses?

The answer, some studies suggest, is “Yes.” A Maine Medicaid PA policy slowed initiation of treatment for bipolar disor­der by reducing the rate of starting non-preferred medications, although the same policy had no impact on patients already receiving treatment.¹⁴ Another study exam­ined the effect of PA processes for inpatient psychiatry treatment and found that patients were less likely to be admitted on weekends, probably because PA review was not avail­able on those days.¹⁵ A third study showed that PA requirements and resulting impedi­ments to getting refills were correlated with medication discontinuation by patients with schizophrenia or bipolar disorder, which can increase the risk of decompensation, work-related problems, and hospitalization.¹⁶

Problems with PAs
Whether they are helpful or counterpro­ductive, PAs are a practice reality. Dr. A’s proposed solution sounds appealing, but it might create ethical and legal problems.

Among the fundamental elements of ethi­cal medical practice is physicians’ obliga­tion to give patients “guidance … as to the optimal course of action” and to “advocate for patients in dealing with third parties when appropriate.”¹⁷ It’s fine for psychia­trists to consider prescribing treatments that patients’ health care coverage favors, but we also have to help patients weigh and evaluate costs, particularly when patients’ circumstances and medical interests militate strongly for options that third-party payers balk at paying for. Patients’ interests—not what’s expedient—are always physicians’ foremost concern.¹⁸

Beyond purely ethical considerations, you might face legal consequences if you refuse or fail to seek PAs for what you think is the proper medication. As Table 1 shows, one key factor is whether you are under contract with the patient’s insurance carrier; if you are, failure to seek a PA when appropriate may constitute a breach of the contract (Donna Vanderpool, written communication, October 5, 2014).

If the prescribed medication does not meet the standard of care and your patient suffers some harm, a licensing board complaint and investigation are possible. You also face exposure to a medical malpractice action. Although we do not know of any instances in which such an action has succeeded, 2 recent court decisions suggest that harm to a patient stemmed from failing to seek PA for a medication could constitute grounds for a lawsuit.^19,20 Efforts to contain medical costs have been around for decades, and courts have held that physicians, third-party pay­ers, and utilization review intermediaries are bound by “the standard of reasonable com­munity practice”²¹ and should not let cost limitations “corrupt medical judgment.”²² Physicians who do not appeal limitations at odds with their medical judgment might bear responsibility for any injuries that occur.^18,22


Managing PA requests
Given the inevitability of encountering PA requests and your ethical and professional obligations to help patients, what can you do (Table 2^9,23,27)?

Some practitioners charge patients for time spent completing PAs.²³ Although phy­sicians should “complete without charge the appropriate ‘simplified’ insurance claim form as a part of service to the patient;” they also may consider “a charge for more complex or multiple forms … in conformity with local custom.”²⁴ Legally, physicians’ contracts with insurance panels may pre­clude charging such fees, but if a patient is being seen out of network, the physician does not have a contractual obligation and may charge.⁹ If your practice setting lets you choose which insurances you accept, the impact and burden of seeking PAs is a factor to consider when deciding whether to par­ticipate in a particular panel.²³

In an interesting twist, an Ohio physi­cian successfully sued a medical insur­ance administrator for the cost of his time responding to PA inquiries.²⁵ Reasoning that the insurance administrator “should expect to pay for the reasonable value of” the doctor’s time because the PAs “were solely intended for the benefit of the insur­ance administrator” or parties whom the administrator served, the judge awarded the doctor $187.50 plus 8% interest.

Considerations that are more practi­cal relate to how to triage and address the volume of PA requests. Some large medi­cal practices centralize PAs and try to set up pre-approved plans of care or blanket approvals for frequently encountered con­ditions. Centralization also allows one key administrative assistant to develop skills in processing PA requests and to build rela­tionships with payers.²⁶

The administrative assistant also can compile lists of preferred alternative medica­tions, PA forms, and payer Web sites. Using and submitting requests through payer Web sites can speed up PA processing, which saves time and money.²⁷ As electronic health records improve, they may incorporate patients’ formularies and provide automatic alerts for required PAs.²³

Patients should be involved, too. They can help to obtain relevant formulary infor­mation and to weigh alternative therapies. You can help them understand your role in the PA process, the reasoning behind your treatment recommendations, and the delays in picking up prescribed medications that waiting for PA approval can create.

It’s easy to get angry about PAs
Your best response, however, is to practice prudent and—within reason— cost-effective medicine. When generic or insurer-preferred medications are clini­cally appropriate and meet treatment guidelines, trying them first is sensible and defensible. If your patient fails the initial low-cost treatment, or if a low-cost choice isn’t appropriate, document this clearly and seek approval for a costlier treatment.⁹
 

BOTTOM LINE
Physicians have ethical and legal obligations to advocate for their patients’ needs and best interests. This sometimes includes completing prior authorization requests. Find strategies that minimize hassle and make sense in your practice, and seek efficient ways to document the medical necessity of requested tests, procedures, or therapies.
 

Acknowledgment
Drs. Marett and Mossman thanks Donna Vanderpool, MBA, JD, and Annette Reynolds, MD, for their helpful input in preparing this article.

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